In the matter of Fortescue Metals Group Limited [2006] FCA 826 | Case summary

In the matter of Fortescue Metals Group Limited [2006] FCA 826 | Case summary | Zoe

8 The procedure set out in Part IIIA of the Trade Practices Act is well known. It is a two-stage process. First there is an application to have a service declared. Service is defined in s 44B to mean "a service provided by means of a facility" and relevantly includes: "(a) use of an infrastructure facility such as a road or railway line". If a declaration is made this creates in favour of any person interested in obtaining access to the declared service an enforceable right to negotiate access to the service. The right is enforceable in the sense that, if the negotiations are not successful, the dispute may be resolved in arbitration conducted by the Australian Competition and Consumer Commission (ACCC). 9 To initiate the first stage any person (there are no rules as to standing) must apply to the National Competition Council (NCC) to recommend that a particular service be declared: s 44F. If the NCC recommends a declaration, the designated Minister (in these cases the Treasurer) must either declare the service or not declare it: s 44H(1). 10 On 11 June 2004 FMG made application to the NCC that it recommend making a declaration in respect of "the service provided by part of the Mt Newman railway line and part of the Goldsworthy line". The application described the service as: [5.1] (1) The use of the Facility, being: (a) that part of the Mt Newman railway line which runs from a rail siding that would be constructed near Mindy Mindy in the Pilbara to port facilities at Nelson Point in Port Hedland, and is approximately 295 kilometres long … and (b) the part of the Goldsworthy railway line that runs from where it crosses the Mt Newman railway line to port facilities at Finucane Island in Port Hedland, and is approximately 17 kilometres long…" (2) Access to the Facility's associated infrastructure, including, but not limited to: (a) railway track, associated structures, over or under track structures, supports (including supports for equipment or items associated with the use of the railway); (b) bridges; (c) passing loops; (d) train control systems, signalling systems and communication systems; (e) sidings and refuges to park rolling stock; (f) maintenance and protection systems; and (g) roads and other facilities which provide access to the railway line route. 11 The Mt Newman railway line was defined as "the railway line of approximately 425 kilometres in length which is currently used to carry ore from the mines in the Mt Newman area, as well as from Yandi, to Port Hedland." The Goldsworthy railway line was defined as "the railway line of approximately 210 kilometres in length which is currently used to carry ore from the mines near Yarrie to Port Hedland." 12 In its application FMG described the purpose for which it sought access as being, among other things, to transport its ore from Mindy Mindy and other mines to Port Hedland. FMG also made reference to its "foundation principle" that Pilbara rail infrastructure should be made more available to third parties. 13 The NCC determined that the use of the Goldsworthy line was not a "service" for the purposes of Part IIIA. Accordingly, it omitted that line from its consideration. 14 On 16 November 2007, The Pilbara Infrastructure Pty Ltd (TPI), a wholly owned subsidiary of FMG, made application to the NCC that it recommend the making of a declaration in respect of "the Facility comprising the Goldsworthy railway from a location near Yarrie, at one end, to a location near Finucane Island within the port of Port Hedland, at the other end, and all points in between." Access was also sought to associated infrastructure. The application stated that TPI seeks access to the Goldsworthy service in order to be able to offer an above rail haulage service to mining companies seeking to move bulk materials between any two points on the Goldsworthy line including, without limitation, interconnection points with other rail networks and railway lines. 15 On 16 November 2007 TPI made application to the NCC that it recommend the making of a declaration in respect of the Hamersley rail network comprising: · The rail line from Paraburdoo to Dampier, including all points in between; · The railway line from Yandicoogina to Rosella siding, including all points in between; and · The railway line from Brockman No 2 to Rosella siding, including all points in between. Access to associated infrastructure was also sought. 16 The application stated that TPI sought access to the Hamersley service in order to be able to offer an above rail haulage service to mining companies seeking to move bulk materials between any two points on the Hamersley rail network including, without limitation, interconnection points with all other rail networks and railway lines. 17 By letter dated 31 January 2008 TPI advised the NCC that, for the avoidance of doubt, the facilities to which TPI sought access were: (a) the railway line from Parraburdoo to Dampier which is approximately 385km long; (b) the railway line from Yandicoogina to Rosella siding (on the Parraburdoo to Dampier railway) which is approximately 195 km long; and (c) the railway line from Brockman No 2 to Rosella siding which is approximately 45km long. 18 On 18 January 2008 TPI made application to the NCC that it recommend the making of a declaration in respect of "the use of the facility comprising the Robe railway from a location near Mesa J to Cape Lambert and all points in between". That service was defined to include the use of all associated infrastructure. 19 The application stated that TPI sought access to the Robe service in order to offer an above rail haulage service to mining companies seeking to move bulk materials between any two points on the Robe railway.

23 Section 44K(1) provides that if the Minister declares a service, the service provider may apply to the Tribunal for a review of the declaration. Section 44K(2) provides that, if the Minister decides not to declare a service (which would include a deemed decision not to declare), the applicant for the declaration recommendation may apply to the Tribunal for a review of the Minister's decision. 24 A review by the Tribunal is a reconsideration of the matter (s 44K(4)) and for that purpose the Tribunal has the same powers as the Minister (s 44K(5)). That means that the Tribunal must reconsider each application afresh. This allows the parties to put before the Tribunal for its consideration any material that may be relevant to the issues raised, whether or not that material was before the Minister. In addition, the rules of procedural fairness require the Tribunal to afford a party which may be adversely affected by its decision, the right to be heard, to be legally represented at a hearing before the Tribunal and to lead evidence and cross-examine witnesses. Speaking very generally, the Tribunal is master of its own forms and procedures. But the rules of procedural fairness act as a strong brake on the Tribunal's ability to control the parties' conduct in a proceeding. One consequence is that proceedings before the Tribunal have every appearance of a court-style hearing. 25 Though there are four applications for review (one in respect of each decision) the parties suggested, and the Tribunal agreed, that in view of the overlapping issues as well as the commonality of much of the evidence, it would be more expeditious and less costly were the reviews to be heard concurrently. 26 At the joint hearing, the parties (applicant and service providers) took the opportunity to present material far in excess of that which had been placed before the Minister. In all, the parties filed 130 affidavits from 73 witnesses, together with a large number of documents. This material took up approximately 70 large lever arch files. The transcript of the hearing runs for over 3,300 pages. Of the witnesses, 15 were expert economists and 29 were, in alphabetical order, bankers, computer simulation experts, engineers, environmental scientists, geologists, metallurgists, quantity surveyors, rail modellers and train schedulers, among others. 27 In light of the number of witnesses and the volume of material, the Tribunal took the view that the parties should be allocated set times within which to question witnesses and present their respective arguments. A suitable arrangement was worked out. By and large the parties adhered to their allotted times. Nonetheless the hearing occupied 42 sitting days. 28 Prior to the hearing the Tribunal read and considered the evidence of all the experts. It decided that it would not be assisted were all of them to be cross-examined. Accordingly, the Tribunal identified those experts who were not to be called. While not called, their evidence has been taken into account. 29 In another step to expedite the hearing as well as to better understand the evidence of the experts, the Tribunal considered that, where possible, the experts should be grouped by topic and called as a group. In addition the Tribunal decided that, in the first instance, it would question the economists and the capacity modellers without intervention by the parties' lawyers, and that cross-examination would only be permitted by leave and then be confined to a short period of approximately one hour per party. The parties largely cooperated in this enterprise. In the result, the evidence of those experts was dealt with in just six days. In the case of some there was little cross-examination. A number were not questioned by some parties at all. 30 The large volume of material that has been placed before the Tribunal led to a submission by FMG, strongly supported by the NCC, that much of it was irrelevant to the first stage (whether or not there should be a declaration), although it may be relevant at the second stage (whether access should be granted and, if so, on what terms). The Tribunal accepts that a large part of the material is relevant to stage two. But, in the view of the Tribunal, almost all the material was, to a greater or lesser extent, also relevant to the stage one inquiry. This view was fortified by the failure of FMG and the NCC to identify, even in broad categories, which material should be disregarded, despite several requests that they do so.

2.2.2.2 Physical properties 48 The mining of iron ore produces particles of varying sizes. Particle size directly bears on the smelting process, since a blast furnace operates with the greatest efficiency where the particle size allows easiest passage of gas between particles (if iron ore becomes compacted, the gas will not circulate properly and the furnace will become choked, adversely affecting output and quality). In a worst case scenario, this can result in the blast furnace becoming inoperable, resulting in a shutdown of the steel mill. 49 Iron ore can be sold as lump, fines or pellets. 50 "Lump" generally refers to iron ore material with a particle size of 6mm or more. Generally, lump can only comprise 20% or less of the iron product being fed into a blast furnace. Lump products can be fed directly into a blast furnace provided they do not exceed maximum size specifications and are as free as possible of adhering fines material that may have been generated during handling. Accordingly, lump products must be well screened to remove such material before shipment and again at the steel plant before feeding into the furnace. Such fines removed during screening will be used to produce sinter (defined below). The percentage of fines that needs to be removed in screening is a significant factor considered by purchasers of lump iron ore. 51 "Fines" generally refers to iron ore material with a particle size of less than 6mm. There are two types of fines product: concentrate fines and sinter fines. Fines cannot be fed directly into a blast furnace; rather, sinter fines need to be agglomerated (fused) into sinter and concentrate fines need to be processed into pellets before they can be used by a steel mill. Fines products are generally sold at lower prices than lump or pellet products, to reflect the need for this further processing. 52 "Sinter" is formed by agglomerating fine particles of iron ore. Agglomerating involves mixing fines, coke breeze (small particles from the coke screening plant), fine fluxes (such as limestone and dolomite), mill scale (small flakes of iron oxide from the steel rolling process), dust from the blast furnace and BOS furnace exhaust gases (containing fine iron oxide particles) to produce new chemical forms of iron. This iron is then broken into lumps, cooled and screened, forming a product ranging from 5mm to 50mm in size, which is then sent to the blast furnace. As sinter tends to disintegrate while being transported, steel mills typically have their own sintering plants. Sinter is usually re-screened at the blast furnace immediately before it is fed to remove fine material generated during transport from the sinter plant. 53 Steelmakers can alter the type of fines used to make sinter, including where there is a supply problem or shortage of a particular fines product, a change in the chemistry or delivered cost of ore, or a change in the burden (the combination of lump, pellets and fines). Sometimes a steelmaker must alter its blend quickly, due to sudden or unexpected changes in supply conditions. In other cases the steelmaker may be able to introduce changes over a longer time frame (ordinarily one to two years). 54 "Pellets" are a form of high-grade feedstock produced from concentrate fines (either hematite or magnetite). Pellets are produced through a process which involves combining finely ground iron ore with binding clay (and fluxes if required), moistening and rolling the mixture into 14mm balls, and baking the pellets. The ore used is normally ground finely due to previous beneficiation processes. This process produces pellets of high-grade, and consistent size, shape and density, which are large enough to be directly fed into a blast furnace. Although pellets are high quality and, because of their metallurgical properties, are efficient in the blast furnace process, the additional processing required to produce them makes them more costly than lump or fines products. 55 Pellets are normally produced near the mine site by iron ore producers. However, some steelmakers also produce pellets (eg OneSteel has a pellet plant at Whyalla in South Australia).

2.2.3.2 Product quality 61 Iron ore producers work to tight product quality specifications for their lump and fines products. Most iron ore products for direct shipping or blending have: (a) iron content of between 60% and 65%, although the highest sales volumes of Pilbara iron ore in the globally traded market have an iron content in the range of approximately 57-58%; (b) alumina values of less than 3%: 2.7% alumina remains the highest alumina content standalone bulk iron ore product in the main contract sector of the globally traded market, although the ability to blend means that small quantities of ore with more than 3% alumina could be sold; and (c) phosphorus content of between 0.02% and 0.08%. 62 Iron ore supply contracts specify strict quality or "grade" requirements for the iron ore supplied. These are the outside limits and customers expect product grade specifications to fall within much tighter specifications. These contracts apply cost penalties if specific quality constraints are not met in individual shipments. If the deviation is not significant, neither is the penalty. During recent strong demand for iron ore exports, there has been a slight, although not major, relaxing of product quality acceptable to the steel mills. However, there have been no significant sales of high impurity products in the contract market. The position is different in the spot market. 63 Lower quality iron ores are generally less attractive to steelmakers, as they contain less iron and inevitably higher levels of impurities, or higher LOI characteristics. Accordingly, lower quality ores impact on the supplier's revenue. Depending on the ore type, lower quality ores will have a lower price than high-grade ores. Further, shipping costs are higher for lower quality ores per tonne of actual iron content (as international bulk shipping costs are tonnage driven) and a further freight cost is imposed as a result of the greater LOI of lower quality ores, as the chemically combined water the LOI represents will be driven off in the steelmaking process. The obvious implication is that there is a "baseline" quality below which ore of lesser quality will not be acceptable to customers. 64 Nonetheless, lower quality iron ores are purchased by steelmakers who then blend it with higher quality iron ore to create their sinter feed. Typically between four and eight different fines or concentrates may be added to a sinter plant feed blend. Thus, small quantities of lower quality ores will be procured.

2.2.3.6 Substitution of lump, fines and pellets 71 Each steel producer is likely to use a blend of lump, pellets and/or fines, depending on the location of its steel mill and the economics of its steelmaking process. The precise combination of lump, fines and pellets which a steelmaker uses is referred to as the burden, and depends on what is most efficient from the perspective of that steelmaker, having regard to a number of factors, including cost, sintering capacity, geographical location, blast furnace efficiency considerations and the quality of the steel to be produced. 72 The location of customers' steel mills also affects the value in use of individual iron ore products. For example, because lump ore breaks down during transport, lump products may not offer the same value to inland steel producers as to customers located at or near ports. Similarly, pellets (which are high cost) tend to be used by steelmakers where the cost of shipping alternative ores is higher, where the steelmaker needs to compensate for lower grade ores, or where the steelmaker's sinter plant has insufficient capacity to meet its requirements. Consequently, pellet products are principally produced in North America and Europe (primarily for local use) and Brazil, both for export (mainly to Europe) and local use. Pellet production in Australia is very low, relative to lump and fines product. 73 If the price of any of lump, fines or pellets was decoupled from the price of the others for any significant period of time, steel producers would consider optimising their processes to reduce the impact of this shift in value, for example, by investing in additional sintering or pelletising plant or altering the use of lump in favour of pellets. 74 It is possible to change the burden by investing in additional sinter capacity, pelletising capacity, and/or improved blast furnace feeding facilities. Any significant changes to the burden are likely to involve considerable investment in new plant, so that decisions by a steel producer about whether to change the mix of iron ore types comprising the burden used in its blast furnaces are not taken lightly. Once steel producers invest in their own pelletising or sintering plant near their steel mills, it will usually be most cost effective to purchase fines in order to take maximum advantage of this capacity. However, a steel producer operating a pelletising or sintering plant at full capacity may find the use of lump more cost effective than investment in additional facilities.

2.3.2 Demand for iron ore 77 Prior to the Second World War, iron ore mining and production tended to be undertaken by small producers located near to, and usually integrated with, steel mills. There are still considerable domestic iron ore reserves in a number of steel producing countries, including Russia, the United States, India and China. 78 The rapid growth in demand from the Japanese steel industry provided momentum for the rapid development of internationally traded iron ore in the 1960s and 1970s (notably, Japan does not have a domestic iron ore industry). This was particularly the case for iron ore sourced from the Pilbara, which is geographically in an advantageous position to service Asian customers, minimising shipping costs compared with longer hauls by competitors from, for example, Brazil. The facilities of BHPB and RTIO were progressively expanded to service the growing demand. 79 China began to import high-grade iron ore from Australia in the 1970s. Since then, Chinese demand for imported iron ore has increased progressively. China has become the world's major importer of iron ore and its iron ore imports have been driven by high compound annual growth rates for crude steel production. 80 The global steel industry is now dominated by a modest number of extremely large steelmakers who often act as a kind of national or regional "champion" in their country or region. On the other hand, in some countries, such as China and India, there are still a considerable number of smaller steel mills who tend to participate at the edges of the global steel market. The development of such large steel producers has been the result of a process of consolidation which has occurred in two phases. The first phase occurred relatively slowly over the last two decades prior to 2004. The second phase was more rapid and involved aggressive market consolidation since around 2004. As well as mergers, a number of steel manufacturers are also looking to exploit economies of scale and scope in other ways, such as through shared production, joint procurement or technology transfer arrangements. 81 The global trade in iron ore can now be understood in terms of two largely distinct sources of demand: China and the "rest of the world" (including other countries which have traditionally had strong and substantial steel industries, such as Japan, Taiwan, South Korea, Western Europe, Australia and the United States). Since 2002, although demand from the rest of the world has grown moderately, China's rapid industrialisation has been the source of remarkable growth in steel consumption and substantial increases in demand for, and the price of, iron ore. 82 Imported iron ore constitutes, on average, around half of the iron ore used globally (across all of lump, fines and pellets). This highlights the important role that continues to be played by domestic production in a number of countries and, in particular, in China. China has significant domestic iron ore deposits, which provide around two thirds of its iron ore supplies. Most Chinese domestic reserves are of relatively low quality ore, which requires beneficiation before it is suitable for use by Chinese steel producers. 83 As well as China, the other country with a sizeable and growing stake in global iron ore consumption and production is India. The total production of iron ore in India during 2008 was approximately 220mt, with production expected to rise to 240mt in 2009. Of this total production, approximately 86mt was exported and the remainder was consumed by domestic Indian steelmakers. The majority of Indian iron ore is hematite ore sold as sinter fines.

2.3.3 Supply contracts and pricing 84 Historically, most of BHPB's and RTIO's customers purchase iron ore under long-term supply arrangements. Approximately 90-95% of BHPB's iron ore products are sold through contracts of between 5 and 15 years duration. RTIO also sells most of its iron ore under long-term agreements. Recently there has been a move towards short-term supply contracts by both firms, reflecting the higher price that is currently being obtained in the spot market. 85 Supply contracts tend to contain: (a) a commitment as to the annual tonnages to be purchased and shipped; (b) the specific chemistry and physical properties of the iron ore to be supplied, usually including at least minimum iron content, physical size specifications, maximum levels of impurities and maximum LOI - these grades are often amended during annual benchmark pricing discussions, typically due to changes in the producer's ore or facilities, the customer's production plans or other iron ore sources; (c) variations to payments to reflect the grade of ore actually provided; (d) a "Base Price" for iron ore supplied during the first year of the contract and a mechanism for varying the Base Price in subsequent years to reflect annual benchmark price negotiations; and (e) provisions dealing with liability in the event of any delay to a shipment. 86 The benchmark price negotiations occur annually between iron ore producers and their customers to determine movements in the benchmark price. Such movements are negotiated through a series of annual meetings held in Europe and Asia between individual iron ore producers and steel producers. The credibility of a supplier or buyer for the purpose of benchmark negotiations reflects the generally held view of the industry based on the size of the parties to an agreement and the significance of the contracts they are negotiating. 87 The steel market has traditionally participated in the benchmark negotiation process as regional "blocks", but in recent years, China has operated to a large extent independently of the other major Asian players. 88 The annual price settlement, in the form of a percentage increase or decrease from the benchmark price of iron ore in the previous year, is usually set by the first agreement reached between a significant buyer and a significant seller anywhere in the world. This first settlement can be reached at any time from December through to as late as September the next year. 89 The pricing signals for iron ore tend to become blurred by a complex range of factors, such as uncertainty about global supply conditions (including the timing of planned capacity expansions), the state of forecast supply and demand, market development considerations (such as placing new or additional tonnage in the market), the relationship dynamics between a supplier and each customer, and changes in the costs of transportation. The result of this uncertainly is that benchmark price negotiations tend to be a mix of backward-looking analysis of the market conditions and forward projections of supply and demand. This complex balancing act is then reflected in a view as to the appropriate benchmark price for the next 12 months.

117 The first documented recognition of the iron ore potential of the Pilbara occurred in 1889. However, in 1938 the Commonwealth government placed an embargo on the export of iron ore, and little exploration for iron ore in the Pilbara is recorded between 1889 and the activity of some independent prospectors in the 1950s (notably, Lang Hancock and his associate Peter Wright, and Stan Hilditch and his associate Charles Warman). 118 The lifting of the export embargo in 1960 enabled the state of Western Australia to grant tenements with rights to explore for iron ore. This coincided with rapidly growing global demand for high-grade iron ore to meet sustained increases in global steel production, primarily due to post-World War II reconstruction and industrial development in South-East Asian countries. These factors led to the establishment of the Pilbara as one of the world's major iron ore production provinces. 119 Between 1962 and 1964, the Geological Survey of Western Australia (GSWA) conducted systematic regional geological mapping campaigns for the whole of the Hamersley Province to record the distribution of the rock-type units, including the iron-rich units (formations) and their associated geological structures (folds and faults). These GSWA geologists discovered additional hematite-goethite enrichment in the Brockman, Marra Mamba, and Boolgeeda Iron Formations and recorded the presence of iron ore deposits in the Brockman Iron Formation at Jimblebar. 120 Agreements were executed between prospectors and large mining companies, whereby the mining companies brought in partners who contributed to risk-sharing or technical know-how, and the prospectors were rewarded with private royalties from the developments. These partnerships supported further exploration and discoveries of iron ore, and led to the founding of new iron ore mining companies such as Hamersley Iron and the Mt Newman Mining Company. 121 Large-scale iron ore mining operations subsequently commenced at Mt Tom Price (mined by Hamersley Iron) and Mt Whaleback (mined by the Mt Newman Mining Company). Those deposits were of exceptional size, and consisted of high-grade, low-impurity iron ore. 122 There were also some limited development of smaller scale, high quality resources in the 1960s. 123 Many mining operations were established pursuant to agreements between the mining company and the State of Western Australia. The agreements (which were approved by Acts of the Western Australian Parliament) granted the mining company the right to a mineral lease to develop a particular iron ore deposit, and to construct a railway between the mining area and the port on special leases of railway corridors to be granted under the Land Act 1933 (WA). In return, the mining company assumed certain obligations with respect to the development of the mining area, including obligations regarding the construction of a mine, the payment of royalties for the ore and rent payments for the lease. The state agreements contained an obligation on the mining company to transport the passengers and carry the freight of the State and of third parties on the railways, provided this could be done without unduly prejudicing or interfering with the mining company's operations. 124 The first State agreement, ratified by the Iron Ore (Mt Goldsworthy) Agreement Act 1962 (WA), was in respect of the Goldsworthy project at Mt Goldsworthy, which is approximately 100km east of Port Hedland. The joint venture parties to the agreement (which did not include BHPB at that time) developed a mine at Mt Goldsworthy and constructed the Goldsworthy railway from Mt Goldsworthy to Port Hedland. In 1984 BHPB acquired an interest in the joint venture. In 1990 it acquired a 100% interest in the venture and sold 15% to two Japanese companies, CI Minerals and Mitsui. 125 The mines developed by the joint venturers are east of the original Mt Goldsworthy mine, at Nimingarra, Yarrie and Cattle Gorge. Of these, Cattle Gorge is the only project currently being mined. The Goldsworthy railway was extended to carry ore from those sites to Finucane Island at Port Hedland. 126 The second state agreement was in respect of the iron ore deposits at Tom Price in the Hamersley Ranges. It was ratified by the Iron Ore (Hamersley Range) Act 1963 (WA). The Hamersley project, which was developed by CRA Limited (now Rio Tinto Limited), involved establishing a mine at Tom Price, a shipping port at Dampier and the construction of an approximately 274km long railway between Tom Price and Dampier. Later the line was extended to haul ore from mines at Paraburdoo, Eastern Range, Marandoo, Yandicoogina, West Angelas, Hope Downs 1, Nammuldi and Brockman No 2. All the mines are either wholly owned by RTIO (or a subsidiary) or by joint ventures in which RTIO has a majority or controlling interest as well as project management rights. 127 The third State agreement ratified by the Iron Ore (Mt Newman) Agreement Act 1964 (WA) was entered into with a consortium of mining companies, excluding BHPB, following the discovery of iron ore at Mt Whaleback, some 400km south of Port Hedland. Pursuant to that agreement, the Mt Newman line was built. 128 The Mt Newman mining area now consists of the mines at Mt Whaleback, Orebodies 29 and 30, Orebody 18, Wheelara (Jimblebar) and Orebodies 23 and 25. A rail spur has been constructed from the Newman mine to Jimblebar. There are mines west of the Mt Newman line at Yandi and Mining Area C which are connected by a rail spur running from the Yandi junction. 129 BHP joined the original consortium in 1966. Now the mines are owned by several differently constituted joint ventures in which BHP has a majority interest. BHPB is the manager of each joint venture and in that capacity operates the Mt Newman line. 130 Finally there is the Robe River Valley deposit that was discovered by Cleveland Cliffs Iron Company (an American iron ore producer). The development of this deposit was approved by the State agreement which was ratified by the Iron Ore (Robe River) Agreement Act 1964 (WA). The Robe River deposit is located near Pannawonica. Pursuant to that agreement, the Robe line was built. In 1998, the Robe River joint venture expanded its operations by developing the West Angelas deposit. Ore from this mine runs up the Hamersley line to Western Creek then joins the Robe line.

2.7.1 The legislative requirements 131 The ability to prospect and explore for minerals, as well as the right to mine (extract) minerals from the land in Western Australia, is governed by the Mining Act 1978 (WA). The Act established three basic mining tenements, two being concerned with investigating claims (a prospecting licence and an exploration licence) and one concerned with production (a mining lease). Prospecting is a smaller-scale, lower cost version of exploration, and is generally undertaken before exploration to assess whether exploration is economically viable. Provision is also made for two ancillary tenements, a general purpose lease (for the use of land in connection with mining operations, eg for storage purposes) and a miscellaneous licence (permitting eg construction of roads, railways, pipelines and bridges). By later amendment, the retention licence was created to preserve the right to apply for a mining lease where mining the site is impracticable for a period of time. 132 A prospecting licence authorises the holder to enter upon land for the purpose of prospecting for minerals except iron ore. Prospecting for iron ore requires Ministerial consent. The licence holder may conduct works necessary for prospecting. The area over which a prospecting licence operates must not exceed 200 hectares. The licence remains in force for four years. If granted prior to 10 February 2006, the licence may not be extended. If granted after that date, it may be extended for a further term of four years. 133 The holder of a prospecting licence must comply with expenditure conditions. The requirement is to spend at least $40 per hectare per year, with a minimum expenditure of $2,000 per year. The licence may be transferred at any time. 134 An exploration licence authorises the holder to enter land for the purpose of exploration for minerals other than iron ore and to undertake all necessary works for that purpose. Exploration for iron ore requires Ministerial consent. 135 An exploration licence is granted for a term of five years. If granted prior to 10 February 2006, the licence may be extended for one or two years and, if granted after that date, may be extended for five years and further periods of two years. 136 The maximum area covered by an exploration licence varies from 196 square kilometres to approximately 231 square kilometres, dependent upon the location of the licence area. 137 Prescribed expenditure conditions require the holder of an exploration licence to expend $1,000 per block (roughly three square kilometres) in the first three years, $1,500 per block in the next two years, $2,000 per block in each of years six and seven and $8,000 per block each year thereafter. 138 In the first year of the term, an exploration licence may not be assigned, save in the case of death or bankruptcy. Thereafter the licence is freely assignable. 139 A mining lease is granted to permit the holder to mine for and dispose of minerals discovered by prospecting or exploration. Once again, permission to mine for iron ore requires Ministerial consent. A mining lease remains in force for 21 years. The holder has an option to renew for a further 21 years and the Minister has discretion to renew the term for further periods of up to 21 years. Expenditure conditions are imposed unless an exemption is granted.

2.8.3 Mineral resources 150 A mineral resource is a concentration or occurrence of material of intrinsic economic interest in or on the Earth's crust in such form, quality and quantity that there are reasonable prospects for eventual economic extraction. Reasonable prospects for eventual extraction requires a judgment by a "competent person" in respect of technical and economic factors likely to influence the prospects of economic extraction. It is a realistic inventory of mineralisation which, under assumed and justifiable technical and economic conditions, might, in whole or in part, become economically extractable. 151 An inferred mineral resource is that part of a mineral resource for which tonnage, grade and mineral content can be estimated with a "low level of confidence". It is inferred from geological evidence and assumed (but not verified) geological and/or grade continuity. 152 An inferred mineral resource is identified with a lower level of confidence that an indicated mineral resource. This category is intended to cover situations where a mineral concentration has been identified, and limited measurements and sampling completed, but where the data is insufficient to allow a confident interpretation of geological and/or grade continuity. Given the speculative nature of the information, technical or economic studies laying claim to inferred resources should be read cautiously. 153 An indicated mineral resource is that part of a mineral resource for which tonnage, densities, shape, physical characteristics, grade and mineral content can be estimated with a "reasonable level of confidence". 154 An indicated mineral resource has a lower level of confidence than that applying to a measured mineral resource, but a higher level of confidence than that applying to an inferred mineral resource. Indicated mineral resources are supported by sufficient data from which geological and/or grade continuity may be assumed (but not confirmed) and economic viability evaluated. 155 A measured mineral resource is that part of a mineral resource for which tonnage, densities, shape, physical characteristics, grade and mineral content can be estimated with a "high level of confidence". 156 Measured mineral resources are supported by sufficient data leaving no reasonable doubt that the tonnage and grade of the mineralisation can be estimated to within close limits. Economic viability may be evaluated with a greater degree of certainty than an evaluation based on an indicated mineral resource. However, mineral resource estimates are not precise calculations and are dependent on the interpretation of limited information. Classification of mineral resources as inferred, indicated or measured will depend largely on the quality, quantity and distribution of data available.

3.1.1 Mt Newman Joint Venture operations 164 The Mt Whaleback mine, together with Orebody 29 and Orebody 30, provides the bulk of the production for the Mt Newman Joint Venture mining operations. Mt Whaleback itself is the biggest open cut iron ore mine in Australia with approximate dimensions of 5.5km long by 1.2km wide. 165 Orebodies 29 and 30 are located in close proximity to the Mt Whaleback mine. There is no operational infrastructure at Orebodies 29 and 30. The ore from these mines is hauled by truck to the crusher at Mt Whaleback. Ore from the mines located at Orebodies 23 and 25, Orebody 18 and Wheelara (Jimblebar) is blended with the ore from Mt Whaleback and Orebodies 29 and 30 at Port Hedland. 166 There is a higher degree of complexity in the production of the Mt Newman joint venture products principally as a result of: (a) the use of different ores from a number of different mine sites spread over 70km; (b) the number of different ore types that make up the Mt Newman joint venture products; (c) the range of ore grades from each of the mines, which includes both iron content and contaminants; (d) wet weather; and (e) the fact that the blending of ores occurs approximately 420km away at Port Hedland. 167 Changes to reduce this complexity are being implemented. They involve relocating the crushing, screening and blending activities from Port Hedland back to the mine. When that occurs ore that is mined in the Mt Newman mining area will be processed into final product size within the Mt Newman mining area, rather than being crushed and screened at the port. The ore will then be blended in stockpiles at Mt Whaleback prior to railing to port for stockpiling and shiploading. 168 In addition: (a) ore from Orebody 18 and Jimblebar will be railed to Mt Whaleback for crushing and screening and blending (along with Mt Whaleback ore and ore from satellite Orebodies 29 and 30); (b) ore from Orebody 25 (along with ore from Orebody 23) will be loaded as lump and fines and then railed to Mt Whaleback for blending in trains with ore from the Mt Newman hub; and (c) once blended with the ore from Orebody 25, Orebody 18, Jimblebar, Mt Whaleback 29 and Mt Whaleback 30, this ore will be railed directly to Port Hedland for stockpiling and shiploading.

178 Iron ore mining at BHPB's operations includes the following tasks (although not necessarily in the sequence below and not all of which are carried out at all of the mines): (a) drilling and blasting; (b) load and haul (of waste and ore); (c) crushing and screening - ore is crushed and screened at various stages of the operations to allow efficient handling and processing downstream in the production process, or to produce ore that is of a size acceptable to customers. All ore undergoes at least primary crushing at the mine prior to railing, and some ore will undergo secondary and even tertiary crushing to meet size specifications. The crushing and screening, for example, at Area C, Yandi and Orebodies 23 and 25 allows for the production of finished product sizes at the mines; (d) beneficiation - beneficiation is a process by which medium-grade ore (where waste is mixed in with high-grade ore) is treated to remove the waste and thereby allow the high-grade ore to be included in the Mt Newman high-grade blend iron ore product. On occasion, high-grade ore is processed through the plant to produce a product with low impurities to overcome specific impurity issues within the overall Mt Newman high-grade product at the time; (e) mine post-crusher stockpiling - after processing at the mines, ore is stockpiled prior to railing. Stockpiling operates as a buffer between mine production and train loading and ensures the continuation of the mine production processes when no rakes (sets of ore cars) are present to be loaded as well as the availability of sufficient ore to load the rake at a maximum rate when it does arrive; (f) stockpiling at the port - the type of stacking process used depends on the amount of blending that has been carried out at the mine and therefore the amount of blending required at the port to produce acceptable inter-cargo variability. When current expansions are completed, the last in first out (LIFO) stacking approach will apply to all ore stacked at port. The LIFO stacking process should significantly reduce the complexity of the stockpiling operations at the port because: (a) dumped ore can be directed to any stockpile containing the correct product for stacking (provided that the stockpile has room to accept ore); and (b) ore can be reclaimed from any stockpile containing the product that is required to be loaded onto the ship; (g) reclaiming - stockpiles are reclaimed utilising three equal horizontal cuts (known as benches) of the stockpile profile at a right angle to the length of the stockpile; (h) ship loading - under current development plans, approximately [commercial in confidence (c-i-c)]% of product will be loaded directly from trains to ships. This will increase to approximately [c-i-c]% of product under Rapid Growth Project 5 (RGP5): see Chapter 3.5.4. The remaining portion of product will continue to be loaded onto ships from stockpiles at port; and (i) blending - BHPB utilises different types of blending, including: (i) dedicated blending (such as when stacking ore), which is used for the purpose of grade control and the reduction of variability; and (ii) inherent blending, which occurs as a result of the handling, movement, stockpiling, processing and shiploading of ore.

3.3.1 The port of Port Hedland 179 Port Hedland is Australia's largest iron ore port, and one of the largest in the world, by annual tonnage throughput. BHPB iron ore constitutes approximately 96% of the cargo (by weight) that is shipped from the port. The port is serviced by a single shipping channel, which is not wide enough to allow ships to pass each other. 180 BHPB operates two separate port facilities on either side of Port Hedland harbour at Nelson Point and Finucane Island. 181 The principal equipment and infrastructure used by BHPB at Nelson Point comprises three car dumpers, two crushing and screening plants, two stockyards, six stackers and four reclaimers, a lump re-screening plant, two berths and various conveyers. At Finucane Island, there is one car dumper, two stockyards, one lump re-screening plant, two berths and conveyors. Car dumpers unload ore from ore cars at the port. 182 There is a 1.4km under-harbour tunnel between Nelson Point and Finucane Island. Some ore that is dumped at Nelson Point is conveyed to Finucane Island through this tunnel. 183 The ability to load and therefore ship iron ore is dependent on the following key constraints: (a) the availability of product at the port; (b) the ability to load product onto ships; (c) the availability of ships; (d) windows for sailing; and (e) channel and port capacity (which includes the availability of tugs and pilots). 184 Any interruptions to the performance of any element of what is substantially an integrated production system (from mine to port) may, for a short period, be addressed through "buffering" at the port; ie maintaining a sufficient inventory of product through the use of existing stockpiles of ore at the port to allow ore to be loaded and shipped in the event of any disruption to one or more of the inflow components of its system. But the mine, rail and port operations do not have an infinite amount of buffering capacity, especially at the port. 185 In practical terms, BHPB's port buffer is equivalent to approximately one to two weeks' demand for each of the products produced by BHPB. 186 The pattern of ship arrivals at Port Hedland is unpredictable and demand for particular products can vary significantly at any particular point in time. A queue with a minimum of six to eight ships must be waiting to enter the port at all times in order to provide BHPB with some flexibility as to the berthing order, which enables BHPB to "smooth" demand for particular products in the short-term. Approximately 50% of vessels are berthed outside their order of arrival in the queue. 187 Any shortage of supply at the port can result in changes being made to BHPB's rail operations at short notice to try and accommodate this demand.

3.4.1 Trains 204 A set of ore cars is referred to as a "rake". Rakes on the Mt Newman line generally comprise 112 ore cars. Rakes on the Goldsworthy rail line vary from 38 to 106 ore cars. More than one rake may be attached to a train, though rakes are always separated by one or more locomotives. BHPB presently uses a combination of two rake and three rake trains on the Mt Newman line. Trains are often referred to as "consists". 205 There are three layers to the train control system: (a) a solar-powered electronic interlocked signalling system, which prevents two trains from entering the same portion of track, distributed along the track in wayside cabinets; (b) a remote train control system, which allows train controllers to remotely monitor and remotely control the signalling and movements of trains and rail vehicles; and (c) wayside asset protection equipment, which monitors passing trains for problems such as overheated wheels and bearings, dragging equipment and wheel impacts, as well as weighing the individual wheels of ore cars and locomotives as they pass (in order to monitor unevenly loaded or overloaded ore cars). 206 These train control systems are supplemented by Automatic Train Protection (ATP), a failsafe microprocessor-controlled collision avoidance system. ATP is composed of three key elements: (a) a train location system, which involves the interaction of hardware located onboard locomotives and equipment mounted adjacent to the track; (b) signal detection equipment; and (c) microprocessors onboard locomotives which monitor the train's location, speed, braking distance and whether the train has authority to proceed from the signalling system. 207 ATP continuously monitors each train's speed, direction and location on the Mt Newman line or Goldsworthy line. Radio frequency tags in the track measure train location. ATP also monitors the permissions that have been given to each individual train from Centralised Traffic Control via the inter-locking signal system, uses this information to calculate the target speed and consequent braking distance for the train under operating circumstances and advises the driver of time to act and target speed.

3.4.2.4 Yarrie section 218 The Yarrie section of the Goldsworthy railway line runs east of the Goldsworthy Junction to the Yarrie mine and is approximately 193km in length. The Yarrie section of the Goldsworthy line is a low volume operation with no more than one train operating on a daily basis. 219 There are four active sidings or passing tracks on the Yarrie section - Hardie, Goldsworthy, Allen and Taplin - and one inactive siding known as Ryan. A siding is a section of track distinct from the mainline. The sidings' uses include marshalling, storing rolling stock and maintenance. The siding may or may not connect with the mainline at both ends; in other words, a train may have to reverse from the siding to re-enter the mainline. A passing siding, passing loop or passing track is a specific type of siding that allows one train to pass another. It therefore connects to the mainline at both ends and trains can move forward to re-enter the mainline. Almost all sidings on the lines appear to connect to the mainline at both ends, which means they can be used as passing sidings (although they may have other uses). The parties and their witnesses often used the terms passing siding, passing track, passing loop and siding interchangeably because most sidings can be used as passing sidings. 220 On the Goldsworthy line, the Hardie and Goldsworthy sidings can accommodate 90 ore cars but trains must reverse in order to exit the siding and move back onto the mainline. Allen and Taplin are passing tracks which can accommodate 120 and 90 ore cars respectively and trains can exit the siding by simply moving forward onto the mainline. Ryan siding was primarily used to store sleeper trains during the installation of sleepers 10 years ago. 221 The Yarrie section is in relatively poor condition due to: (a) the use of timber sleepers instead of concrete sleepers; (b) approximately two thirds of the Yarrie section being made from inferior rail, as opposed to the higher grade rail used on the Mt Newman railway line; (c) original bridges with varying degrees of metal fatigue; (d) parts of the Yarrie section being subject to severe seasonal flooding; (e) the signalling and communications systems as well as the ATP and control systems being relatively basic on the Yarrie section; and (f) less maintenance being carried out on the Yarrie section than the Mt Newman railway line. 222 A review conducted in October 2003 noted that several of the existing bridges could be approaching their fatigue life by 2013. 223 BHPB imposes restrictions on the axle loads and speed of trains operating on the Yarrie section. In particular: (a) ore cars are restricted to a maximum axle load of 25-26 tonnes; (b) the maximum tonnage per ore car is roughly 75 tonnes; (c) trains are single rake only, consisting of one locomotive and no more than 110 ore cars (the usual number of ore cars per train being 50-60); and (d) train speed on the Yarrie section is kept to a maximum of 60km/h, in contrast to the maximum speed of 75km/h on the Mt Newman line. Further speed restrictions are currently in place for 11 segments of railway and bridges because of poor condition. Four of these restrictions require trains to travel at speeds of less than 25km/h, while the remainder require speeds of no more than 45km/h. 224 From 2000 to 2006, BHPB transported approximately 7mtpa on the Yarrie section of the Goldsworthy line. Since 2007, BHPB has reduced its usage of this part of the Goldsworthy line to roughly 2mtpa. BHPB plans to maintain this level of usage for the next five years.

3.4.12 Wheel and rail interface 249 The management of the wheel and rail interface is of critical importance to rail operations. In particular, technology regarding the profile, specifications and material used for locomotive and ore car wheels directly impacts upon the level of wear on the rail (and vice versa). The management of the wheel and rail interface involves grinding the rail to a particular profile. Profiling the track reduces energy loss between the wheels of ore cars and the track by ensuring that each ore car wheel runs along the same contact point on the track along the length of the railway line. 250 In addition to carefully managing track profile, BHPB also machines ore car wheelsets to an optimal profile (based on BHPB's selected rail profile) on a scheduled basis. This: (a) improves the stability of rolling stock, as worn wheelsets and misaligned components can result in rolling stock experiencing dynamic problems such as tracking bias or "hunting". Tracking bias describes the tendency for one wheel to rub on the flange (lip) of one rail track whilst the other wheel sits away from the other rail track as a result of wheels not being symmetrical with the rail tracks. "Hunting" describes side-to-side movement (lateral instability) of ore cars as they travel along the track; (b) improves the steering of the fleet's bogies (a bogie is the frame attached to the bottom of the ore car which carries the wheel and axle assemblies and which pivots to allow the wheels of the ore car to follow track curvature), reducing wear due to flange contact and consequently reducing the need for rail grinding; (c) improves ore car productivity and lowers maintenance costs by reducing wheelset wear; (d) avoids the need to lubricate the railway track (the need for track-based lubrication is quite common on other heavy haul railways) and increases tractive effort, thereby enabling longer rakes to be hauled; (e) increases the track lifespan by reducing wear; and (f) increases system capacity by reducing maintenance, other things being equal.

Area C Project and PACE Project Area C: Spur from Yandi to Mining Area C 10 - 15mtpa (dry) - 89mtpa (dry) PACE: development of rail capability to enable a system production capability of 89mtpa Interim Growth Project Increasing Mt Newman line rail capability to 100mtpa (wet) ~ 11mtpa - 100mtpa (wet) RGP1 Passing sidings on the Mt Newman line; further rolling stock 10mtpa (wet) 130m 110mtpa (wet) RGP2 Rail works associated with OB18; additional rolling stock; modifications to Finucane Island rail yard; siding at Boodarie and rail for new Car Dumper 4 at Finucane Island 8mtpa (wet) 575m 109mtpa (wet) (previously 118mtpa (wet)) RGP3 New and extended mainline sidings; extension of loop at Area C; new locomotive preparation facility at Nelson Point; additional locomotives, rakes and track maintenance equipment 21mtpa (wet) 1.5bn 129mtpa (wet) RGP4 Construction of track sidings and acquisition additional rolling stock 26mtpa (wet) 2.1bn 155mtpa (wet) RGP5 Mt Newman line: substantial double tracking between Bing siding and Yandi mine (approximately 260 km); construction or modification of bridges; construction of a rail "wye" between Yandi 1 and Yandi 2; additional rolling stock; upgrade of signalling and controls system; Goldsworthy line: extension of the Boodarie siding at Finucane Island; reconfiguration of the Finucane Island loop 55mtpa (wet) 4.8bn 205mtpa (wet) in first three years following commission, reducing to 195mtpa (wet) thereafter. RGP6 Subject to further study. - - 240mtpa (wet) Quantum 1 Subject to further study. - - 300mtpa (wet) Quantum 2 Subject to further study. - - 350mtpa (wet)

3.5.3 RGP4 257 Rapid Growth Project 4 (RGP4) was approved by the Board on 22 March 2007, and was designed to increase system capacity from 129mtpa to 155mtpa by around January 2010. 258 The increase in capability is to be achieved by increasing tonnage from the Mt Newman mining area from approximately 45mtpa to approximately 68mtpa, and increasing the volume of Yandi product by approximately 3mtpa (wet). The key elements of RGP4 are: (a) construction of crushing and train loading facilities at the Jimblebar mine, a new car dumper, gradual relocation of the Nelson Point crushing and screening operations from Port Hedland to Mt Whaleback, which involved the construction of a new crushing and screening plant at Mt Whaleback, and minor modifications at Orebodies 18, 25 and the Yandi mine; (b) construction of track sidings and acquisition of additional rolling stock; (c) construction of various infrastructure such as offices and workshops; and (d) construction of a second stockpile row in the East Yard at Finucane Island. 259 Assuming RGP4 has been completed: (a) ore from Orebody 18 and Jimblebar is being railed to Mt Whaleback for crushing, screening and blending (along with Mt Whaleback ore, which includes ore from satellite Orebodies 29 and 30); (b) ore from Orebody 25 (along with ore from Orebody 23) is base loaded as lump and fines and railed to Mt Whaleback for blending in trains with ore from the Mt Newman hub; and (c) once blended with the base loaded ore from Orebody 25, Orebody 18, Jimblebar, Mt Whaleback 29 and Mt Whaleback 30, is railed directly to Port Hedland for stockpiling and shiploading. 260 The transfer of processing from the port to the Mt Newman mining area will mean that fewer stockpiles will be used at the port to blend Mt Newman ore; those stockpiles will act primarily as a demand buffer to meet fluctuations in shipping demand. These stockpiles will also continue to contribute to the blending of the ore, which inherently occurs through the port stockpiling process. This change will reduce the current level of complexity surrounding the port operations, thereby increasing shiploading rates and total system capacity.

3.5.4.1 Pre-approval of funding for RGP5 263 The request for pre-commitment funding was based on the scope of works considered in the RGP5 Selection Phase Study. This Study determined that an increase in system capacity to 205mtpa would require the duplication of a large part of the Mt Newman railway, the expansion of the port and the modification of the operating philosophy at Yandi hub to bring it into alignment with the other mining hubs. 264 The RGP5 design was refined following the approval of pre-commitment funding. For example: (a) the double tracking of the [c-i-c], including the sections of track just north of [c-i-c] and just north of [c-i-c], was taken out of the proposed RGP5 works due to the significant capital expenditure that this rail work would have required; and (b) the construction of the [c-i-c] sidings and rail duplication [c-i-c] were also removed from the proposed RGP5 works. 265 A further change from the proposed RGP5 works is that it is now proposed to use two rake trains of about 120 ore cars each (rather than single rake trains, as previously proposed). One proposal is to operate trains consisting of two rakes of about 120 ore cars each, with two locomotives per rake. An alternative proposal is to operate two rake trains with only three locomotives per train. While the use of only three locomotives is less capital intensive, these trains require "banking" (the use of additional locomotives at certain portions of the line to traverse certain conditions of track), which causes congestion and increased interactions on the line. A change to [c-i-c] rake trains would have a [c-i-c] impact upon BHPB's ability to achieve the targeted 205mtpa production rate. TSG estimated this [c-i-c] production to be approximately [c-i-c]mtpa. The advantages of two rake trains are that they carry larger quantities of ore, and that less drivers and driver accommodation is required. 266 RGP5 will see the adoption, system wide, of a "pull" approach to BHPB's infrastructure and operations. 267 Under RGP5, BHPB is proposing to double track a large part of the section of track between the Bing siding on the Mt Newman railway and Yandi Junction. However, a part of that section will not be duplicated (being certain bridges). Between Yandi Junction and Jimblebar Junction, [c-i-c] have been excluded from the scope of works. Only part of the Mt Newman to Jimblebar Junction section will be duplicated; there will not be any duplication of the Goldsworthy line west of the Goldsworthy Junction. 268 BHPB also plans to construct a new section of track referred to as the Chichester deviation. This new section of track will reduce the grade of the track from 0.55% to 0.33% by running the new length of track around a hill rather than over it. The new track is a continuation of the dual tracking on the rest of the line, but instead of constructing a dual track at a 5m offset to the existing track, the new track for the Chichester deviation will travel around the 0.55% graded hill. The benefit is that the new section will not require additional power for loaded trains to climb the 0.55% grade. The Chichester deviation is expected to increase cycle time and sectional running time, as it takes a longer route through the area than the existing track. 269 In addition to the double tracking, the works to be undertaken under RGP5 are as follows: (a) substantially increasing production from Yandi (including construction of additional crushing, screening and stockpiling facilities at Yandi), and introducing in-train blending of product from Yandi 1 and Yandi 2, and changing Yandi from a "push" to "pull" philosophy, as well as incrementally increasing production from Mining Area C and Mt Newman; (b) the construction or modification of bridges, construction of a rail "wye" (a triangular-shaped arrangement of rail tracks) between Yandi 1 and Yandi 2, acquisition of additional ore cars and locomotives and upgrading the signalling and controls system, as well as shifting to revised rail operations (which were originally expected to be single rake trains comprising 160 wagons and 2 locomotives, but are now likely to be 2 rakes of 120 ore cars, as discussed above); and (c) constructing additional shipping berths and shiploaders at Harriet Point near Finucane Island and increasing the proportion of direct shipping ore, upgrading Car Dumper 4 by improving the rail loop services and introducing additional equipment The berths at Harriet Point will be dedicated to shipping product from the west yard; the existing berths at Finucane Island will be dedicated to shipping product from the east yard. All berths at Finucane Island will operate on a "two berths/two shiploaders" basis - that is, instead of having each berth serviced by a dedicated shiploader, both berths at Harriet Point will be able to be serviced by both shiploaders at Harriet Point and both berths at Finucane Island will be able to be serviced by both shiploaders at Finucane Island. 270 On completion of RGP5, products from Yandi will be on specification (both physical and chemical) when trains leave the mine for the port. No additional crushing, screening or blending will be required at the port. Once the Yandi mine moves from a "push" to a "pull" philosophy, the only blending that will occur outside the mine hubs will be the blending of [c-i-c] products at Port Hedland. 271 There will be no change to the Goldsworthy mining operations under RGP5; approximately 2mtpa will continue to be produced from the Goldsworthy mining area. 272 Under RGP5, the constraint at Finucane Island will be the capability of the car dumpers: both car dumpers servicing Finucane Island will be required to operate at a rate of more than [c-i-c]mtpa. The increased emphasis on direct ship loading operations is expected to lead to an increase in the shiploading rate able to be achieved from Port Hedland, because the operation of reclaiming activity will represent less of a constraint on the system capacity. It is anticipated that this shift will increase port capability and decrease maintenance requirements (for example, on the stacking and reclaiming equipment).

296 The natural distribution of ore across RTIO's mining areas is irregular, both across ore bodies and within an individual ore body. In order to maximise value, trade-offs must be made between quality (absolute grade and grade variability), volume of ore and cost. For example, mining high grade material in the short-term may be at the expense of volume and long-term product consistency. The relative contribution of each of these elements to value is influenced by the market conditions at the time, and RTIO adjusts its production mix accordingly. 297 RTIO must blend the ores from its mines in order to create lump or fine ore products which meet contract specifications, while at the same time permitting an efficient exploitation of its resources. Pilbara Blend products are the mainstay of RTIO's production. The development of Pilbara Blend products has enabled RTIO to significantly extend the life of its ore reserves, and to mine and sell a far greater amount of the ore in its resource portfolio. 298 Blending processes take place as ore is transported from individual pits to mine stockpiles and from mine stockpiles to port stockpiles. These processes are required in order to reduce the grade variability of the product shipped to customers. 299 RTIO's Pilbara Blend products are built by blending ore railed from a number of mines in a carefully managed sequence so as to achieve specification product when that stockpile is reclaimed. As no single mine which contributes to Pilbara Blend actually meets RTIO's specification for the product, RTIO's Pilbara Blend products cannot be blended to specification until multiple trainloads, typically eight, are brought together in a specific sequence to enable blending of ore onto port stockpiles. 300 Due to the high variability of ore from each of the mines which contribute to Pilbara Blend, the relative contributions of individual mines to Pilbara Blend product stockpiles differ significantly from stockpile to stockpile. Rarely will ore from a single mine conform. 301 On occasions, it may become apparent that one or more mines will not produce ore of the quality and volume anticipated in the Medium Term Production Plan (described below), which could lead to port stockpile builds not meeting grade specifications without remedial action. To address this issue, RTIO may change the blending recipe for the stockpile, in order to reduce the impact of the "offending" ore. This involves a revision to the Production Schedule (also described below) which could result in trains being sent to different dumpers or ports. Alternatively, RTIO could split trains in order to spread the grade deviation over a larger volume of finished product. A further option is that RTIO may try to find a buyer for the "offending" ore as a spot sale (ie sell it as a RTX product). Selling the ore as a separate product will require scheduling the ore to the port at a time when there is space to create a separate stockpile.

302 RTIO's products are primarily sold under long-term contracts of up to 20 years duration (although prices are negotiated annually). 303 Each of the five core products has a typical specification that is agreed with RTIO's customers. The specifications for each product include the following chemical specifications: (a) a typical and a minimum iron content; and (b) a typical and a maximum content for each of the impurities. 304 Several physical qualities are also included in the product specifications. Examples are size and moisture content of particular ore types. 305 There is no contractual penalty if the grade of the product falls within specified non-conformance limits. However, outside these limits, there are penalties or bonuses applied for variations in iron levels, and penalties if impurity or moisture levels are above the maximum requirements specified. The penalty for failing to achieve base specification usually is far greater than the bonus for exceeding the standard. If impurities reach the "no-ship limit", suspension of shipments could result. 306 Creating products that consistently meet product specifications along multiple dimensions is a complex task given the variability that is inherent in RTIO's integrated production system. For example: (a) the characteristics of the material presented at the mine faces, including the delineation between ore and waste, are highly variable; (b) the quality of the ore differs between pits within the same mine; (c) the quality of the ore across RTIO's mines is also variable; (d) not only is the in situ ore quality variable, but there is typically some "prediction error" between the actual quality of the ore mined and the expected quality of the ore mined, based on drilling samples obtained just before placing explosives to blast the rock; and (e) a high degree of process variability is inherent in the nature of RTIO's Pilbara operations: planned and unplanned maintenance and equipment failures can impact the production capability of parts of the system, often at short notice.

4.8.5 Train control 330 RTIO's train control systems control and monitor train movements, ensure safe separation between trains and warn about unsafe conditions. 331 The wayside signalling system used in the RTIO rail system is known as the Integrated Control and Signalling System (lCSS). ICSS is an integrated system where the signalling forms a component of the track and is incorporated into the locomotive systems. 332 Unlike traditional signalling systems, ICSS does not require any trackside coloured light signals. Transponders attached to concrete sleepers along the railway track are read by an antenna under the front of the locomotive as a train passes that point on the track. These transponders are programmed with information about the track ahead. Trackside computers manage each adjacent section of track to ensure that it is safe for the train to proceed past a signal location. 333 ICSS utilises the track circuits that detect broken rails to send signal information to the lead locomotive of the train. This is the means of communicating an authority to the train, such as "stop", "proceed" or "caution". If there is an interruption to the flow of this information, such as a broken rail ahead of the train or an unauthorised train movement into the same signal section, the onboard computers and the driver are immediately notified that the track ahead may not be safe. 334 An important component of ICSS is the ATP sub-system, which is an onboard computer system which monitors information such as the current status of the train, signal authorities and the information from the transponders attached to the railway track. It is also an important safety system which prevents the train from exceeding the speed limit for an area of track and, if the driver is not applying enough braking force to stop the train, ATP will intervene and apply the brakes. 335 Assuming safety requirements are satisfied, a train controller's decisions about which train will be given priority are guided by the requirements that have been mandated by the integrated planning division, which are designed to achieve the required stockpile levels and blending requirements at the ports, and to manage stockpile levels at each of the mines. Provided the train controllers meet these priorities, they may alter train priorities by reference to other considerations, such as minimising track wear and managing other internal rail priorities. 336 The status of each track section is displayed at the operations centre on computer screens that are used by the train controllers to set routes for trains.

4.8.6.2 RTIO's rail scheduling 343 The RTIO rail system is also operated using the "run when ready" principle. 344 The Mesa J system conveys ore from the Mesa J mine pit and the crushing and processing plant at Cape Lambert. There is very limited stockpile capacity at the mine to provide a buffer for operation by the railway between the mine and port facilities. This means that if a loaded train is at the mine and ready to depart, but the next empty train has been delayed en-route, digging and hauling of ore must stop. Continuous movement throughout the system is therefore essential to maximising production output. The times when trains are run on one day often bears no relationship to the times when trains run on another day. 345 Pooled Fleet trains are "scheduled" according to the requirements for the stockpiles that are to be blended in the port yards for loading onto ships that are travelling towards the port or waiting at anchor in the queue at the time that the stockpile is constructed. The "schedule" must also accommodate the stockpile levels and grades at the various mines. The "schedule" is more accurately described as an order of trains that need to be run to the mines and to the dumpers to achieve these objectives. While a "schedule" is published and contains nominal times when each train is planned to run, the run when ready principle applies to departure. If the train is ready to depart it will not be held back because it happens to be ahead of that "schedule". For example, depending on the priorities of the day, a train may depart as soon as the first section out of the rail yards no longer has another train on it, or it may depart and then wait at the first station in the single track or dual track while another train passes it. 346 The Pooled Fleet train schedule is published for a particular week, commencing at 0700 hours on Monday morning. However it is constantly adjusted, commencing as soon as it is published. More often than not the mine or dumper to which an empty or loaded train is sent will differ from the original scheduled mine or dumper. The destination mine or dumper will also often change while the train is en route. 347 This method of scheduling and operating the trains maximises the tonnage of ore which can be transported and blended to the required quality. Efficient operation of the system relies on all operating trains being scheduled and operated in the same manner.

4.8.7 The management of variability and unplanned events 348 The Production Schedule is executed by train controllers in conjunction with schedulers and other personnel, including the inload scheduler, outload schedulers, the demand chain coordinator and the demand chain scheduler. The train controllers and the other scheduling personnel work together to minimise the impact on the Production Schedule of variability and unplanned events. 349 The rail schedule is produced by Integrated Planning in consultation with relevant personnel from many other areas of RTIO - it provides the link between the build and composition of the stockpiles of ore needed to load the forecast ship arrivals and the forecast production of ore at the mines. 350 In the management of the RTIO rail system flexible scheduling is adopted to cope with the inherent variability of the production system, particularly variability that is the result of unexpected events. Flexibility to re-schedule across the mines, rail, ports and shipping is essential to minimise the impact of these events and their flow-on effects. 351 Unexpected mine stockpile grades are a significant cause of schedule changes. Mine stockpile grades are measured as the stockpiles are being constructed. If there are deviations from the scheduled grade, there are options to rectify the problem through flexible mine, rail and port operations. For example: (a) If time and resources permit, it is sometimes possible to meet mine stockpile grade targets by mining ore from a different face within the same pit or mining ore from a different pit. (b) If the mine is not able to rectify the problem, the quality scheduler will look for opportunities to change the recipe of the port stockpiles by changing the ore contribution from the individual mines. This is a common outcome where the flexibility of the RTIO rail system is crucial to achieving the production targets. (c) If either of the above options is not possible, and conditions permit the sale of a spot cargo, an RTX port stockpile could be constructed. 352 Although the Production Schedule is developed on a weekly basis, it is continually refined to accommodate delays and variability in the system. Unplanned maintenance activities are a primary reason for changes to the Production Schedule. For example, if a load-out at Mt Tom Price is out of service for several hours for unplanned maintenance, Rail Operations might bring forward maintenance planned on the Mt Tom Price railway line to synchronise the planned and unplanned maintenance. If the trains that were scheduled for Mt Tom Price are able to be re-scheduled to other mines, the overall impact of the unplanned maintenance can be minimised. 353 Another example is a car dumper breakdown. A dumper unexpectedly breaking down for the time it takes to dump three trains could create the need to reschedule 10 trains (in terms of which mines and which dumper at which port the trains are sent to). Changes such as these can only be achieved with full flexibility across the mines, rail and ports.

4.8.10 Technological developments 362 RTIO relies heavily on technology and new technological developments to assist it to achieve maximum throughput and to expand its production. RTIO is continually investigating the use of new technologies. 363 The RTIO rail system operates at or near world's best practice on engineering and technological fronts. A key measure of track integrity in a heavy haul rail system is the rate of broken rails experienced by reference to the net tonne/kilometres hauled. Between 2004 and 2008 RTIO has maintained its rate of broken rails at a fraction of the North American benchmark, while also achieving significant increases in its average axle loads and tonnage railed over the same period. 364 One of the principal explanations for this high level of performance and efficiency in its rail system is the continual development of technology. The closed system nature of the rail system, with its single management structure for "above" and "below" rail, has enabled RTIO to undertake rapid development of technology to meet its business needs. 365 A research and development program is implemented through an internal approval process either as part of the annual planning process or through an ad-hoc process where the improvement idea is generated outside the annual planning cycle. Approval for funding of the project is required at a particular management level within RTIO set by financial approval limits. Any test work or trial as part of the research and development work is generally conducted by the project team liaising with the relevant operations and maintenance teams. 366 Many technological improvements to the system involve a close relationship between track and train. The optimal outcome can be achieved by having the flexibility to invest less in trains and more in track (or vice versa) in order to achieve an overall throughput target. For example, if RTIO wished to change its allowable wheel wear limits to improve overall system wear rates or change its ATP hardware this could require changes of equipment and costs in respect of all rolling stock or locomotives operating on the RTIO rail system.

4.9.1 The expansion process 367 The process by which an expansive plan is adopted is like that of any other RTIO project. First it must pass through a series of study stages: (a) Conceptual; (b) Feasibility; and (c) Implementation, including: (i) detailed engineering, procurement & construction management; and (ii) commissioning and ramp-up (to steady state production). 368 There are internal and external approvals which must be obtained during the various project development stages. Projects are only progressed to the next stage when sufficient engineering, social, community, environmental, legal and commercial investigations for that study stage have been completed. 369 Conceptual studies are typically done as a desktop analysis, and tend to rely on existing information and in-house experience and comparisons with similar past projects and environments in order to assess options and gauge value accretion for the company. It is the study team's role to ascertain whether the project is sufficiently credible to warrant a more detailed Order-of-Magnitude technical and economic study. 370 Following completion of a Conceptual study a more detailed Order-of-Magnitude study may be undertaken. In some circumstances, such as where sufficient (quality) data is available, it may be possible to accelerate the project to Pre-Feasibility. 371 The key objective during the Order-of-Magnitude stage of project development is to determine whether a business case exists to justify the cost of the project progressing to Pre-Feasibility. During this phase, a number of different options for project development are identified and ranked in order of preference for more detailed evaluation. With significant resource definition work having been completed prior to the commencement of the Order-of-Magnitude stage, limited additional fieldwork is undertaken during this stage in order to establish the economic, technical and commercial viability of the various options. The size and composition of the study team may be expanded to include people with appropriate experience to address these matters. 372 A formal Order-of-Magnitude study report is produced during this stage. Formal, in-house reviews also take place. Any project which involves potential capital expenditure in excess of US$250m must be submitted for a Scoping Review by the Technical Evaluation Group and Business Evaluation Department. A Scoping Review is also required for projects that may involve significant risks, for example, if it is proposed to adopt new technology, or to enter a new market. 373 The Scoping Review will then be referred to the Investment Committee, comprising Rio Tinto's Chairman, Chief Executive, Finance Director and Group Executive Technology and Innovation. Referral to the Investment Committee at this stage ensures that the Committee understands the complexities of the proposed project at an early stage and has an opportunity to consider (and endorse) the planned project strategy. Subject to any necessary additional instructions or guidance it considers appropriate, the Investment Committee may approve the scope and budget for the next stage. 374 The decision by RTIO to proceed to Pre-Feasibility is significant because it marks an increase in commitment and the resources allocated to the project. Pre-Feasibility defines the project scope, typically by evaluating the business's requirements and assessing a number of options. At the conclusion of Pre-Feasibility, a choice will be made on a single preferred option and whether or not to commit funds and resources to take that option to full Feasibility. 375 The aim of the Feasibility stage of project development is to optimise and comprehensively define the project scope and value. At the end of this stage, the Feasibility study report and an Investment Proposal are provided to the Investment Committee for either its approval and authority to implement the project, or refer it to the Rio Tinto Board. 376 Investment Committee approval is required for any project estimated to cost over US$100m. Projects which are controversial, outside of normal business, involve new technology or which may face major political, reputation, environmental or community issues must also be submitted to the Investment Committee. 377 The Investment Committee makes a decision whether or not to approve the proposal by reference to its level of authority. It must refer the following matters to the Rio Tinto Board: (a) proposals involving a commitment of over US$250m, following review by the Investment Committee; and (b) projects that are potentially controversial or that may have unusual ramifications will be brought to the Rio Tinto Board for discussion and decision, as decided by the Chief Executive (even though the financial implications are within Investment Committee levels of authority).

PROPOSED JOINT VENTURE BETWEEN RTIO AND BHPB 408 On 5 June 2009, Rio Tinto announced that it had entered a non-binding agreement with BHP Billiton to establish a production joint venture of both companies' Western Australian iron ore assets. The proposed joint venture will encompass all current and future Western Australian iron ore assets and liabilities, and will be owned 50:50 by BHP Billiton and Rio Tinto. 409 Rio Tinto has entered into a framework agreement with BHP Billiton which requires them to negotiate in good faith legally binding agreements necessary or desirable to establish the joint venture and regulate its ongoing operation. Completion of the joint venture is subject to certain conditions, including approvals from Rio Tinto and BHPB Billiton shareholders and regulatory authorities. Regulatory approval is required in both Australia and overseas. 410 BHP Billiton and Rio Tinto have been unable, for legal reasons, to discuss the nature and costings of infrastructure to be developed for the joint venture. Nonetheless, BHP Billiton and Rio Tinto anticipate that a joint venture will create substantial synergies in the following areas: · resource optimisation, optimisation of mine development across the combined portfolio, and blending of products to maximise resource recovery; · the shared use of rail, port and power infrastructure in relation to the Hamersley Iron, Robe River, Yandi, Newman and Goldsworthy areas, and capital expenditure savings from using the closest rail infrastructure for each East Pilbara development; · use of the lowest cost and most efficient infrastructure expansion pathways; · improved berth utilisation from joint operation of the ports, blending of product and joint scheduling of ships between the ports; · savings in operational expenditure and capital expenditure from the joint operation and expansion of BHP Billiton's and Rio Tinto's Yandi operations; · operating and procurement efficiencies arising from the transfer of best practices, joint procurement, shared services and shared equipment across the Pilbara; and · the reduced duplication of corporate overheads in BHP Billiton's and Rio Tinto's Perth offices. 411 The joint venture contemplates some BHPB mines using RTIO's rail system and vice versa. If it goes ahead, it may be necessary to transport significant amounts of iron ore to Cape Lambert. 412 [c-i-c]. If that occurs it would increase the tonnage of iron ore being transported down the Hamersley and Robe lines to Cape Lambert.

427 The Mindy Mindy project is owned by Pilbara Iron Ore Pty Ltd (PIO), a joint venture vehicle in which each of FMG and Consolidated Minerals Limited has a 50% interest. PIO holds mining tenements at Mindy Mindy, located approximately 295km south east of Port Hedland, 50km east of BHP's Mining Area C mining area and approximately 17km from the Mt Newman line. 428 There was considerable debate about the viability of the Mindy Mindy project and, if viable, the options PIO has for transporting ore to port. We do not think that the resolution of this debate will assist the resolution of this case. Nonetheless, in deference to the time devoted to the issue, we will address some of the debate. 429 First there is the quality of ore at Mindy Mindy. In May 2007 Mr Tapp, Head of Government Relations at FMG, estimated that the Mindy Mindy resource will produce a commercially recoverable reserve of around 56mt capable of supporting a 5mtpa operation for around twelve years. Mr Tapp did indicate that PIO had not undertaken a drilling program (which he claimed would cost approximately $10m) that would establish the full extent of the resource. He said that without access to transport infrastructure the deposit was effectively stranded. 430 Two witnesses, Mr Derek Miller (a mining engineer) and Mr Harmsworth (a geologist), said that Mr Tapp had been far too optimistic in his assessment, overstating the likely level of saleable ore. They relied upon the level of iron (Fe) content and the level of impurities identified in exploration to date. 431 Dr Clout (a geometallurgist and head of Resource Strategy at FMG) disagreed with their assessment. Nonetheless, it was accepted that if Dr Clout's target grade for the Mindy Mindy iron ore was correct the ore, while saleable, is at the lower end of product acceptability. 432 Mr Tapp remains confident that the operation is viable, notwithstanding the relatively low product quality. He produced a memorandum of intent entered into with a Chinese steel mill which indicated that the ore was of sellable grade. 433 Assuming that Mindy Mindy ore is saleable, the second issue is what transport options are available. Three potential options were put forward. The first is for PIO to build a road to truck iron ore to Cloud Break, and then use the Chichester line to haul the ore to Anderson Point. The second is to build a rail spur to Cloud Break (or some point to the north) and then use the Chichester line. The third is to use the Mt Newman service. 434 As regards building a rail spur to Cloud Break, Mr Tapp says that the size of the Mindy Mindy deposit does not justify the cost of a spur from Mindy Mindy to the Chichester line (which he estimates would cost approximately $500m). This, at least, appears to be uncontroversial. 435 As regards trucking to Cloud Break, it would be necessary for PIO to build a road to Cloud Break. Complicating the matter is the Fortescue Marsh, which lies between Mindy Mindy and Cloud Break. The Fortescue marsh is an important ecological habitat for water birds, listed as a nationally important wetland and is an "indicative place" on the register of the National Estate. Any proposal to construct a road across the marsh is unlikely to obtain environmental approval. Hence a road would need to take a somewhat longer, indirect route. 436 One route is to follow the Mt Newman line to the northern extremity of the Fortescue marsh at the Cowra line camp and then proceed along the northern edge of the Fortescue marsh to Cloud Break. Mr Tapp suggested that even this route would face difficulties with obtaining environmental approval. In a report commissioned by BHPB, GHD Australia Pty Ltd estimated that the cost of this 126km road would be around $100m. 437 There is a live issue whether the option of trucking to Cloud Break would be more cost-effective than using the Mt Newman line. BHPB says that trucking would be cheaper, relying on costings by Mr Miller. Mr Miller made a "conceptual" estimate of the costs of each option, with a degree of error of ± 25%. He estimated that for a resource of 56mt it would cost $40.45 per tonne to transport ore via the Mt Newman line compared with $43.35 per tonne if the Cloud Break trucking alternative was employed. Mr Tapp disputed Mr Miller's haulage costs analysis in numerous respects. Mr Miller said that Mr Tapp's criticisms did not cause him to revise his cost estimate. He put it this way: "Given its conceptual nature, the cost comparison between the proposal to use the Newman railway and the Cloud Break trucking alternative was too close to provide a basis for determining which proposal is likely to be more cost efficient." Nonetheless he remained of the view that the Cloud Break alternative is "likely" to be the more practical and efficient option. 438 During the hearing the Tribunal indicated that one of several issues it was inappropriate (if not impossible) to rule upon was whether Mindy Mindy was a viable operation. We remain of that view. Still, in light of the growing demand of iron ore from China and the likely demand from India over the next two decades (the maximum life of a declaration), it would be unsafe other than to proceed on the basis that Mindy Mindy could be turned into a viable operation. Similarly, given that Mr Miller himself concedes that the cost effectiveness of the trucking and Mt Newman service options is too close to call, we assume that Mindy Mindy may well be a source of demand for the Mt Newman service.

439 The Solomon group comprises a group of mining tenements straddling the Hamersley railway north of Rosella. The two major projects are the Solomon East deposit located east of the Hamersley line and the Serenity deposit located to the west. Within the Solomon East area there are three key sites: Valley of the Queens; Valley of the Kings; and Firetail. 440 On 29 April 2009, FMG announced that the total (JORC-compliant) mineral resource estimate for Solomon was 2.2 billion tonnes with an average Fe grade of 56.2%: 552mt is classified as an indicated resource and 1.7 billion tonnes as an inferred resource. Since then, that resource estimate has increased steadily to 2.45 billion tonnes, and FMG expects it to increase with further exploration. 441 The Solomon Group area project consists of a variety of ore types, including a number of CIDs and a medium to high phosphorous Brockman deposit. Some of the ore is relatively high quality with low alumina contest. On the other hand, a significant proportion of the Solomon group area deposits will require upgrading or blending to be saleably exploited. Mr Tapp explained that the Solomon Group area deposits might variously be exploited by selectively mining the high quality ore, blending low quality and high quality ore, upgrading lower quality ore through processes such as washing and beneficiation, or some combination of the above. 442 FMG expects that some Solomon area deposits (for example, its Firetail deposit) will be suitable for blending with lower quality product from the Chichester and Christmas Creek area to produce a saleable product. 443 The Solomon group is a key part of FMG's future strategy. In a presentation made on 12 October 2009, FMG stated that the Solomon project was under fast-track management and provides potential for enormous growth. Notwithstanding the need for blending or upgrading, the Solomon project consists of very large resources, with the potential to support very significant levels of production.

446 FMG proposes to triple the size of its operations. FMG has announced plans that contemplate increasing its production rate in three stages, from 55mtpa to 95mtpa, then to 155mtpa, and then to 255mtpa or more. 447 The planned 95mtpa expansion will be achieved by the Christmas Creek rail extension and an increase in production from the Christmas Creek operation to 55 or 60mtpa. It will also require an expansion of port and rail infrastructure. The expanded facilities at the Anderson Point port facility will include three berths, two ship loaders, two stackers, two reclaimers and two train dumpers. 448 The second stage involves the construction of a rail spur from the Solomon area to connect with the Chichester line. This spur has previously been referred to as the Kennedy line and we will use that term for simplicity. 60mtpa will be hauled from the Solomon area. Combined with 95mtpa from the Christmas Creek and Cloud Break area, a total of 155mtpa will be hauled down the Chichester line to be shipped from Anderson Point. This will require significant upgrading of both the Chichester line and Anderson Point. The Chichester line will be double tracked, except perhaps for bridges. The port will be upgraded to create five berths, three rail loops, three train unloaders and increased stockpile capacity. 449 Expansion beyond 155mpta will require additional rail and port infrastructure because Port Hedland is unlikely to have sufficient capacity. FMG investigated the development of a multi-user port between Cape Lambert and Cape Preston to ship at least 200mtpa of iron ore. An engineer called by FMG, Mr Neil Miller, considered the available options - Cape Lambert, Dixon Island, Intercourse Island, Dampier and Cape Preston - and concluded that the Dixon Island location (ie Anketell Point) was the preferred option, having regard to the capital expenditure, the likelihood of obtaining the necessary approvals and the establishment of rail access. 450 FMG appears to have selected Anketell Point as its preferred option. In June 2009, it entered into a co-operation agreement with Aquila Resources Ltd to investigate the joint development of a port at Anketell Point. Planning is relatively well progressed. Aquila has indicated that the port has an initial design capacity of 40mtpa, with provision to allow for the progressive expansion of facilities for other users up to 350mtpa. A conceptual plan has been submitted to the Western Australian Government, with a definite feasibility study due shortly. In March 2010 FMG indicated that it intends, as part of the third stage of its expansion plans, to ship 100mtpa of iron ore from Anketell Point in about 2013/2014. 451 To transport iron ore to Anketell Point will require a railway. FMG's stated preference is to access RTIO lines for that purpose. If it cannot obtain that access, it may build its own line from the Solomon area to Anketell Point. In a March 2010 investor presentation, FMG noted that if it does not obtain access to the RTIO lines, it will build a 220km railway. Mr Miller's report identified a preliminary route for the line running from the Solomon Group area to the port. He has estimated the cost of the construction of that line (the Dixon line), based on the actual cost of the Chichester line, to be $950m for infrastructure plus $350m for rolling stock. 452 If FMG were to combine the proposed Kennedy and Dixon railways and port facilities at Anketell Point, iron ore from all of FMG's operations could be delivered to either the existing Anderson Point port facility or to the planned port facilities at Anketell Point. 453 To the extent it has foreshadowed developments beyond 95mtpa, Mr Tapp said that the development of the Solomon area is only at an early stage of conceptual study. In respect of the proposed Kennedy railway, Mr Tapp said there is no prospect of construction commencing within the next two years and it is by no means certain that construction will ever go ahead. If it does, Mr Tapp says, it will be at least 5 years before it is completed. 454 In the Tribunal's view, if FMG does not obtain access to the Hamersley line it is highly likely to construct both the Dixon line and the Kennedy line. The Dixon line will be constructed once the Solomon output increases beyond 60mtpa and FMG's port facilities at Port Hedland reach capacity. The Kennedy line will be constructed first to handle the initial Solomon output. 455 There are several reasons for us reaching this conclusion. First, FMG's recent public announcements all suggest that the lines will be built if there is no access to the Hamersley line. In fairness to Mr Tapp, these announcements supersede his evidence. 456 Second, the Solomon deposits are a very large, valuable resource. FMG's latest investment announcements suggest that the area can produce 160mtpa, which is more than enough to justify building both lines. 457 Third, all the evidence suggests that FMG's long-term expansion is likely to be closely tied to use of the port at Anketell Point, given capacity constraints at Port Hedland. 458 Fourth, in its closing submission FMG acknowledged that if the costs of access to the Hamersley line were to exceed the cost of building the Dixon line, then "Fortescue might be expected to proceed down that alternative course". This suggests that the lines will be built in the absence of access to the Hamersley line. 459 Fifth, FMG has expressed a firm intention of blending products from its Solomon mines with products from its Chichester hub mines. This will necessitate the construction of the Kennedy line. 460 Indeed, it is likely that the Kennedy line would be built even if the Hamersley service were declared. This is because: (1) FMG intends to blend Solomon mine and Chichester mine products; (2) FMG will not need to wait for the construction of Anketell Point before it can exploit its Solomon mines; and (3) FMG's stated preference is to have an integrated, Pilbara-wide rail network. 461 There may be other FMG expansions in the pipeline. FMG is contemplating developing a "Western hub" in the Western Pilbara which may, in the fullness of time, produce ore which will be shipped from Anketell Port. 462 Ultimately, FMG says that "its medium term management target" for the Pilbara region is 355mpta. On any view, this will involve large expansions.

467 The availability of port facilities is an important factor affecting the transport options of junior miners. If there are no port facilities available for them to ship their ore, then there is no point in transporting the ore in the first place. Although port availability is limited, the evidence suggests that port options will soon be opened up. 468 The current port facilities in the Pilbara consist of BHPB's ports in the Port Hedland area, RTIO's ports at Dampier and Cape Lambert and FMG's port of Anderson Point at Port Hedland. BHPB's and RTIO's ports are not available for third party use. 469 FMG, on the other hand, is obliged to put in place an access regime for third party use of its Anderson Point port. This is a condition of its State agreement. Currently FMG processes ore at Anderson Point for Atlas Iron, and has entered into a joint venture agreement with BC Iron under which ore will be processed at Anderson Point. 470 Anderson Point's current capacity is 55mtpa. In an investor presentation on 28 May 2009, FMG announced that it would enlarge the port facility to increase its capacity to 155mtpa to cater for its own demand. It is unclear to what extent there will be capacity to accommodate third party demand. 471 In the next few years a number of new ports will be constructed to cater for junior miners. The first is a new public berth at Utah Point in Port Hedland. Utah Point is located on the western perimeter of Port Hedland harbour, between Harriet Point to the south and Finucane Island to the north. The construction of the port commenced in March 2009 and is due to be completed in mid 2010. 472 It is planned that the Utah Point public berth will have a capacity of approximately 18mtpa. Most of the capacity will be used for the export of iron ore. The capacity has been allocated to BHPB (6mtpa), Atlas Iron Limited (3mtpa), Consolidated Minerals Limited (1.7mtpa), Aurox Resources Limited (6mtpa of magnetite slurry), Process Minerals International PL (1.2mtpa for iron ore and manganese) and High Tech Energy (0.2mtpa of manganese products). Atlas Iron and Aurox have entered into a cooperation agreement that, among other things, gives Atlas Iron a first right to use any of Aurox's unused or surplus berths. The companies have announced their intention to merge. If the proposed Atlas Iron and Aurox merger is approved, the merged entity will control 50% of the port's capacity. 473 The cost of the Utah Point public berth will be approximately $225m, of which $105m is to be provided by the foundation customers. 474 The Utah Point berth can be accessed by road. With the exception of BHPB and Aurox, the users of the berth will use road transport to deliver their ore to the public berth stockyard. BHPB will deliver ore by a conveyer from its existing port facilities and Aurox will deliver ore by a slurry pipeline. There are no plans for rail access to the Utah Point public berth. Rail access would require additional facilities, including a car dumper. Given the cost of a car dumper (approximately $500m) it is not likely to be feasible to install a dumper for the load tonnages that will be shipped through the berth. 475 Another port which is being considered for the Port Hedland area is at South-West Creek. This is being jointly investigated by the PHPA and a group called the North West Iron Ore Alliance (NWIOA). The NWIOA was formed in 2007 to represent the interests of a group of iron ore companies operating in the Pilbara. The NWIOA's core objective is to promote the development of the junior iron ore industry. The NWIOA currently has three members: Atlas Iron, Brockman Resources Limited and FerrAus Limited. 476 The NWIOA, in conjunction with the Port Authority, has recently released its pre-feasibility study for the development of berths and associated facilities at South West Creek. The study found that a 50mtpa two-berth development costing an estimated $2.1bn could be operational by the second half of 2013. The berths will be dedicated to NWIOA members and other junior miners. A definitive feasibility study is now being undertaken, which will assess engineering proposals and funding arrangements. Various regulatory approvals are already being sought. 477 The PHPA is also considering the development of the outer harbour at Port Hedland. If the outer harbour is developed, an additional 400mtpa could be shipped from the port. 478 In addition, there are well progressed plans to develop a new port at Anketell Point. Aquila is interested in this port because it is developing a mining project in the West Pilbara. 479 A new port is currently under construction at Cape Preston. Cape Preston is some 100km west of Cape Lambert. The port is being constructed pursuant to an agreement between the State of Western Australia and Mineralogy Pty Ltd (and others), ratified by the Iron Ore Processing (Mineralogy Pty Ltd) Agreement Act 2002 (WA). Under that agreement, the firms building the port are required to provide access to the State and third parties, provided their use shall not, in the reasonable opinion of an independent expert, unduly prejudice or interfere with the incumbents' operations: cl 21(3). It is unclear whether any third party will seek access to that port.

7.3.1 Trucking 482 It seems to be uncontroversial that mining companies that use trucking services usually acquire the service from third party haulage operators who use their own equipment. Unless capital works are required to establish a private road (as opposed to using public roads), the costs associated with trucking are largely operating expenses. 483 There is conflicting evidence regarding haulage charges in the Pilbara. Evidence of actual quotes received, and actual rates paid, by miners indicate a rate of around 12 to 13c per tonne per km. This was also the figure assumed in a NWIOA submission to the NCC in 2008. Mr Tapp estimated the cost of trucking to be slightly higher (at 15c), although this is inconsistent with quotes for lower rates received by FMG. Mr Fisher, an engineer, nominated a rate of 9c per tonne per km, but this was based on an informal, indicative quote from his brother, a haulage provider, and Mr Fisher had not confirmed whether that rate was actually used. By most accounts, the rate per tonne paid seems to be relatively constant regardless of (i) the volume and (ii) the distance being trucked. In this regard, we note that trucking quotes received by Brockman for its Marillana project are very similar to those paid by Atlas for its Pardoo project, which trucks shorter distances and involves smaller volumes. 484 The Tribunal considers that a rate of 12-13c per tonne-km is most credible. This rate is relatively constant over longer distances, as miners further away from port are charged (or quoted) the same rate as those closer to port. 485 An important limitation of trucking is that it is only viable for transporting relatively low volumes of ore. This limitation arises for a number of reasons. First, there are only so many trucks that can practically be put on the road at any time. Mr Richards of Brockman said that in order to truck 1 to 1.5mtpa, Brockman would require a fleet of approximately 14 to 19 road train trucks to operate continuously on a 24 hour, 7 day a week basis. Mr Flanagan of Atlas made a similar point, noting that at some point "it ends up being like caterpillars chasing each other down the road". 486 There may also be legal limits on the volume of ore that can be trucked. Under the Road Traffic Vehicle Standards Regulations 2002 (WA), trucking by road train requires a permit issued by a road authority, Main Roads Western Australia. Various witnesses said that permits have been issued with a condition capping volumes that can be trucked. The legal basis for imposing such a cap, if they are in fact imposed, is unclear. It certainly is the case that a permit can be issued with conditions (see regs 31 and 35) and some of those conditions (eg conditions restricting the time of day when the truck can be driven) could have the practical effect of limiting the volume that may be trucked. In any event, there is little doubt that trucking volumes are an issue of concern for Main Roads, as BHPB's witnesses (such as Mr Fisher), acknowledged in relation to Port Hedland and as Dr Fitzgerald, an economist, observed in respect of the use of rail and trucking in the Mid-West (Geraldton) region of Western Australia. 487 Apart from permits, there are other reasons why trucking volumes need to be limited. Safety may be an issue at higher volumes, as Mr Tapp and Mr Richards indicated. There are also potential legal constraints on trucking large volumes of ore through towns. By way of example, Mr Flanagan said that the volume of ore which Atlas is entitled to truck through the Utah Point facility is limited by local regulations. 488 It is difficult to quantify the precise volume of ore which can viably be trucked. Of the projects to which the Tribunal was referred, those which use trucking tend to transport volumes of 5mpta or less. The Tribunal has evidence that Atlas proposes to truck 6mpta (from various projects) to Utah Point. Mr Flanagan suggested that the total tonnage contemplated to be trucked by Atlas through to Utah Point is 9mtpa. This is the largest volume of ore contemplated to be trucked in the material before the Tribunal. It is, we think, safe to say that it is unlikely that trucking would be viable for volumes in the order of 10mtpa or higher. 489 Another limitation on trucking is the distance over which iron ore can viably be trucked. Given that trucking is charged on a per km basis, at some distance trucking ceases to be feasible. The contentious issue is when this limit is reached. 490 There is no universal rule in this regard, as the economics of trucking will depend on a number of variables such as the quantity of the ore being shipped, the price of the ore, the costs of mining and other business costs, and the rate of return required. Of the mining projects which use trucking to which the Tribunal was referred, a large majority appear to truck for distances of 150km or less, and occasionally up to 200km or so. There are a few instances of trucking for longer distances. In one case referred to by Dr Fitzgerald, the Crossland Resources project at Jack Hills in the Mid-West, ore is trucked a distance of 640km to port. 491 The Tribunal is of the view that trucking such a distance in the Pilbara would not be sustainable in the long-run (at least based on current projections for iron ore prices). As Dr Fitzgerald notes, Crossland Resources intends to move to rail. It also appears that, based on trucking rates for the Pilbara at least, trucking distances of 640km would be prohibitively expensive. At 12c per tonne per km, the trucking cost alone would be around $77 per tonne. Various experts have made projections regarding the long-term price of iron ore, ranging from around A$60-70 per tonne (Dr Fitzgerald) to A$110 per tonne (Mr Taylor, an economist and management consultant). Even at the higher projected prices, once costs other than trucking and a reasonable rate of return are taken into account, trucking distances of 640km is unlikely to be viable. 492 The weight of the evidence is that trucking distances of up to around 200km will often be viable, but trucking longer distances will not be viable, although the individual circumstances of the miner may permit longer distances. 493 In all cases, if trucking is to be used in conjunction with rail, the distances which can be viably trucked are less than if trucking alone were used. It is one thing to truck a certain distance to port; it is another to truck the same distance to a loading point on a railway, and then bear the additional cost of rail haulage.

7.3.2 Rail haulage 494 Rail haulage is much more capital intensive than trucking. Capital costs become particularly significant where it is necessary to construct dedicated (as opposed to shared) track. 495 On the other side, rail operating costs appear to be significantly lower than trucking. For example, a 2008 Metalytics study commissioned by BHPB surveyed the operations of various iron ore producers (including BHPB, RTIO and FMG). The study concluded that the weighted average cost for rail haulage was 1.35 cents per tonne per km. 496 Despite the relatively high capital costs associated with rail haulage, it should not be assumed that it is too expensive for small junior miner operations. There are several reasons for thinking that rail may well be viable for relatively small operations. First, there is empirical evidence of this. As noted above, Dr Fitzgerald surveyed the use of road and rail in the Mid-West region of Western Australia near Geraldton. He observed that a number of small projects (around 2-3mtpa) were using a combination of road and rail. In the case of Midwest Corporation, a company with a volume of less than 1mtpa, its operations manager suggested that a move from road to rail transportation would reduce the company's per tonne transport costs. 497 Second, transport studies done in the Pilbara suggest that rail is affordable and cost-competitive with trucking for relatively small volumes. In May 2009 [c-i-c] commissioned a consulting group called Beyond Rail Solutions (BRS) to study rail options for several of its projects. The study noted that a number of measures could be taken to reduce the capital costs of rail operations, such as by leasing rolling stock or using "bottom dumping" wagons to reduce loading and unloading infrastructure costs. The study ultimately concluded that in the case of [c-i-c], rail would not be cost competitive with trucking for volumes of [c-i-c]mtpa, but was quite likely to be cost competitive for volumes of [c-i-c]mpta. It did not take into account certain loading and unloading costs. This is because the loading equipment could vary widely in cost, ranging from basic front end loading through to more sophisticated batch weighing train loaders. There is other evidence giving a sense of these loading and unloading costs. Mr Tapp estimates that for Mindy Mindy, a 5mtpa operation, it would cost $25m to build rail loading facilities comprising the cost of a siding and a run around loop, an unloading ramp for road trains, a stockpiling area for road trains to be flipped and front end loaders. Mr Derek Miller says that this figure is reasonable. In the context of the large expenditure otherwise associated with developing a mine, $25m or so for loading facilities does not appear to be a major cost.

502 Having covered the background facts, the first important issue that must be considered is the "essential facility" problem. It has been said that Part IIIA is an attempt to deal with that problem. Before looking at Part IIIA in detail, it is necessary to understand the nature of the problem and regulatory responses to it. 503 The efficient functioning of key infrastructure sectors of the economy - transport, energy and communications - is vital to sustaining the economic development of a country. One reason is that infrastructure provides inputs into a wide range of services that support distinct (so-called "dependent") markets. There is, however, a tendency toward monopoly in infrastructure sectors. This arises because the key sectors are characterised by large capital costs, the immobility of the incumbent(s) and strong economies of scale. For this reason, such infrastructure has sometimes been described as "natural monopolies". Those features have also historically led governments to provide infrastructure services which were shielded from direct competition. 504 Indeed, Australia has a long tradition of public ownership of infrastructure services and network industries (eg mail services, telecommunications, electricity and many transport services) and other natural monopolies (eg ports and airports). There are many reasons for this, the principal among them being: (1) the ideological views of the state; and (2) the historical background to the grant of public monopolies as a revenue raising device. 505 Public ownership of infrastructure and the existence of natural monopolies lead to ineffective competition (actual and potential). At the extreme, the market is supplied by a single firm protected by barriers to entry (especially statutory barriers) which, in the absence of regulation, faces no check on the price it can charge consumers. 506 In the 1970s in England, and later in other countries including Australia, there began a review of the role of the state in economic activity. Policy-makers started to question whether the state should exercise control over infrastructure through ownership or regulation or whether economic activity should be guided by the market, subject only to general antitrust (competition) laws.

507 To understand the causes of ineffective competition that may arise in infrastructure industries it is necessary to have some technical knowledge of natural monopolies. 508 The traditional formulation of a natural monopoly, best expressed by the late Alfred Kahn, is that a monopoly is "natural" when there "is an inherent tendency to decreasing unit costs [ie average cost] over the entire range of the market. This is only so when the economies achievable by a larger output are internal to the individual firm - if, that is to say, it is only if more output is concentrated in a single supplier that unit costs will decline." (Alfred E Kahn, The Economics of Regulation (1988) 119). This is the definition of a natural monopoly resulting from economies of scale. 509 Put simply, on this view a natural monopoly exists when production technology involves relatively high fixed costs, which are usually sunk (ie irretrievable), and there are relatively low operating costs, as a result of which long-run average total cost (found by dividing total cost by total output) and marginal cost (found by taking the derivative of total cost at a specific level of output) decline as output expands. The reason why long-run average total cost declines is because the firm can spread its fixed costs across an ever-larger number of units as production increases. Hence the monopoly is "natural". Higher costs will result if more than one producer supplies the market, as each producer will have to spread its fixed costs over smaller output volumes. Moreover, competition will be wasteful if duplicative facilities are constructed and the lowest possible average cost cannot be achieved. In other words, it is most efficient for one firm to supply all of the market's demand. 510 The existence of a natural monopoly resulting from economies of scale depends upon a precise definition of the market, a proper identification of the nature of demand in that market, and an underlying assumption that the firm can continually expand its capacity to meet that demand at a lower average cost than a new firm entering the market. As Posner put it, natural monopoly "does not refer to the actual number of sellers in a market, but to the relationship between demand and the technology of supply": Richard Posner, "Natural Monopoly and its Regulation" 21(3) Stanford Law Review 548(1969). 511 The definition of a natural monopoly resulting from economies of scale is applicable to firms that supply a single product. Where multi-product firms are involved, economies of scale are not a sufficient condition for the existence of a natural monopoly. Further, the existence of economies of scale is a sufficient, but not necessary, condition for a natural monopoly. A different test should be applied in the multi-product firm scenario or when a single-product firm does not exhibit economies of scale. That test involves a "technical condition" which requires that the entire output of the market can be supplied by a single firm at a lower total cost than by any combination of two or more firms. This condition is expressed mathematically as "subadditivity of costs". It stems from the work of Baumol, Panzar and Willig, who were especially critical of the application of the traditional formulation to situations of multiple output production. 512 The cost subadditivity test can be explained in the following way. Assume that there are two firms each producing the same two products. Let A and B represent the products, 1 and 2 represent the firms and Q represent quantity of output. Therefore, (QA,1, QB,1) and (QA,2, QB,2) represent the multi-product output of each firm. Also assume that the entire output could be produced by one firm. If TC designates total cost and TC(QA, QB) represents the total cost of this one firm producing the current output, subadditivity requires that, for the current output level, TC(QA, QB) < TC(QA,1) + TC (QB,1) + TC(QA,2) + TC(QB,2) for any division of the total output of A and B among the firms. That is, a single firm produces more efficiently. 513 To put the matter somewhat generally, where multi-product firms are involved, not only economies of scale but also economies of scope must be considered. The subadditivity test captures both economies of scale and economies of scope. The latter exist when the total cost of the joint output of a single firm is less than the combined total cost of several firms, each of which is producing a single product. For example, a single firm may be able to produce lamb and wool at a lower total cost than one firm producing wool and one firm producing lamb. 514 Moreover, economies of organisation may arise because transaction costs for organising supply within an industry are minimised with a single supplier. Such economies are part of what can be called "economies of joint production", of which scope economies are a component. There are many examples of the development of an overall system which requires planning and coordination between industry segments. If there are difficulties and costs associated with establishing effective commercial contracts between diffuse firms, the industry may be better organised as a single vertically integrated firm. 515 The focus of a natural monopoly is on the efficiency of a single firm supplying the market's total demand. However, the efficiency of the firm is, as Posner notes, a function of production technology. In this sense, it is possible to think of certain technology or facilities as having natural monopoly characteristics. Strictly speaking, a facility cannot be a natural monopoly. The term describes an optimal industry structure. Nonetheless, it is not uncommon to see references to certain technology or facilities as having natural monopoly characteristics; the reference is to the facility that can most efficiently supply total market demand.

516 A firm operating as a natural monopoly is, in one sense, efficient. By definition, it is able to supply market demand at a lower cost than two or more firms could. But when there is no competition, there can, and usually will, be market failure. 517 Firms with natural monopoly characteristics often exhibit poor economic performance: ie there can be market failure. We accept that in some markets a firm with natural monopoly characteristics will not have much market power. Another supplier may emerge and, at least for a time, trade profitably. At the other end of the scale, a firm with natural monopoly characteristics can, and often does, set prices above marginal cost. 518 Another instance of market failure in the case of industries characterised by a natural monopoly is the likelihood of strategic behaviour by the monopolist to deter entry and to protect its monopoly position. This may entail building excess capacity and other spending of resources in socially wasteful ways. 519 Although, by definition, a natural monopoly involves a production condition that it is less costly to produce output via a single firm, a monopolist may not have the incentive to keep production costs down: ie it may not produce efficiently. There is also the possibility that a monopolist firm will not, as would occur in a competitive market, adopt innovations in production and product. That is, a monopolist is not so much concerned with quality, the adoption of innovations, and investment in research and development as would be a firm facing rivalry from other firms. 520 This brings us to the "essential facility" problem. The problem arises when a firm is vertically integrated: ie where the firm occupies successive levels in the supply chain that leads from raw materials to the final consumer. Indeed, within boundaries which are constantly shifting due to technological change, many firms produce for their own account goods and services that could be acquired from other firms. The negotiating and administering of contracts by which a firm can acquire inputs involves costs. It may be more efficient (ie less costly) for a firm to produce some or all of its inputs. 521 Sometimes technological interdependence implies that vertical integration is more efficient than if two different firms were involved in the production of the relevant product. But it need not be interdependence that results in efficiency; it can arise simply from the limitations on, and the costs of, dealing with a second firm. 522 Although vertical integration brings about efficiencies, it has great potential for adverse effects on competition and therefore on the allocation of resources. There is no doubt that vertical integration can create or raise barriers to entry. Either forward or backward (downstream or upstream) integration by all incumbent firms will require potential competitors to enter two stages of the supply chain simultaneously if the incumbents will not supply new entrants with product at one stage of the process. This increases the capital necessary for entry and the managerial skills required to operate efficiently. The capital input is particularly significant if it would be sunk (irretrievable) if entry fails. Forward (downstream) integration can facilitate the raising of price above marginal cost. Of course, this would not happen if the downstream industry were competitive. 523 An extreme case of market dominance exists where the vertically integrated firm is a network industry or a natural monopoly (eg controls a facility which has natural monopoly characteristics). Assume that access to the facility is necessary to compete in a dependent (upstream or downstream) market. It is these facilities which are referred to as essential facilities. Where a vertically integrated monopolist controls the essential facility, vertical integration will not be social welfare enhancing. This is because the rational profit-maximising firm will take full advantage of all output-restricting and profit-enhancing opportunities in forward or backward markets. At the extreme, management may withhold supply of the good or service produced by the essential facility.

524 In the United States the courts developed the so-called "essential facilities" doctrine, sometimes known as the "bottleneck monopoly" doctrine, to deal with the anti-competitive effects of a monopolist firm that withheld access to important facilities. Put rather simply, the doctrine provides that facilities that cannot practically be duplicated by would-be competitors must be shared on fair terms by those in possession of them: AD Neale, The Antitrust Law of the United States, (2nd ed, 1970) at 67, where the term "essential facilities" doctrine is first used. 525 It is a controversial doctrine with many critics, the chief among them being the late Phillip Areeda, whose article "Essential Facilities: An Epithet in Need of Limiting Principles", 58 Anti-Trust Law Journal 841 (1990), featured prominently in the 2004 decision of the US Supreme Court, Verizon Communications, Inc v Law Offices of Curtis V Trinko, LLP 540 US 398 (2004), which in dicta strongly criticised the doctrine. 526 The foundation for the doctrine is to be found in s 1 of the Sherman Act (15 USC (1994)), which prohibits "[e]very contract, combination … or conspiracy in restraint of trade or commerce" (ie concerted horizontal combinations) and s 2, which condemns "[e]very person who shall monopolise … any part of the trade or commerce" (ie unilateral monopoly misuse). What the statute prohibits is both concerted and unilateral refusals to give access to a facility. 527 The first and most important case is United States v Terminal Railroad Association of St Louis 224 US 783 (1912). Rail access to St Louis required the use of either of two bridges spanning the Mississippi River. An association of fourteen railroad companies which controlled all rail terminals in St Louis acquired control of the two bridges. The association allowed all railroad companies (members and non-members) to use their terminals. They charged new members the same price for access that they charged themselves. This price, however, constituted monopoly rents that disadvantaged non-members. 528 The Supreme Court held that the combination of terminal facilities was unlawful. It found that "as a practical matter it was impossible for any railroad company to pass through, or even enter St Louis, so as to be within reach of its industries or commerce, without using [the Association's] facilities": 224 US at 379. The court emphasised that "in ordinary circumstances, a number of independent companies might combine for the purpose of controlling or acquiring terminals for their common but exclusive use. In such cases other companies might be admitted upon terms or excluded altogether. If such terms were too onerous, there would ordinarily remain a right and power to construct their own terminals": 224 US at 405. Hence the court said that the Association's unified ownership was "an obstacle, a hindrance, and a restriction upon interstate commerce": 242 US at 405. 529 As to relief, the court entered a decree allowing non-member competitor railroad companies access to the facilities essential for the St Louis interchange. The court ordered access to the ten remaining terminals "on such just and reasonable terms as shall place such applying company upon a plain of equality in respect of benefits and burdens [incurred by Association members]": 242 US at 401-411. 530 The second case is Associated Press v US 326 US 1 (1945). Approximately 1200 newspapers created the Associated Press News Organisation (AP). The AP was a vehicle for gathering and exchanging news reports created by local and foreign newspaper members. The AP bylaws allowed existing members to block membership of competing newspapers and thereby remain the exclusive AP outlet in its locality. Blocked entrants had limited alternative news-gathering organisations. 531 The Supreme Court held that the concerted effort of AP members to exclude competitors violated s 1 of the Sherman Act. The Court said that the Sherman Act was: "specifically intended to prohibit businesses from becoming 'associates' in a common plan which is bound to reduce their competitor's opportunity to buy or sell the things in which the groups compete": 326 US at 15. The relief granted required AP news "to be furnished to competitors of members without discrimination": 326 US at 21. 532 The third case, Otter Tail Power Co v United States 410 US 366 (1973), is a unilateral refusal to deal case in violation of s 2. Otter Tail was a wholesaler and retailer of electricity. It had a monopoly in a transmission system. It refused to sell wholesale electricity to municipal systems and refused to wheel (ie transfer by direct transmission or displacement) power on behalf of other suppliers. In the Federal District Court, something resembling the essential facilities doctrine was articulated. The Court held that Otter Tail's "conduct is prohibited by the Sherman Act … [since] the unilateral refusal to deal with another, motivated by a purpose to preserve a monopoly, is illegal": 331 FSupp 54, 61 (D Minn, 1971). As an alternative basis for attacking Otter Tail's conduct, the Court stated that "cases expressing the 'bottleneck theory' of antitrust law" were also relevant. According to the Court, these cases reflect the view that "it is an illegal restraint of trade for a party to foreclose others from the use of a scarce facility" (id at 61) citing, amongst other authorities, the text by A D Neale. 533 The Supreme Court accepted the District Court's determination that Otter Tail had "a strategic dominance in the transmission of power in most of its service areas and that it used this dominance to foreclose potential entrants into the retail area from obtaining electric power from outside sources of supply": 410 US at 377. The court affirmed the District Court's remedial decree, which enjoined Otter Tail from "[r]efusing to sell electric power at wholesale to existing or proposed municipal electric power systems in cities and towns located in [its service area]" or refusing to use its transmission lines to wheel electric power: 410 US at 375. 534 The first case to use the term "essential facilities" was Hecht v Pro-Football Inc 570 F2d 982 (D.C.Cir, 1977); cert denied 436 US 956 (1978). In Hecht, a group of promoters who sought an American Football League franchise challenged a restrictive covenant in the 36-year lease of the Robert F Kennedy Stadium to the Washington Redskins. The contract provided that the stadium could not be rented to any other football team. 535 One issue on appeal was whether the District Court judge had erred in failing to give an instruction to the jury concerning the essential facilities doctrine. The requested instruction would have only related to s 1 of the Sherman Act, but the Court of Appeals expressed the opinion that the "essential facilities doctrine will also support an allegation that the Redskins' refusal to waive the restrictive covenant constituted illegal monopolisation under s 2": 570 F2d at 992. 536 Before reaching that conclusion, the court noted that "[t]he essential facility doctrine, also called the 'bottleneck principle', states that 'where facilities cannot practicably be duplicated by would-be competitors, those in possession of them must allow them to be shared on fair terms. It is an illegal restraint of trade to foreclose the scarce facility'. This principle derives from [Terminal Railroad] and was recently reaffirmed in [Otter Tail]; the principle has regularly been invoked by the lower courts": at 993 n 44. 537 In holding that the District Court had erred in failing to instruct the jury as to the elements of the essential facilities doctrine, the Court of Appeals observed that: "To be 'essential' a facility need not be indispensable, it is sufficient if duplication of the facility would be economically infeasible and if denial of its use inflicts a severe handicap on potential market entrants. Necessarily, this principle must be carefully delimited: the antitrust laws do not require that an essential facility be shared if such sharing would be impractical or would inhibit the defendant's ability to serve its customers adequately": at 992-993 (footnotes omitted). 538 The most important decision in the development of the essential facilities doctrine is MCI Communications Company 708 F2d 1081 (7th Cir, 1983). At issue in that case was the extent to which AT&T had to open up the long-distance telephone market to competitors. MCI had built a long-distance telecommunications network and wanted to connect its network to local telephone facilities in each of the cities that were served by MCI. It also sought access to AT&T's long-distance network for the cities not reached by its own network. 539 The Seventh Circuit set out a four-part test to deal with the essential facilities claim that has been followed in many later cases. According to this test, courts must examine the following factors: (1) control of the essential facility by a monopolist; (2) a competitor's inability, practicably or reasonably, to duplicate the essential facility; (3) the denial of the use of the facility to a competitor; and (4) the feasibility of providing the facility. 540 The second element is part of the definition of what is an essential facility. That is to say, if the facility can be reasonably or practicably duplicated, it is highly unlikely or even impossible that it will be found to be essential at all: City of Anaheim v Southern California Edison Company 955 F2d 1373 (9th Cir, 1992). The fourth element raises the question whether there is a legitimate business justification for the refusal to provide the facility: Metronet Services Corporation v Qwest Corporation 383 F3d 1124 (9th Cir, 2004). 541 The Seventh Circuit explained the rationale behind the doctrine. It accepted as a general proposition that antitrust laws do not require a monopolist to deal with anyone: see eg United States v Colgate and Co 250 US 300, 307 (1919). On the other hand, the doctrine requires a monopolist controlling an essential facility to make it available to competitors on non-discriminatory terms to prevent the monopolist from extending "monopoly power from one stage of production to another, and from one market into another": MCI 708 F2d at 1132 (7th Cir). 542 Hovenkamp, Federal Antitrust Policy, The Law of Competition and Its Practice (3rd ed, 2005) at 7.7a, lists three categories of facilities which the courts have found to be essential: (1) natural monopolies or joint venture arrangements subject to significant economies of scale; (2) structures, plants or other valuable productive assets that were created as part of a regulatory regime, whether or not they are properly natural monopolies; and (3) structures that are owned by the government and whose creation or maintenance is subsidised. 543 The following observations may be made about the MCI test. First, there is no definition of the "facility" mentioned in the first element. Ordinarily one would envisage a structure of some sort. But, then again, there was no "structure" in Aspen Highland Scheme Corp 738 F2d 1509 (10th Cir, 1984). 544 Second, to be essential, a facility need not be indispensable. A facility will be essential if an alternative facility is impossible or unduly expensive to construct. It is not sufficient for the plaintiff to show that the facility is merely "more economical" than the alternatives. Further, the facility must be central to the competitor's viability in the relevant market or at least present a severe handicap to market entry if it is not made available to the would-be entrant. In substance, the test of economically infeasible duplication, which is derived from Hecht, inquires: if the plaintiff is able to duplicate the facility, will the plaintiff obtain a reasonable rate of return on the cost that will be incurred? 545 Third, whilst there must be a monopolist, the monopolist need not exist because it is a natural monopoly. While it is true that most of the cases involve a natural monopoly, this is not a necessary condition. 546 Fourth, the third element (the feasibility of providing the service), which is also derived from Hecht, ensures that the monopolist is only required to make (presumably current) excess capacity, not capacity it is using itself, available to competitors. 547 Fifth, for the most part, the cases involve vertically integrated firms where access to the facility is required so that a firm can compete in another, usually downstream, market. 548 Finally, the doctrine requires a monopolist to give access to a competitor. The competitor is a competitor not in the market for the provision of the facility where the monopolist (being a monopolist) will have no competitor, but in a dependent market, where the essential facility is a necessary input to the provision of products or services.

552 On 4 October 1992, the then-Prime Minister announced the establishment of an independent inquiry into competition policy, to be chaired by Professor Fred Hilmer, Dean of the Australian Graduate School of Management at the University of New South Wales. The Hilmer Committee Report was delivered on 25 August 1993. It made a number of recommendations in relation to the nature of competition policy as well as the principles that should guide the implementation of such a policy. 553 One of the policy principles, provision for third party access to nationally significant infrastructure, is the genesis of Part IIIA. This topic is dealt with in Chapter 11 of the report. But, to put that chapter into its proper perspective, it is necessary to consider another policy principle, the restructure of public sector monopoly businesses, which is covered by Chapter 10. 554 In that chapter, the report discussed what conditions were necessary for establishing effective competition in markets traditionally supplied by public monopolies. It suggested that it was necessary to restructure these monopolies. Among the proposals considered was the separation of regulatory functions from commercial functions, and that natural monopoly elements may need to be separated from potentially competitive activities and potentially competitive activities may also have to be separated from each other. 555 As regards the second possibility (the separation of natural monopoly elements from competitive activities), the report described "natural monopoly characteristics" as being present when "a single firm can supply the entire market most economically", giving as examples electricity transmission grids and rail tracks. It described (at 218) what it considered to be the two competing policy concerns with public monopolies having "natural monopoly characteristics". The first was that irrespective of whether the natural monopoly element was integrated vertically or horizontally with the potentially competitive element, a natural monopoly presented opportunities for cross-subsidisation. The second concern arose when there was a vertical relationship between the two activities. In that event, integration of the natural monopoly element and a potentially competitive activity may give rise to abuse by stifling or preventing competition in the potentially competitive sector. 556 The report suggested that there were two broad regulatory alternatives for addressing these concerns. The first was that the natural monopoly element be separated from the potentially competitive element. The second was to impose intrusive regulatory controls to guard against cross-subsidisation. 557 The report then turned to consider access. Chapter 11 began by identifying the problem with essential facilities. The report stated (at 239) that: "In some markets the introduction of effective competition requires competitors to have access to facilities which exhibit natural monopoly characteristics, and hence cannot be duplicated economically." This description of the consequence of a monopoly being "natural" is very important. So also is the definition provided of an "essential facility". It was defined (at 239) as "a monopoly, permitting the owner to reduce output and/or service and charge monopoly prices, to the detriment of users and the economy as a whole". The report explained that where the owner of the facility is also competing in markets that are dependent on access to the facility, the owner can restrict access to the facility to eliminate or reduce competition in dependent markets. 558 The report then described what it saw as the "essential facilities problem". These facilities, which it said (at 240-241) exhibit natural monopoly characteristics, "occupy strategic positions in an industry, and are thus 'essential facilities' in the sense that access to the facility is required if a business is to be able to compete effectively in upstream or downstream markets. … Where the owner of the 'essential facility' is not competing in upstream or downstream markets, the owner of the facility will usually have little incentive to deny access, for maximising competition in vertically related markets maximises its own profits. … Where the owner of the essential facilities is vertically-integrated with potentially competitive activities in upstream or downstream markets … the potential to charge monopoly prices may be combined with an incentive to inhibit competitors' access to the facility." 559 The report acknowledged that it was difficult to define precisely the term "natural monopoly", but instanced electricity transmission grids, telecommunication networks, rail tracks, major pipelines, ports and airports as examples. The report referenced the IAC Government (Non-Tax) Charges vol III (1989), where there is a detailed discussion of the concept of natural monopoly. 560 The Hilmer report suggested two possible methods of mandating access to essential facilities. One was to rely on the general rules governing the misuse of market power. But, as the report noted, that was a difficult proposition in light of the prevailing case law on s 46 of the Trade Practices Act, the section which proscribes the misuse of market power. 561 In this context, the report mentioned the so-called "essential facilities doctrine" developed in the United States through the application of the Sherman Act. The report observed (at 244) that the limits of the US doctrine were not clear and that commentators had observed that "the doctrine has not developed with clarity, coherence or consistency, let alone with strong economic foundations". 562 The second method of mandating access was to establish a special statutory access regime. In the end, this was the preferred option. The report (at 251) recommended that access should only be declared where "the facility in question is essential to permit effective competition in a downstream or upstream activity". In other words, access to the facility should be essential, rather than merely convenient. The report also recommended (at 251) that making a declaration should be in the public interest, having regard to: (a) the significance of the industry to the national economy; and (b) the expected impact of effective competition in that industry on national competitiveness. 563 The report (at 261) recognised that the owner of an essential facility required protection. It recommended that a right of access only be created if "the legitimate interests of the owner of the facility will be protected by the imposition of an access fee and other terms and conditions that are fair and reasonable".

564 The Council of Australian Governments considered the Hilmer Report and agreed to a reform package comprising legislative and non-legislative action. The Competition Principles Agreement is a key element of the package. It contains the guidelines for the development of competition policy. Among other things, the agreement required the Commonwealth to enact legislation to establish a regime to provide access to "significant infrastructure facilities". Relevantly, cl 6 of the agreement provides: (1) Subject to subclause (2), the Commonwealth will put forward legislation to establish a regime for third party access to services provided by means of significant infrastructure facilities where: (a) it would not be economically feasible to duplicate the facility; (b) access to the service is necessary in order to permit effective competition in a downstream or upstream market; (c) the facility is of national significance having regard to the size of the facility, its importance to constitutional trade or commerce or its importance to the national economy; and (d) the safe use of the facility by the person seeking access can be ensured at an economically feasible cost and, if there is a safety requirement, appropriate regulatory arrangements exist. 565 The reform package contained draft legislation setting out the criteria that should be satisfied before there would be a declaration of a service. The relevant criteria were: (a) that access (or increased access) to the service could promote competition in at least one market (whether or not in Australia), other than the market for the service; (b) that no other facility exists that can economically provide the service; (c) that it would be uneconomical for anyone to develop another facility to provide the service; (d) that the facility is of national significance, having regard to: (i) the size of the facility; and (ii) the importance of the facility to constitutional trade or commerce; (e) that access to the service can be provided without undue risk to human health or safety; (f) that access to the service is not already the subject of a single, effective access regime; (g) that access (or increased access) to the service would be in the public interest.

566 The legislation that was enacted, the Competition Policy Reform Act 1995 (Cth), which introduced Part IIIA, substantially watered down the Hilmer Committee proposals. The criteria that had to be satisfied to enable a declaration to be made were that: (a) access (or increased access) to the service would promote competition in at least one market (whether or not in Australia), other than the market for the service; (b) it would be uneconomical for anyone to develop another facility to provide the service; (c) the facility is of national significance, having regard to: (i) the size of the facility; or (ii) the importance of the facility to constitutional trade or commerce; or (iii) the importance of the facility to the national economy; (d) access to the service can be provided without undue risk to human health or safety; (e) access to the service is not already the subject of an effective access regime; (f) access (or increased access) to the service would not be contrary to the public interest. 567 Three points are noteworthy. First, criterion (a) required only that competition be "promoted" by access, as long as this was not contrary to the public interest. This was a much lower threshold than the Hilmer Committee's recommendation. Second, the public interest test was stated differently from that suggested by the Hilmer Committee, although it is possible that it was to the same, or to a similar, effect. Third, the draft reform package had included as a criterion "that no other facility exists that can economically provide the service". This was dropped in the legislation, so the criteria can be satisfied even where there is an existing facility that can provide the service. 568 It is also worthwhile noting that no objects clause was contained in the Bill, nor in Part IIIA when originally enacted. Nonetheless, in the Senate, the Second Reading Speech to the Bill explained that the purpose was: "to establish a legal regime to facilitate third parties obtaining access to the services of certain essential facilities of national significance. The notion underlying the regime is that access to certain facilities with natural monopoly characteristics, such as electricity grids or gas pipelines, is needed to encourage competition in related markets, such as electricity generation or gas production": Commonwealh, Parliamentary Debates, Senate, (1995) Vol S107, 2434. 569 Once a declaration is made following a finding that access to a facility satisfies the statutory criteria, the terms and conditions of that access must be worked out. Following the Hilmer Committee recommendations, Part IIIA requires the parties (the facility provider and access seeker) to enter into negotiations concerning access. If an agreement cannot be reached, there is a compulsory arbitration process with the ACCC as the arbitrator. 570 The ACCC's powers in an arbitration are found in s 44V. In the arbitration, the ACCC may make or not make a determination concerning access. That is, the arbitration process need not necessarily result in access being granted. 571 In making a determination, s 44V(2) provides that the ACCC may "deal with any matter relating to access by the third party to the service, including matters that were not the basis for notification of the dispute". By way of example, the determination may: (a) require the provider to provide access to the service by the third party; (b) require the third party to accept, and pay for, access to the service; (c) specify the terms and conditions of the third party's access to the service; (d) require the provider to extend the facility; (da) require the provider to permit interconnection to the facility by the third party (this was not originally provided for in s 44(V)(2): see Chapter 9.5.) (e) specify the extent to which the determination overrides an earlier determination relating to access to the service by the third party. 572 The Hilmer Committee's recommendation that the legitimate interests of the owner be protected was given effect by s 44W(1). That section provides that the ACCC must not make a determination that would have any of the following effects: (a) preventing an existing user obtaining a sufficient amount of the service to be able to meet the user's reasonably anticipated requirements, measured at the time when the dispute was notified; (b) preventing a person from obtaining, by the exercise of a pre-notification right, a sufficient amount of the service to be able to meet the person's actual requirements; (c) depriving any person of a protected contractual right; (d) resulting in the third party becoming the owner (or one of the owners) of any part of the facility, or of extensions of the facility, without the consent of the provider; (e) requiring the provider to bear some, or all, of the costs of extending the facility or maintaining extensions of the facility.

573 The National Competition Policy provided for the review of the access regime following five years of its operation. That review was conducted by the Productivity Commission in 2000-01 and published on 28 September 2001. The Commission's report recommended retention of the access regime but made several suggestions for its improvement. Most suggestions were adopted by the government. It is only necessary to mention some of them. 574 First, the Commission recommended that the legislation should specify the objectives of access as there was scope for divergence between the intent of the legislation and the interpretation of the operational criteria (at 124). The Tribunal in Re Sydney International Airport [2000] ACompT 1 ("Sydney Airport (No 1)") at [107] had taken the view that the purpose of access was to "unlock a bottleneck so that competition can be promoted in a market other than the market for the service". The Productivity Commission, however, observed that the promotion of competition is desirable only when it is efficient. Accordingly it proposed the following objects clause (at 130): The objective of this part is to: (a) enhance overall economic efficiency by promoting efficient use of, and investment in, essential infrastructure services; and (b) provide a framework and guiding principles for industry-specific access regimes. 575 Second, the Commission considered the declaration criteria. It observed, correctly, that criterion (a) set a lower standard than initially envisaged by the Hilmer Committee. It recommended that this criterion be amended such that access (or increased access) to the service would promote a substantial increase in competition in at least one market (whether or not in Australia) other than the market for the service. Yet it suggested it was best to wait for its next review to examine further whether the criterion needed to be strengthened. 576 Third, the Commission considered the "uneconomical to develop" test in criterion (b), observing that it was generally considered to test for the existence of a natural monopoly. Submissions made to the Commission about this criterion were divided. Some suggested that criterion (b) should be strengthened, others that it should be weakened, and others argued that it should be made clearer, either by defining what was meant by natural monopoly or by introducing a substitute concept. 577 The Productivity Commission saw an in-principle case for focussing more explicitly on monopoly power and efficiency issues in the criterion. Once again, it thought it best to wait for its next scheduled review to assess whether further strengthening of any particular criterion, or recasting of that criterion to focus explicitly on market power and efficiency considerations, was required. 578 Fourth, the Commission gave consideration to the ACCC's power under s 44V to make a determination requiring an owner to extend the facility. In a position paper published in 2001, it raised for consideration the question whether there was a need to include an expansion power. The position paper (at 166) emphasised the rationale for such a power being that: · Some facilities have been constructed at a sub-optimal size and/or capacity intentionally in order to avoid an access regime; · Demand for the service has increased, but it still would be uneconomic to build another facility. The Commission did not, however, see a need for including an expansion power in ss 44V or 44W. 579 With respect to facility extensions, the Commission's view was that requirements to extend facilities would generally go beyond the legitimate reach of an access regime. Hence it suggested that when arbitrating a dispute for a declared service, any scope for the ACCC to require extensions of facilities should be removed. 580 On the other hand, the Productivity Commission did say that the ACCC should be able to require that a service provider permit interconnection to its facility by an access seeker. 581 In its final report, the Productivity Commission acknowledged (at 224-226) that difficulties and complexities may make directed capacity expansions inappropriate in many circumstances. Nevertheless, it said that instances could arise where there would be a case for the relevant regulator to direct a facility owner to expand capacity. It did, however, indicate that providing for such expansions should not leave the service provider at a financial disadvantage. In the end, the Commission's report concluded (at 226): "While the scope for directed capacity expansion is a feature of some industry-specific regimes, the Commission considers that such a provision is not warranted in Part IIIA given that it is a residual access route." The Commission seemed to be of the view (which can also be seen at p 224) that the ACCC had implicit power to order expansions, and did not recommend that this power be made express. 582 The government's responses to the recommendations were as follows. It agreed that there should be an objects clause. By the Trade Practices Amendment Act (No 1) 2006 (Cth), s 44AA was introduced and reads: The object of this Part is to: (a) promote the economically efficient operation and use of, and investment in, essential infrastructure services, thereby promoting effective competition in upstream and downstream markets; and (b) provide a framework and guiding principles to encourage a consistent approach to access regulation in each industry. Amendments were also made requiring the relevant decision-maker (the NCC, the Minister, the ACCC and the Tribunal) to have regard to the objects clause. 583 The government initially agreed in-principle with the recommendation that criterion (a) be changed so that access would result in a substantial increase in competition. In its response, however, it stated that: "In this context, the term 'substantial' may exclude situations where a small supplier is prevented from gaining access to nationally significant infrastructure. The government therefore will include the word 'material' to ensure access declarations are only sought where the increases in competition are not trivial." Accordingly, criterion (a) was amended to read: (a) that access (or increased access) to the service would promote a material increase in competition in at least one market (whether or not in Australia), other than the market for the service. 584 The amendment brought no change to the existing law. The Tribunal had always taken the position that criterion (a) required a non-trivial increase in competition: Re Duke Eastern Gas Pipeline Pty Ltd [2001] ACompT 2and Sydney Airport (No 1) are cases on point. 585 As regards the recommendation to amend s 44V(2) to cover interconnection, a new subsection (da) was added, which provides that a determination may: (da) require the provider to permit interconnection to the facility by the third party.

586 There were two aspects of Part IIIA that were of concern to the Western Australian government. The first was the role of the ACCC in setting the terms and conditions of access. The second was that Part IIIA allows the NCC to assess whether a state or territory access regime is "effective" and thus outside the operation of Part IIIA. 587 The Western Australian government decided that access to government owned infrastructure should be determined in Western Australia rather than by federal agencies. Hence it resolved to enact legislation that provided an effective access regime for those parts of the state rail network that fell within the ambit of Part IIIA. 588 The legislation is the formerly titled Government Railways (Access) Act 1998 (WA). The Access Act applied only to railways under the control of the state. In broad outline the Access Act operated in the following way. It provided for the establishment of a Rail Access Code to govern the use of government railways for rail operations to persons other than the state: s 4. It designated a regulator for monitoring and enforcing functions relating to the implementation of the code: s 20. It specified the kind of administrative arrangements that the state was to have in place for the purposes of that implementation: Part IV. 589 The Rail Access Code, which is subsidiary legislation, establishes the parts of the railway network and associated infrastructure opened to access; outlines the process and procedures to negotiate access, including avenues for dispute resolution; specifies the matters to be considered in access agreements; identifies the information requirements of the regulator; and outlines the pricing principles to be applied in determining prices to be paid for access: s 4(2). It requires the access provider to submit for regulatory approval a variety of policies regarding, for example, segregation of access related functions from other parts of the access provider's business (s 42), train management guidelines (s 43), train path policies (s 44), and costing principles (s 46). 590 To add new or remove existing routes under the Code, the Minister is to consider criteria which, in all substantial respects, are the same as the criteria under Part IIIA, save that they are railway specific: s 5. 591 In 2004 the Access Act was amended. Instead of applying only to state railway networks its coverage was broadened to include privately owned railways including, in particular, the Chichester line: see the definitions of "railways network" and "TPI Railway and Port Agreement" in s 3(1). In acknowledgment of its broader operation the title of the Access Act was changed to Railways (Access) Act 1998 (WA).

PROTECTION OF THE INCUMBENT 592 This and the following chapters deal with issues which concern the capacity of the lines and related matters. An orderly way to proceed is to start by considering an incumbent's entitlement to use the capacity of its own line. This takes us to s 44W(1)(a) (the "requirements safeguard"). 593 The requirements safeguard provides that the ACCC must not, in an access dispute, make a determination which has the effect of preventing an existing user (which would include the owner) obtaining a sufficient amount of the service to be able to meet its reasonably anticipated requirements, measured at the time a dispute is notified. 594 There are a number of points to note about this provision. First, the phrase "reasonably anticipated requirements" is forward looking, and not confined to an existing user's current requirements. Second, the requirements safeguard protects "existing users". An existing user is a person (including the owner) who is using the service at the time the dispute is notified: s 44W(5). There may be two or more existing users at any one time. Suppose, for example, that both the owner and a third party who has gained access are using the line. Both would be existing users. 595 Third, the phrase "requirements" relates to requirements in respect of the service. The amount (quantum/quantity) of the service required will often depend on the manner in which the service is used. Take the use by BHPB and RTIO of their lines. They adopt a flexible, "run when ready" use of the lines. That usage varies from day-to-day. On some days, the owner's demand may be particularly acute. Obviously, on those days the amount of the service the owner requires is significant. Moreover, in the case of both BHPB and RTIO, the day-to-day requirements are unpredictable and always subject to regular change. For the requirements safeguard to apply sensibly in these circumstances, the owner must have priority to a sufficient amount of the service to enable it to transport its target volume of iron ore using its run when ready system, and any third party user must "fit in" with the owner's requirements for flexibility. This, we should say, is the basis on which the parties proceeded. 596 Fourth, the safeguard protects reasonably anticipated requirements measured at the time when the dispute is notified. It does not address what is to happen if those requirements change. This is a limitation which will be discussed when considering criterion (f). 597 It must be said that the requirements safeguard does not work well in a number of situations. One is where both the owner and a third party are existing users and another third party notifies the ACCC of an access dispute requiring arbitration. When the existing third party user originally applied for access, the requirements safeguard gave priority to the owner's use. When a new access seeker raises a dispute, the requirements safeguard now applies to both existing users. As against the new access seeker, both existing users have priority. As against each other, neither appears to have priority. The practical effect, if one were to adopt a strict reading of the requirements safeguard, is that each existing user has equal priority: ie the owner's initial priority has been lost simply because there are two existing users at the time of the new dispute. 598 Such an anomalous result cannot, we think, have been intended. But, as s 44W(1)(a) is currently drafted, it is a result that is difficult to avoid. We had given consideration to reading the provision in a way that a third party could only "reasonably" anticipate its requirements for use of the line if that use does not displace or obtain priority over the owner's requirements. We think, however, that this reading stretches the language of the requirements safeguard too far. An existing user's requirements will principally reflect its own business needs. To say that it is unreasonable to anticipate those requirements because another existing user also requires the line is too strained a construction. In any event, such a construction would be difficult to apply. Assume there are multiple existing third party users. Each third party's "reasonably anticipated requirements" would presumably depend on both the owner's and other third party users' requirements and whatever order of priority applied among them. The reasonably anticipated requirements of the user with lowest priority could only be determined after the reasonably anticipated requirements of higher priority users have been determined. 599 A similar analysis applies where a third party gains access to the service and then, some time later, that third party notifies the ACCC of a fresh access dispute. Assume, for example, that six months before its current term of access expires, the third party raises a new dispute regarding access to the service for a further term. Assume also that at the time of the dispute both the owner and the third party user are "existing users". The anomalous result, which cannot be avoided, is that the owner loses its priority merely because of the existence of a new dispute. 600 The operation of the requirements safeguard is also problematic where there is a new dispute between an existing third party user and the owner in respect of the earlier determination. In this instance, s 44W(2) is triggered. That subsection provides that the requirements safeguard and s 44W(1)(b) do not apply in relation to the requirements and rights of the third party and the owner when the ACCC is arbitrating a dispute relating to an earlier determination of an access dispute between the third party and the owner. 601 Subsection (2) prevents the operation of both the requirements safeguard and s 44W(1)(b). The latter provides that the ACCC must not make an order having the effect of preventing a person from obtaining, by the exercise of a pre-notification right, a sufficient amount of the service to be able to meet the person's actual requirements. "Pre-notification right" means a right under a contract, or under a determination, that was in force at the time when the dispute was notified. An existing third party user's right under an earlier determination would be a pre-notification right. 602 The apparent intention of subsection (2), according to the explanatory memorandum to the Competition Policy Reform Bill 1995 (at [231]), is to allow the ACCC to "over-ride the rights and obligations of the provider and the third party under the earlier determination". For this reason, the ACCC may make an order preventing the existing third party user exercising its pre-notification rights under the earlier determination. Subsection (4) makes provision for compensation due to loss of pre-notification rights. 603 Subsection (2) gives rise to an anomaly, similar to that discussed earlier, where an owner who has priority over a third party user at the time of an earlier determination, may lose that priority merely because of a new dispute in relation to the earlier determination. We do not think this anomaly can be avoided as the legislation is presently worded. In saying this, we appreciate that subsection (2) operates so that, if an owner is deprived of its priority rights under an earlier determination, it might be entitled to compensation under subsection (4). This will give the owner some comfort. But it will not be of much assistance in the case of losses suffered because the owner no longer has priority in respect of its reasonably anticipated requirements as at the time of the new dispute, requirements which may have increased significantly since the earlier dispute. 604 Section 44W is not the only provision which is relevant to whether the owner has priority to the service. The objects clause and s 44Z(1)(a) are also relevant. One object (to which we regularly refer) is to promote the economically efficient investment in the infrastructure by which services are provided. Section 44X(1)(a) provides that in making a final determination, the ACCC must take into account the legitimate business interests of the provider and the provider's investment in the facility. FMG suggests that while s 44X(1)(a) does not guarantee the owner protection in respect to its demand for the service, it is nonetheless an important discretionary factor. It refers in this regard to BHP Billiton Iron Ore Pty Ltd v National Competition Council (2008) 236 CLR 145 at [43]. There, the High Court states the uncontroversial proposition that in making an access determination, the ACCC must take into account the incumbent's legitimate business interests. We do not think that this passage can be read, as FMG would have it, as implying that special, if not overwhelming, weight will be given to this factor relative to other discretionary factors. 605 Nonetheless, there are other reasons for thinking that the ACCC would not readily make a determination which did not enable the owner to meet its reasonably anticipated requirements. First, a contrary determination would drastically undermine the owner's incentives to invest in infrastructure in the future. It is one thing to give access to a service which the owner is not using; it is another to force the owner to forgo its own usage. Second, a determination which prevents the owner from meeting its requirements could be an acquisition of property. Section 44ZN provides for the Commonwealth to pay compensation if a determination would be constitutionally invalid because it does not sufficiently compensate a person for an acquisition of property. Notably, the pricing principles in s 44ZZCA do not envisage the payment of compensation for such an acquisition. The price for access is based on efficient costs of providing access and a return on investment commensurate with the regulatory and commercial risks involved. There is, therefore, risk that s 44ZN would be triggered, a clumsy outcome at best. 606 To summarise, the Tribunal is of the view that the requirements safeguard will give the owner priority use of the service as against a new access seeker, but that protection may be lost in a variety of circumstances. Still, even if the requirements safeguard is lost, there will generally be powerful discretionary factors for favouring the owner in any arbitration. For this reason, the Tribunal will proceed on the basis that an owner will generally have priority of use over third parties, noting limitations to this rule where relevant.

607 There has been considerable debate over whether the four railway lines have capacity to accommodate third party usage. Before dealing with this debate, it is necessary to mention the legal context in which it arises. 608 Part IIIA does not expressly refer to the capacity of a service to accommodate the requirements of an access seeker. Nonetheless, it is clear that in deciding whether to declare a service, it is necessary to consider whether the facility providing the service has spare capacity for third party use. This is because of the requirements safeguard: s 44W(1)(a). Hence, if there is no spare capacity for third party use then, unless the facility can be expanded, a declaration would be futile. 609 Putting to one side the existence of an expansion power, what is the position if there is some capacity for third party use but not sufficient capacity to cater for all third party demand? Most infrastructure access regimes have a mechanism to allocate the capacity of the infrastructure between users on some fair and reasonable basis. The access regime established by Part IIIA gives priority to the reasonable requirements of the incumbent (who is not necessarily the facility owner) but makes no provision for the allocation of the remaining capacity between third parties. 610 Obviously allocating spare capacity will create significant problems, made worse because the demand from third parties is unlikely to arise at the same time. Part IIIA's intention is to foster 'regulation by negotiation' at the access stage. This intention will be undermined when, as is likely, potential access seekers do not think it worthwhile to take the time to negotiate because other access seekers have 'beaten it to the chase' by obtaining earlier access. Moreover, by s 44W(1)(a) access seekers who are 'existing users' of the service have priority over new access seekers. 611 Even if insufficient capacity can be overcome by a power in the ACCC to direct the facility owner to expand the facility, other problems will arise. The most obvious will be: Who pays the cost of the expansion works, a cost that could be upwards of hundreds of millions, if not billions, of dollars? This problem is particularly acute in the case of successive access seekers, some of whose requirements can be met out of existing spare capacity and others only by an expansion. Once again, Part IIIA is silent. 612 In due course, it will be necessary to consider this issue. In the meantime we will examine whether each railway line has sufficient spare capacity to accommodate the demand of potential access seekers, bearing in mind that access seekers must effectively 'fit in' with the incumbent's requirements.

613 When assessing capacity it is necessary to do so by reference to a particular point in time. The extent to which the subject railway lines can accommodate third party usage will vary over time. The reasons are, first, the incumbent's requirements are constantly changing and, second, the physical configuration of the lines (apart from the eastern section of the Goldsworthy line) will change significantly as the incumbents expand their operations. These forces will continue for at least the period for which the declarations are sought. 614 The Tribunal has decided to focus (though not exclusively) on the incumbents' anticipated usage and on the configuration of the lines around 2014/2015. An in-depth examination of capacity at other points in time would require detailed evidence, most of which is not to hand, as well as a large degree of speculation. In any event, such a task would be unworkable and inconsistent with the administrative character of the access declaration enquiry. 615 There are several reasons why the Tribunal has chosen 2014/2015 as its reference point. First, it is unlikely that if a declaration were made now any third party access would occur before then, given (1) third parties would require some time to establish train and port infrastructure; (2) the need to negotiate or arbitrate terms of access, a process that will inevitably take time; and (3) the possibility of time-consuming appeals and other legal challenges against this decision, and any arbitrated determination. Second, and just as important, it is around 2014/15 when the incumbents' RGP6 (BHPB) and 330mtpa (RTIO) expansion plans will be implemented. These changes to the railway lines will be significant and there is little point looking at the lines at any earlier point. 616 The Tribunal understands that any assessment of the most likely physical state of an incumbent railway line at 2014/2015 is fraught with difficulty. What is thought to be an important aspect of the design of a railway on one day can be significantly altered the next, which means that any conclusion drawn from the evidence can be tentative at best. But that is better, in the Tribunal's view, than an assessment based on soon-to-be-outdated facilities.

617 How capacity is defined, let alone calculated, is not straightforward. The starting point for defining capacity is to ask: Capacity to do what? In 2004, the International Union of Railways (UIC) published a leaflet - The UIC 406 Capacity Leaflet - which states that "capacity as such does not exist. Railway infrastructure capacity depends on the way it is utilized." The service the subject of each declaration application is use of the lines. All of the parties accepted that the lines are to be exclusively used for the haulage of iron ore. It is the capacity to perform that task which is the subject of the inquiry. 618 What is meant by "capacity" to haul iron ore? Railway capacity is a complex, but elusive, concept. There is no single, universal definition of rail capacity. Many definitions are collected in Alex Landex's doctoral thesis "Methods to Estimate Railway Capacity and Passenger Delays" (PhD thesis, Technical University of Denmark, 2008). Some view capacity as simply the maximum number of trains which can be operated on a line in a given period. Others view capacity in terms of maximum performance taking into account system constraints. Those constraints might relate to, for example, service levels and cost, punctuality, or a given set of resources under a specific service plan. Others again view capacity in terms of the ability of the railway to handle timetables, such as the range of timetables which can be accommodated, or the number of trains which can reliably be fitted into a viable and commercially attractive timetable. 619 It will be observed that many of these meanings are concerned with timetabling. Here, where the incumbents haul iron ore on a "run when ready" basis, timetabling is not particularly relevant. Regardless of whether the railway is timetabled or not, all definitions of capacity acknowledge that the capacity of a railway will depend on railway infrastructure, including rolling stock, and the manner in which trains are run. 620 UIC 406 makes the obvious point that railway capacity may be viewed differently, dependent upon the perspective of the investigator. For instance, from the consumer's (ie customer's) viewpoint, capacity will take into account the expected number of train paths, the expected mix of traffic, the speed of the train, the infrastructure quality, whether journey times are as short as possible, and that all short- and long-term demand is met. Those involved with infrastructure planning will have regard to the expected number of train paths (on average per specified time period), the expected mix of traffic and speed (on average), the expected condition of the infrastructure, and time supplements for expected disruptions and maintenance. Those planning the timetable will consider the requested number of train paths, the requested mix of traffic and speed, existing condition of the infrastructure, time supplements for expected disruptions, time supplements for maintenance, connecting services in stations, requested out of regular interval timetables (system times, train stops, etc). Those operating a railway will have regard to the actual number of trains, the actual mix of traffic and speed, the actual condition of the infrastructure, delays caused by operational disruptions, delays caused by track works and delays caused by missed connections. 621 FMG has opted for the definitions of capacity suggested by Abril et al. in their paper "An Assessment of Railway Capacity" Department of Informational Systems and Computation, Department of Applied Statistics and Operational Research, Technical University of Valencia, Spain, (20 April 2007). These definitions are by no means universally accepted in the academic world, but were adopted for the purposes of argument in this review. Those definitions are: · Absolute capacity (sometimes referred to as theoretical capacity), being the number of trains that could run in a perfect environment (trains running permanently at minimum operational and safe headway - ie the time intervals between trains). This measure is theoretical because it ignores the effect of unknown and unpredictable events such as variations in traffic and operating conditions that occur in reality. · Practical or sustainable capacity, being the practical number of trains that can be moved on a line at a reasonable level of reliability. That is, it is the capacity that is available under normal operating conditions, taking into account the system and operational variability which inevitably occurs. · The capacity used by the actual volume of traffic occurring over the rail network. Available spare capacity, in a practical sense, is the difference between used capacity and sustainable practical capacity. It indicates the additional traffic volume that can run on the railway over the specified time period. 622 While the nomenclature used by the rail experts and other witnesses differed somewhat, it was largely agreed that capacity could be viewed at these alternative levels. 623 As all definitions of capacity make clear, railway capacity is not static, even in a defined time period. For present purposes it is important to understand how variability affects capacity. Abril et al. explain that capacity is extremely dependent on how the railway is used. It is dependent upon the mix of trains and the order in which they run. It varies with changes in infrastructure and operating conditions. Some of the more important factors that affect capacity include: (1) infrastructure parameters - eg block and signalling systems which affect train speed, single or double tracks, the number of halts along the line, network effects (ie crossing and overlapping lines), track structure and speed limits, and the length of sections; (2) traffic parameters - such as new or existing lines, the mix of trains, train scheduling problems, traffic concentration and the priority of trains; and (3) operating parameters - including planned events (maintenance), unplanned events (train failures, broken rails, etc), train stop times, and quality of service. 624 Complicating matters further is the distinction between "physical" or "choke" capacity on the one hand and what is most accurately described as "economic utilisation" of a line on the other. Capacity is influenced by the level of traffic on the line. Where a line accommodates trains travelling in opposite directions, the greater the volume of traffic there is, the greater the possibility for "meets" (ie where one train must move off the line in order to allow the other to pass). Meets create congestion and slow down travel times for trains. Professor Ferdows, a production systems expert, describes how there is a non-linear relationship between the number of trains running on a line and the extent to which the line's capacity is utilised. The marginal value of additional ore carried by each extra train decreases as a result of ever-increasing congestion. Professor Ferdows represents this relationship diagrammatically: 625 "Physical" or "choke" capacity is the point where the addition of an extra train does not result in any additional ore being carried on the line. The train simply adds more congestion. In the diagram it is the point at which the curve in the graph asymptotes (ie where 100% of the capacity is utilised). The "economic utilisation" of the line is some point lower down the curve where the marginal value of the extra ore carried by an additional train is less than the marginal cost of adding that train. 626 When determining capacity it is necessary to choose an appropriate unit of measurement. Two units of measurement were suggested. One is to measure capacity in terms of tonnage transported over a defined period. This is the measure proposed by BHPB and RTIO, which they use for their own internal purposes. 627 FMG says this is not a fair benchmark and that the appropriate unit of measurement should be train paths over a defined period. With regard to the BHPB and RTIO approach, FMG points out that train configurations (the number of rakes in a consist and the number of ore cars in a rake) can, and often do, change. Any one particular configuration may result in more ore being transported than another in a given period of time. FMG says that a change in train configuration may be made for entirely subjective reasons unconnected with any constraint imposed by the railway itself and may reflect, it is argued, inefficient practices of the incumbent. Hence it contends that it is unfair to use tonnage as a unit for measuring capacity, because it is "self evident that the train paths measure reflects a much fairer and objective view of railway line capacity than tonnage". 628 In this connection, FMG relies on a report prepared by BRS which was prepared for [c-i-c]. The report says: Expressing the capacity in terms of tonnes … depends on the net load of the trains travelling across the railway. The real capacity of a railway is the assessed number of train movements that can occur over the railway in any given time period. This is equivalent to nominating a roads [sic] capacity as the number of vehicle movements per day. While the net load of trains is a function of the network's design, it is also influenced by other factors such as the capacity of terminals to handle particular train sizes. The road equivalent is to express the road's capacity in terms of passenger movements per day. Again this measure is very much influenced by exogenous factors such as the number of people per vehicle. 629 FMG also refers to Kozan & Burdett, "A Railway Capacity Determination Model and Rail Access Charging Methodologies" (2005) 28(1) Transportation Planning & Technology Journal 27, who state (at 28): It is not easy to define or quantify capacity. A simplest approximation is that the capacity of a single line is the total number of standard train paths that can be accommodated across a critical section in a given time period, where a standard train type is the more prevalent type to traverse the corridor. 630 But, as the authors point out, this "simple definition", which is not their preferred measure, assumes, among other things, a standard type of train. If different trains, or differently configured consists, run on the line, they say there can be no "standard train" and a more complex measurement of capacity is required. 631 The Tribunal is of opinion that the debate about how to measure capacity is of no practical consequence. If the measure is expressed in tonnage it can readily be converted to figures expressed in train paths and vice versa, although, perhaps, with some margin for imperfect correspondence. In any event, the standard train configuration which FMG's measure must assume will reflect the incumbent's allegedly inefficient practices as much as a tonnage measure would.

632 It was agreed that, for purposes of determining the spare capacity of a line, practical capacity was the most useful yardstick. Where the parties disagreed was on how practical capacity should be calculated. Their disagreement revolved around two interrelated issues: (1) the appropriate technique or methodology for determining practical capacity; (2) what variables that technique should take into account. 633 There are several techniques which can be used to calculate practical capacity. They may be classified as analytical methods, optimisation methods and simulation methods. Optimisation methods are best applied to timetabled services and can be put to one side. 634 Analytical methods usually calculate absolute capacity, and then determine practical capacity as a proportion of absolute capacity based on historical performance. One common approach is to calculate absolute capacity using analytical methods to identify the capacity of a bottleneck section of the line. Logically, the absolute capacity of the line can be no greater than that of its bottleneck section. Burdett & Kozan ("Techniques for Absolute Capacity Determination in Railways" (2006) 40(8) Transportation Research Part B: Methodological 616) prefer this model. They explain: The simplest definition and the most prevalent encountered in the literature is that the capacity of a single line is the total number of standard train paths that can be accommodated across a critical section in a given time period (i.e. the time period duration divided by the trains [sic] sectional running time), where a standard train is defined as the most prevalent type to traverse the corridor. This implies that a single bottleneck section limits the total flow of trains throughout the entire corridor and consequently the analysis is called a bottleneck approach. As an estimation of capacity, this approach is especially useful as an indicator of where additional infrastructure could best be placed. It is also useful because of its mathematical simplicity. 635 In their paper Burdett & Kozan also explain how the bottleneck approach can be improved by incorporating additional factors such as accommodating a variety of train mixes with differing speeds and making allowance for signals, crossing loops and dwell times. 636 Simulation methods imitate the operation of the line over a period of time. Simulation examines the dynamic behaviour of the system. It attempts to examine capacity in an environment as close as possible to reality. 637 A key difference between analytical and simulation methods is the manner in which they take variability into account. Simulation models are typically built with stochastic inputs. That is, random events such as breakdowns are programmed into simulation models using probability functions. The modelling is often described as "dynamic", in that it simulates the performance of the system over time, with all of its fluctuations and random variations. Because of this variation, modelling outputs are best represented as a distribution of predicted results, rather than a definitive or single point answer. In contrast, an analytical method is static, in that it takes a "snapshot" of a system. By definition, it does not take into account variability over time. It may make allowances for events which cause variability - for example, by assuming that unscheduled maintenance will reduce performance of the system by a given amount - but these allowances are single-point estimates usually based on averages. 638 There is no consensus as regards the preferred model. There are experts who prefer analytical models as they can provide an accurate and low-cost alternative to event-oriented simulation. For example, Kraft points out that simulation, by its nature, is a sampling technique and may contain hidden statistical sampling errors: Kraft, "Analytical Models for Rail Line Capacity Analysis" (1988) 29(1) Journal of Transportation Research Forum 153. According to him, those relying on simulation must use appropriate statistical techniques that involve careful specification of the true underlying probability distributions in order to generate values of the relevant the random variables so as to avoid misleading results. It is clear, however, that when conducted by experienced modellers, dynamic simulation is reliable enough to be regularly used to model large capital expenditure decisions. 639 The differences in approach to modelling capacity disclosed in the literature spills over into these proceedings. Each of FMG, BHPB and RTIO called witnesses who have modelled the capacity of the subject railway lines. Most of the witnesses favour dynamic modelling. For instance, Mr Hoare and Dr Dallimore from The Simulation Group Pty Ltd (TSG), a consulting firm specialising in simulation modelling, have modelled the capacity of BHPB and RTIO's railway lines using simulation modelling for BHPB for over 15 years and for RTIO for a shorter period. For these proceedings Mr Hoare and Dr Dallimore, for BHPB and RTIO respectively, conducted simulation modelling to determine the future capacity of the lines.

640 TSG has constructed a detailed simulation model of BHPB's Pilbara iron ore operations that captures all significant activities from mine to ship loading, including rail operations, port operations and shipping activities within Port Hedland harbour and the shipping channel. This model is used by BHPB to test the operational impacts of major investment alternatives and the effects of adjustments to operational protocols and procedures. 641 Mr Hoare (who is TSG's General Manager) explained that the model is based on discrete event simulation. Each physical item (eg truck, rake, reclaimer, ship) is modelled as a discrete entity, with its own unique defined set of properties or attributes (eg speed, material type, reliability, carrying capacity). These entities represent the operational activities that make up the processes being modelled. Each activity involves discrete periods of time and incurs delays that can be logically induced (eg ore bin empty, no rake) or randomly induced (eg breakdown, failures), all of which are dependent on the data and rule settings that are defined for that process. The random variables are assigned the probability distributions that are thought best to model their occurrence. The operation of the system as a whole is then simulated. The simulation is then repeated many times. The results of the simulations are then plotted, and a "line of best fit" is generated. 642 The gravamen of Mr Hoare's evidence is that to determine the capacity of the BHPB railway lines (ie, the Mt Newman line and the Goldsworthy line) it is necessary to take into account the whole of BHPB's mine, rail and port systems. This is because of the interdependence of the mine, rail and port operations, which operate in a "closed loop". Mr Hoare emphasised that there are many sources of variability which affect this integrated system, and that the rail sub-system plays a key role in allowing the system as a whole to respond to such variability. For example, there may be a need to quickly get ore to port from certain mines in order to meet unanticipated customer demand or a surge in ship arrivals, or to complete the blending of a stockpile to specification at port. There may be a need to divert trains to avoid queuing where equipment failures have occurred at a particular mine or port. According to Mr Hoare, the efficiency of the system as a whole is dependent on the ability of the rail sub-system to "flex" - ie to respond dynamically - to meet both supply and demand variability, taking account of "buffering capacity". Mr Hoare therefore considers it necessary to take into account "end effects", being variables off the line (generally at mine and port) which affect utilisation of the line. 643 In late 2007, Mr Hoare was asked to undertake simulation modelling of the Mt Newman line by including end effects. Mr Hoare was first asked to consider how the addition of 10mpta of third party ore would impact on the performance of the Mt Newman line. The configuration of the line and data settings used were, with a few exceptions, based on those used by Mr Hoare for an RGP4 feasibility study. The third party ore was assumed to enter the line near Mindy Mindy, and other assumptions about its operations were based on those provided by FMG at BHPB's request. Mr Hoare concluded that adding the third party ore would increase the variability of the performance of the BHPB closed loop, increase train travel times and reduce opportunities for maintenance. 644 Mr Hoare was then asked to determine the capacity of the Mt Newman line if (1) it were double tracked and (2) if it were triple tracked and BHPB introduced automatic trains. The latter scenario has not since occurred and it is not suggested that BHPB is planning to implement such measures. Thus, it is of limited relevance. For the double tracking scenario, Mr Hoare assumed track configurations and data settings based on his (then) understanding of RGP5 expansion plans and beyond. He tested the capacity of the double tracked line by assuming a given train configuration, with stepped increases in mine production and shipping demand, together with a variable number of train sets. He then plotted the results of this testing to produce a rail capability curve which indicated that the economic utilisation of the line would vary between 250-280mtpa, and that (what can be described as) the choke capacity of the line was in the vicinity of 290-305mtpa. 645 Mr Hoare produced a second report on 30 March 2009. For this report Mr Hoare was asked to base his reference case upon RGP5 settings which had recently been approved by the BHPB board. A critical difference from his first report was that he was asked to assume that, in order to allow for third party access, BHPB would operate as a scheduled operation rather than the current "run when ready" operation. He was also asked to make other assumptions, including that trains would always depart according to schedule (even if not fully loaded in time). On these assumptions, Mr Hoare concluded that introducing third party access would result in a 20% decrease in sustainable iron ore sales by BHPB. 646 The Tribunal makes the following observations about Mr Hoare's second report. First, most experts in this proceeding (including Mr Baunach, a rail modeller called by FMG) agreed that the lines would carry ore more efficiently on a run when ready basis rather than a timetabled basis. One expert called by FMG, Mr Clements, initially suggested otherwise but ultimately, we think, accepted that the run when ready approach adopted by the incumbents is an efficient means of operating its lines. Second, BHPB is not considering moving to timetabled operations and is highly unlikely to do so in future. Third, the introduction of a third party does not necessarily mean that the incumbent would need to adopt timetabling. The third party might be required to operate flexibly and take up available rail slots opportunistically. Even if the third party has scheduled times for entering the line, BHPB itself should be permitted to continue operating on a run when ready basis (just as it currently does notwithstanding other scheduled events such as scheduled rail maintenance). Fourth, without needing to express a final view, the Tribunal has grave doubts that, in the event a line were declared and an access dispute arose, the ACCC would have power to require the incumbent to fundamentally change its operating practice by moving to a timetabled operation. For those reasons, the Tribunal does not obtain much assistance from Mr Hoare's second report.

647 RTIO called Dr Dallimore. TSG's model for RTIO is not as detailed as that which exists for BHPB. The model does not attempt to incorporate every individual process in RTIO's production system; rather, boundaries of the model are drawn at the car dumpers at port and the mine load-outs. While the model takes into account some end effects such as variability due to loading and unloading and in mine processing plant production rates, it excludes other end effects associated with operations at port (eg stockyards, ship loaders and shipping channels), mine (eg pit or equipment constraints) and processing plants (eg process plant capacity). Dr Dallimore said that the excluded end effects are an important part of capacity analysis. Notwithstanding its limitations, the model is regularly relied upon by RTIO. 648 Dr Dallimore was requested to investigate the capacity of RTIO's rail network (combining both the Hamersley line and the Robe line), adopting the infrastructure that would exist at the port, mine and rail lines when RTIO completed its (then) proposed expansion to produce 320mtpa. His report dated 23 March 2009 listed his assumptions in detail, such as in relation to the allocation of ore shipping between RTIO's ports at Dampier and Cape Lambert. 649 Dr Dallimore modelled the maximum rail system capacity (his term for choke capacity) by following a process similar to that adopted by Mr Hoare - namely, to hold the rail configuration fixed and progressively introduce additional volume to the system until it reached a point where the target volume could not be achieved, no matter how many trains were run. Dr Dallimore concluded that the maximum rail system capacity was 350mtpa. He stressed that the economically viable rail system capacity (which he defined as the point at which money is better spent on extracting additional capacity by adding track rather than extra rolling stock) would be lower. 650 Dr Dallimore then considered the effect of adding to the system a new hypothetical mine located near Rosella junction. He assumed that the new mine's additional tonnes were dumped solely at a new car dumper based at Cape Lambert. The effect of this assumption was that the new mine's ore would be railed along double tracked sections for almost all of its trip. Dr Dallimore also assumed that the new mine's loading and unloading facilities, and the new car dumper's facilities, would have more than enough capacity to handle the new mine's tonnes, thereby eliminating end effects which otherwise might have been created. The results of his modelling was that, with this extra hypothetical mine, the maximum theoretical capacity of the system could increase by 10mtpa, but not 20mtpa or more. On this basis, Dr Dallimore concluded that the Emu-Rosella section of the line would be the most significant bottleneck on the RTIO system in 2014/15.

651 Mr Baunach is an experienced computer simulation modeller and has undertaken modelling work for railroads. 652 Mr Baunach was asked by FMG to calculate the "absolute maximum capacity" of (1) the Yandi-Goldsworthy section of the Mt Newman track; and (2) various sections on RTIO's rail network. In contrast to TSG's modelling, Mr Baunach assessed capacity by looking particularly at a section of the line: the Yandi-Goldsworthy Junction section. The logic of modelling the Yandi-Goldsworthy Junction section was, as Mr Baunach explained, that this section was the bottleneck or "choke point" section through which all traffic on the line would travel. He said the capacity of that choke point section would effectively define the capacity of the line as a whole. Mr Baunach appears to have applied a similar logic in respect of RTIO lines. Dr Dallimore suggested that the true bottlenecks on the line might well be junctions on the line, which Mr Baunach did not specifically model. 653 Viewing each relevant section in isolation, Mr Baunach conducted dynamic modelling taking into account some, but not all, of the sources of variability incorporated in TSG's models. Critically, Mr Baunach was instructed to ignore the impact of "end effects", ie the reduction in the efficiency of the rail system due to delays caused by loading and unloading operations. He also ignored variation in trains waiting to get on the relevant section by assuming that there was an infinite stream of trains waiting to enter at each end of the section. On the other hand, Mr Baunach took into account interruptions to normal operations by reason of the following events: rail grinding, rail defect repairs, major planned maintenance, random unplanned maintenance and train breakdowns (factors which might be thought of as "endogenous" to the line). 654 Initially Mr Baunach modelled the Yandi-Goldsworthy Junction section of the Mt Newman line assuming track configuration and data settings provided by his instructors. The assumed track configuration appears to have been based on the actual track configuration existing at that time. 655 Later Mr Baunach remodelled the capacity of the Yandi-Goldsworthy Junction section under three new scenarios: (1) after completion of upgrades to the track under RGP5, which would involve double tracking of the entire Yandi-Goldsworthy Junction section; (2) post-RGP5 but with the use of automatic trains; and (3) if the section were triple tracked and automatic trains were used. In light of criticisms raised by BHPB about the use of FMG assumptions in his first report, Mr Baunach adopted many of the data settings used by Mr Hoare in his first report, with two significant exceptions. First, his instructions remained to ignore end effects. Second, he assumed three rake trains for all of the new scenarios being modelled, whereas Mr Hoare had assumed smaller trains, capable of carrying less ore, for his modelling of RGP5 and triplication of the track. 656 The result of Mr Baunach's modelling contained in his report of 11 September 2008 was that with the duplication of the Yandi-Goldsworthy Junction track and without the use of automatic trains, the track was able to support 39 trains on average in each direction per day (equivalent to 516mtpa on his assumed three-rake train configuration). With the use of automatic trains the capacity of the track increased to 62 trains in each direction per day (or 820mtpa) on average. Mr Baunach did not model the triplication scenario specifically but estimated, as an order of magnitude estimate, that the capacity of the line would be 1,098mtpa. 657 Mr Baunach explained the differences between the results he obtained and those obtained by Mr Hoare. He said that the differences were due to: (1) the modelling of end effects by Mr Hoare; and (2) Mr Hoare's assumption of smaller trains, resulting in a significant reduction in tonnage of iron ore moved by each train trip. 658 Mr Baunach was later asked to model RTIO's rail operations. In particular, he was asked to model the maximum theoretical capacity (as defined by Abril et al.) of the following sections of the Hamersley and Robe lines: (1) Dampier to Emu; (2) Emu to Rosella; (3) Rosella to Wombat junction; (4) Wombat junction to Paraburdoo; (5) Wombat junction to Tom Price; (6) Rosella to Juna Downs junction; (7) Junior Downs junction to Yandicoogina; (8) Cape Lambert to Western Creek; and (9) Western Creek to Mesa J. That modelling was to be undertaken under three scenarios: (1) the then current configuration of the Hamersley and Robe lines; (2) the configuration of the Hamersley and Robe lines adopting the infrastructure enhancements proposed by RTIO for its expansion to 320mtpa; and (3) where the section of the Hamersley line between Emu and Rosella is triplicated. 659 The modelling results appear in the following tables which are taken from Mr Baunach's report of June 2009: Modelling Results for Scenario 1 (Current) Track Current Rio Tonnage (mt/y) Modelled Capacity Utilised Capacity (%) Av. Trains/ Tonnage (mt/y)

661 Although FMG called Mr Baunach, a dynamic modeller whose evidence was that dynamic modelling gives a better indication of capacity than static analysis, FMG preferred to rely on static modelling to show there was spare capacity in the lines. To this end, it first relied on a report prepared by Evans and Peck (E&P) which had been submitted to the Western Australian Department of Industry and Resources in August 2004. That report estimated the carrying capacity of the Mt Newman line at more than 400mtpa if dual tracked. The estimate was made on the assumption that "only the rail system's 'trunk line' capacity was under consideration". End effects were not considered, although other sources of variability such as maintenance and track shut-downs were. 662 In December 2005 E&P were asked by BHPB to review their assessment of the carrying capacity of the Mt Newman line. E&P defined its task as being to determine "the maximum achievable level of production for [the] system". It described the approach it would adopt as "primarily based on static (ie spreadsheet-based) models". The report referred to a static model developed by BHPB, on which E&P's static model was based. Unlike in its first report, E&P took into account end effects (which it defined as "random delays arising from loading and unloading operations") based on 2004-05 data. As a result of its modelling E&P estimated that BHPB's dual track production capacity was 206mtpa. The Mt Newman dual track production capacity was between 234 and 247mtpa (taking into account the impact of end effects) but overall production was constrained by the "weakest link" in the system, the port. 663 FMG contends that BHPB "undermined" the revised assessment by instructing E&P to adopt an incorrect approach to the assessment of capacity (the maximum achievable level of production) which, by definition, required E&P to take into account end effects. 664 The attack on the revised assessment called in aid a 2009 BRS report, produced at the request of [c-i-c], which considered the use of rail transport by [c-i-c] for its various iron ore projects. The authors of the report had available to them the E&P report. They noted that: The factors described by Evans and Peck as "end effects" are not relevant in assessing the practical capacity of a rail network. End effects relate to terminal issues and delays around train loading and unloading etc, and are a measure of how the train operator manages trains within the terminal and should not be part of the assessment of a rail network's capacity. In some circumstances, trains queue on the network while waiting to enter a terminal and have the potential to impact on the railway's capacity. Observations at Port Hedland indicate that trains do queue at Bing and just outside the Nelson Point terminal. This has the effect of limiting the ability of trains travelling over the railway, but, as noted before, relates to the terminal capacity rather than the network capacity. Providing additional staging roads either within terminals or at some location outside the terminal can eliminate the influence of terminal constraints on train movements on a network. 665 BRS performed its own static analysis which considered the theoretical capacity of the Mt Newman line using several different assumed figures for train headway. The results suggested that the theoretical capacity of the line was sensitive to changes in headway. BRS suggested that the practical capacity of the line would likely be 70-80% of the theoretical capacity on a single track, and could be expected to be in the range of 80-90% of theoretical capacity in a double track scenario. 666 Finally, and importantly, FMG relies on the static modelling undertaken by the incumbents. Apart from the BHPB spreadsheet model referred to in the E&P report, it relies heavily upon a "high level rail model" prepared for the purposes of an RTIO order of magnitude study. That model suggested that RTIO would only utilise around 50% of the theoretical capacity of the track. In cross-examination, Mr Ranson accepted that the high level model provided a reasonably accurate measure of the RTIO lines' utilisation and capacity today. That said, Mr Ranson stressed that he was not a rail modelling expert, and that more detailed dynamic modelling would be used once a planning process moved beyond the order of magnitude stage. 667 In the Tribunal's view static modelling has significant limitations which affect its value as a tool to determine spare capacity. First, static modelling may be vulnerable to inaccuracy. In their 2005 report, E&P said that "static models can lead to very inaccurate estimates if a consistent set of input parameters is not used. A static model is by necessity a simplification of the real system and can ignore the "flow-on" effects that changing one parameter can have on another (for example, reducing the spacing between trains will always increase production capacity in the static model but may actually decrease capacity in a real system if the loading/unloading facilities are unable to keep up with the extra trains)." Second, and perhaps as a result of this, the modelling experts agreed that dynamic modelling is more reliable than static modelling as a basis for major capital expenditure decisions. BHPB and RTIO appear only to use static modelling for preliminary planning purposes. Once planning is more advanced, dynamic modelling is used exclusively. This is consistent with Mr Baunach's view. In his experience, while railway owners do some degree of static modelling, ultimately they base their decisions on dynamic modelling. Indeed, FMG itself is developing a dynamic model for the Chichester line. 668 None of this is to suggest that static modelling does not have its uses. It is just that for purposes of determining the spare capacity that is likely to exist in a line, static modelling will not provide the most reliable information.

669 To better understand (as well as to test) what many witnesses, including Mr Hoare and Dr Dallimore, said about the need for the incumbents to maintain flexibility, the Tribunal requested BHPB and RTIO provide data in respect of their train operations on randomly selected days or weeks over the period from late October 2008 to September 2009. The information requested related to the number and time of actual train departures, and the time and place of train arrivals. The object was to see how train movements differed from what had been scheduled. The "schedule" (something of a misnomer in a run when ready system) was in BHPB's case based on its rolling 72 hour rail schedule, and for RTIO the schedule for the upcoming 24 hour period set at 9 am on the relevant day. 670 The information which was produced shows that deviations from the schedule are commonplace in both incumbents' operations. Of the sampled days, the number of actual train departures was often less than those scheduled. On a few days a relatively high number of actual departures eventuated, but the number of actual departures on other days was significantly lower. Using the information provided by RTIO as the example, FMG calculated the average "gap" (in minutes) between train departures (on both a scheduled and actual basis). According to those calculations, on all of the sampled days the average gap between actual train departures was at least 58 minutes and there was at least one gap between actual train departures of more than two hours. 671 Based on this analysis, FMG contended that the incumbents' actual utilisation of the line is in practice often less than its scheduled utilisation of the line. Hence it should be concluded that the line has spare capacity. 672 The response by the incumbents was two-fold. First, they adduced evidence to the effect that, on sampled days where the incumbents ran fewer trains than scheduled, it did not follow that a third party would have been able to "take up the slack". Relying on train graphs showing the progress of trains on certain of the sampled days, the incumbents showed that disturbances on the lines which prevented the incumbent from running its scheduled trains would equally have prevented third parties from running trains, or there would have been opportunistic maintenance carried out. Second, the incumbents argued that even where a third party might have used a spare rail slot, the existence of the slot was only known with hindsight. In practice, they said, it would be difficult to identify those opportunities in advance, or at least sufficiently in advance to realistically allow a third party time to ready its train to take up the opportunity. 673 For reasons which we will explain when dealing with criterion (f), we think that protocols could be developed which would allow the third party sufficient notice to take up rail slots generally. But, as regards rail slots which suddenly become available due to an unexpected event such as a car dumper failure at port, it is not likely that such slots would be capable of being exploited by a third party. Further, it is reasonable to assume that the level of unexpected incidents may vary (and indeed decrease) from time to time. In the case of the Mt Newman line, for example, the train graph analysis was significantly affected by the fact that (1) critical sections of the line are currently single tracked and (2) BHBP is using a certain train configuration which has experienced some technical problems. Under future expansion plans, the line will become double tracked and different train configurations will be used. We think, therefore, it is difficult to reach firm conclusions about the likely spare capacity on a line based on train graph analysis. A more rigorous approach is required.

674 An alternative basis for assessing the existence of capacity is based upon the way in which all lines, with the exception of the eastern section of the Goldsworthy line, are to be expanded. FMG contends that the manner in which the expansions are to be carried out means an opportunity for access will arise from time to time. 675 This argument is based in part on the fact that rail investment can sometimes be "lumpy", in the sense that the extra capacity added by an expansion may necessarily be greater than the capacity which is needed in the short-term. A good example of "lumpiness" is dual-tracking a line. With a single track line, additional capacity may be achieved by relatively inexpensive incremental upgrades such as passing loops and sidings. At some critical point, however, it will become necessary to double track the line. The capacity added by dual-tracking may be more than is immediately required for the incumbent's use. This might occur when the expansion of the line is part of an overall expansion of the whole operation (mine, railway and port) and the line expansion is completed first. Thus, if the incumbent's use of rail is constrained because expansions at the mine or port are lagging behind, there will be spare capacity on the line. 676 The extent of these "windows", in terms of their duration and the volume they can accommodate, is necessarily uncertain. The availability of capacity windows depends on future expansions plans which are speculative and constantly changing. Spare capacity also depends on the sequence in which capacity expansions are undertaken (ie whether port and mine expansions keep in step with rail expansions and vice versa). 677 In any event the Tribunal believes there is not much in the windows of opportunity point. Any windows of spare capacity will likely be limited in number and duration. In part this is because the incumbents' expansion plans are usually designed to be fit for purpose and to avoid investing in capital works earlier than is needed. 678 There must also be some doubt about whether short-term access would be viable for a third party. Relatively large up-front capital costs are required to establish rail operations which may only be paid off over longer periods of use. Mr Tapp, for example, accepted that if FMG were to seek access to carry iron ore from the Solomon area along the Hamersley line, a short-term access arrangement in the order of four years or so would not be commercially practical. Notwithstanding this, FMG's counsel suggested that short-term rail access might be viable for two reasons. First, it appears that some of the costs of establishing rail operations can be ameliorated. For example, the BRS Report suggested that consists could be leased. Second, it was put that a third party might use a short-term window to "ramp up" production to enable it to develop its own rail alternative, transferring rail infrastructure onto its new line at the appropriate time. We accept that this is a theoretical possibility. However, even in that scenario, the access seeker might incur significant costs in accessing the incumbent's line that could not be recovered through subsequent use of its own facility (for example, the cost of building spurs connecting with the incumbent's line). 679 Having regard to what was said by Mr Tapp, and the Tribunal's own assessment of the position, the Tribunal considers an approach based on short-term windows of capacity as too speculative to be a basis for proper decision-making. 680 Nor is the Tribunal prepared to base its decision on the suggestion there may be spare capacity if the proposed BHPB-RTIO joint venture in the Pilbara does not proceed. FMG relies on the fact that the calculations of the synergies for the joint venture which have been undertaken to date do not appear to include costs of expanding capacity on the incumbents' rail systems. This shows, so the argument goes, that the incumbents assume that their rail systems have sufficient capacity without any further expansion to handle extra traffic generated by the joint venture. Hence, if the joint venture does not proceed, there will be spare capacity on the lines. 681 There are a number of problems with this approach. No one knows whether a joint venture proposal will proceed. Detailed joint planning regarding capacity has not been undertaken to date for regulatory reasons. If the joint venture does not proceed, rail expansion plans may well change. It is, we think, far too early to reach any meaningful conclusion about capacity based upon a joint venture which may or may not proceed.

682 A critical difference between the parties is the extent to which end effects should be taken into account in determining the spare capacity on the line. According to FMG, the sources of variability affecting utilisation of capacity should be divided into two groups: · factors endogenous to the railway line - these are the factors which constrain the ability to run trains on the railway which are imposed by the railway itself (eg layout and infrastructure, failures and disruptions, practical operating speed, time lost due to weather, expansions and upgrades and maintenance); and · factors exogenous to the railway line - these are factors which constrain the operator's ability to run trains on the railway which are not imposed by the railway itself (eg availability of rolling stock, availability of railroad personnel and management of railroad operating conditions and end effects. 683 As described above, the modelling undertaken by TSG takes into account end effects, whereas the static modelling evidence and the modelling undertaken by Mr Baunach does not take into account end effects. 684 FMG contends that exogenous factors (and in particular end effects), while an important constraint on the incumbent's ability to use the railway line, are not relevant to an assessment of the practical capacity of the line. The rationale is that third party trains will not be using substantial parts of the railway lines near mine and port, and will never use the incumbent's loading, unloading and terminal facilities at mine or port. Hence loading and unloading and terminal facilities do not constrain (and therefore are not relevant to a calculation of) the number of trains that can be operated on the line by all users on a practical basis. In contrast, the incumbents both argue that a railway must be viewed as part of an integrated system and end effects cannot be ignored. 685 The Tribunal considers that to most accurately answer the legal query that it faces, end effects cannot be ignored. The Tribunal must determine the extent to which the incumbents' reasonably anticipated requirements would be affected by third party usage. The approach to the issue which FMG advocates was as follows: · the practical capacity of the line should be assessed ignoring the end effects of both the incumbent and any third party users; · the capacity utilised by the incumbent (ie to meet its reasonably anticipated requirements) should then be assessed based on the incumbent's current actual use or future anticipated use. This use is use in the absence of any third party; and · the difference between practical capacity and the capacity utilised will be available for third parties to use. 686 The problem with this approach is that if a third party were introduced the incumbent might need to use more train slots than it previously did in order to meet its throughput requirements. What was once a "free slot" without third parties may no longer be free. 687 The Tribunal accepts that the ability of the incumbents to meet throughput requirements depends heavily on the flexibility of the rail system to respond to variability. For example, there is ample evidence which shows that the incumbent is often required to reorganise train operations to respond to unanticipated events. The introduction of a third party, no matter how flexibly it might be willing to operate, could reduce operational flexibility for the incumbent. It would certainly introduce additional variability into the system. 688 This is not to suggest that a third party can never be accommodated. It may be shown that, on careful assessment, a third party's impact on the incumbent's operations is minimal. But whether this is so or not requires the combined use of the line by the third party and incumbent to be dynamically modelled. 689 The Tribunal's view is certainly shared by Mr Hoare, who suggested that to determine accurately the maximum throughput of a common section of the main line rail, it would be necessary to model a combined, fully integrated scenario which took account of both BHPB's and each third party's end effects and requirements for optimising each party's output. Dr Dallimore gave like evidence. So also did Mr Baunach. This is what he said at the modellers' conference: Q: I am talking about working out how it is going to work in a real world of real operations. At the moment, maximum capability might tell me something but it might not tell me enough. A: Yes. Q: Or think of it like this: assume that one rule which Parliament has imposed, … is, the incumbent cannot be adversely affected in its operations. So, the question becomes, can we run a - can we use some slots, if they are available, in a manner and by which the incumbent's operations are not affected? Because Parliament says they are not allowed to be affected. Now, the question is, how do I go about testing that? A: Then you need to model. Q: I know. And not in a way that anybody has modelled so far. I know ‑ ‑ ‑ Q: Just coming in there, do you believe you could model it if we gave that as one of the critical inputs? A: I believe it could be modelled.

690 In light of all the evidence, the Tribunal formed the view the best evidence of capacity was that based on dynamic modelling, taking into account end effects. But, the Tribunal had reservations about TSG's evidence on this issue. 691 As to Mr Hoare's work, the most obvious problem (not of Mr Hoare's making we should add) was that by the time of the hearing BHPB had in place a new plan (RGP6) to expand the capacity of its mine, rail and port system that would produce an output of 240mtpa. Mr Hoare's modelling was based on soon-to-be-outdated infrastructure. 692 A second potential problem was that it was unclear whether Mr Hoare's modelling took into account the fact that an access seeker would have its own mine and port facilities. As Mr Hoare explained, his analysis of the choke capacity of the line under RGP5 assumed BHPB as the sole user of the line. To test the choke capacity of the line he said in his first report that he had modelled stepped increases in "mine production and shipping demand". Discussing this modelling exercise in his second report, he also said that he had included "extra ports". It is not entirely clear exactly what new infrastructure Mr Hoare assumed as part of this exercise. The practical importance of this is the capacity of the line could be affected by whether the additional ore was to be transported using the incumbent's existing infrastructure or new infrastructure, and by where any new infrastructure was assumed to be located. 693 Dr Dallimore's modelling also had aspects that were of concern. First, as was the case with BHPB, RTIO was no longer planning to increase its output to 320mtpa, as Dr Dallimore's model assumed, but to 330mtpa, and with quite different infrastructure. Second, Dr Dallimore had modelled RTIO's two railway lines as a single system. The Tribunal, however, is required to give separate consideration to the declaration of each railway line. Third, Dr Dallimore assumed the existence of a new mine (he located it near Rosella) but then proceeded to model the system on the basis that ore from that mine would be taken to a new dumper at Cape Lambert. But there is no capacity for a new entrant at Cape Lambert. 694 Another concern which the Tribunal had in respect of Mr Hoare's and Dr Dallimore's reports is that both assumed that all trains in the system would have equal priority. It had been suggested by FMG, however, that in order to minimise its impact on the incumbent, a third party's trains could operate flexibly, taking lower priority than the incumbent's trains and waiting for free slots. If so, the third party might be able to run more trains without displacing the incumbent's output. 695 As a result of its concerns, the Tribunal invited Mr Hoare and Dr Dallimore to conduct further modelling based on an updated and more appropriate set of instructions. 696 In adopting this course, the Tribunal was mindful of its role at the declaration application stage. When simulating the incumbent's operations, the impact of the third party can differ significantly depending on a number of variables such as the volume of iron ore being moved, the number of proposed third party trains and the points at which the third party enters and exits the line. At the declaration stage, it is not possible to predict the precise nature of the access which might be sought by a third party or several third parties in the event of a declaration. Hence it is not possible to be certain about the effects on the incumbent of third party access. 697 Nonetheless, the Tribunal was of the view that it would derive important information from additional dynamic modelling. First, a number of different scenarios could be modelled which, while not covering the entire field of possibilities, could form a robust basis for arriving at some general conclusions about capacity. Second, there was utility at the access declaration stage to model "optimistic scenarios" of third party access - ie, scenarios which, while realistic, are likely to have the least impact on the incumbent. If, even in those scenarios, the third party could not use the line without significantly displacing the incumbent's use, this would at least suggest that the line, without expansion, could not accommodate a third party. 698 Accordingly, the Tribunal, in consultation with the parties, developed a set of instructions for dynamically modelling a series of "optimistic" access scenarios. The modellers were asked to assume a track configuration which is likely to exist in or around 2014 - namely, BHPB's track configuration under RGP6 and RTIO's track configuration under its "330 plan". The modellers were required, to the extent possible, to use the latest model inputs they had developed to assist their clients' internal planning decisions (ie for purposes other than the proceedings). Where those inputs needed to be amended or supplemented for the Tribunal's modelling exercise, the modellers were instructed to record the changes. 699 Mr Hoare was asked to assume that the third party would enter the line near Mindy Mindy (at the point identified in FMG's application for declaration of the Mt Newman line) and would exit at a point 25km south of the Goldsworthy junction. Because of the "all points" nature of the applications seeking access to RTIO lines, Dr Dallimore was asked to model four different sets of entry (ie near mine) and exit (ie near port) points. Two scenarios involved entry and exit on the Hamersley line, the first being from Pelican to Dingo (the Dingo case) and the other being from Pelican to Emu (the Emu case). Another scenario involved entry at Pannawonica and exit at Harding, both on the Robe line (the Robe Line case). The final scenario involved entry on the Hamersley line at Pelican and exit off the Robe line at Harding (the Harding case). In all cases, the third party train would initially have lower priority than the incumbent's train when waiting to enter the track. If it had not entered the track after 12 hours, the third party train was to be accorded equal priority with the incumbent's trains. Once on the line, the third party's train was to be treated in the same way as the incumbent's trains. 700 The nature of the third party operation which the modellers were instructed to assume reflected, in large part, the parties' suggestions. There were a number of suggested assumptions which were contentious. For example, FMG suggested that the third party would build a "flyover" (analogous to a road overpass) to minimise the interference of its trains entering and exiting the track. The incumbents claimed this would be prohibitively expensive. To deal with contentious assumptions, the modellers were asked to model a base case involving one set of assumptions, and then model alternative scenarios to see whether the results were sensitive to a change in each of the contentious assumptions. 701 The objectives of the modelling, which were amended after correspondence with the modellers, were as follows (assuming a fixed track configuration): · to identify the choke capability of the rail line from the relevant entry and exit points; · to assess the capability of each rail line to haul additional throughput for a third party without reducing the incumbent's anticipated throughput, in particular an additional 3mtpa, 5mtpa, 10mtpa, 20mtpa and 50mtpa; and · to assess, for each relevant level of additional third party throughput, (1) the minimum number of consists which the third party would need to haul the relevant throughput and (2) how many extra consists, if any, the incumbent would require to enable it to haul its anticipated throughput (as compared to where no third party is present). 702 Mr Hoare produced his final report on 17 December 2009. In the time available to undertake the modelling and prepare his report Mr Hoare was not able to carry out all the requested sensitivity analysis. Nonetheless, he was able to test the sensitivity of the results to the flyover assumption, and his key findings are important. The results are summarised in the following table: Third Party ore (mtpa) Displaced BHPB ore (mtpa) Third party consists required Additional BHPB consists required 3 Not provided (but logically ≤ 1) Not provided, (but logically 1) Not provided (but ≤ 1)* 5 ≤1 1 Not provided (but ≤ 1)* 10 ≈ 1 2 Not provided (but ≤ 1)* 20 ≈ 3 3 2 50 = 12 (without flyover) 6 (with flyover); 7 (without flyover) 5 (with flyover))

704 The results for an assumed third party tonnage of 50mtpa were more variable. Some of the more notable features were as follows: · The third party generally required 8 to 11 consists to move its ore; · The displacement of the incumbent's ore was generally around 12mtpa for the Dingo case, around 4 to 10mtpa for the Harding case and Emu case, and around 2mtpa for the Robe Line case; · In the majority of sensitivity scenarios, the additional RTIO consists required were 9 for the Dingo case, 3 to 4 for the Harding case and the Emu case, and 1 to 2 for the Robe Line case; · In some scenarios, it was not possible to remedy the displacement by adding consists to RTIO's fleet. This was a particularly important issue in the Robe Line case, because that line must cater not only for trains from RTIO's pooled fleet but also a special fleet servicing its Mesa A and Mesa J mines. In several sensitivity testing scenarios, it was not possible to remedy displaced throughput for this special fleet by simply adding consists, due to infrastructure limitations associated with the fleet. 705 It is important to appreciate that the work performed by Mr Hoare and Dr Dallimore at the Tribunal's request has limitations. The modellers themselves qualified their findings in several respects. One limitation is that there is a margin for error built into the model which may be significant at lower volumes. Mr Hoare expressed concern that the assumptions he was asked to make regarding third party use were simplifying complex processes and potentially understating variability, and Dr Dallimore made similar comments in regards to the RTIO model he used. Both point out that the models do not take into account the economic utilisation - as opposed to choke capacity - of the lines. 706 The Tribunal accepts that these qualifications are justified and to some extent reflect limitations of simulation modelling in general. Still, the Tribunal gives significant probative weight to the work performed by these two modellers at the Tribunal's request, because it is the best, albeit imperfect, evidence tackling an issue which is inherently difficult to resolve.

Expansion Power 715 As the modelling evidence shows, expansions to the lines might be necessary to carry larger volumes of third party demand. Suppose the incumbent is unwilling to agree to an expansion. The question which arises is whether Part IIIA provides for a power to compel the expansion? 716 The only possible source of the power is s 44V(2) where the arbitral powers of the ACCC are found. The subsection provides: A determination [by the ACCC] may deal with any matter relating to access by the third party to the service, including matters that were not the basis for notification of the dispute. By way of example, the determination may: (a) require the provider to provide access to the service by the third party; (b) require the third party to accept, and pay for, access to the service; (c) specify the terms and conditions of the third party's access to the service; (d) require the provider to extend the facility; (da) require the provider to permit interconnection to the facility by the third party; (e) specify the extent to which the determination overrides an earlier determination relating to access to the service by the third party. 717 The NCC argues that the power under paragraph (d) to order an "extension" of a facility encompasses a power to order an expansion (or, put another way, an extension is synonymous with an expansion). FMG says the plenary power to deal with "any matter relating to access by the third party to the service" would enable the ACCC to order an expansion, even if it were not considered an extension. 718 Before dealing with these arguments, it is necessary to make reference to other provisions that deal with "extensions". Subsection 44W(1)(d) provides that the ACCC cannot make a determination which results in a third party becoming the owner (or one of the owners) of any part of the facility, or of extensions of the facility, without the consent of the provider. Subsection 44W(1)(e) prevents a determination requiring the access provider to bear some or all of the costs of extending the facility or maintaining extensions of the facility. The combined effect of these provisions is that where an the owner will not voluntarily carry out an extension, the ACCC can only direct an extension if (1) the third party is willing to pay for the work and the maintenance; and (2) the owner takes full title to the extension. On the other hand, s 44X(1)(e) provides that in making a determination, the ACCC must take into account the value to the owner of an extension whose cost is borne by someone else. This suggests the ACCC must take into account the value to the owner of an extension when assessing access charges. 719 As regards the first argument (the meaning of "extension"), not surprisingly the word is not defined. Not being a term of art it must be given its ordinary meaning. Dictionaries give several meanings. One is stretching out or lengthening in a particular (or various) direction. This meaning can easily be applied to facilities such as railway lines, roads, electricity transmission grids and gas reticulation systems. It contemplates work which would enable a third party to gain access to the facility. But the work need not result in an increase in the capacity of the facility. Another meaning is to increase the size, or enlarge the scope of a thing. On this meaning an "extension" would include works that would increase the capacity of a facility. 720 In a submission made to the Productivity Commission on its review of Part IIIA, the NCC distinguished between three meanings of extension:

geographic extension (extending the geographical range of the facility);
interconnection (the ability of the access seeker, which constructs the extension itself, to connect to an existing facility); and
capacity expansion (increasing the capacity of the existing facility). 721 The Productivity Commission recommended that Part IIIA be clarified to ensure that the ACCC has a power to order an interconnection. That recommendation was adopted: see s 44V(2)(da). As regards "capacity expansions", the Commission was equivocal. Ultimately it did not recommend the enactment of a power to order capacity expansions, but noted that "the ACCC's power to direct a facility owner to expand a facility is implicit rather than explicit". 722 The meaning to be given to "extension" can be determined, we think, by deciding which of the available meanings best gives effect to the objects of Part IIIA. 723 Those objects (found in s 44AA) support a power to order capacity expansions. The objective of promoting the efficient investment in infrastructure must be understood in light of Part IIIA's concern with facilities which exhibit natural monopoly characteristics. That is, the only facilities to which the access regime may be applied are those whose output can, or when expanded will, satisfy market demand more efficiently than if provided by more than one facility. It would be odd to hold that a facility displays natural monopoly characteristics and, at the same time, deny to the ACCC the power to compel the owner to increase the facility's capacity so that it can in fact satisfy that demand. 724 Second, it is clear that, whatever "extension" means, it is more than merely permitting a third party to connect to the facility, as the incumbents argue. The problem with the incumbents' argument is that there is a separate power for ordering interconnections. This suggests that extensions, which must be different in nature, could potentially be more extensive than mere interconnection. 725 Third, if extend means merely to elongate a facility, then, as a practical matter, some perverse or arbitrary results might arise. Assume there is a port where an extra berth could be built, or an airport which could handle extra traffic if a runway were extended. In those examples, it would be possible to increase the capacity of the facility through elongating the facility. On the other hand, it is not difficult to imagine other facilities where capacity could not be increased through elongation. It seems to be somewhat arbitrary for the ACCC to have a power which, when exercised, would in some cases result in capacity expansions but not in others. In any event, the distinction between an extension and an expansion in the rail context may be blurred. Strictly speaking, building a siding extending from a single track line is an extension, but will usually have the effect of expanding capacity. 726 Fourth, we accept that a power to order a capacity expansion is potentially a dramatic power. Consider a major railway expansion which could easily cost hundreds of millions of dollars or more, and take years to complete. Once a direction is given and the works are commenced many things could, and often will, go wrong. If the parties do not resolve those matters, the ACCC will be called upon to arbitrate the outstanding dispute. The existence of an expansion power would result in the ACCC effectively becoming a party to the access arrangement and having a significant project management role. Such a power would be a serious interference with a firm's right to go about its affairs, and undertake its own style of operation, as it thinks fit. 727 The other side of the coin, however, is that (1) the power to order capacity expansions is discretionary and (2) there will be many circumstances where the exercise of the power will not have a significant adverse effect. Here we are dealing with railways, but for other facilities, capacity expansions may be more straightforward. 728 Fifth, we think that the reference in the preamble to the ACCC making orders in respect of "any matters relating to access" indicates a legislative intention that the ACCC should have a broad, plenary power to deal with access disputes. A narrow reading of "extend" would be inconsistent with this. 729 Sixth, there is the decision of the Privy Council to which the NCC made reference, Municipal Council of Shanghai v McMurray [1900] AC 206. The case concerned a regulation which dealt with the "extension of the lines of roads". The true construction of the word "extension" was in issue. In the course of giving the advice, Lord Hobhouse said (at 210): The word 'extension' is very commonly used in connection with railways and tramways, both in legal documents and by people at large. When an extension of the Great Western Railway is spoken of no one supposed that the thing meant is merely to prolong the existing line or to increase its breadth for laying down more rails. Branches are contemplated as well as the original main line when extensions are spoken of. That is certainly a common use of language, nor can their Lordships see that in point of etymology or philology it is incorrect. 730 It is for these reasons that the Tribunal is of opinion that a capacity expansion falls within the extension power. It follows from this that the limitations regarding extensions (eg in s 44W) will apply. In contrast, on FMG's preferred approach, the limitations would not apply, except as matter of discretion. We do not think s 44V was intended to interfere with the rights of an owner to that extent. 731 Finally, we should note an argument put by RTIO regarding s 44V(2). RTIO points out that the power under s 44V(2) must relate to access to "the service". Expanding the capacity of the facility, RTIO argues, would change the nature of the service and must be beyond power, because the ACCC can only make determination regarding the service as it exists at the time of the dispute. We disagree with this argument for two reasons. 732 First, s 44V(2) expressly envisages an extension of the facility. Even if the power is confined to directing an elongation, that might just as easily alter the nature of the service. 733 Second, and more fundamentally, we think that RTIO's argument ignores what the Tribunal sees as a fundamental distinction between, on the one hand, the service which is being provided by the facility, and, on the other hand, the facility which provides the service. It will often be the case that the physical characteristics of the facility itself may change from time to time. Given the possibility that a declaration may continue for 20 years, it is very likely that the facility will change during the life of a declaration. It does not follow that a change in the characteristics of the facility results in a change in the nature of the service. On the contrary, the service remains constant - in this case, use of the facility. If one were to accept RTIO's argument, it would be a simple thing to undermine the regime. 734 We note, in passing, that the discussion regarding the "essential nature of the service" by the Full Court in Rio Tinto Ltd v Australian Competition Tribunal (2007) 246 ALR 1, 13-14, concerns a different point. There the issue was whether the service FMG actually sought was the same as the service described in its application to the NCC.

13.2.1 Haulage by BHPB and RTIO 737 Attempts to date by junior miners and FMG to have BHPB and RTIO haul iron ore have been futile. This is despite the fact that, under the relevant State agreements pursuant to which BHPB's and RTIO's lines were built, they have an obligation to carry the freight of third parties provided this could be done without unduly prejudicing or interfering with their operations. What went wrong? 738 The Hon AF Griffth, the Minister for Mines in the WA Brand government, said on the Second Reading in the Upper House of the Bill to approve the first of the State agreements, the Mt Goldsworthy agreement, that he was certain it was a "fair and equitable agreement": Western Australia, Parliamentary Debates, Legislative Council, 28 August 1962, Vol 161, p 670. During the debate (most of which was taken up with the question whether, by virtue of the Public Works Act, permission to construct the Mt Goldsworthy railway required a special Act), the Hon EM Heenan said that: "I have read through the agreement with a good deal of interest and there again I think that whoever was responsible for the drafting of this difficult measure is deserving of congratulations. I assume, of course, it was the chief parliamentary draftsman. The Bill strikes me as being a splendid example of his craft, and I think he is worthy of our congratulations": 29 August 1962, Vol 161, pp 726-727. 739 The Member of the Pilbara, Mr Bickerton, was not so sure. He referred to the clause dealing with the construction of a new railway and new roads. That clause (which in one form or another appear in each State agreement) provides that the joint venturers were to: 9(2)(a) operate their railway in a safe and proper manner and where and to the extent that they can do so without unduly prejudicing or interfering with their operations hereunder allow crossing places for roads stock and other railways and transport the passengers and carry the freight of the State and of third parties on the railway subject to and in accordance with by-laws (which shall include provision for reasonable charges) [in force]. … (b) … allow the public to use free of charge any roads (to the extent that it is reasonable and practicable so to do) constructed or upgraded under this clause PROVIDED THAT such use shall not unduly prejudice or interfere with the Joint Venturers' operations hereunder; 740 Mr Bickerton went on: "I thought that this agreement, from the publicity received and the statements made about it would mean that the company would be beneficial to the area generally because it would assist people living there in the transportation of their goods and materials." After referring to the provision which exempted the joint venturers from liability as a common carrier, he enquired of the Minister for Lands, the Hon Mr Bovall, how he (the Minister) would get on if he fell off the train. The Minister assured the Member for the Pilbara that he: "Would not find [himself] in that predicament". He went on to inform the members that he did not think that the joint venturers would be unreasonable in regard to their obligations, noting that the "joint venturers are there to foster and develop this [iron ore] industry": 23 August 1962, Vol 161, p 652. 741 The Minister's assurance proved to be false. The main reason is that clause 9(2) is particularly ill-suited for imposing an obligation on a mining company to carry freight for third parties. Many problems could easily have been overcome by careful drafting. Principal among the problems are: (1) The inexplicable failure to require the joint venturers to make available capacity at their port (in this agreement the port at Port Hedland) for third party goods; (2) The equally inexplicable failure to require the joint venturers to construct a railway line with capacity to carry goods for third parties; (3) The failure to specify how a third party's goods would be carried, eg in the case of iron ore, whether the goods would be carried in third party ore cars or in ore cars owned by the joint venturers; (4) The failure to make provision for loading facilities; and (5) The failure to the require the joint venturers to make storage and holding areas available for third parties. 742 As a result, in the 50 years or so since the railway lines were constructed, the owners have not carried a single good on behalf of a third party. Nor have they carried a single passenger. This is despite efforts by several junior mining companies to negotiate haulage arrangements. For instance, Mr Flanagan of Atlas Iron met with representatives of BHPB on a number of occasions between February 2005 and January 2007 to discuss the provision of a haulage service on the Goldsworthy railway on commercial terms. Those discussions were fruitless. [c-i-c] was involved in negotiations between 2007-2009 with senior officers of BHPB concerning haulage on the Mt Newman line. Those discussions came to an abrupt end when, in these reviews, FMG filed an affidavit sworn by [c-i-c]. 743 Other attempts to gain access under the WA regime have led to expensive litigation. In a case against Hancock Property, BHPB contended that it was not required to enter into access negotiations with a third party unless that third party was actually operating a mine that produced iron ore products. This contention was rejected out of hand by the court. Nevertheless, these kinds of cases can send an important message that a great deal of expense will be incurred whether or not access is eventually obtained.

13.3.1 The Marillana spur 757 Brockman is an ASX 300 listed company. Its principal activity is the development of the Marillana project, located 100km north of Newman and 290km south of Port Hedland, less than 3km from the Mt Newman railway. It is located approximately 30km from FMG's Mindy Mindy deposit. 758 The Marillana project has a total indicated and inferred resource of 1.63 billion tonnes. The project comprises 1.528 billion tonnes of beneficiation feed ore at average grade of 42.6% Fe and 101.9mt of CID at 55.6% Fe. 759 Brockman had initially intended to commence a small scale staged mining operation but decided in late 2008 to aim at commencing a larger mining operation with a production rate of 25mtpa over a forecast mine life of up to 25 years. 760 A scoping study released in April 2008, based on the initial small scale operation, considered two options for hauling ore to port. One option was the construction of a spur line from Marillana to Cloud Break. The second option was to negotiate access to the Mt Newman railway. In each case ore would be hauled to the berth at Port Hedland secured by the NWIOA, of which Brockman is a member. 761 On 6 March 2009 Brockman announced the completion of a scoping study for the development of a multi-user berth at Port Hedland. It stated that it was completing a pre-feasibility study on the construction of a 15mtpa iron ore mining and processing operation, commencing production in 2012 and subsequently expanding to 25mtpa in 2015. 762 On 10 August 2009 Brockman announced completion of its pre-feasibility study on the development of the full scale Marillana Project. The study confirmed the viability of the project based on targeted tonnages of 17 to 20mtpa with a minimum mine life of 20 years. The Brockman board subsequently approved the decision to proceed to a definitive feasibility study. 763 The pre-feasibility study considered two rail transport options. The first was to transport iron ore from Marillana via BHP's Mt Newman railway to the NWIOA's port infrastructure. The second was to transport iron ore from Marillana via the TPI railway to NWIOA or FMG port infrastructure. This involved the construction of a 110km spur line, which would connect to the Chichester line at a point north of the Chichester Ranges. The proposal acknowledges the difficulty of obtaining environmental approval to build a line through or along the northern part of the Fortescue Marsh, which would provide the most direct route for connecting to the Chichester line. Brockman is presently investigating several different routes for the construction of this rail spur as part of its definitive feasibility study. 764 Access to the Mt Newman railway is Brockman's preferred option. Mr Richards, the Managing Director of Brockman, said the Marillana project could not support the construction of independent rail infrastructure from the mine to Port Hedland or any other port. He also said it would not be economically viable to truck ore from Marillana to Port Hedland. A scoping study report conducted in 2007 estimated road transport costs for a distance of 360km to be $43 to $45 per tonne, which Mr Richards considered prohibitive. 765 There is no evidence that Brockman has investigated the use of the Hamersley or Robe line for the development of its Marillana project. The Hamersley line is located approximately 40km west of the project. Its use has not been investigated presumably because Brockman's plans for use of Port Hedland port facilities are well advanced. 766 In the Tribunal's view if access is not granted to the Mt Newman line, it is most likely Brockman will build a spur line to connect to the Chichester line. In his evidence, Mr Richards confirmed what appeared in his company's reports, namely that access to the Mt Newman line and a spur to the Chichester line were the only two haulage options under investigation. Both options have a positive NPV. The biggest hurdle facing Brockman is the ability to raise the finance. While this is presently being investigated, the Tribunal is of opinion, given the positive NPV of the project, that finance will become available. 767 This conclusion is confirmed by two factors. First, there are a number of projects other than Brockman in the Marillana area which would benefit from the spur line and might well be willing to contribute to its funding - Mindy Mindy, for example. Second, taking Mr Tapp's estimate that a 30mtpa project could fund a 260km line, it seems highly likely that a 20mtpa project could fund a 110km line.

13.3.3 The Aquila line 769 If the Robe line is not declared, it is likely that the API joint venture will build a railway from south of the Pannawonica area to Anketell Point. 770 API is an unincorporated joint venture, with Aquila holding a 50% interest and being responsible for the management of the project. AMCI Holdings Australia Pty Ltd holds the other 50% interest. 771 API's West Pilbara project is located south west of the Pannawonica project (owned by the Robe River Joint Venture between RTIO and Robe), approximately 150km from the coast. The project consists of a series of ore bodies, spread over a distance of about 800 square kilometres. 772 In March 2010 Aquila announced a JORC compliant resource in the west Pilbara of 742mt, comprising 586mt of CID resources with an average quality of 56.5% Fe and 156mt of bedded iron deposit resources at 61.47% Fe. The channel island deposits are clustered in the area to the south of Pannawonica. The bedded iron deposits lie further south east at a deposit known as Hardey Deposit, located approximately 50km west north west of the Paraburdoo township and within approximately 50km of the terminus of RTIO's Tom Price to Paraburdoo railway. 773 Aquila has been investigating the development of its West Pilbara project for some time. In 2008 a pre-feasibility study was completed which confirmed the technical and financial viability of a 25mtpa iron ore project, with a 160km railway being built to a new port at Cape Preston. Since then, its target production rates have grown and it has shifted its attention to the new port at Anketell Point. 774 On 6 February 2009 Aquila announced that the joint venture partners had agreed on Anketell Point as the preferred port site for a feasibility study for the west Pilbara project. The feasibility study was to be based on a production target of at least 30mtpa of channel island deposits, using new dedicated rail and port facilities. Since Aquila entered into a cooperation agreement with FMG regarding Anketell Point in mid 2009, it has made several announcements regarding the strong progress of its port plans. 775 On 28 August 2009 Aquila announced a strategic cooperation with Baosteel group Corporation, China's largest steel mill and one of the world's biggest steel makers, to fast-track the development of Aquila's key steel raw materials projects, including its iron ore projects. Baosteel is to invest up to $285.6m in Aquila via a placement of up to 43.95m shares. On 30 October 2009 Aquila announced that the proposed investment of $285m by Baosteel by way of subscription for shares in Aquila had been approved by the Foreign Investment Review Board and by Aquila's shareholders. 776 Aquila also announced that it had signed memoranda of understanding with 13 steel mills in China, Japan and South Korea for the testing of ore products. 777 Aquila's current plans appear to be for an initial planned production capacity of 30mtpa of channel iron ore. This will involve the construction of a 275km railway to Anketell Point. Importantly, the railway will extend a considerable distance south of Pannowonica and south of the Robe line. There will be an opportunity to later increase the rate of production through the addition of bedded iron products from the Hardey deposit. Aquila is undertaking a pre-feasibility study of developing the Hardey deposit, which will consider the construction of a 140km line to connect with the southern part of the proposed Aquila line. 778 The evidence suggests that Aquila should have little, if any, difficulty raising funding for its project. First, Mr Taylor estimated the NPV value of Aquila's project to be in the vicinity of $1.7bn. Second, there is the significant investment by Baosteel in Aquila. 779 Based on this material we think it to be inevitable that if Aquila cannot obtain a haulage service on the Robe line, it will construct its own line. 780 It is also necessary to consider whether Aquila would construct its own line even if access to the Robe line were available. If it could access the Robe line, Aquila would almost certainly build the southern part of the Aquila line which would connect its deposits with the Robe line. Whether it would build the remainder of the line is less obvious. Aquila has made many public announcements which discuss the means by which Aquila will transport its ore. The only means mentioned is to haul along its own line. 781 There has been no suggestion that Aquila wishes to obtain access to the Robe line. However, if Aquila could avoid the significant expense of building the Aquila line north of Pannowonica and instead access the Robe line, it might choose that option. 782 As for the 140km spur from the Hardey deposit, it is possible that it will be constructed. But, given that the Hardey deposit is not far from the Hamersley line, Aquila may seek access to the Hamersley line to transport ore from this deposit.

785 The parties have referred to "stranded deposits". We take a stranded deposit to refer to a tenement where the junior miner has no viable option for transporting its ore. 786 Obviously junior miners, whose deposit will otherwise be stranded, will benefit most from access. Other junior miners, whose deposits are not technically stranded, may also benefit from access. For example, a fairly large iron ore project will benefit if its only other option is to truck some of its ore to port. So also will the junior miner who, for reasons of cost, prefers to rail ore rather than truck it. Yet another example is a deposit which is closer to an incumbent's line than another line. The junior miner might be able to arrange to transport its ore on the other line, but would prefer for it to be hauled on the incumbent's line, again for reasons of cost. 787 To determine whether any tenements will be stranded, it is necessary to compare the areas that could be serviced if haulage was available on the incumbents' lines with the area that would be serviced by lines which will be built if there is no declaration. The map at Schedule 5 shows all the existing and potential new lines, except for the Marillana spur, which would run from roughly Marillana to connect with the Chichester line north of Cloud Break. 788 Beginning with the Goldsworthy line, it is notable that if the Goldsworthy line were not declared, there would be no alternative rail transport for the junior miners in the vicinity of the line. It is true that some projects south of the Goldsworthy line might, at least theoretically, be able to access the Chichester line. For those projects, chances are that Port Hedland would be a similar distance from the project as the Chichester line, in which case it would obviously be more practical to transport ore to Port Hedland directly. This is not to suggest that all the tenements would be stranded. Given their proximity to Port Hedland, it may be viable for some to truck ore to port. But for those with larger volumes, or located some distance from Port Hedland, road haulage may not be practical. 789 As regards the Mt Newman line, it is important to appreciate that the service applied for is point to point. Access seekers cannot enter the line at any point; they must enter at the proposed Mindy Mindy access point. The Marillana spur would extend almost all the way to Mindy Mindy. It follows that the Chichester line, in combination with the Marillana spur, would offer a rail service which would provide at least the same coverage as the Mt Newman line. Indeed, for some junior miners, the extended Chichester service may offer a better rail option. This is because access to the extended Chichester line will be to an "all points" rather than "point to point" service. Further, the Chichester line may be a more attractive option, at least from a cost perspective for those who are closer to the line than to Mindy Mindy. 790 The situation with the Hamersley line is less straightforward. For tenements along the line as it nears the coast, trucking directly to port may be viable, but tenements along the inland side of the line are likely to be too far from port to truck. From the Solomon area north to Emu, the Dixon line would cover the same territory as the Hamersley line. Projects in this area would not be stranded in the absence of access to the Hamersley line. Looking south of the Solomon area, the analysis is more complicated. There are some projects along the Brockman Spur and parts of the Rosella-Paraburdoo and Paraburdoo to Marandoo sections which might be close enough to the use the Dixon line (or even the Kennedy line in some cases). Of course, if they are required to construct roads, this may make transportation unviable. For projects further away from the Solomon area, trucking becomes less feasible (particularly if roads need to be built). East of Marandoo, and particularly to the south of the Hamersley line, there may be tenements that will be stranded without access to the Hamersley. Overall, of all of the lines, the risk of tenements being stranded or seriously inconvenienced (ie put to extra cost) if there is no access is greatest on the Hamersley line. 791 Dealing finally with the Robe line, it is clear that tenements to the north of the Robe line may be close enough to the coast to truck ore directly to port. The port might be Anketell Point or Cape Preston. Comparing the Aquila line and the Robe line, several observations can be made. First, the paths which the Aquila line and the Robe line follow are different. Although they intersect near Pannawonica and terminate in the Cape Lambert area, the Aquila line takes a more northern route. The routes diverge to the extent that the lines, at one point, are some 70km apart. Second, the Aquila line extends considerably south of the Robe line (at least 60km). Third, it is likely that many potential access seekers would be within trucking distance of either the Aquila line or, for some deposits to the east of the Robe line, the Hamersley or Dixon lines. Whether it is viable to truck to these lines will depend on whether new roads must be built. Of course, this is not to deny that for some junior miners who could access alternative lines, access to the Robe line would be at a significantly lower cost. Equally, for some junior miners, the Aquila or Hamersley/Robe lines may be less costly, because either (1) it is closer to them and/or (2) it is easier to access Cape Preston from the Aquila line. 792 In summary, the following broad conclusions can be stated:

if the Goldsworthy service is not declared, there is some prospect of deposits being stranded, and a greater likelihood of some deposits being inconvenienced;
if the Mt Newman service is not declared, there does not appear to be any project which would become stranded or would be inconvenienced. If anything, it may be more convenient for some to access the Chichester line;
if the Hamersley service is not declared, some tenements are likely to be stranded, and a significant number may incur greater cost to access or have their ore hauled on an alternative line; and
if the Robe line is not declared, a significant number of tenements are likely to be able to access alternative rail services, but some may incur greater cost to have their ore hauled on another line. 793 The foregoing analysis is subject to two important qualifications. First, we have not considered the characteristics of every tenement in the vicinity of the lines. We simply do not have the evidence needed to determine the likelihood of tenements being developed to the point of requiring rail transportation. A second, related, qualification is that it may be possible for junior miners (either individually or collectively) to build their own railway lines or a spur to a line to transport their iron ore. We cannot predict the likelihood of that occurring. Even with these qualifications, we think our analysis is robust and provides a good basis for understanding some implications were there to be no access.

The criteria 794 A service cannot be declared unless the Tribunal is satisfied that each criterion is met. If the criteria are met that is not the end of the matter. The Tribunal still retains a discretion whether or not to grant the declaration sought. 795 In these reviews all parties contend (and the Tribunal accepts) that (subject to a qualification that we will encounter later) two criteria are satisfied: criterion (c) - that the facility which provides the service is of national significance - and criterion (d) - that access to the service can be provided without undue risk to health or safety. Criterion (e) is irrelevant. Nothing more need be said about these criteria for the time being. The real dispute between the parties is about the remaining criteria. 796 When considering both the meaning and application of each criterion and when deciding how its discretion is to be exercised, the Tribunal must have regard to the objects in s 44AA(1): (1) efficient use of, operation of, and investment in, infrastructure; and (2) promotion of effective competition in dependent markets. The Tribunal must also have regard to s 44AA(2), which is not relevant here. We should explain, in a word or two, what we understand by these objects. 797 If there is efficient investment in infrastructure and competition in dependent markets, welfare is maximised. At a simple level, both efficient use of infrastructure and competition maximise welfare because they result in lower prices, better products and greater choice. More technically, competition in particular ensures efficient market outcomes. When competition is not inhibited by exclusionary practices or anti-competitive agreements, firms' rivalry for customers by offering lower prices, superior quality or new functions requires them to adopt more efficient means of doing business. There are different types of efficiencies, some more important than others. 798 A market achieves allocative efficiency when resources are allocated, through price, to their highest-value use among all competing uses. Then price will not rise far above the marginal cost of production. Monopoly, on the other hand, cannot achieve allocative efficiency as the monopolist's price will exceed marginal cost. 799 Productive efficiency results when all goods are produced at the minimum possible cost under existing technology. This means that as little of society's wealth as is necessary is expended in production. It also means that the producer is unable to sell above marginal cost because consumers will move to another producer. Monopolists, on the other hand, who are free from the constraints of competition, may be higher cost producers. 800 Dynamic efficiency arises because rivalry between firms encourages innovation to develop new and improved products. Schumpeter, with whom the dynamic efficiency principle is most closely associated, acknowledged the advantage of large firms to finance substantial research and development, but held that new firms would also be a constant source of supply of new ideas and innovations. Some economists contend that innovative efficiency provides the greatest enhancement of social wealth, suggesting it is the single most important factor in the growth of real output in industrial countries. 801 It should not be assumed that efficient markets, which produce an increase in aggregate social wealth, always produce benefits for consumers if those benefits are measured by price and quality. For example, competition laws may eliminate discriminatory pricing. But discriminatory pricing may lead to increased output, thereby increasing allocative efficiency, though such pricing may reduce consumer benefits by transferring wealth from the consumer to the producer. Competition laws also seek to eliminate most forms of collaboration between competitors. Collaboration will, however, lower production costs and achieve production efficiencies but will also enable those firms to raise prices and to take wealth from consumers. 802 Another category of efficiency is what might usefully be called transactional efficiency. Reference has already been made to the efficiencies of vertical integration. More broadly, economists such as Williamson have explained that firms design their business practices, contracts and organisational forms to minimise transaction costs (including information costs) and reduce their exposure to opportunistic behaviour or "hold ups". Firms will design contracts, create joint ventures, vertically integrate their operations, or propose mergers to minimise transaction costs. Allowing these efficiencies enhances allocative, productive and dynamic efficiencies. 803 The objectives identified in s 44AA contemplate the traditional categories of economic efficiencies, ie efficiencies in production, allocation of resources and innovation. The result to be achieved is the enhancement of aggregate local wealth. 804 We will now turn to the criteria, keeping in mind the economic goals of Part IIIA. For ease of discussion and to maintain a logical approach, we will deal with criterion (b) before criterion (a). We will finally deal with criterion (f) and discretionary considerations.

808 To identify what other facility may provide the service it is necessary, first, to consider what is contemplated by an equivalent service. Assume that there is a road and a canal side-by-side a railway. Each is capable of being used for haulage via, respectively, a truck, a barge and a train. Is each providing an equivalent service? At a very general level the answer is in the affirmative. Each is a means by which persons or goods may be transported. 809 Yet this is too simple an approach. Take the instant case. FMG applies for a declaration in hopes of obtaining access to a below rail service for a particular purpose - to haul iron ore. Not all railway lines are capable of providing that service - they must be specially constructed as heavy haulage lines. Not all roads or canals can be used to haul iron ore. Some roads are in a geographic location where haulage of iron ore is not permitted. In the case of a canal, its physical attributes, especially its depth, may make iron ore haulage impossible. In the Tribunal's view an equivalent service is one which is capable of satisfying the particular need for which the service is sought. 810 Turning to another aspect of equivalence, the Tribunal's position is that an equivalent service must allow for use between the same or similar locations. This was the approach adopted in Duke Eastern. There the service for which a declaration was sought was for the delivery of natural gas in a pipeline extending from Longford in Victoria to Sydney in New South Wales. The Tribunal (at [70]) rejected an argument that other gas pipelines originating or terminating at different locations provided the same "service". The Tribunal explained that there were two possible approaches: (1) identification of the service by reference to the markets they serve; or (2) identification of the service in terms of both start and end points. The Tribunal favoured the latter approach. That approach applies here, in a case where the start and end points are of real importance. What FMG has in mind is for the "inland" location to be proximate to a mine and the termination point to permit easy access to a particular port. 811 RTIO argues in the case of the Hamersley line that an alternative rail service would be FMG's proposed Kennedy line (ie connecting its mines in the Solomon area to the Chichester line, to haul iron ore to FMG's port facilities at Anderson Point). The Tribunal does not accept that this project, if it eventuates, would provide the same service as that provided by the Hamersley line. First, the port location is different. Second, such a line would only enable haulage of ore from the Solomon area, whereas the Hamersley line extends well beyond that area. 812 It is clear that iron ore can be transported in a number of ways apart from rail haulage. One way is trucking. In the case of magnetite iron ore, slurry pipelines are often used. In other parts of the world, use of barges by canals is also common. In this case, the incumbents (particularly BHPB) argue that trucking should be considered the same service. The Tribunal does not accept this argument. We will explain why. 813 The evidence is that for many mining companies who may seek access or to avail themselves of the service, the use of slurry pipelines or trucking is not a realistic alternative to rail haulage. Slurry pipelines can only be used for magnetite iron ore, when many mines in the Pilbara produce other types of ore. For reasons we explained in Chapter 7.3.1, trucking by road is not likely to be viable for transporting iron ore over longer distances and at large volumes. It is true that for some miners, transporting by road or slurry pipelines may substitute for rail. It does not follow that transporting by road or slurry pipelines is providing the same service for purposes of criterion (b). Those facilities do not provide the same service because they are not a substitute for the many mining companies that require haulage by rail as the only practically possible or economically feasible means of hauling their desired throughput of iron ore. 814 The last aspect of the meaning of "another facility", which is whether it must be a new facility or an existing facility that provides some or all of the service, will be considered later.

15.4.1 Privately profitable test 816 To understand the debate, it is necessary to appreciate the different results that might be reached depending on whether or not a privately profitable test is adopted. It was accepted by all the economic experts that the existence of a natural monopoly does not necessarily preclude the profitable development of a second facility. For example, suppose that an incumbent and potential access seeker occupy infra-marginal positions in a related market; ie, their marginal cost is below the prevailing market price. In that circumstance, it may be profitable for a second facility to be built, notwithstanding that it would be more efficient to share an existing facility. 817 The question is whether Part IIIA is intended to apply in circumstances where it is profitable - albeit less profitable, and potentially less efficient from society's perspective - for a second line to be built. The incumbents say Part IIIA is not intended to apply in those circumstances, because it is concerned with removing "bottlenecks" and criterion (b) should be seen as a bottleneck test. In contrast, FMG and the NCC argue that while bottleneck considerations may be relevant to criterion (a), criterion (b) is concerned with efficiency. 818 In resolving this debate, the place to begin is with the objects of Part IIIA. There are a number of features of the objects clause which should be noted. First, it refers to "effective competition" rather than competition per se. This is to be contrasted with s 2 of the Trade Practices Act, which relevantly provides that the object of the Act generally is to enhance the welfare of Australians through the promotion of competition. A very useful shorthand description of effective competition is proffered by Professor Hausman. He said: "By "effectively competitive" economists mean that no individual firm (or group of firms) is exercising significant market power nor is the price above the competitive price." A privately profitable test does not sit easily with the object of achieving effective competition. If viewed as a "bottleneck" test, as the incumbents would have it, then it simply tests whether a person could compete in a related market without access. It does not ask whether that person could compete effectively. It is not hard to conceive of circumstances in which a market is less than effectively competitive because third parties, relying on marginally profitable alternative facilities, cannot truly compete with an incumbent using (a much more profitable) facility with natural monopoly characteristics. 819 Another feature of s 44AA is that it is concerned with two distinct but related concepts - efficiency and effective competition. The references to efficiencies in "operating", "using" and "investment" in infrastructure connote concepts of productive, allocative and dynamic efficiencies. On any view, the scope of criterion (b) must take into account both effective competition and the efficiencies contemplated by s 44AA(1). 820 One issue raised by the privately profitable test is whether it ignores efficiency considerations, in particular, the allocative efficiency associated with the use of a natural monopoly facility. The proponents of the privately profitable test contend that if it is privately profitable to develop an alternative facility, there is a strong incentive for the access seeker and facility owner to voluntarily implement a socially efficient sharing arrangement. Professor Willig, for example, explains why it is that the credible threat of a new facility will result in "strong individual incentives" for the incumbent facility owner to enter into an efficient facility-sharing agreement. In substance, his approach suggests that once an incumbent realises that refusing to share a facility will not prevent an access seeker from competing in a related market, the incumbent would rationally share the facility if it is efficient to do so. If the incumbent cannot prevent a competitor entering a related market, the incumbent may as well profit from that entry, if possible. Indeed, Professor Willig argues, if the incumbent still refuses to share the facility in the face of a credible threat, then this strongly suggests that there must be inefficiencies and costs associated with sharing that outweigh any efficiencies. His point is that where it is privately profitable, an entrant will come into the market because it will be competitive (ie there is no market failure) and there is no need for regulation. If an alternative facility is not privately profitable, then regulation might be warranted. 821 A variation of this argument is that even if the market does not always achieve an efficient outcome when there is a credible threat of a new facility, it still gets it right most of the time and does a better job than regulation. For example, Professor Ordover draws a distinction between a technical natural monopoly test and evidence of market behaviour indicating an "independent [ie third party] commitment to enter/construct/duplicate" the facility. If there is such a commitment, Professor Ordover argues that the technical assessment whether it is or is not economic (or "uneconomical") to build another facility is "trumped" by the revealed behaviour of market participants. 822 Similarly, Professor Kalt, who deals with the US essential facilities doctrine, notes that under the second criterion expounded in the MCI case (the "impractical to duplicate" criterion), a would-be entrant could not succeed if it were privately profitable to duplicate the facility. In other words, says Professor Kalt, criterion 2 of MCI uses the market to perform the social cost test on the efficiency of new facilities. Professor Kalt acknowledges that market forces do not always produce satisfactory outcomes. But he says that "it is a false standard to place such market forces up against some vision of a court/regulatory process that is costless and without error of its own". He points out that the operative assumption of the US approach is that firms are rational profit-making concerns and this underlies the stringent US policy toward mandating access. His view is that, from a policy perspective, the US approach has the distinct advantage of using competing parties' respective self-interests to compel them to assess accurately the costs and benefits of their alternatives and for each to act on those alternatives in accord with their own individual self-interest. 823 In essence, the economists supporting the privately profitable test assert that where an alternative facility can credibly be built, private negotiations will necessarily result in efficient outcomes or, at least, will achieve efficient outcomes more readily than regulation. There are several problems with this assertion. First, it assumes that firms always, or usually, behave in an economically rational manner but from empirical observation we know they do not - especially when it comes to dealing with potential competitors. Second, there are often reasons for an incumbent owner who is behaving rationally to deny access to a potential competitor even when sharing would be socially optimal. Forcing the competitor to use a less profitable alternative facility may harm that competitor. The incumbent may seek to exploit the fact that it will take some time to build the alternative facility. It may be that it is only profitable to build an alternative facility with limited capacity (which is lower than the spare capacity on the existing facility which would otherwise be available). The incumbent may be mindful of not giving a fledgling competitor a "leg-up" to facilitate its growth into a larger player. Third, given the potential for market failure, it is far from clear that market forces achieve a better result than regulation as a general rule. In any event, it is doubtful that the Tribunal is entitled to assume that the decision-makers regulating Part IIIA will make errors. 824 There is another reason why a privately profitable test may not lead to the efficient use and operation of a facility. Suppose that an existing facility is a natural monopoly, ie it can satisfy society's total demand at a lower cost than two or more facilities. Suppose also that it is privately profitable to build an alternative facility which can only satisfy some, but not all, potential demand. On a privately profitable test, a declaration could not be made, even though many potential users would not be able to use the second facility, but could use (and use more efficiently) the incumbent's facility. It is difficult to see how such an outcome is consistent with the efficient use and operation of infrastructure - and for that matter, the achievement of effective competition - contemplated by s 44AA. 825 Another factor which tells against a privately profitable test is the emphasis on natural monopolies in the background materials to Part IIIA: see Chapter 9, where this material is discussed in detail. For present purposes, it is sufficient to note that the two consistent themes which emerge from this material are that: (1) facilities requiring access exhibit natural monopoly characteristics; and (2) the notion of being "uneconomical" has been linked to the definition of natural monopoly. In Chapter 11 of the Hilmer Report, which concerns access to "essential facilities", there is repeated reference to a facility being uneconomical to duplicate because it exhibits natural monopoly characteristics. Indeed, the Hilmer Committee's concern was with what it described as the "essential facilities problem". They defined an essential facility as one which satisfied two key conditions. First, the facility must exhibit natural monopoly characteristics and "hence cannot be duplicated economically". Second, access to the facility is required if a business is to be able to compete effectively in a related market. That said, the final recommendations of the Report did not refer to natural monopolies, instead recommending that a necessary condition for access was that access be essential to permit effective competition in a related market. This reproduced only the second element of the Committee's "essential facility" definition. 826 For reasons which are unclear, but probably relate to drafting style, the legislation and Competition Principles Agreement that followed the Hilmer Report adopted a more elaborate series of criteria for declaring access than those which were originally recommended. Criterion (a) seems to correspond, albeit couched in different terms, to the "second limb" of the essential facilities definition adopted in the Hilmer Report. Criterion (b) seems to correspond to the "first limb" of that definition. It is significant that in "reintroducing" this first limb, COAG has continued to associate the first limb with natural monopolies. In September 1994, a draft legislative package was released by COAG for public comment. One of the documents in the package was the explanatory memorandum that described an essential facility as one which "exhibits a high degree of natural monopoly characteristics; ie, a competitor could not duplicate it economically". In the second reading speeches for the Competition Policy Reform Billin the House of Representatives (Commonwealth, Debates, House of Representatives (1995) Vol HR202, p 2799) and in the Senate (Commonwealth, Debates, Senate (1995) Vol S170, p 2438), it was said that the notion underlying the proposed Part IIIA regime is that access to certain facilities with natural monopoly characteristics is needed to encourage competition in related markets. 827 While the background material makes repeated reference to infrastructure that displays natural monopoly characteristics, there is no reference to a privately profitable test in terms or in similar language. Nonetheless, the incumbents argue that the Hilmer Committee was principally concerned with bottlenecks, rather than natural monopolies. In this regard, they place particular emphasis on the passage from the Hilmer Report which recommended that a necessary condition for a declaration be that access is essential to permit effective competition in a related market. The Report then observed (at p 251): "Clearly, access to the facility should be essential, rather than merely convenient." 828 What must be borne in mind is that this comment was made in the context of access being essential for effective competition. It was not about whether access is essential to be able to compete per se. A "bottleneck" approach, as advocated by the incumbents, only tests the latter. 829 Another point against the privately profitable test is that it would lead to a significant degree of overlap between criterion (a) and criterion (b). All parties accept that whether or not it is privately profitable to build an alternative facility would be a relevant consideration for criterion (a). This is not to suggest that there is complete overlap between criterion (a) and a privately profitable test under criterion (b). There may be circumstances where, for example, it is not privately profitable to build an alternative facility but, for separate reasons (eg the existence of an alternative facility which provides part of the service), criterion (a) is not satisfied. Conversely, one can conceive of circumstances in which it would be privately profitable to build an alternative facility but criterion (a) would nonetheless be satisfied, such as when the use of a less profitable alternative facility would not allow for effective competition. The point is that the existence of a bottleneck (or whether it is privately profitable to build an alternative facility) is not in itself determinative of whether access would promote effective competition and, for that matter, whether access is socially efficient. It is, therefore, unclear why criterion (b) should separately test for a bottleneck on a stand-alone basis. An alternative approach is that criterion (b) is concerned with efficiency (ie the efficient use of existing infrastructure) and criterion (a) considers effective competition. This approach has the merit of avoiding overlap between criteria (a) and (b) and, more importantly, directly addresses the objects in s 44AA(1). 830 Were the interpretation of criterion (b) governed solely by the forgoing considerations, the Tribunal would have no doubt in concluding that the privately profitable test should be rejected. There are, however, other considerations which muddy the waters. The first is criterion (e), which is whether access to the service is already the subject of an "effective access regime". When considering this criterion, s 44H(5) provides that the minister must have regard to the Competition Principles Agreement. Clause 6(1)(a) of that agreement refers to whether it is "economically feasible to duplicate the facility." Putting these points together, the incumbents argue that the reference to "economically feasible" indicates a privately profitable test. They then argue that criterion (b) must be interpreted consistently with clause 6(1)(a), or otherwise different results may emerge, depending on whether one is assessing an application under a State access regime or under Part IIIA. 831 The Tribunal acknowledges that this argument has force. Perhaps the most natural meaning of the phrase "economically feasible" connotes private profitability. However, it is not too strained to read "economically feasible" as economically efficient, in the sense that something that is inefficient may be economically unfeasible when looked at from society's perspective. 832 Another feature supporting the privately profitable test is the reference in criterion (b) to it being uneconomical "for anyone" to develop another facility. The incumbents argue that this reference does not sit well with a natural monopoly or net benefit test. Both of those tests, it was said, ask what is best for society, rendering the reference to "for anyone" otiose. 833 There is also force in this argument. But, equally, the application of the phrase to the privately profitable test creates difficulties. It was suggested by the incumbents that "anyone" means "any particular individual who can be identified", including, for example, mining companies whose iron ore is sufficiently valuable to subsidise the cost of building the alternative rail facility. An alternative view is that "anyone" means anyone at all, asking whether any hypothetical person could build a rail line and make a profit from providing below rail services. 834 We think that under either view, practical problems arise. On the incumbent's view, how is one to identify the relevant individual who might profitably build the line? For example, in this case, is one required to consider whether a Chinese steel mill might fund a line in order to facilitate obtaining long-term supply contracts? If the alternative view is adopted, how is one to assess the potential demand for the services and the price for below rail services that "anyone" would be willing to pay? 835 In the end, the Tribunal does not consider that criterion (b) should be interpreted as a privately profitable test. That test is inconsistent with the enacting history and does not adequately meet the objectives of Part IIIA.

15.4.2 A net benefit or natural monopoly test 836 Having rejected the privately profitable approach, it is necessary to consider whether a natural monopoly approach or net social benefit approach is to be adopted. The Tribunal has in the past favoured the net social benefit approach. The Tribunal first considered the meaning of criterion (b) in Sydney Airport (No 1). There the Tribunal said that "uneconomical" did not inquire whether duplication was privately profitable. It said the criterion had to be construed in a broader social cost-benefit sense, in which the total costs and benefits of constructing another facility are to be taken into account. The Tribunal went on to explain (at [205]) that: If 'uneconomical' is interpreted in a private sense then the practical effect would often be to frustrate the underlying intent of the Act. This is because economies of scope may allow an incumbent, seeking to deny access to a potential entrant, to develop another facility while raising an insuperable barrier to entry to new players (a defining feature of a bottleneck). The use of the calculus of social cost benefit, however, ameliorates this problem by ensuring the total costs and benefits of developing another facility are brought to account. 837 The construction of criterion (b) was again considered in Duke Eastern. It was pointed out that the criterion appeared to describe a facility which exhibits natural monopoly characteristics. That led the Tribunal to say (at [64]): "[I]f a single [facility] can meet market demand at less cost (after taking into account productive allocative and dynamic effects) than two or more [facilities], it would be "uneconomic", in terms of criterion (b), to develop another [facility] to provide the same services." The Tribunal concluded (at [137]) that the "test is whether for a likely range of reasonably foreseeable demand for the services provided by means of the [facility], it would be more efficient, in terms of costs and benefits to the community as a whole, for one [facility] to provide those services rather than more than one." 838 With respect, we consider that a natural monopoly approach is preferable to a net social benefit approach adopted in previous Tribunal decisions for several reasons. First, the background material to criterion (b) consistently links the term "uneconomic" to the notion of a facility exhibiting natural monopoly characteristics. Natural monopoly rests upon a production cost function which does not take into account social benefits or net social benefits. Second, natural monopoly characteristics are concerned with the costs of production based on the available technology. Third, a net social benefit test gives criterion (b) a role which overlaps substantially, and perhaps usurps, the role of criterion (f). Both would involve a weighing up of many of the same social costs and social benefits. Importantly, in weighing up those costs and benefits, the criteria might arrive at different results. It must be borne in mind that many social costs and benefits are necessarily difficult, and sometimes impossible, to quantify. Accordingly, it may be difficult to conclude, at least in quantifiable terms, that there is or is not a "net social benefit". A requirement to be positively satisfied of such a matter - which would be a requirement if criterion (b) were a net social benefit test - would create a threshold which may, in practical terms alone, be difficult to satisfy. This is to be contrasted with criterion (f), which is framed in the negative. 839 Moreover, criterion (f) looks at the issue in a different setting. For criterion (f) to be satisfied (although it is expressed as a negative), it is not sufficient for the net costs of access to exceed net benefits, ie even if that is the result of the inquiry, the making of a declaration may yet not be contrary to the public interest. Other factors might carry the day. We will explain why this is so when dealing with that criterion. There is, however, no such latitude given to the regulator were all social costs and benefits to be brought into account in criterion (b).

840 The question is what costs are to be brought to account when deciding whether a facility has natural monopoly characteristics. Here several sub-issues arise. For one thing, we know that the costs cannot be those incurred by the "facility", for a facility incurs none. The relevant costs are those of the firm. In saying this, we recognise the concept of a firm having a number of plants, one of which may exhibit cost characteristics akin to that of a natural monopoly. This is a concept which some economists (eg Sharkey) call "plant subadditivity". Components of firms, like plants or facilities, cannot be natural monopolies as strictly defined because they are not a firm but only a component of a firm. Nonetheless, criterion (b) requires the notion of a natural monopoly to be applied to facilities. 841 Whether a firm's facility has natural monopoly characteristics involves determining whether the firm's cost function in relation to that facility is subadditive at all levels of output. This is a purely technical inquiry which looks at a firm's production costs. That is, the cost function to which regard must be had is the cost of the inputs (eg labour, operating costs, capital costs, co-ordination costs) incurred in producing the relevant good or service. 842 The expert economists put a different proposition. Most supported the position that "costs" are not confined to costs of inputs but should include all social costs. Dr Fitzgerald put it this way: "'Costs' should clearly, as a matter of economics, be defined comprehensively - i.e. not be confined to comparative capital and operating costs, but encompass all costs that differ as between the alternatives, including e.g. diseconomy costs of multi-user c.f. single-user operation" (his emphasis). With respect to the social cost test, Professor Willig points out that if some version of the test is to be applied, it would be inconsistent with basic economic logic just to take into account construction costs and operating costs. He says it would be necessary in the comparison to include all the impractical costs "engendered by the inevitability and the need to handle as well as possible the conflicts, disputes and paralyses that accompany the mandated sharing of the single set of facilities". It would also be necessary, he says, to include the significant social costs of the disincentives to invest caused by sharing. These costs, Professor Willig rightly says, are difficult to estimate and predict. 843 The Tribunal has given careful consideration to the views of the experts but is of the firm view that a natural monopoly test under criterion (b) should not take into account all social costs, for three reasons. First, almost all of the extensive literature on natural monopolies and the extensive works that consider whether particular railways exhibit the characteristics of a natural monopoly suggest that social costs should not be taken into account. The literature shows that economists have tested for natural monopoly by only taking into account costs of production: see by way of example only David Evans and James Heckman, 'A Test for Subadditivity of the Cost Function with an Application to the Bell System' (2006) 74(4) The American Economic Review 615, 622; Paul Joskow, 'Regulation of Natural Monopolies', in A Mitchell Polinsky & Steven Shavell (eds), Handbook of Law and Economics (2007), 11; John Bitzan, Railroad Cost Conditions - Implications for Policy (10 May 2000), prepared for the Federal Railroad Administration, US Department of Transport, 41-46; Marc Ivaldi and Gerard McCullough, 'Subadditivity Tests for Network Separation with an Application to US Railroads' (2008) 7(1) Review of Network Economics 159, 165. 844 The second reason why a natural monopoly test should not take into account all social costs is that, by the nature of the inquiry, many of those costs would not be taken into account even if known. A natural monopoly test is a static test. It assesses the state of an industry at a given point in time by taking a set level of demand and technology. Indeed, a common criticism of the natural monopoly approach is that it fails to take into account dynamic issues such as, for example, the social benefits of facilities-based competition, where competition is enhanced by each firm having its own facility, encouraging it to innovate (and hopefully to lower its costs) to capture market share. Many of the social costs which the incumbents say will be caused by access are dynamic in nature - for example, delays to expansions or the retardation of technological development. It may be possible to annualise some of these costs, but leaving aside the difficulty involved, there remain many dynamic costs which are incapable of being quantified. 845 The third reason is that, in this case, the social costs which the incumbents urge should be taken into account are, speaking rather loosely, the cost of the diseconomies and the inefficiencies that are said would result from access, those costs being largely in the form of lost production associated with activities in a downstream market. In the Tribunal's view, this confuses the cost of production of the service with the cost of providing access. The diseconomy and inefficiency costs, if they are incurred, are only incurred because of the incumbent's participation in a downstream market. But that is just an historical accident. In other cases, the incumbent owner may have no involvement in a dependent market. It could hardly be supposed that whether or not a facility displays natural monopoly characteristics depends upon whether or not the owner is engaged in a particular trade. 846 It should be stressed that just because a natural monopoly test does not take into account social costs does not mean that those costs are irrelevant. The costs are clearly relevant to criterion (f) and, perhaps, as discretionary factors. It is just that they are not relevant to criterion (b). 847 A related issue is the extent to which a natural monopoly test should take into account above rail costs. Criterion (b) is concerned with the economics of developing another facility to provide the service. There are some costs which are unambiguously costs of providing a below rail service (eg capital and operating costs of the track and coordination costs for scheduling and negotiating the allocation of "rail slots"). There are other costs that are not strictly below rail costs, but are costs which are consequential on whether or not a second line is built. For example, if a second line is built it is likely that newer, more efficient, trains would be used, whereas this may not be possible on an existing line. Those costs are properly characterised as above rail costs. The Tribunal considers that taking into account above rail costs in criterion (b) conflates the provision of below rail and above rail services. These costs may, of course, be relevant to other criteria. 848 There are three main ways to test for natural monopoly: engineering cost analysis, survivor analysis and econometric analysis. The first involves analysis of firm-level data. Survivor studies involve examining data on the growth of industries to determine those firms, assumed to be efficient, which increase market share over time and those firms, assumed inefficient, which decrease market share over time. Econometric studies involve examining cost and output data on firms in a particular industry over a period of time and applying statistical techniques to estimate the cost function for the industry. 849 Testing for a natural monopoly is notoriously difficult. Mr Sundakov, an economist, recognised the problem with implementing the test because of the difficulty in obtaining relevant cost information. Nonetheless, he said in the course of the experts' conference that "a traditional natural monopoly test, which is a re-examination of the costs of production with and without duplication, is a test that can be universally applied." As regards the difficulties in application, he explained that: "We have a choice here of having a test [the privately profitable test] that is easy to apply but looks at the wrong thing and doesn't quite give us the right number, versus the [natural monopoly] test that we recognise is difficult to apply but tries to answer the right question". Mr Sundakov suggested that it did not seem to be right to "go with the simplicity of the answer if it doesn't quite try to answer the right question". We agree. 850 In the present context, the question comes down to this: Can each line provide society's reasonably foreseeable demand for the below rail service at a lower total cost than if provided by two or more lines? The relevant costs are, as we have said, the costs of producing the below rail service. 851 An important assumption of this enquiry is that an existing line can, if necessary, be expanded to meet the reasonable foreseeable demand for the service. This is consistent with the economic theory of a natural monopoly, which takes into account the ability of the facility (or, more classically, the firm) to expand the relevant output: see eg Carl Kaysen and Donald Turner, Antitrust Policy - An Economic and Legal Analysis (1959); Richard Posner, Natural Monopoly and Its Regulation (30th Anniversary ed, 1999). In the case of an incumbent's line, the additional costs to be taken into account are of operating the line on a shared basis plus the capital cost of any expansion that is necessary to meet the demand. Those costs are to be contrasted with the sum of the costs of operating the incumbent's line (plus the cost of any expansion) for its own use and the cost of constructing and operating a new line(s) to meet third party demand. 852 It is necessary to determine a point in time at which to calculate whether it is uneconomic to develop an alternative facility. The obvious answer is that the facility must have that characteristic at the time of declaration. However, an access declaration may be made to continue for up to 20 years. It is appropriate, therefore, to consider what the future holds. The problem with that approach is that as cost structures change with ever-changing demand and, as technology changes, what is a natural monopoly today may not be one tomorrow. Provided the future is predictable with some measure of confidence, that future should be taken into account. 853 In deciding whether a line exhibits natural monopoly characteristics today, and whether it is likely to do so in the medium-term future, it is, we think, appropriate to begin by looking at what the position would be around 2015. We think this is appropriate for several reasons. First, this is the earliest an access seeker could most likely take up access having regard to the likelihood of appeals, the time consumed in negotiations and arbitrations, and the time it would take to construct the appropriate infrastructure. Second (and partly with the first reason in mind), the Tribunal has detailed capacity modelling of the lines (except for Goldsworthy) at around 2015. Third, if a line exhibits natural monopoly characteristics at that time, it is likely to do so now. Our reason for this proposition is simple enough. If in 2015 (or thereabouts) society's demand can be more efficiently accommodated on one line than on two or more, third party demand before 2015, which is likely to be significantly lower, will be more efficiently accommodated on one line due to the high upfront capital costs of constructing a new line, regardless of the level of demand. 854 Given that plans for the lines beyond 2014/2015 are still at a very early stage, it is impractical to conduct a detailed assessment of the lines beyond then. 855 In summary, to determine whether a facility is a natural monopoly, it is necessary, first, to determine the reasonably foreseeable potential demand for the facility (strictly the service provided by the facility), and then compare the capital and operating costs of a shared facility to the sum of the capital and operating costs of an existing facility (or an expanded existing facility) and a new facility.

15.5.1 Identifying reasonably foreseeable demand 856 Reasonably foreseeable demand consists of the incumbent's expected demand plus reasonably foreseeable third party demand. The material produced by the parties did not provide sufficient information about the activities of mining companies to assess demand. Accordingly, pursuant to s 44K(6), the Tribunal requested the NCC to prepare a report updating the information. For the report, the NCC was asked to consider only information made publicly available by or on behalf of the junior miner (and for public companies, only information published through the ASX). 857 The specific issues the NCC was asked to address on a project-by-project basis were the following: (1) The nature and extent of the resource for each project, specifying, in the case of a mineral resource, whether the JORC classification is inferred, indicated or measured, and in the case of an iron ore reserve, whether the JORC classification is probable or proved; (2) The chemical characteristics of the resource or reserve (eg its Fe content); (3) Any target production rate which is proposed or being investigated for the project; (4) Any transport arrangements that are proposed or being investigated for the project; and (5) Any use of port facilities that is proposed or being considered for the project. 858 Following the receipt of the report, the Tribunal invited the parties to comment on any errors or omissions. Comments were received from RTIO and BHPB. 859 Based on the information that was initially provided to the Tribunal, the information more recently provided by the NCC, and RTIO's and BHPB's comments, the Tribunal has been able to assess the demand for each service. In that regard the Tribunal has applied a number of broad rules to "filter out" tenements which it regards as not being treated as an appropriate source of demand. 860 First, in the absence of a concluded joint venture between BHPB and RTIO, the Tribunal has assumed that neither RTIO nor BHPB will demand use of the other's lines. This is consistent with a key theme in both their cases - that each operates its Pilbara iron ore business in a way which requires exclusive use and control of its rail lines. 861 Second, contrary to submissions by FMG, the Tribunal has excluded tenements at an earlier stage of development. 862 Third, only tenements which have JORC certified mineral resources and/or ore reserves are included. A mineral resource under the JORC can only be certified if there are "reasonable prospects for eventual economic extraction." Mr Harmsworth said that when an iron ore project is developed, there is usually a "reconnaissance" stage whereby some limited initial drilling is conducted. If it is thought that further exploration is warranted, it is most appropriate for further drilling to be undertaken to establish an (inferred) mineral resource. The Tribunal considers that the results of exploration activity prior to reaching JORC standard are too speculative to be relied upon. 863 The Tribunal appreciates that further JORC certified deposits will emerge. But the Tribunal must at least arrive at a "broad indicative estimate" of third party demand. No reliable estimates can be made of potential demand from mineral deposits which have not been proven to JORC standard. The result is that the Tribunal's estimates for third party demand are likely, if anything, to understate potential future demand. 864 Fourth, several tenements were excluded because of their proximity to port. The tenement include those where the deposit lies closer to the port than to the rail line (or in the case of Mt Newman, the loading point) and where the deposit lies so close to the port that rail access is highly unlikely to be uneconomic. 865 Fifth, several tenements were excluded on the basis that they have entered into long-term "mine gate" sales or similar arrangements. 866 Sixth, where a junior has committed itself to alternative means of transport (eg as where there is an agreement to use the Chichester line, or where the junior is planning to construct a slurry pipeline) its tenements have been excluded. 867 Seventh, junior miners with no access to port facilities were included. This was for the reason that new port facilities are likely to be developed to cater for third party demand. There are plans to expand and develop facilities at Utah Point and South-West Creek. There are also plans to develop Anketell Point, with estimates of the commencement of that port's operations varying from 2013 to 2015. 868 Eighth, the Tribunal has not, as suggested by BHPB, disregarded tenements with a low quality resource. Given the growth in demand predicted in years to come, we think it would be unsafe to proceed on any basis other than to assume that so called "low quality" resources could be developed. 869 Ninth, only juniors who have commenced production or who have announced they will commence production have been included. For the latter group the Tribunal has assumed that the project will be operational around 2014/2015. The evidence shows that once a decision is made to move to production the start-up time can be relatively short. Mr Richards said that even projects which are at the stage of being conceptual targets could be operational within 12 months. Mr Tapp said that it would take approximately 18 months to develop Mindy Mindy to the point of being fully operational. 870 To assess the volume of the demand the Tribunal has relied upon the production figures in the NCC's report and the incumbents' comments on that report. If a junior miner is considering a range of target production levels the Tribunal has used the highest figure. 871 A difficult issue with which the Tribunal had grappled is how to deal with demand for a particular line which is contingent on another service being declared. This issue arises in two ways: (1) third party ore hauled on the Mt Newman line may use the western part of the Goldsworthy line to get to port; (2) ore coming off the Hamersley line may use the Robe line north of Emu junction to get to port. We do not think that potential demand coming off other lines can be ignored. For this reason, we have considered the position both with and without other lines being declared.

15.5.1.1 Demand for the Mt Newman service 872 The service applied for on the Mt Newman line is point to point. Demand for the service will be limited to those who can realistically make use of the loading point near Mindy Mindy. FMG identified numerous tenements in the vicinity of the loading point, the owners of which might potentially demand the Mt Newman service. It assessed the likely reasonably foreseeable demand to be between 5mtpa and 15mtpa. 873 Mr Harmsworth considered the demand from tenements which he grouped according to geographic location. The first group, which he described as the "western group", is located between 170km and 425km from the Mindy Mindy loading point. The Tribunal believes this group is too far from the loading point to be considered part of the demand. 874 The next group is the "north-eastern group", being tenements that lie significantly to the north of the Mindy Mindy loading point. In some cases, they are closer to Port Hedland than to Mindy Mindy. These tenements cannot be considered part of the demand given that it would involve transporting the ore inland (ie away from port) for extended distances, which is both impractical and unnecessarily costly. 875 The next group is the "North Fortescue Valley" group. These tenements lie between 60km and 135km from the Mindy Mindy loading site, separated from it by the Fortescue marshes. One group of tenements belongs to Atlas but has not been proven to JORC standard. The other group is owned by BC Iron. BC Iron has expressed an intention to use the Chichester Line, and has awarded a road haulage contract for transportation of its iron ore to Cloud Break. These tenements are not a source of demand. 876 Another group identified by Mr Harmsworth is the "south-eastern group". These tenements lie close to the loading point. Some were owned by United Minerals Corporation, which was taken over by BHPB and are to be treated as part of BHPB's demand. Iron Ore Holdings Limited's (IOH) prospect at Phil's Creek has been excluded, as it has entered into a mine-gate sale agreement. A number of other tenements in the south-eastern group have not been proven to JORC standard or have not considered production targets. 877 Some tenements in the south eastern group should be included in the demand for the Mt Newman service. The largest is Brockman's Marillana project. According to Brockman's March 2010 Quarterly Report, the company is considering a production target of 20 mtpa. Next is the Mindy Mindy deposit. FMG envisages a production rate of 5mtpa. 878 IOH has commissioned a scoping study for its Iron Valley project and anticipates that a production rate of approximately 5 to 10mtpa is achievable. Although IOH has recently announced it has entered into an exclusive six month negotiating period with RTIO to sell the Iron Valley deposit, the deposit has not been sold and we think it should remain a potential source of demand. 879 FerrAus has just completed a scoping study for its prospects at Robertson Range and Davidson Creek. The scoping study concluded that the project could produce up to 15mtpa. FerrAus' project lies a considerable distance south of the Mindy Mindy loading point. Ideally, it would access the Mt Newman line from a point south of the Mindy Mindy loading point. The company is also considering building a rail spur to the Mindy Mindy loading point. FerrAus' project is a potential source of demand. 880 Talisman Mining Ltd has conducted a scoping study for its Wonmunna project. The study concluded that the project is "potentially economic based on high-grade and low-grade options with ore sold to nearby mines at production rates of 2 Mpta to 5 Mtpa". In its most recent annual report Talisman stated the results of the scoping study were promising, and the company was continuing to evaluate different transport alternatives including, among others, infrastructure sharing and other arrangements. The project should be treated as a potential source of demand for 5mtpa. 881 A further potential source of demand for the Mt Newman service is Dynasty Metals Australia Ltd's Prairie Downs project. This project is located south of Mt Whaleback. Mr Weekes, a geologist and engineer who is the Manager of Strategic Assets for the Pilbara Iron Company (Services) Pty Ltd, has provided evidence regarding tenements in the vicinity of the RTIO lines. He estimates that the Prairie Downs project is 30km west of the Mt Newman line and 46km south-east of the Hamersley line. It is located some distance from the Mindy Mindy loading point. On 18 March 2010, the company announced that the project has a JORC certified inferred resource of around 452mt at 23.1% Fe. The company believes that simple beneficiation processes can increase iron content and reduce silica to near-commercial direct shipping ore grades, with further improvements through optimisation likely. The company also believes that the project is "similar to, with the potential to match, [Brockman's] 17Mtpa mine and 1.6bn tonne Marillana Detrital Project". It identifies use of RTIO, BHPB and FMG railways as potential transport options. Although the project is at a relatively early stage of development, the Tribunal considers it should be taken into account as a potential source of demand. 882 The following table summarises the demand for a Mt Newman service: Project Demand (mtpa) Brockman - Marillana 20 FMG/Consmin - Mindy Mindy 5 IOH - Iron Valley 10 FerrAus - Robertson Range/ Davidson Creek 15 Talisman - Wonmunna 5 Dynasty Metals - Prairie Downs 17 Total 72

15.5.1.2 Demand for the Goldsworthy service 883 The Goldsworthy service is an all points service. It is necessary to distinguish between demand on the eastern (Yarrie) section of the line and demand on the western (Finucane) section of the line. 884 As regards the eastern section, Mr Harmsworth considered six tenements in the vicinity of the Goldsworthy line. Two do not have mineral resources and we put them aside. Two are Atlas projects at Pardoo and Ridley. In its 2009 Annual Report, Atlas indicated it would increase production from Pardoo to 3mtpa. It is true, as BHPB points out, that on 17 May 2010 Atlas clarified that Pardoo only has proved and probable ore reserves for 3.5 years based on a production target of 2.4mtpa. But Atlas says it has more extensive inferred and indicated resources. We do not think this means that Atlas' long-term production rate of 3mtpa from Pardoo should be discounted. Atlas' 2009 Annual Report also referred to a pre-feasibility study for the Ridley magnetite deposit which gave it an assumed rate of production of 15mtpa. Whether Ridley can be considered a realistic source of demand is questionable. There are large costs associated with magnetite iron ore production. The Ridley pre-feasibility study estimates the costs of establishing production to be $2.7bn (including the cost of constructing a power station). On the other hand, Mr Flanagan has said that the Goldsworthy service might be an option for the Ridley project. For this reason, the Tribunal considers it appropriate to include the Ridley project as a potential source of demand. 885 The Consolidated Minerals' (ConsMin) Woodie Woodie manganese project is located approximately 170km from the Goldsworthy railway at its closest point. ConsMin has been trucking ore to Port Hedland at a rate of less than 1mtpa since 1999 (it appears to be viable to truck manganese for longer distances because it is more valuable). [c-i-c]. The Tribunal will treat this project as a potential source of demand for 1mtpa. 886 Muccan Mines' Spinifex Ridge project is now owned by Moly Mines Limited. In its 2009 Annual Report, Moly Mines advised that a pre-feasibility had concluded that the project would be viable at 1mtpa. [c-i-c]. This project is a source of 1mtpa of potential demand. 887 The NCC's report provided information about Giralia Resources Ltd's McPhee Creek project. A scoping study for the project considered a base case for road haulage to Port Hedland, initially at 2mpta. Given the relatively low volumes involved, and the distance from the Goldsworthy line (at least 100km), the Tribunal does not consider this to be a realistic source of demand. 888 The following table summarises the demand for the eastern section of the Goldsworthy line: Project Demand (mtpa) Atlas - Pardoo 3 Atlas - Ridley 15 Consolidated Minerals - Woodie Woodie 1 Moly Mines - Spinifex Ridge 1 Total 20

15.5.1.3 Demand for the Hamersley service 891 The most obvious source of demand for the Hamersley line is from FMG's projects in the Solomon area. FMG has recently announced that its Solomon hub strategy will take place in two stages. It is anticipated that 60mtpa will initially be transported east to the Chichester line and ultimately to an expanded FMG port (with a capacity of 155mtpa) at Anderson Point. The second stage involves an additional 100mtpa being transported to Anketell Point. The target timeline for the second stage is 2013/14. The Tribunal considers this 100mtpa to be reasonably foreseeable demand. 892 Beyond the Solomon area projects, there are a number of other projects which represent potential demand for the Hamersley service. IOH's Iron Valley project (discussed previously in relation to the Mt Newman service) could transport 10mtpa on the Hamersley service. 893 Murchison Metals recently announced that a scoping study for its Rocklea project contemplates production of up to 10mtpa. 894 Talisman Mining's Wonmunna project (also discussed in relation to the Mt Newman line), is located within 25km of the Hamersley network. It is a source of demand for 5mtpa. 895 Similarly, Dynasty Metals' Prairie Downs project is clearly a source of potential demand (for 17mtpa) for the Hamersley service, having expressed an interest in the service and in using port facilities at Anketell Point. 896 Flinders Mines Ltd's Hamersley project lies relatively close to the Hamersley line: a distance of approximately 20 to 30km. Flinders Mines say that "a range of development options up to 15Mtpa have been financially modelled and give robust economic returns". 897 Aquila's Hardey deposit is located some 50km from the Hamersley line. An Aquila presentation made on 23 July 2009 nominates a production rate of 8-10mtpa for this deposit. Aquila said that a pre-feasibility study for the Hardey deposit will consider a 10mtpa mining operation with a 140km extension of a railway contemplated for API's West Pilbara deposits. The Hardey deposit should be considered a potential source of demand for 10mtpa. 898 Brockman's Marillana project has been excluded. A pre-feasibility study for the Marillana project considered various options for transporting ore, but did not consider use of the Hamersley line. Brockman's plans for securing port facilities in South West Creek are relatively well advanced which explains why the Hamersley line has not been considered. 899 The following table summarised the demand for the Hamersley line: Project Demand (mtpa) FMG - Solomon group projects 100 IOH - Iron Valley project 10 Murchison Metals - Rocklea 10 Talisman - Wonmunna 5 Aquila - Hardey 10 Flinders Mines - Hamersley 15 Dynasty Metals - Prairie Downs 17 Total 167

15.5.1.4 Demand for the Robe Service 900 The Robe service is an all points service. We propose to consider potential demand for (1) the south-west section of the line from Mesa J to near Emu and (2) the northern section of the line between Emu and Cape Lambert. 901 For the south-west section the main source of demand is API's West Pilbara project. API's demand is up to 40mtpa. This includes 30mtpa from its West Pilbara deposit and 10mtpa for the Hardey deposit. We think it quite likely that, if the Hamersley line were also declared, iron ore from the Hardey deposit could use the Hamersley service (thereby avoiding the cost of constructing a 140km spur to the West Solomon deposits). In that event, API's demand for the south-west section would only be 30mtpa. 902 There are two large magnetite projects near the coast southwest of Karatha, Australasian Resources Limited's Balmoral South Iron Ore Project and CITIC Pacific Limited's Sino Iron Project. Both projects have well-advanced plans to transport ore to the proposed new port facilities at Cape Preston. These projects have not been taken into account as potential sources of demand. 903 For the northern section of the Robe line, the potential demand includes third party ore that could be transported on the Hamersley line (if that line were declared) and then to Anketell Point. The total potential demand for the northern section is 197mtpa, comprising 167mtpa from the Hamersley line and 30mtpa from the south-west section of the line. 904 The analysis of the Robe line is complicated somewhat if the Hamersley line is not declared. Recall that the Tribunal considers that, if the Hamersley line were not declared, the Dixon line would be built. Theoretically, the Dixon line could be built to near Emu and then traffic coming off the Dixon line might seek to use the Robe line for the last part of the journey to Cape Lambert. We think this is unlikely. The evidence suggests that FMG would build the Dixon line as part of an integrated, Pilbara-wide network. For reasons which we will explain in Chapter 17.5.2, trains which run on the FMG railways are not compatible with the RTIO network. Accordingly, to obtain access to the RTIO network FMG would require a dedicated fleet of trains. It would make more sense to build the Dixon line an extra 65km to Cape Lambert. This would avoid the cost of double-handling and extra consists. For the same reason, we do not think that any junior miners using the Dixon line would choose to use the Robe line from Emu. 905 For completeness, the Tribunal notes it has excluded China Metallurgical Group Corporation's Cape Lambert Iron Ore project as a source of demand. This is a large-scale magnetite project. It is located between Karatha, Cape Lambert, Roebourne and Wickham, lying some 20km from Cape Lambert. Given the volumes involved and the mine's proximity to port, it is likely that this project will use its own facilities to carry ore to nearby Anketell Port.

15.6.2 Differences in capital cost 910 Ideally, the Tribunal would have detailed evidence regarding the respective capital costs of expanding the incumbent's line and developing an alternative line to accommodate third party demand. Engineers would describe which sections of track require expansion and would describe the nature of the works needed. Quantity surveyors would then estimate the cost of those works. Similar evidence would detail the cost of developing a new line. 911 The parties did not provide much evidence on these topics. They left it to the Tribunal to work out as best as it could what the costs would be. In the result, the Tribunal was left with these choices: (1) attempt to calculate the costs as best as it could on the available evidence; or (2) simply proceed on the basis that there is no comprehensive costing information and let the parties take the consequences. The Tribunal will, where possible, adopt the first approach, knowing that, in some instances, the result can at best be described as approximate or indicative. From the Tribunal's perspective, the thought of deciding such an important point on a "no evidence" basis is quite unsatisfactory and will be adopted only as a last resort. 912 To calculate expansion costs it is necessary to assess what are the required expansion works. We will embark upon that task, once again with little assistance from the parties. In doing so we will apply the following rules: (a) a single line with an appropriate number of sidings is capable of carrying up to approximately 140-155mtpa; (b) a double track line is capable of carrying up to 320-350mtpa; and (c) a triple track line can carry in excess of 320-350mtpa but the precise extent of the excess does not appear from the evidence. 913 These rules are based on broad, static estimates of track capacity. The capacity of a line is sensitive to factors such as the length of the line, the type of trains being run, the number of users on the line and the manner in which trains are run (eg on a scheduled vs flexible basis): see Chapter 11. For this reason, to most accurately assess the capacity of a line requires dynamic modelling, a task which has not been undertaken. 914 Nonetheless, the Tribunal has evidence about the capacity from which its rules are derived. First, there is the evidence of Dr Dallimore, Mr Hoare, Mr Ranson and RTIO's and BHBP's future expansion plans. There is also evidence from Mr Tapp about the capacity of the Chichester line. The evidence suggests that BHPB's, RTIO's and FMG's lines would reach their capacity limits at roughly the same point. While the capacity of a single track line can be increased through relatively small incremental expansions such as sidings, the limit of a single track line is approximately 140-155mtpa. The limit of a double track line is 300-350mtpa, depending on train configurations. 915 The only evidence regarding the capacity limit of a triple tracked line is an estimate of [c-i-c]mtpa in Mr Hoare's first affidavit. This was his estimate of the capacity of the Mt Newman line if automatic trains were used and the line were triple tracked. We believe the figure is not reliable because the assumption of triple tracking was not accompanied by an assumption about upgrading car dumping and port capacity. It is likely, therefore, that Mr Hoare's figure of [c-i-c]mtpa would underestimate the actual capacity of a triple tracked line. We note BHPB intends to haul [c-i-c]mtpa on the Mt Newman line without triple tracking the line as part of the Quantum expansions. 916 It is important to appreciate that the estimates of track capacity are based on a single user operation. The capacity limits when there are multiple users may be lower. Nonetheless, we think they provide a useful guide for determining the kind of expansions that would be required for certain sections to accommodate third party demand. If the total demand for a particular section is well below the limit, then it is safe to assume that whatever expansions are required would be significantly less than double tracking or triple tracking that section. Conversely, if the addition of third parties would bring total demand to or above the limit, then double tracking or triple tracking would clearly be required. 917 By the same token, when considering what kind of alternative facility would be required to accommodate third party demand in the no-access case, the capacity limits provide a useful guide as regards whether the line would need to be single tracked or double tracked. 918 We propose now to analyse what expansion works (if any) are required to each line on a section by section basis, applying our rules for capacity. Estimates of third party demand are based on our conclusions regarding potential third party demand for the lines. Estimates of incumbent demand are based on the incumbents' planned production schedules, and in RTIO's case, the production schedule used by TSG for its modelling of the 330 project.

15.6.2.2 The Robe line 924 Once again, we will consider two distinct sections of the line - the 122km south-west section between Mesa J and Western Creek and the 60km northern section of the line between Western Creek and Cape Lambert. 925 We have identified potential third party demand for the south-west section to be 30-40mtpa. A new section to cater for this demand would need to be single tracked with some sidings. If the existing section were to be shared, total demand would be around 60-70mtpa (comprising RTIO demand of around 30mtpa and third party demand of 30-40mtpa). This is well below the limit for a single track line, suggesting that any expansions required would be significantly less costly than double tracking. This conclusion is in line with Dr Dallimore's modelling of the Robe line, which shows that the impact of 50mtpa of third party demand on the Robe line could be ameliorated by relatively inexpensive measures (although some works might be required near the port). There is, therefore, likely to be large capital savings from sharing the existing south-west section rather than building a new section. 926 Turning to the northern section, under RTIO's current plans this section will be double tracked by 2014. RTIO's demand for the section is 180mtpa. Third party demand will depend on what happens with the Hamersley line. 927 If the Hamersley line is not declared, we have assessed third party demand to be 40mtpa. A new section would need to be single tracked with possibly a few sidings to handle the demand. If the existing section is shared, total demand would be 220mtpa. This is well below the limit of a double track line. Whatever expansions are required would, we think, be less expensive than building an entirely new section. 928 If the Hamersley line were declared, there would be 167mtpa of third party demand for the northern section. A new section would need to be double tracked to handle that volume. If the existing section was shared, it is quite probable it would need to be triple tracked to handle the demand. It is likely that whatever expansions are required would cost less than constructing a new double track section. 929 Based on this analysis, we conclude that the Robe line is a natural monopoly.

15.6.2.3 The Hamersley line 930 Demand for the Hamersley line varies from section to section. There are some sections where it is clear that there would be capital savings from sharing the line, and there are others where this is not so clear. 931 The most congested section is likely to be the 173km Emu to Rosella section. This reflects the fact that (1) all RTIO traffic on the Hamersley line flows through this section and (2) a large part of potential third party demand originates at or near the Solomon area. Total demand for the Emu to Rosella section is 465mtpa, comprising of 167mtpa of third party demand and 298mtpa of RTIO demand. If a new line were to be built, this section would need to be double tracked. Substantial expansion would be required if the existing section was to be shared - triple tracking and, possibly, quadruple tracking. In the end, it is unclear whether there would be any capital saving from sharing this section. 932 Another section where there may be limited, if any, capital saving from sharing is the 54km Juna Downs to Marandoo section. Total demand for that section is 187mtpa, comprising 155mtpa of RTIO demand and 32mtpa of third party demand. A new section would need to be single tracked, with some sidings, to accommodate third party demand. Sharing the existing section would require the line to be double tracked. In both scenarios, the section would effectively need to be duplicated. 933 There are, however, a number of sections where there would be capital savings from sharing. One is the 50km Brockman spur. Total demand for the spur is 89mtpa, comprising 79mtpa of RTIO demand and 10mtpa of third party demand (from Aquila's Hardey deposit). This demand is well below the notional limit for a single track. Whatever expansion is required to handle an extra 10mtpa of third party traffic will be significantly cheaper than building a new section. 934 Another section is the 134km section between Paraburdoo and Rosella. Total demand for that section is 69mtpa, comprising 59mpta of RTIO demand and 10mtpa of third party demand (from Murchison Metal's Rocklea project). A new section would need to be single tracked. Expansions required to cater for third party demand on the existing single track section would likely be relatively minor. 935 A similar analysis applies to the 193km section between Yandicoogina to Rosella, although that section must be broken down into smaller parts. Total demand for the 58km Marandoo to Rosella section is 202mtpa, comprising 170mtpa of RTIO demand and 32mpta of third party demand. The existing line is to be double tracked, and the expected demand is well below the notional capacity limit for a double tracked line. Expansion will cost less than building a new section. Similarly, total demand for the 81km Yandicoogina to Juna Downs section may be up to 92mtpa, comprising 60mtpa of RTIO demand and up to 32mtpa of third party demand (it is uncertain whether Dynasty's Prairie Downs project would require use of this section). The line will be single tracked with sidings. It will cost less to expand the existing section than build a new section to handle third party demand. 936 For completeness, it is necessary to consider the 63km Emu to Dampier section. It will be single tracked to cater for 150mtpa of RTIO demand. Third party demand will depend on whether the Robe line is declared. If the Robe line is declared, there is likely to be no third party demand for the section, as all third party ore would utilise the Robe line. If the Robe line is not declared, third party demand is 167mtpa. A new section would need to be double track. If the existing section is shared, total demand would be 317mtpa. That would require at least double tracking, or, more likely, triple tracking. We do not believe there would be capital savings from sharing this section of the track. 937 While there are sections of the Hamersley line where there may be no capital savings from sharing, there are substantial portions of the line where there will be large capital savings. We are satisfied that the Hamersley line is a natural monopoly.

15.6.2.4 The Mt Newman line 938 This is the most difficult line of the four to deal with. The declaration is sought in respect of only part of the Mt Newman line. We will break down that section into two parts for our analysis. 939 The first section is the "main" section of the line from Yandi junction to the termination point near Port Hedland, a distance of approximately 280km. By 2014/2015, this section will be double tracked. All of BHPB's throughput on the Mt Newman line traverses this section. All third party traffic would also use this section (up to the termination point for the Mt Newman service). Total demand forecast for the section in 2014/2015 is 312mtpa, comprising 240mtpa BHPB demand and 72mtpa third party demand. However, under its Quantum projects (which are likely to be completed by about 2015), BHPB demand for this section will increase to 350mtpa. Importantly, no expansion work to the section is included in the Quantum projects. Total demand is therefore likely to be at least 422mtpa, comprising 350mtpa BHPB demand and at least 72mtpa third party demand. We think it would be necessary to triple track this section to cater for total demand. 940 The other section of the Mt Newman line is from the Mindy Mindy loading point to Yandi junction, a distance of some 32km. This section will still be single tracked under RGP6. It is unclear whether it is necessary to double track this section to cater for third party demand, although given the large volume of third party demand (72mtpa), we think this is likely. 941 There are two approaches to determining what is an alternative facility for the Mt Newman line. One is to assume that the alternative facility is a new line between Mindy Mindy and the exit point immediately south of Port Hedland. On this approach, the alternative facility would be a single track line with sidings. The cost is likely to be similar to the cost of expanding the Mt Newman line to meet the forecast demand. Both would require the construction of an additional line, whether by triple tracking the Mt Newman line or by building a new line. Perhaps the only difference would be that some additional sidings may be required on the new line. The cost of those sidings will not be a significant proportion of the total cost. We could not be satisfied that the limited, if any, capital saving from sharing the line would warrant the conclusion that the Mt Newman line is a natural monopoly. 942 The alternative approach is to consider whether there is already in existence a facility which, if expanded or extended, could satisfy third party demand. There is such a facility, the Chichester line. It is possible to build a spur from Marillana to the Chichester line, a distance of approximately 110km. Although this spur would be originating from a slightly different point to the Mt Newman service, it is so close that we treat it as providing the same essential service. As for the Chichester line, Mr Tapp said that the Chichester line would most likely be double tracked, apart from the portions which go over bridges - although even those portions may have to be double tracked to accommodate an additional 72mtpa. 943 The cost of this alternative facility may or may not be less than the cost of the Mt Newman line. It depends on whether the cost of double tracking the entire new spur line, including double tracking over bridges, is less than the cost of double tracking the Yandi junction to Mindy Mindy section. However, given the increased volume that is expected by 2016, it would likely be cheaper to build a spur to the Chichester line and expand that line than expand the Mt Newman line. We say this because demand for a double tracked Chichester line would likely be well below the notional capacity limit of a double track line, meaning that limited expansions would likely be required and, in contrast, it would be necessary to triplicate a large part of the Mt Newman line. 944 On the facts, whichever alternative facility is assumed, we could not be satisfied that the Mt Newman line is a natural monopoly. 945 There will be cases where the nature of the alternative facility might lead to a different result. For that reason the Tribunal thinks it appropriate to express its view on the correct approach. 946 Mr Sundakov favours an approach ignoring any existing facility. He describes that approach as the only one which satisfies "economic logic". He explains that considering presently existing facilities could produce an absurd result. Assume, despite the fact that one facility would efficiently satisfy market demand, two identical facilities are constructed. Assume also that both facilities fall into the ownership of one firm. Finally, assume that a request for access to each facility is refused. If the existing line is taken into account, criterion (b) could never be satisfied because there is always in existence an alternative facility. The result is that there will never be access to either facility. 947 The Tribunal, on the other hand, has, in the past, adopted the approach of considering existing infrastructure. In Duke Eastern the Tribunal said that no economic reasoning could support a finding that infrastructure that is actually duplicated should be covered under the access code for national gas pipelines (which in relevant respects is the same as Part IIIA), unless for some reason the duplicated infrastructure was not an acceptable substitute. The Tribunal said that ignoring the existence of infrastructure would be to proceed in a "blinkered way" (at [57]). 948 Mr Sundakov is very critical of this view. He said that the economic logic of Part IIIA is to promote economic efficiency through the promotion of competition and the efficient use of infrastructure. In this regard, even if there has been wasteful duplication of infrastructure resulting in the construction of substitute facilities, the question to be examined under criterion (b) is not which of the existing facilities should be used to serve the demand but rather whether it is uneconomical from a social point of view to construct yet another facility. 949 As a matter of economic principle, we agree with Mr Sundakov. From a purely economic standpoint, what he says is plainly correct: a natural monopoly test considers whether one firm (or in the case of Part IIIA, one facility) could satisfy market demand at a lower cost than two firms (or facilities). It is a normative concept in the sense that an industry may be a natural monopoly, yet multiple firms may exist. 950 Whether that view should be applied to criterion (b) depends upon how, as a matter of construction, that criterion is to be applied. Is it intended to be given a practical, rather than a theoretical, operation? If it is and there is in existence an alternative facility, that facility must be considered in deciding whether access to existing infrastructure is required. We are of the view this is the correct way to proceed. That is, we think it best to adopt an approach the effect of which will be to consider what is the most cost effective (ie efficient) way to proceed in a "real world" scenario. Such an approach will give proper effect to the objects of Part IIIA. 951 Finally, we should note that Mr Sundakov's approach, if applied, does not negate the relevance of an existing facility to our overall inquiry. It would be relevant to whether access would increase competition in a dependent market (as to which see criterion (a)). It would also be relevant to the discretion.

15.7.1 General principles 953 Three questions arise: (1) What is meant by "profitable"?; (2) What is meant by "profitable … to develop another facility to provide the service"?; and (3) What is meant by "anyone" in the expression "uneconomical for anyone"? 954 As to (1), to an accountant profit means sales revenue minus the costs incurred in producing that revenue (eg wages, rent, raw materials, etc). To an economist profit is the minimum dollar value necessary to attract a firm to an industry or to induce the firm to remain in it. This will require credit to be given to all the opportunity costs incurred, including the cost of capital. An activity will be profitable if the return on a unit of output, less expenses (including depreciation and the opportunity cost of capital), is equal to or greater than the marginal cost of producing it. This is in line with the approach of Dr Fitzgerald, who said that an activity will be privately profitable if the entrant could cover "the full private costs of the substitute facility (ie both operational and capital costs), including a normal rate of return on capital, so as to compete in the dependent market." We prefer the approach of Dr Fitzgerald. 955 As to (2), the question raised is whether the profit must be generated directly by the provision of the service on the new facility, or whether it is sufficient that a downstream activity generates a profit for which the service is an input. The incumbents favour the latter view. They argue that the private profitability test would be satisfied if there is a mining company for which building a railway would be profitable, whether or not the cost of constructing and operating the railway is subsidised by profits from a downstream activity (eg the sale of iron ore). The incumbents' position is to be contrasted with a requirement that a person could profitability develop a new railway as a stand-alone business. 956 As to (3) (what is meant by "anyone"), there are two views. One, favoured by the incumbents, is that "anyone" means "any particular individual who can be identified", including, for example, mining companies whose iron ore is sufficiently valuable to subsidise the cost of building the alternative rail facility. In other words, criterion (b) will not be satisfied if only one person (eg FMG with its special characteristics of a mining company with a large resource) may find it profitable to construct a railway line to transport its iron ore to a port. An alternative view is that "anyone" means anyone at all, asking whether any hypothetical person could build a rail line and make a profit directly from providing below rail services. In this sense, the approach to the meaning of "anyone" is linked to whether the relevant profit-making activity must be the provision of the service, or some downstream activity for which the service is an input. 957 There are conflicting arguments regarding which approach is correct. The incumbents' approach is favoured, they say, because it tests for whether a facility is truly a bottleneck. While we have previously rejected the bottleneck approach, it is clear that, if criterion (b) is about bottlenecks, then the incumbents' approach tests for this. The alternative approach does not - it tests the viability of establishing a stand-alone below rail business. It may be unprofitable to establish a stand-alone business, but if it were worthwhile for a miner to build a new railway in order to generate profits from downstream activities, the existing railway would not be a bottleneck. 958 On the other hand, the plain words of the statute support the alternative view. The statute only refers to it being uneconomical for anyone to develop a facility to provide the service. In other words, the return that would render the construction of an alternative facility profitable is the return from the provision of the service produced by the use of the facility. On this approach, it is not relevant that a reasonable return can be obtained from the provision of some other service or the supply of goods for which the criterion (b) service is an input. 959 In the end, the Tribunal does not accept the incumbents' approach. The alternative approach accords with the plain meaning of the words in criterion (b) and sets up an objective standard. Whether or not criterion (b) is satisfied is not to be viewed from the perspective of any particular firm. Rather, criterion (b) looks at the position of a hypothetical firm and asks whether that firm could obtain a reasonable rate of return on capital if it duplicates the facility.

15.7.2 Application to the facts 960 Little evidence was put forward regarding the profitability of establishing a stand-alone alternative rail haulage business. The profitability of such a business would no doubt depend on a variety of factors, most fundamentally the demand for the service, the price which could be charged and the margins to be expected. One would need also to take into account that if too high a price is charged, customers may go elsewhere or look to build their own line. In the end, there is not enough evidence for us to be satisfied that it would be unprofitable to build an alternative railway (ie as a stand-alone business) to any of the lines for which a declaration is sought. 961 There is, of course, considerable evidence about the profitability of miners building railways as part of their integrated mining operations. We will summarise the conclusions that can be drawn from that evidence. 962 Nothing indicates that it would be privately profitable for a miner to build an alternative railway to the Goldsworthy line. BHPB suggests that it would be profitable to truck iron ore from tenements in the vicinity of the Goldsworthy line. This may be the case, but for reasons we have explained, trucking is not providing the same service as below rail. Given the relatively limited amount of identified third party demand in the area, we are satisfied that it would not be profitable to build an alternative railway to the Goldsworthy line. 963 In relation to the Mt Newman line, it would be profitable for Brockman to build a spur to the Chichester line. This would be developing another facility to provide an equivalent rail service to the Mt Newman service (if "develop" includes connecting to another facility). In any event, we are certainly not satisfied that it would be unprofitable for Brockman to do so. 964 In relation to the Hamersley line, RTIO argues, and we accept, that it is profitable for the Dixon line to be built. This, however, would be developing a facility which provides only part of the Hamersley service. Whether it is profitable to duplicate the remaining parts of the line - which extend for quite a considerable distance - is debatable. Mr Taylor has assessed the prospect of various projects in the vicinity of Yandicoogina building their own railway. However, Mr Taylor only assessed the profitability of building a railway for those projects to connect with the Chichester line. Such a railway, which would be going in the opposite direction to the Hamersley line, would provide a different service to the Hamersley service. We have assessed potential demand from that area as 32mtpa. That might be enough to justify building the Rosella to Yandicoogina part of the railway to connect with the Dixon line (given Mr Tapp's suggestion that a 30mtpa project would be able to attract financing for a 260km line). Similarly, it is possible, but ultimately quite unclear, whether there would be sufficient demand for other parts of the line to be replicated to connect with the Dixon line. Ultimately, we cannot be satisfied that it is unprofitable to develop an alternative railway to provide an equivalent service to the Hamersley service. 965 As regards the Robe line, it is likely to be profitable for Aquila to build its proposed line. Given that the Aquila line follows a quite different course to the Robe line, the Aquila line would not be providing the same service as the Robe line. Nonetheless, because it appears to be profitable to build a line in the vicinity of the Robe line, we could not be satisfied that it would be unprofitable to build a new line to provide the Robe service.

16.2.2 The consist proxy approach 971 This method compares the number of consists that the incumbent and third party would use on their own lines with the number of consists that the incumbent and the third party would use to achieve the desired output if they were to share the incumbent's line. The difference multiplied by the cost of a consist could be, in the Tribunal's view, a useful proxy for short-run below rail expansion costs. 972 When seeking to increase rail throughput, an incumbent could undertake below rail expansions, above rail expansions or (as is usually the case) a combination of both. An above rail expansion involves running extra trains. The extra trains might carry additional iron ore but will increase congestion (unless there are also below rail expansions), so there are diminishing marginal returns from such expansions. When the line is at its economic capacity, the marginal return of adding a consist will be less than the marginal cost, and it would be necessary to undertake below rail expansions to increase throughput. Nonetheless, if a line is not yet at its economic capacity, additional consists may well be a useful proxy for below rail expansions that would otherwise be necessary. It must be borne in mind, however, that as lines approach their economic capacity, the consist proxy is likely to overstate the cost of expansions due to diminishing marginal returns. 973 While a useful proxy, this method cannot be applied to any of the lines except the Robe line, if the Hamersley service is not declared. The Tribunal does not have any information from which to perform the calculation for the Goldsworthy line. With respect to the other lines, the Tribunal only has evidence of the number of extra consists required if there is up to an additional 50mtpa of third party ore. The number of extra consists required increases in a non-linear fashion with the number of additional tonnes that must be accommodated. This relationship has not been estimated. Thus it is not possible to use the consist proxy for tonnages significantly greater than 50mtpa and each of these lines (other than the Robe line if the Hamersley line is not declared) must accommodate, in at least some sections, well in excess of 50mtpa.

16.2.3 The Metalytics approach 974 BHPB commissioned an iron ore industry consultancy group, Metalytics, to conduct a survey of various costs in the Australian iron ore industry. Metalytics assessed the average cost of expansions to iron ore projects, using two methods. The first involves averaging the costs of 21 greenfield (new line construction) and brownfield (existing line expansion) projects completed or scheduled for completion over the period 2005-2011. The projects include a number of RTIO, FMG and BHPB projects, as well as projects from the Mid-West region of Western Australia and the Pilbara region. The second method involves a smaller sample of large-scale capacity expansions completed or planned for completion over the period 2006-2011 in the Pilbara only, largely comprising a number of RTIO, FMG and BHPB projects. 975 Metalytics prefers the results of the second method for several reasons. First, the figures are slightly less sensitive to inflation because they encompass a narrower time frame; second, they are Pilbara-specific, thus more accurately accounting for terrain; and third, they only relate to large-scale projects, reflecting the nature of RTIO and BHPB expansions. Within the second method, Metalytics estimated an overall cost of expansion, taking into account rail, port and mine expansions and also provided a breakdown of each element, including below and above rail costs. Metalytics concluded that the cost of below rail expansion is $18 per tonne per annum ($18m/mtpa). 976 The Tribunal is not willing to apply the Metalytics approach. Below rail expansions are a function of track length and terrain. A per tonne calculation does not account adequately for the length and terrain of a specific line. For this reason, Metalytics concluded that "the average capital cost breakdown estimates and percentages … may not be applicable to specific iron ore capacity expansion projects". A further problem with the Metalytics approach is that it is potentially difficult to apply where there is different demand for different sections of a line. The problems with applying the approach can be illustrated with a simple example. Take the Goldsworthy line. There is 20mtpa of potential third party demand for the 191km eastern section of the line, which could be accommodated with minimal expansion works. The only significant expansion works required would likely be on the 17km western section of the line, to accommodate 87mtpa potential third party demand. The Metalytics figure can only be applied assuming a single rate of demand - should the assumed demand be 20mtpa or 87mtpa? If the 87mtpa figure were used, the Metalytics figure would suggest the cost of the expansion would be $1.566bn. This is farfetched. Mr Neil Miller has estimated that the cost of expanding the western section is likely to be around $50m. He has also estimated the cost of duplicating the Goldsworthy line to be $657m - less than half the estimate calculated using the Metalytics figure.

16.4.3 The Hamersley line 992 It will be useful to calculate the maximum capital savings from sharing this line. The approach is the reverse to that for the Goldsworthy line. Where it is unclear what expansion works are required for a particular section, we assume the minimum expansion works which might reasonably be expected. This approach will likely overstate the true capital savings. 993 Expansions required on the Emu to Rosella section range from expanding to triple tracking or to quadruple-tracking. To estimate the maximum capital saving, we have assumed that triple tracking would suffice. Mr Miller estimated the cost of the double tracking between Rosella and Emu to be $668m. We assume that the incremental cost of triple tracking (ie moving from double track to triple track) would cost the same as moving from single track to double track. 994 The Juna Downs to Marandoo section requires double tracking. The distance which requires double tracking based on Mr Miller's estimate is 44.4km (Mr Miller estimated 9.6km of the 54km track to be double tracked already). This is multiplied by Mr Miller's average cost per kilometre estimate for expanding the Yandicoogina to Rosella spur (which we have applied to this section of the line). The final estimated cost is around $162m. 995 The Yandicoogina to Juna Downs section is likely to require the addition of sidings. To estimate the maximum capital saving, we have ignored the cost of sidings for two reasons. First, it is too difficult to estimate the number of additional sidings required. Second, sidings cost a small amount of money relative to other capital expansions: a siding only costs approximately $8m according to Mr Miller's estimate. 996 The Paraburdoo to Rosella section and the Brockman spur are likely to require either no expansion or the addition of sidings. To estimate the maximum capital savings, we assume that no expansions are required. 997 If the Robe line is declared, the Emu to Dampier section will likely require no expansion. If the Robe line is not declared, the Emu to Dampier section will require double tracking to handle potential third party demand. Mr Miller estimated this to cost $180m. 998 The minimum cost of expansions to the Hamersley line are summarised in the tables below: HAMERSLEY EXPANSIONS (WITH ACCESS TO ROBE) ($m) Triple tracking Emu-Rosella 668 Double tracking Marandoo-Juna Downs 162 Total 830

16.4.4 The Robe line 1001 Once again it will be useful to consider the minimum capital savings to be had from sharing the Robe line (ie, the same approach as that for the Goldsworthy line). Calculating the minimum capital savings for the Robe line is complicated by the possibility of the Hamersley line being declared. 1002 If the Hamersley line is not declared, we have assessed third party demand to be 40mtpa for both the south-west and northern sections of the Robe line. In this scenario, it is possible to gain some insights by using the "consist proxy method" based on Dr Dallimore's evidence. Dr Dallimore concluded that a maximum of two extra incumbent consists and two extra third party consists would be required to accommodate 50mtpa of third party demand, while one extra incumbent consist and one extra third party consist would be required to accommodate 20mtpa. In his report, Dr Dallimore estimated the number of consists a third party would require if it had access to the Robe line. In his oral evidence, he estimated the number of consists a third party would need if it had its own line. Comparing these estimates gives the number of additional consists a third party would require if sharing the Robe line. To calculate the minimum capital savings, we have adopted Dr Dallimore's estimates based on 50mtpa, although our assessed level of potential demand is only 40mtpa. Again, this approach may well overstate the capital cost of expansions. The difference is 4 consists. 1003 What is the cost of a consist? The Tribunal was provided with a variety of estimates. In the end, in its closing submissions, FMG adopted a figure of $56m for the cost of a consist. RTIO argued that the cost of an additional consist would be $105m, comprising $65m for the consist, $20m for additional drivers and $20m for extra maintenance. This estimate was provided by Dr Dallimore based on unsubstantiated, second-hand evidence, at a very late stage of the proceedings. FMG hotly contested this figure, particularly in respect of the $40m cost for drivers and maintenance. We have significant misgivings about RTIO's figure. Nonetheless, for the sake of this analysis, we will consider the potential capital savings if the RTIO figure is assumed. Based on the need for four additional consists to cater for 40mtpa third party demand, the additional consists would cost from around $224m (using FMG's figure) to $420m (using RTIO's figure). We stress that the consist method is only a proxy for the cost of below rail expansions required. The actual below rail cost may well be less than this, although we note that according to Dr Dallimore, some minor additional below rail expansions may be required near port to avoid third party use displacing throughput coming from the Mesa J area. 1004 Mr Miller's estimate of duplicating the Robe line is $875m. Mr Miller said his estimate was likely to understate the cost of duplicating the Robe line because some of the terrain that the line runs over is steeper than any of the terrain over which the Chichester line runs. 1005 The consist proxy method suggests that the minimum capital savings to be had from sharing the Robe line are likely to be somewhere in the vicinity of $455-651m. This is arrived at by taking the cost of duplicating the Robe line ($875m) less the cost of additional consists (ranging from $224m to $420m) used as proxies for expansion costs. 1006 The analysis is more complicated if the Hamersley line is declared. In this scenario, there would be 30mtpa of third party demand for the south-west section and 197mtpa for the northern section. Dr Dallimore's consist proxy figures cannot be used, as they do not contemplate 197mtpa of demand, and they assume the same level of demand across the whole line. It therefore becomes necessary to approach the scenario on a sectional basis. It is unclear whether there would be any capital savings from sharing the northern section of the line with 197mtpa of demand. It is quite possible that there would be savings. But even if it be assumed that there were no savings, there would be large capital savings from avoiding duplication of the south-west section of the line. We do not know precisely what would be required to accommodate third parties on the south-west section, but we are mindful that the total demand for that section would be well below the notional capacity limit of a single track line. We are confident that the cost of expansion would be well below the cost of duplicating the south-west section, which on Mr Miller's figures would be around $650m.

1009 In Re Queensland Co-operative Milling Association Ltd (1976) 8 ALR 481 ("QCMA") the Tribunal said (at 517) that "the concept of a market [is] basically a very simple idea". So it might be. But it has not always been that easy to define or locate, nor will it ever be. First of all, to a businessperson, a market is a place or area where goods may be sold or, more broadly, where there are people who are sufficiently aware of a firm's product to consider buying it. This concept of a market concentrates its attention on buyers rather than sellers. 1010 We are not here concerned with the businessperson's understanding of a market but rather with the analytical definitions developed by economists. Several classical economists have offered definitions. Cournot defined a market as: "The entire territory of which parts are so united by the relations of unrestricted commerce that prices there take the same level throughout, with ease and rapidity": Augustin Cournot, Researches into the Mathematical Principles of the Theory of Wealth (1971), 51-52. Marshall said that: "The more nearly perfect a market is, the stronger is the tendency for the same price to be paid for the same thing at the same time in all parts of the market: but of course if the market is large, allowance must be made for the expense of delivering the goods to different purchasers": Alfred Marshall, Principles of Economics (8th ed 1920), 325. According to Stigler: "A market for a commodity is the area within which the price tends to uniformity, allowance being made for transportation costs": George Stigler, The Theory of Competitive Price (1942), 92. 1011 This economic (or relevant) market, then, consists of groups of buyers and groups of sellers in a geographic region who seek each other out as a source of supply of, or as customers for, products. The interaction of the buyers and sellers determines the price for the products. 1012 We have not referred to a "group" of products because implicit in the classic economists' definition of a market is the assumption that there is only a single homogeneous product and that the firms in the market produce perfect substitutes. 1013 In the real world it is not only homogeneous products of rival sellers that affect price; price is also affected by the products of rival sellers that are close substitutes. Hence it is necessary to expand the definition of a market to include not only identical goods but also close substitutes. 1014 The output of the process of defining the relevant market - the identification of the participating firms, a description of the products exchanged and the borders within which the exchange occurs - is critical to an assessment of the behaviour of firms in the market (ie whether or not they impose competitive constraints upon one another) and, importantly, whether or not a firm has, or a group of firms have, power to control price or reduce competition (ie to shift the price away from that which would be obtained in a competitive market, namely the marginal cost of the product). 1015 In QCMA (at 517), the Tribunal defined a market to be a "field of rivalry" between firms in which there is "substitution between one product and another, and between one source of supply and another, in response to changing prices." 1016 Section 4E was introduced in 1977 to give statutory recognition to the concept of substitution. It provides that "for the purposes of this Act, unless the contrary intention appears, 'market' means a market in Australia and, when used in relation to any goods or services, includes a market for those goods or services and other goods or services that are substitutable for, or otherwise competitive with, the first-mentioned goods or services." 1017 It is often difficult, and sometimes impossible, to define with any precision the relevant dimensions (product, geographic, functional and temporal) of a market. On occasion, it can be particularly difficult to describe the relevant product market and its geographic borders. The process often involves judgments as to matters of degree that can be difficult to measure. 1018 As regards the product market, the notion of substitution refers on the demand side, to a customer's practical ability to switch from one product to another and, on the supply side, to the capacity of a supplier to switch production from one product to another. There are various conventional approaches to determining substitutability. 1019 The first significant approach developed by the courts and leading economists in the US is the "reasonable interchangeability of use or the cross-[price] elasticity of demand between product itself and substitutes for it": Brown Shoe Co. v United States 370 US 294 (1962) at [15]. Cross-price elasticity of demand measures the extent to which consumers will change their consumption of a product in response to a price change in another product. A high cross-price elasticity value suggests that products are good substitutes and are probably in the same product market. 1020 Reasonable interchangeability of use is established by looking at actual and potential buyer substitution patterns. Relevant evidence will include product characteristics (including differences in grade or quality), price differences (including price trends), past buyer responses, the views of firms regarding who their competitors are, and the existence or absence of different distribution channels. 1021 On the supply side, cross-price elasticity is also relevant. Products will be in the same market if a firm can readily switch production from one product to another. What is important is the ease with which the switch can take place. It may be immaterial that consumers do not regard the products as substitutes, that a price difference exists, or that the prices are not closely correlated. 1022 The geographic market is the area of effective competition in which sellers and buyers operate. What is relevant, as a starting point, are actual sales patterns, the location of customers and the place where sales take place, and any geographical boundaries that limit trade. But it is not sufficient to measure only historical and current market behaviour. It is also necessary to consider whether customers would readily turn to more remote suppliers in response to a price increase by local suppliers or whether remote suppliers would choose to enter the local market. 1023 In the United States, geographic markets are occasionally defined based on shipment flows. Under this method, commonly termed the Elzinga-Hogarty approach after the economists who proposed it, the geographic market is defined as the smallest insular region into which few products have come from the outside and from which few shipments go outside: Kenneth G Elzinga and Thomas F Hogarty, The Problem of Geographic Market Definition in Antimerger Suits, 18 Antitrust Bulletin 45 (1973). This is a neat approach, provided one bears in mind that it can either understate or overstate the regions that would be defined based on buyer substitution. 1024 A more recent (and increasingly popular) approach is to define a market as a group of products and a corresponding geographic area within which a hypothetical monopolist would be able to raise prices profitably. This approach dates back to the US Justice Department's 1982 Merger Guidelines and has its roots in the work of G Werden, an economist who worked in the Antitrust Division of the US Department of Justice. The guidelines defined an antitrust market as follows (at II and IIA): "A market is a group of products and an associated geographic area … [where] a hypothetical firm that was the only present and future seller of those products could raise price profitably." 1025 The definition in the current US Merger Guidelines is slightly different (at 1.0): A market is defined as a product or group of products and a geographic area in which it is produced or sold such that a hypothetical profit-maximizing firm, not subject to price regulation, that was the only present and future producer or seller of those products in that area likely would impose at least a "small but significant and nontransitory" increase in price, assuming the terms of sale of all other products are held constant. 1026 The hypothetical monopolist test has been adopted by the ACCC. Its 2008 Merger Guidelines state: 4.19 The HMT [hypothetical monopolist test] determines the smallest area in product and geographic space within which a hypothetical current and future profit-maximising monopolist could effectively exercise market power. In general, the exercise of market power by the hypothetical monopolist is characterised by the imposition of a small but significant and non-transitory increase in price (SSNIP) above the price level that would prevail without the merger, assuming the terms of sale of all other products are held constant. 1027 The test works this way. One looks at the effect of a price increase on a single product. If so many buyers would shift to alternative products that the monopolist would find the price increase to be unprofitable, the group of products is too narrow to constitute the market. The market should include all those products for which the hypothetical monopolist's price increase would be profitable. 1028 The factors that the guidelines use to determine whether the test is satisfied are conventional. The factors include buyers' perceptions, similarities and differences in price movements between different sets of products over a period of years, similarities or differences in product characteristics and evidence of sellers' perceptions. 1029 The test for the existence of significant market power is phrased in terms of the magnitude of the price increase that could be imposed by the hypothetical monopolist. It is generally accepted by both US and Australian regulators that a price increase is significant in size and length if it is at least 5% (sometimes 10%) and for at least one year, but there is flexibility in both the magnitude and the time period. It should be noted that these are not tolerance levels for anti-competitive price increases; they are merely a benchmark for assessing substitution. 1030 The merger guidelines approach is a proxy for examining the cross-price elasticity of demand for the products in the relevant market. It is an approach that ignores the supply side, although supply side substitution is taken into account separately (but similarly) to identify the participants in the market. 1031 Geographic markets are defined in an analogous manner. One identifies tentatively a small geographic area such that a hypothetical firm that is the only present producer of the relevant product or service is not able to profitably impose a price increase. If the product or service could be obtained elsewhere, an attempt to raise the price could not be profitable and the tentative geographic area would be too small. The geographic area is expanded until the hypothetical monopolist's price increase would be profitable. 1032 The hypothetical monopolist definition was designed to satisfy three objectives: (1) To connect the relevant market in antitrust case law to economic policy justifications in merger cases; (2) To arrive at a standardised definition of a market; and (3) To develop a test that fell within existing jurisprudence that relied on the product and geographic dimensions of a market. 1033 The hypothetical monopolist test has its critics. Stigler and Sherwin refer to it as "completely nonoperational": "The Extent of the Market", 28 Journal of Law and Economics 555, 582 (1985). Harris and Jorde say it is "simply devoid of any commercial reality: 71 California Law Review 464 (1983), 481. The test is, in essence, a "thought experiment". Initially it had a mixed reception in US courts. It was adopted in some cases but rejected in others. But the trend is towards acceptance of the test (a recent example is FTC v Whole Foods 548 F3d 1028 (2008) at 1038) not only in merger cases, but wherever it is necessary for competition purposes to define the boundaries of a relevant market (see eg Kentucky Speedway, LLC v National Association of Stock Car Auto Racing, Inc., 588 F3d 908 (2009)). 1034 Notwithstanding the criticisms, or in spite of it, the hypothetical monopolist test has gained currency in Australia: see eg Seven Network Ltd v News Ltd (2009) 262 ALR 160 at [353]; Australian Competition and Consumer Commission v Liquorland (Australia) Pty Ltd [2006] FCA 826 at 204, 209; Re Qantas Airways Limited [2004] ACompT 9 at 66-67; and Australian Competition and Consumer Commission v Australian Medical Association Western Australia Branch Inc (2003) 199 ALR 423 at 480.

1035 Vertical integration raises difficult issues about market definition and the identification of market participants. Assume, for example, that technological interdependence is the cause of integration. That is to say, a firm carries out successive processes which dictate a particularly efficient manufacturing structure. The usual example is the integration of iron and steelmaking. One issue is whether the output of each successive process, which may be said to be supplied by a firm to itself (sometimes referred to as in-house or self-supply), is a product or service for which there is a market. Another issue is whether the in-house produced product or service is in the same market as similar or substantially similar goods or services produced by third parties. Yet another issue is the extent to which the in-house supply can be treated as an indirect constraint on suppliers of similar or substantially similar goods or services produced by third parties. Here it is necessary to deal only with the first of these issues, but passing reference will be made to the other two. 1036 In her important paper, "Market Definition Issues in Australia and New Zealand, Trade Practices Litigation" (1990) 18 Australian Business Law Review 86, Professor Brunt (a former member of the Tribunal) said (at 122) that: "Where there are such efficiencies of vertical integration [within the relevant organisation(s)] … market coordination between buyers and sellers is superseded by in-house coordination. There would, in such a case, be no functional split to create market transactions between stages of production". It is necessary to examine this proposition closely and in different circumstances. In all circumstances, it is assumed that the good or service supplied in-house is a good or service which could, without major adaptation, be sold to a buyer for at least marginal cost. 1037 First, assume there are existing vertically separated sellers (ie non-in-house sellers) of that good or service. The presence of vertical separation may be sufficient to conclude that there is a separate functional market. Evidence of actual market transactions by vertically separate firms is strong evidence of the existence of separate functional markets. But some care must be taken in reaching that conclusion. There may be unusual circumstances, such as where there are very few market transactions, to negative the conclusion of a separate market. 1038 Accepting there is a separate functional market, the question that then arises is: Should the in-house producer be included in that market? The in-house producer should be included in the dependent market if a hypothetical monopolist of vertically separated supply could not profitably increase its price. This is frequently the case with end products, where consumers do not consider whether firms are vertically integrated or not when making their consumption choices. The same analysis may also apply in upstream input markets. If a vertically separated supplier of an input increases its price, the increase is likely to be passed through to consumers of the end product. The in-house producer may help to defeat the price increase by selling the input to vertically separated suppliers or, alternatively, it may continue to supply it in-house but increase its production of both the input and the end product. In that way, the in-house producers will either directly (by selling) or indirectly (by increasing in-house supply) constrain the behaviour of vertically separated sellers in the upstream market. 1039 There is another way in which the vertically integrated producer can be treated. It can be excluded from the market but taken into account when analysing competition in the market because it acts as a constraint on market participants. The better view is that if the vertically integrated producer responds directly or indirectly to a price increase, it should be included in the market because it is in competition (whether directly or indirectly) with the other firms in the market. 1040 Another question that arises is: Should two stages of the supply chain be "collapsed" into the same functional market? This question comes up when the constraint is indirect. For example, vertically integrated retail firms may consume only their own wholesaling services, while vertically separated retailers may consume only vertically separated wholesaling services. Although there is no direct competition between vertically integrated and vertically separated wholesalers, they indirectly constrain each other's behaviour. To explain, if a vertically separated wholesaler increased its prices and the vertically separated retailer passes this price increase on to consumers, the vertically separated retailer would lose market share to the vertically integrated retailer, who would consume more of their in-house wholesale services. 1041 It is when activity at one functional level (eg retail) constrains activity at another level (eg wholesale) that one may ask whether transactions at the two levels should be treated as being in the same market. Although the approach in merger cases has not been uniform, the answer has generally been in the affirmative. We believe, however, that it is preferable to find there to be separate markets, otherwise the market would consist of some firms that compete with each other (eg the vertically integrated retailer and the vertically separated retailers) and others that do not (eg the vertically separated wholesalers and the vertically separated retailers). This would be at odds with a commercial view of markets, as vertically separated suppliers at the two levels would not see themselves as part of a combined market. It would also be at odds with the approach taken to determining product and geographic markets, where the aim is to include substitution possibilities within the same market and exclude other products and geographic areas. In any event, as the two functions are usually complements and not substitutes, they ought not to be included in the same market. Instead, each functional market should be treated as relevant to the way in which vertically integrated firms constrain the market power of vertically separated firms, and vice versa. 1042 Why does it matter whether vertically integrated firms are included in the market or whether a functional split is created between two processes in the supply chain? It is true that in merger cases and abuse of market power cases, whichever approach is taken is unlikely to change the result. In Part IIIA cases, however, the distinction is critically important because the promotion of a material increase in competition will only satisfy criterion (a) if it occurs in a market other than the market for the service. We have explained why, in our view, it is preferable to place only "competitors" (ie those who act on and react to a direct or indirect constraint) in the market. While the purposive approach to market definition means that it is possible for divergent tests to be applied to different sections of the Trade Practices Act, the purpose of market definition in Part IIIA, s 46 and s 50 cases is to allow for an analysis of competition. This rather suggests that the approach to market definition should be uniform. 1043 The second circumstance in which to consider whether there is a functional split is where there are no existing vertically separated sellers of the good or service. Can it then be said there is a market for that good or service? Economists start with the proposition that without transactions there can be no market. We will in due course show that this proposition, correct though it might be, cannot survive the legal analysis that has been applied under the Trade Practices Act to the theory of a market. Under that theory, if there is evidence that a good or service could profitably be sold, there is a market for that good or service. Thus, there will be a functional split in markets if: (1) there is a demand at a particular functional level; (2) there is a real possibility that the demand can be satisfied; and (3) the supplier is able to recover at least the marginal cost of supply at that functional level. 1044 The Tribunal's approach in the past has been to find "no functional split" when vertical integration is "inevitable" or "overwhelmingly efficient". Thus, in Sydney Airport (No 1), the Tribunal explained (at [97]) that activities are economically separable unless the complementarities of vertical integration are such as to dictate that the service must be performed within the same economic entity. Also, in Re Services Sydney Pty Limited [2005] ACompT 7, the Tribunal adopted (at [117]) a test "which asks whether the complementarities of vertical integration are such as to dictate vertical integration". The Tribunal went on to say (at [120]) that in the particular case, it was "not satisfied that the costs associated with market transactions … [were] such, or conversely the efficiencies of vertical integration [were] such, that the existence of separate markets is precluded". 1045 This approach is consistent with the NCC's view that for a functional split, "vertical integration … [must not be] inevitable" (Recommendation for Declaration of the Goldsworthy Line (at 36)), and also with the ACCC's 2008 Merger Guidelines, where it said there will be no functional split if "there are overwhelming efficiencies of vertical integration": Merger Guidelines, November 2008 (at 22). 1046 It is easy enough to ask whether the in-house provision of a particular good or service is always more profitable than if it is purchased from a third party. If the answer is yes, then there are no potential transactions and hence there is no separate market. Resolving the question, however, involves a detailed analysis of the costs of production, including joint costs and common costs. It also involves analysis of revenues from production and of any efficiencies of specialisation resulting from vertical separation. The problem is that the information required is rarely available. 1047 For that reason, it may be helpful to examine a number of proxies that can be used to help come to a decision. First, evidence of functional splits in similar situations will be relevant to show that a functional split is possible in some circumstances. Second, examining firms' demand preferences will help to ascertain whether firms believe that it is possible to participate at one level of the supply chain and not the other. As firms are profit-motivated, absent any ulterior motive, firms' intentions are likely to be a reliable proxy.

17.4.2 Competition 1049 In considering the meaning of competition it is, we think, necessary to draw a distinction between, on the one hand, the process of competition and, on the other, the extent of competition, which is the outcome of that process. 1050 In economics, the word "competition" (as a process) has many meanings. Stigler's well-known definition is "rivalry between individuals (or groups or nations) and it arises whenever two or more parties strive for something that all cannot gain". Under this definition, competition refers to behaviour, and especially to patterns of business behaviour. Relevantly, it relates to behaviour that may affect the price or quality or conditions of sale of goods exchanged in the relevant market. Hence competition may be described as rivalry that amounts to a process that leads to an increase in economic efficiency. 1051 The extent of competition is another matter. Economists describe markets as perfectly competitive, effectively competitive or imperfectly competitive (such as monopolies or oligopolies). The state of competition in a market is the result of internal and external factors which bear upon the nature and extent of the rivalry. Hence, in Application by Chime Communications Pty Ltd (No 2) [2009] ACompT 2, the Tribunal said (at [48])): In the Tribunal's view a market is sufficiently competitive if the market experiences at least a reasonable degree of rivalry between firms each of which suffers some constraint in their use of market power from competitors (actual and potential) and from customers. The criteria for such competition are structural (a sufficient number of sellers, few inhibitions on entry and expansion), conduct-based (eg no collusion between firms, no exclusionary or predatory tactics) and performance-based (eg firms should be efficient, prices should reflect costs and be responsive to changing market forces). 1052 There are other descriptions. Professor Brunt says there is effective competition where "the availability of substitutes, both in demand and supply … act as constraints on each individual firm's market power." (Brunt, supra, 96). Professor Hausman said that "by 'effectively competitive' economists mean that no individual firm (or group of firms) is exercising significant market power nor is the price above the competitive price." The Report of the US Attorney General's National Committee to Study Antitrust Laws (1955) (at 320) states that no firm can choose its level of profit by giving less and charging more. This can only occur if there is a lack of availability of substitutes in supply and demand.

17.4.3 Access 1053 In Re Virgin Blue Airlines Pty Ltd [2005] ACompT 5, the applicant (Virgin Blue) applied to have declared various "airside services" it needed for its domestic flights to and from Sydney airport. Virgin Blue already had access to the services, but was dissatisfied with the terms, including the price. It assumed it would obtain better terms under Part IIIA. 1054 One issue the Tribunal was required to consider was whether access on better terms amounted to access for the purposes of Part IIIA. The Tribunal said that "access" was an ordinary English word and meant the right or ability to use the service. It went on to say that, though Virgin Blue already had the right to use the service, "access" also included access on better terms. The Tribunal put it this way (at [139]): "[T]he notions of access and increased access include the terms and conditions upon which such access or increased access is made available". 1055 The Tribunal then considered the "future with or without" test. It said (at [153]) that the relevant comparison was between the position if (1) the service is declared and (2) the service is not declared (taking into account Virgin Blue's rights of access on the existing terms). 1056 The Tribunal went on to say that whether competition would be promoted in the "future with" access depended upon the extent to which SACL, the service provider, had, in the absence of a declaration, the ability to use its market power to affect adversely competition in the dependent market. If it could, then, according to the Tribunal, the declaration would likely improve the environment for competition. 1057 In order to assess SACL's market power in the dependent market, the Tribunal looked in some detail at SACL's past conduct by reference to the following topics: SACL's use of its monopoly power; whether there were any constraints on the manner in which SACL could exercise its monopoly power; and, based on the analysis of these topics, whether increased access to the relevant service would promote competition in the dependent market. Following that analysis, the Tribunal concluded that a declaration would (as the legislation then read) promote an increase in competition. 1058 On review to the Full Court, this approach was successfully challenged although, for reasons that are not necessary to go into, the Tribunal's decision was not set aside. The Full Court (French J (as he then was), Finn J and Allsop J (as he then was)), found that the Tribunal had erred in law in various respects. First, the Tribunal erred in that the relevant future "with and without" was "with and without access", not "with and without declaration". Second, the Full Court said that the correct comparison is "of the future state of competition in the dependent market with a right or ability to use the service and the future state of competition in the dependent market without any right or ability or with a restricted right or ability to use the service … [T]he terms of s 44H(4)(a) do not incorporate the requirement for comparison with what is factually the current position in any given circumstances": Sydney Airport (No 2) at [83]-[84]. 1059 Given the Full Court's approach, "access" in criterion (a) means access simpliciter ie the right or ability to use the service. It does not mean the right or ability to use the service under Part IIIA. Indeed, the Full Court (at [82]) noted that it would be inappropriate for the Tribunal to surmise what the terms and conditions of access might be as a result of an arbitration under Part IIIA.

17.4.4 Promote competition 1060 How does an act, for example access to infrastructure, promote an increase or, now more correctly, a material increase in competition? In Sydney Airport (No 1), the Tribunal said (at [107]) that "the promotion of competition involves a consideration that if the conditions or environment for improving competition are enhanced, then there is a likelihood of increased competition that is not trivial" (an approach approved by the Full Federal Court on appeal (at [57]). In Application by Chime Communications Pty Ltd (No 3) [2009] ACompT 4(at [18]), the Tribunal proceeded on the basis that any increase in competition must be "socially-useful". 1061 A particular act will have the tendency to promote a material increase in competition in a socially useful way if sellers are given greater freedom to engage in rivalrous behaviour, or if the act will cause an increase in the number of rivals coupled with a move to more independent behaviour. Often the inquiry will come down to this: Will the act (eg an alteration to an aspect of market structure or a change in a firm's conduct) increase the constraints on the market power of sellers or, more directly, will it increase their rivalry in a way that will produce greater efficiency? If the answer is in the affirmative, the act will promote an increase in competition. 1062 What are the practical indicia that would lead to the conclusion that an increase in competition will be promoted? They would include a likely increase in the number of sellers (although that may not necessarily be a sufficient condition), a restraint on anti-competitive conduct such as price fixing or predatory pricing or the creation of incentives for innovation. That is, one is looking for a change that will force sellers to outsell, or attempt to outsell, their rivals in the relevant market, on pain of losing market share if they do not change their competitive strategy. One might also look for a stream of rivalrous initiatives that compel responses. 1063 The ability to examine these indicia has been reduced by the Full Court decision. On one view (which is of the NCC's view), the decision means that satisfaction of criterion (a) is to be determined without regard to factors such as whether access will be taken up; the terms and conditions that might be imposed on the grant of access; and the number of firms that will take up access and the extent to which they will utilise access. 1064 We do not believe that the Full Court intended to ignore any consideration of whether access would be taken up. While the Full Court made clear that the approach to criterion (a) is not to infuse "an overly elaborate body of considerations into that criterion", it could not have intended for the approach to be entirely theoretical. The Full Court itself considered factual considerations such as the substitutability of other specific airports, as well as SACL's right of first refusal to build and operate a second major airport for the purpose of determining whether access would promote an increase in competition. They considered these facts to be essential to their inquiry. Whether or not access will be taken up is also, we think, an essential consideration. 1065 It is possible that the Full Court did not consider that issue because the access seeker had already taken up access. It is also possible the Full Court proceeded on the basis that the mere possibility of access would be sufficient to promote competition. That approach is not dissimilar to what is known as the "contestable market theory". This theory holds that in certain circumstances, the mere threat of entry will constrain even a monopolist from exercising its market power to the detriment of consumers. The necessary condition for a contestable market is that entry is free and exit is costless (ie no sunk costs). It goes without saying, but say it we must, that the markets with which Part IIIA is concerned are as far removed from being "contestable" as may be. By way of example, Professor Willig, one of the authors of the contestable market theory has, in a co-authored paper, said that "the railroad industry is one in which contestability analysis cannot conceivably apply": Ioannis Kessides and Robert Willig, Restructuring Regulation of the Rail Industry for the Public Interest, Policy Research Working Paper 1506, The World Bank (1995). We would add that neither could a related or derivative theory. 1066 Putting that theory to one side, the Tribunal does not adhere to the view that mere access of itself will promote a material increase in competition. What matters is the likelihood of access, the sufficiency of access and the likely timing of access. For the Full Court to hold that it is impermissible to consider whether, when and to what extent, access will be taken up could easily lead to the result that criterion (a) is not satisfied although a close examination of the facts may show otherwise. This result could hardly have been intended. If it was, it is necessary for Parliament to intervene. That said, we accept that in assessing the extent to which access would be taken up, the Tribunal should assume that access is on reasonable terms and conditions, without speculating about any particular terms that might be imposed by arbitration under Part IIIA. 1067 A second issue that arises from the Full Court's decision is whether access must be "essential" or "necessary" to permit effective competition in a related market for criterion (a) to be satisfied. There are passages in the joint judgment which suggest that this is the approach. For example, the Full Court noted (at [86]) that "the essential precondition discussed [in the enacting history] was that access…was necessary to permit effective competition in a downstream or upstream market." Earlier (at [37]), the Full Court had said that the "essential notion" to be derived from the Hilmer Report, the outline of legislation issued by COAG, the Explanatory Memorandum and the Competition Principles Agreement is that "it is necessary for the fact of access (in its ordinary meaning) to be relevant to effective competition in another market". 1068 It is worthwhile pausing at this point to discuss yet another aspect of the Full Court's decision. It is apparent from the passages just cited that the Full Court considers that criterion (a) is concerned with effective competition. To reach this view, the Full Court relied on the legislative background. If anything, the view has been strengthened with the subsequent introduction of s 44AA, which now expressly provides that an object of Part IIIA is to promote "effective competition". The position we take is that if a dependent market is already effectively competitive, intervention is not called for. That is, we read criterion (a) as having no application to a market which is effectively competitive. In any event, even if we are wrong in this approach that the dependent market is already effectively competitive, it would be an important consideration under criterion (f) and the discretion. 1069 Coming back to the "essentiality" requirement, this does no more than reflect the comparative nature of the "with or without" test. If the state of competition with access is the same as the state of competition without access, access is not essential for promoting competition. In most cases there can be no quarrel with this approach. There will, however, be some situations where the result may be unsatisfactory. 1070 Assume that a facility is a natural monopoly but not a bottleneck. Assume also that the facility could profitably be duplicated, although that would be highly inefficient. Assume finally that either access to the natural monopoly facility or the construction of a substitute facility would equally promote an increase in competition. We accept that this assumption would not hold in all cases, ie where in one future, competition would be promoted to a greater extent in one scenario than the other. In the assumed case, criterion (a) would not be satisfied. 1071 We have some reservations about this outcome. It has the arguably perverse result that a declaration must be refused because a highly inefficient duplication of an existing facility would have the same competitive effect as sharing the facility. It is doubtful whether this is what the legislature intended. But we accept that this result cannot be avoided as criterion (a) is presently worded. Still, we think that this is an issue that warrants parliament's close attention.

1072 One dependent market identified by FMG is the global iron ore market. The existence of this market has produced the least disagreement among the economists. Interestingly, all the economists agreed the market was global in geographic dimension. Further, no economist suggested that the product dimension of the market comprised three separate markets: one for fines, one for lump and the other for pellets. Mr Tapp put the argument that there were three separate product markets, none of which was a global market, but this was not pressed in closing. 1073 What received surprisingly little analysis is whether or not the market was for seaborne iron ore or for all iron ore production. Professor Gans said there was a seaborne iron ore market, justifying his view on the proposition that for Pilbara producers it is the only means of transporting Pilbara iron ore to consumers. Professor Gans did not consider the demand side, in particular whether consumers had access to non-seaborne supplies of iron ore. Professor Hausman argued that the market should include all iron ore produced as it competes directly with seaborne iron ore, although he did not analyse why this was so. 1074 Mr Tapp provided a justification for a seaborne iron ore market. He said that iron ore that is produced inland with no access to transport infrastructure is usually located in close proximity to a steel mill that is its only customer. Often the mine and mill are owned by the same firm. This "captive iron ore" is sold in separate smaller markets and does not compete with seaborne iron ore. That is, if the price of seaborne iron ore increases, mills that consume seaborne iron ore will not substitute captive iron ore. Generally this is because transport is unavailable without significant investment and because a vertically integrated firm may not wish to supply its competitors with a cheaper input. Thus, Mr Tapp contends that the seaborne market consists of iron ore that is located close to a port or is located close to infrastructure that can transport it to a port. 1075 In its submissions to the NCC, FMG argued that the vast majority of countries that import iron ore have only negligible domestic iron ore production potential and that Chinese production is confined to low grade magnetite which has high extraction costs and does not constrain seaborne prices. The NCC, on the other hand, considered that the market includes non-seaborne iron ore even though it noted that a significant percentage of domestic iron ore is mined for vertically integrated and co-located steel mills and is thus not available for international sale in the short-term: Mt Newman Final Recommendation at 7.42. It said that the price of global iron ore constrains the price of domestically produced iron ore. 1076 The debate could be resolved by applying the hypothetical monopolist test. If a hypothetical monopolist of global seaborne iron ore could profitably increase its price, then the market does not include non-seaborne iron ore. If the monopolist could not profitably increase its price, then the market needs to be expanded to include possible substitutes. One substitute could be non-seaborne iron ore. Regrettably, the Tribunal does not have accurate data to draw appropriate conclusions. 1077 The precise boundaries of the market are likely to lie somewhere between the two positions. We accept that iron ore produced by vertically integrated firms and iron ore located in mines near to a steel mill are unlikely to be released into the seaborne market in response to a price increase because of the lack of transportation opportunities, the likely differences in quality and the attitudes of vertically integrated firms. Not all domestic sources will fall into this category. Some will be substitutable for seaborne iron ore. Accurate data on the rates of production in the global seaborne iron ore market and global production of iron ore are readily available. But there is no data for the market as we have defined it. 1078 In this circumstance, in order to assess the effect access will have on the iron ore market, we propose to base our analysis on the information we have concerning the size of the seaborne iron ore market. That is likely to be a closer proxy for the true market than the global iron ore market. We appreciate, though this may overstate the benefits of access to competition in the iron ore market. 1079 The following table lists the volume of iron ore exported in the global seaborne market for each of the years 2001 to 2007. The data is derived from the Annual Outlook for Iron Ore (2008) published by CRU International Ltd, and produced by Mr Bristow, the Chief Analyst of BHPB's Carbon Steel Materials business. Global seaborne iron ore exports Year 2001 2002 2003 2004 2005 2006 2007 Volume (mtpa) 444 474 535 595 659 712 781

1081 The first question we must resolve is whether the global seaborne market is already effectively competitive. The market has three main suppliers, Vale, BHPB and RTIO, which supply approximately 30%, 20% and 15% of the market, respectively. In addition, there are many smaller firms. Ten firms supplied more than 20mtpa of iron ore in 2008. 1082 There is no evidence that firms collusively set prices for the major buyers (steel mills in China and elsewhere in the East). To the contrary, the evidence indicates that a large proportion of iron ore is sold under medium to long-term contracts under which a benchmark price is agreed as a result of individual negotiations and increases to the benchmark are also agreed following further individual negotiations. Dr Williams, an economist, described the global iron ore market (perhaps only slightly tongue-in-cheek) as being as close as can be to perfectly competitive. 1083 The Tribunal accepts that the market is effectively competitive and is likely to continue to be so. First, any attempt to give less or charge more by a major supplier would be constrained by the other major suppliers or by a combination of the smaller firms (demand-side substitution). Second, that action may lead to marginal mines coming online or additional ore being produced by existing mines (supply side substitution). 1084 Strictly speaking, that is the end of the inquiry: a market in which there is effective competition is not one in which, relevantly, there can be the material increase in competition contemplated by criterion (a). We do not understand FMG to oppose the conclusion that the seaborne iron ore market is a competitive market. Nevertheless it says that access will increase competition in that market and that is sufficient to pass through criterion (a). While we do not accept this approach, we will, in any event, consider whether it is correct. 1085 FMG bases its case on the framework constructed by Mr Houston, an economist. First he accepts that what is required is the traditional forward-looking analysis of making a comparison between competition in a future with access and competition in a future without access. Then Mr Houston says that for the future state of competition to be materially different (by which we take him to mean where it is effectively competitive), two conditions must be satisfied: (1) the service to which access is sought is a "bottleneck" to competition in the relevant market; and (2) there must be no other significant barrier to entry to that market. Mr Houston says that if access to the service is not a "bottleneck" then access is not necessary for competition to be promoted. Mr Houston views a bottleneck as a service without close substitutes. 1086 Having laid down the two conditions, Mr Houston then applies them to the seaborne iron ore market. He concludes that trucking is not an effective substitute for rail haulage, except where the distances to be trucked are relatively short, eg 75km or less. On this basis, without access there would be mines which would be left stranded. He was asked to assume that, if all four services were declared, additional iron ore would be exported from Australia into the seaborne iron ore market. This would promote competition, Mr Houston reasoned, because the additional iron ore would affect market price. 1087 Even if Mr Houston's approach were correct as a matter of theory, there are at least two factual respects in which we disagree with his analysis. First, while it is clear that some tenements in the Pilbara may not have an effective substitute to the incumbent's lines, there are many which would have effective substitutes in the form of alternative rail facilities. If there is no access, alternative rail facilities are likely to be available for (1) users of the Mt Newman service (via a spur from Marillana to the Chichester line); (2) for a significant proportion of users of the Robe line (on the Aquila line); and (3) for some users of the Hamersley service (particularly those located near the Solomon area and the proposed Dixon line). 1088 The second respect in which we disagree with Mr Houston is with his contention that with access the additional iron ore that would find its way into the market would affect price. 1089 Mr Houston approached the matter this way. He was asked to assume that if all the applications for declaration were successful and access were granted, an additional 62.5mtpa of iron ore would be sold into the global seaborne iron ore market. The assumed volume of additional ore is not one that is supported by the facts. It is also impossible to approach the issue on the basis of all declarations being granted. 1090 However that may be, Mr Houston expressed the opinion that even a modest increase in supply by a low cost producer "will displace ore that would otherwise have been provided by higher-cost suppliers, thereby reducing the world price". He explained that: even modest increases in output by Australian producers may lead to material reductions in global iron ore prices and have wide-reaching implications for the resources deployed worldwide. These effects become especially profound if the equilibrium price is struck at a point on the market supply curve that is 'steep'. In such circumstances, small increases in output by Australian producers may sharply reduce the market clearing price, and supplant many smaller high-cost mining operations until such time as growth in demand causes prices to increase again. (his emphasis) 1091 The Tribunal does not accept this contention. Access might increase the total number of competitors in the iron ore market. But, in that market, that is not sufficient to admit to the conclusion that there is any potential for those new entrants to further constrain the price or non-price decisions of the incumbent producers. Perhaps if the new entrants were low-cost producers able to supply a significant quantity of high-grade iron ore (ie a significant proportion of the total market), the incumbent producers might be further constrained in their decision-making, but even that is not clear. 1092 FMG's case is that the additional volume will affect competition in a market the annual volume of which is currently in the region of 850mtpa. The price for iron ore is determined by the interaction of supply and demand in the market. We agree with both the NCC and Dr Williams that a relatively small addition to supply - particularly additional supply that will come on line not at one time but over a period of years - cannot affect the global iron ore price. 1093 Those views were endorsed by Professor Hausman in relation to the Mt Newman line with output from Mindy Mindy in mind. He said: "When many firms exist in a market selling a similar product (iron ore) the addition of one more (small) firm is unlikely to affect competition. In the current situation the world iron ore market has hundreds or thousands of firms, where the top ten producers each produce more than 4 times as much iron ore per year as Mindy Mindy is expected to produce. With many firms and Mindy Mindy only increasing output by about 0.3%, no increase in competition will occur. … Mindy Mindy is much too small to cause a significant increase in competition." Professor Hausman derived his percentage from the total value of global iron ore production. When the percentage is recalculated based on the total seaborne trade, it is around 0.6%. At that level, it still makes good his point: no single potential access seeker is likely individually to have any meaningful affect on the global iron ore price.

1094 FMG contends that there is an iron ore tenements market, the dimensions of which are: the right to prospect or explore for iron ore on certain land and the resultant knowledge gained about the value (scale and grade) of any iron ore reserves (the product); the activities of prospecting and exploration by drilling for iron ore and its certification in accordance with the JORC Code of any reserve (the function); and the tenements are those located in the corridor surrounding each line or in the Pilbara region (the geographic area). 1095 There are various types of mining tenements that can be granted under the Mining Act, ranging from prospecting and exploration licences to mining leases: see the description in Chapter 2.8. FMG's alleged market includes most tenements, other than mining leases. 1096 The principal reason for contending that there is an iron ore tenements market is that there are transactions in tenements in the Pilbara, including those which have Ministerial consent to prospect and explore for iron ore. The transactions are of various kinds, some direct and some indirect. A direct transaction includes a transfer of the whole or a part of a licence (eg a farm-in or sublease agreement or a joint venture arrangement). Examples of indirect transactions include the acquisition of a significant shareholding in a corporation which holds a tenement. As Mr Sundakov said, it is important to consider the economic content of a transaction, not its legal form. 1097 As regards the number of participants in this putative market, it is convenient to look at what has occurred since 2000. Prior to 2000, five companies (including BHPB and RTIO, which acquired many of their holdings by transfer) held the majority of the iron ore tenements in the Pilbara. Since then there has been a significant change in holdings, coinciding with the worldwide increase in demand for iron ore and the consequent rise in its price. 1098 The Tribunal has evidence from Mr Weekes on the frequency of acquisitions of tenements (other than by RTIO) only for the approximately three year period from 1 January 2006 to 19 March 2009. This evidence is summarised in the following table: (1) (2) (3) (4) (5) (6) Grants (including pending grants) Outright purchase Farm-in/Joint Venture Other Total No. of licences 490 56 24 8 578 Percentage 84.8% 9.6% 4.2% 1.4% 100%

1099 The table includes transactions still pending as at the date of compilation of the database (16 January 2009). Regarding one transaction, it was uncertain whether the transaction in fact went ahead, regarding another transaction, only the iron ore rights were sold (rights over other minerals were retained by the vendor), and regarding another, only 80% of ownership was sold, meaning that, strictly speaking, it should not have been included in column (2). Column (2) includes the sale of nine granted tenements which were not endorsed for iron ore. 1100 Granted tenements were included if they were endorsed for iron ore or if other (external) evidence existed showing that the tenement owner had an interest in iron ore. Some mining rights for iron ore were granted by specific state legislation and these were also included. Pending iron ore tenements cannot be distinguished from those relating to other minerals. 1101 Mr Weekes confined his evidence to tenements within the vicinity of the RTIO railway system, which he described as a corridor extending approximately 50km from RTIO's rail system facing east and 75km facing west. This covers under half of the Pilbara. There is no evidence of the volume of tenement activity in the vicinity of the BHPB lines. No doubt there have been a number of dealings. We are not able to assess to what extent. 1102 One can see what has happened as a result of the dealing from the following diagram which shows the relative surface area covered by licences held by the major companies in the Pilbara as at 30 June 2007. The Tribunal does not have the evidence to bring the diagram up to date. 1103 The companies included in the diagram are those owning at least one granted (or "live") iron ore endorsed tenement as at 30 June 2007. All the tenements (live and pending) of each company are counted in calculating the surface area which that company owned, regardless of whether the tenement was endorsed for iron ore or not. 1104 Mr Clegg, who consults to the mining industry and whose clients include both senior and junior miners, said that only about two per cent of tenements are converted into mining leases. 1105 On the basis of this evidence, the following trends are evident: (1) In 1998, there were only three major tenement holders - BHPB, RTIO and Hancock Prospecting; (2) Notwithstanding the significant increase in issued tenements in the last 12 years, the areas held by BHPB, RTIO and Hancock Prospecting have not changed significantly; and (3) On the other hand, FMG plus a large number of junior miners have acquired a significant number of tenements in respect of large areas in the Pilbara. 1106 Mr Harmsworth has analysed the expenditure on tenements based on figures from the Mineral Titles Online (MTO) database. The result of his analysis (as corrected) is set out in the following table: Pilbara Iron Ore MTO Exploration Expenditure by Group Year BHPBIO RTIO Hancock FMG Juniors Total $ $ $ $ $ $ 2003-04 1,341,957 5,661,633 974,560 137,609 465,936 8,581,695

1107 The figures are not precise. First, they do not incorporate expenditure required by State agreement mineral leases. Second, they include expenditure which is not related to iron ore exploration. 1108 Does the evidence support the existence of an iron ore tenements market? The sole proponent of the contrary view is Dr Williams. He said that iron ore tenements should only be considered as a separate tenements market from the iron ore market if the activities involved in prospecting, initial exploration and development are undertaken by different firms from those that subsequently mine and market iron ore. He explained that his principal problem with the notion of a separate market: is that no business sees themselves as doing that; no business considers itself engaged only in the activity of prospecting or exploring for iron ore within the neighbourhood of the railway lines to which access is sought. Now, I think, as economists, we should take a lot of notice of the way businesses organise their business. If a business doesn't confine itself to the activity of exploring … we should ask seriously: why doesn't a business do that; why is it not a sensible business strategy to confine your activities to this particular thing? … If no business organises itself to do just that we really have to ask why; and the reason almost certainly is because there are synergies associated with undertaking this activity along with other activities. So to define a market to confine our competition analysis to an activity across which there are very likely to be strong synergies is often to conduct a very misleading competition analysis. 1109 To test Dr Williams' view, it is necessary to leave out of account prospecting and exploration licences that lack ministerial consent to explore for iron ore. Those licences do not have the characteristics which FMG relies upon to define the relevant product market. Thus, a key piece of empirical evidence upon which a market may be found to exist is that between 1 January 2006 and 19 March 2009, there were at least 88 (but probably a few more) transactions in tenements other than by grant. 1110 Another piece of evidence is the assumption (we think properly made) that when an exploration licence is granted, the holder has in mind the possibility that it might deal with its interest in the licence, at least by way of subdivision of the interest such as, for example, by a partial sale, joint venture or a farm-in arrangement. The assumption is properly made because of the frequency of those arrangements in the Pilbara and elsewhere. It may also be assumed that the value of the interest depends upon the value of the resource, which is itself affected by the world price for iron ore. 1111 In addition, there is evidence that mining companies explore tenements for the purpose of their exploitation by disposition rather than by mining. Examples include Cape Lambert Iron Ore Ltd, Chrysalis Resources Ltd and AusQuest Limited. 1112 Cape Lambert Iron Ore Ltd is a listed company. In its Annual Report for 2009, the heading 'Principal Activities' states, "[t]he principal activity of the Consolidated Entity during the financial year was mineral investment, exploration and evaluation". The following statement appears under the heading "Review of Operations": The Board intends to continue to follow its strategy of acquiring and investing in undervalued minerals assets and adding value through a hands on approach to management, exploration and evaluation. This approach will enable Cape Lambert to receive a significant return on the divestment of these assets. The Board intends to distribute any surplus cash generated from successful divestments to the Company's shareholders. 1113 Chrysalis Resources Ltd is a listed company. In its Annual Report for the year ended 30 June 2009, the Chairman described the company as being "incorporated on 13 June 2007 for the purpose of acquiring and exploring prospective mineral projects primarily focussed on gold, iron, uranium and nickel". The Report also described the company's principal activity as "exploration for minerals". 1114 The company has been active in acquiring and exploring interests across a range of mineral types. The company's interests include: · an 80% interest in Pioneer Gold Project and a 100% interest in Sheeza Gold Project; · West Angeles Iron Ore Project, Isabella North Iron Ore Project; · Gregory Isabella Uranium Project and Windarra Nickel Sulphide Project; and · new acquisitions: Youanmi Uranium Project, Spring Creek Iron Ore Project and further iron ore tenements in the Gregory Range. 1115 Notes to the financial statements disclose that the company's exploration expenditure is to be capitalised, provided the rights to the tenure are current. The expenditure is expected to be recouped through successful development of areas of interest, or by sale. 1116 AusQuest is a listed company. An extract from the company's website describes the company as "an exploration company targeting new deposits and mineral provinces with major ore potential and the ability to significantly enhance shareholder wealth". 1117 What these dealings in tenements and the statements of the mining companies show is that there is both a demand for tenements and sellers willing to meet that demand, including sellers who do not carry out mining operations. It is difficult in those circumstances to deny that there is a functionally distinct market for iron ore tenements. We agree with the NCC's contention that it is sufficient that companies sell tenements, whatever their intention: Mt Newman Final Recommendation, 7.66-7.67. 1118 The geographic scope of the market must be determined. There is little detailed evidence on the subject. FMG submits that there is a separate market around each line. Examining a map of tenements in the Pilbara, it is obvious that tenement purchasers (whether outright or partial purchasers) have not confined themselves to buying tenements around a specific railway line. This provides strong evidence that a monopoly seller of tenements around a specific line could not profitably apply a price increase because buyers would substitute to other tenements. Further, a monopsony buyer of tenements around a specific line could not profitably decrease price because sellers would easily find a purchaser outside this geographical area. 1119 Most of the experts accept that the market for tenements is at least Pilbara-wide. Dr Fitzgerald supported a global market and pointed to the prevalence of international investors in joint venture arrangements. By the same token, many investors in tenements only participate in Australia. Further, as Mr Houston pointed out, differences in the scale and quality of resources, and different regulatory requirements and business environments, mean that businesses most likely characterise their operations on a region-by-region basis, rather than a global basis. We believe that the market is most likely Pilbara wide, and not global for the reasons given by Mr Houston. 1120 It is now necessary to consider whether the tenements market is already effectively competitive. We think it clear that this is an effectively competitive market. There is no evidence that any seller has market power. If FMG continues to acquire tenements as it has in the past six or seven years, it may be open to conclude that it has acquired power in the tenements market. We do not think that position has yet been reached. None of the evidence suggests that RTIO or BHPB has market power by reason of it owning a railway line. Even if BHPB or RTIO had some power in the tenements market, FMG would act as a constraint on that power. In addition, there are several juniors that purchase tenements, and do so in such volumes as would prevent BHPB, RTIO or FMG from exercising any market power. 1121 Notwithstanding this conclusion, we will, as before, consider whether access to any one railway (not access to four railways, it should again be emphasised) will further promote competition in the tenements market. The two bases upon which it could be said that competition will increase are first, access to rail would encourage tenement holders to incur further expenditure in exploration and so improve what is known about the resource or second, if the quantity of tenements sold increases. Either outcome would result in an increase in competition, because it could produce a better quality or a greater quantity of traded tenements. 1122 The argument against access to any one line promoting competition is as follows. What we have witnessed in the last nine years is a significant growth in exploration expenditure. That expenditure is running parallel with (and no doubt directly caused by) the increase in the worldwide demand for, and therefore price of, iron ore. That demand is projected to continue for the next 10 to 20 years, as depicted in the following graph. 1123 The increase in the global price of iron ore can be seen in the next table, which gives the price (in US cents per dry long tonne unit (DLTU) - expressed in 2007 US dollars) of Hamersley's fines. The data is derived from the Annual Outlook for Iron Ore (2008). The price of fines constitutes the benchmark price in iron ore products, and Hamersley's fines can be taken as representative. Global price of Hamersley fines

1125 No doubt the anticipated increase in world demand for iron ore and the projected increase in price will lead to new tenements coming online and old tenements being further explored in hopes of upgrading the status of a resource, thus increasing its value and making it more "sellable". The question is whether access will add anything to this incentive. 1126 The NCC contends that potential purchasers of iron ore tenements will have less incentive to purchase a tenement if there is no viable means of transporting the mined ore for processing or export: Robe Final Determination at 4.80. The NCC also argues that without access to rail infrastructure services, tenement owners would have a reduced ability to develop their tenements because returns would be less without access to rail transport and tenement owners would be less likely to be able to raise the funding necessary to develop a deposit. 1127 It may be accepted, as has been put by witnesses such as Mr Young from BC Iron and Mr Richards from Brockman, that it is not viable for the owner of a tenement with a small resource to construct its own rail infrastructure and to exploit the iron ore deposits without gaining access to rail infrastructure. Hence, if there is access, there will be, at least for those tenement owners, a greater incentive to spend more on exploration. 1128 It may also be accepted that it is possible that access to any one line will lead to an increase in tenement prices and thus an increase in the incentives to increase expenditure and tenement development, and bring more tenements online in the vicinity of that line. It is difficult to avoid this conclusion without rejecting the evidence of Mr Young and Mr Richards. This we are not prepared to do. 1129 Still, when we have regard to what would happen in a future without access, it is not possible to conclude that access to any one line would promote a material increase in competition. What stands in the way of that conclusion is the lines that will likely be constructed if there be no declaration (ie the Aquila line, the Dixon and Kennedy lines, and the spur to the Chichester line). With the construction of these lines, alternative rail facilities will be available for (1) users of the Mt Newman service (via a spur from Marillana to the Chichester line); (2) for a significant proportion of users of the Robe line (on the Aquila line); and (3) for some users of the Hamersley service (particularly those located near the Solomon area and the proposed Dixon line). 1130 We accept that not all tenements in the vicinity of the Robe line and Hamersley line will be able to access alternative rail options. But we are not satisfied that providing these tenements with rail options would promote a material increase in competition in the Pilbara-wide tenements market. In this regard, we note that (1) the market is already effectively competitive, with levels of activity likely to increase in any event; (2) the majority of tenements in the vicinity of the Robe line and the Hamersley line are likely to have alternative rail options without access; and (3) many of the tenements in the Pilbara-wide market would be too far away from the Hamersley line or the Robe line (as the case may be) to be affected by access to either line. 1131 In the case of the Goldsworthy line, it has the fewest tenements surrounding it, many of which are within trucking distance to the port. Access to that line will have only a minor effect on the tenement market.

1132 The third dependent market proposed by FMG is for rail haulage services for iron ore on each line (the product market), the geographic market being the vicinity of each line. The existence of a separate product market for rail haulage services immediately gives rise to the question whether, in a vertically integrated operation of which rail services are but one part, that part can be treated as functionally discrete for market analysis purposes. This also raises the question whether there can be a market when there are no market transactions and, on the supply side, the putative seller is unwilling to enter into any transaction. 1133 It is convenient to deal first with the second issue: Can there be a market for a product or service absent any market-based dealings in that product or service? In QCMA (at 517), the Tribunal accepted that a market was delineated by "the field of both actual and potential transactions between buyers and sellers among whom there can be strong substitution, at least in the long run" (emphasis added). We think that the statement that market participants include both actual and potential buyers and sellers is uncontroversial. The existence of potential buyers and sellers can in many cases affect the price and non-price decisions of actual buyers and sellers. In those cases, it would be wrong to exclude potential buyers and sellers from consideration. 1134 Here, the issue comes about because there is no actual buyer or seller of the particular service. All that exists are potential buyers and sellers. Is that sufficient to conclude there is a market? There are many economists (probably most) who would answer this question in the negative. Professor Hausman put the point quite starkly. He said: "For a market to exist there must be both willing sellers and willing buyers. If no willing sellers exist, a market does not exist." Professor Hausman went on to differentiate between a "market outcome" and a "command economy", where resource allocation and firm conduct is required to follow a state's economic dictate. In that state of affairs, a market does not exist except by appropriation (ie by modifying the rights of an individual by the exercise by the state of its sovereign power). That, he said, will not create an economic market. 1135 The High Court has laid down that in some circumstances, a relevant market will exist even though there is an absence of transactions between a willing seller and a willing buyer. In Queensland Wire Industries Proprietary Limited v The Broken Hill Proprietary Company Limited (1988) 167 CLR 177, the question was whether BHPB's refusal to supply wholesalers with products known as "Y-bars" contravened s 46 of the Trade Practices Act. That section proscribed a firm from taking advantage of its substantial power in a relevant market. The Full Federal Court had held that because there never had been a market for Y-bar, it was not possible to attract the operation of s 46. The High Court disagreed. Deane J said (at 196): [A] market can exist if there be the potential for close competition even though none in fact exists. A market will continue to exist even though dealings in it be temporarily dormant or suspended. Indeed, for the purposes of the Act, a market may exist for particular existing goods at a particular level if there exists a demand for (and the potential for competition between traders in) such goods at that level, notwithstanding there is no supplier of, nor trade in, those goods at a given time - because, for example, one party is unwilling to enter any transaction at the price or on the conditions set by the other. Dawson J said (at 200): "But the existence or non-existence of sales of a product cannot conclude whether a market exists or not. It must be sufficient to constitute a market that there is a product for exchange, regardless of whether exchange or negotiation for exchange has actually taken place." Toohey J said (at 211-212): "But even if it was necessary, the absence of existing buyers does not mean that there is no market for Y-bar. It would be a curious consequence if the offering by B.H.P. of a limited supply of Y-bar established a market for that product but the withholding of supply altogether meant that there was no market." 1136 The consequence of Queensland Wire is that Professor Hausman's position, no doubt correct from an economist's standpoint, cannot survive the established legal framework. 1137 There is, in any event, another point to make. Professor Hausman's underlying premise is that the incumbents are properly described as "unwilling sellers". We do not accept that premise. Under the State agreements by which they are each bound and, we would add, to which they and their predecessors voluntarily submitted themselves, BHPB and RTIO agreed to provide haulage services provided that to do so would not unduly prejudice or interfere with their operations. Part IIIA only operates to require access if access will not unduly interfere with the incumbent's operations, although the precise language of s 44W(1)(a) is in somewhat different terms. If it will not cause undue interference, each of BHPB and RTIO cannot properly be described as unwilling sellers. 1138 Turning now to the question whether there is a separate functional market, it is our view that the evidence of a separate haulage market is made out. First, the various examples of below rail and above rail separation in other regions of Australia demonstrate that a functional separation is possible in some circumstances. 1139 Second, although neither BHPB nor RTIO has so far hauled iron ore for a third party or allowed below rail access, at least BHPB has been in negotiation (at least with [c-i-c]) to provide a haulage service. While those negotiations have not led to a concluded agreement, there is no suggestion that the negotiations were an artifice. Not only is there no suggestion that BHPB engaged in sham negotiations, senior BHPB personnel were involved. Assuming that BHPB is rational - that is, profit motivated - it would only have engaged in those negotiations if it believed it could make a profit from providing a haulage service. Likewise, assuming that [c-i-c] is rational, it would only have negotiated with BHPB if outsourcing haulage was regarded as more profitable that building its own line. 1140 Third, there are willing buyers and hence there is demand for the service. Demand comes from FMG (which would be a more-than-willing provider of haulage services on some lines) and, at least, the various juniors (including Atlas and BC Iron) with whom FMG has agreed that it would provide, or has entered into a memorandum of understanding by which it has undertaken to enter into negotiations to provide, haulage services. 1141 Fourth, once again assuming that it is appropriate to treat a haulage market as functionally separate from the incumbent's other activities, the only element of an economist's definition of a market that is missing, actual transactions, is missing because the potential supplier, a monopolist, has power to withhold supply. 1142 Fifth, we do not believe that transaction costs are so large that no actual transactions will ever take place. BHPB and RTIO point to large transaction costs that include the costs to compensate the incumbents for any lost production, administration costs, costs of negotiating and enforcing contracts and the cost of arbitrating decisions. [c-i-c] and BHPB's negotiations have demonstrated that neither party believed this to be the case. Dr Williams points to "hold up" costs: that BHPB and RTIO would be held up from making an efficient investment in track infrastructure if it expects that once it has made the investment, the access seeker will refuse to pay more for access. But the hold up costs need not be so significant. Assume that access is sought for purposes of transporting only a relatively low volume of iron ore (eg 5mtpa). The evidence is that this volume could be accommodated on all lines with little disruption to the incumbent's activities and little, or no, displacement of the incumbent's throughput. Thus, we disagree with the factual premise upon which Dr Williams expresses his opinion. It may be accepted that much higher volumes of third party demand could result in hold up costs. Dr Williams' conclusion would, in our opinion, only be valid if in all circumstances the transaction costs were so high as to prevent access being taken up. 1143 So far as the product dimension of the haulage market is concerned, we have explained that trucking is not generally an effective substitute for rail haulage: see Chapter 7.3.3. Whether the product market extends to haulage on other lines will depend on the nature of the other line. Clearly, another haulage service that ends at a different port is unlikely to be a substitute for many miners. On the other hand, haulage services on the Mt Newman line and Chichester lines (including the proposed Marillana spur) are effective substitutes because both would originate and terminate within close proximity to one another. 1144 The geographic market for haulage is best described as a corridor around each line which is the subject of an all-points application, and a circumference around the access point of the line the subject of a point-to-point application. How wide the corridor is (and what length the radius should be) depends on the distance a miner can viably build a spur or truck ore to the line. It is impossible to give an exact distance or radius for any line. 1145 The next question is: Will access to a line promote a material increase in competition? Clearly enough the market is presently not effectively competitive as the only potential provider of haulage services, the incumbent, is able to exert market power by refusing to provide haulage. In Services Sydney (at [135]), the Tribunal said that criterion (a) applied to a market in which there was no competition because of the vertically integrated monopolist. The Tribunal also said that "the facilitation of any competition in any such market is of significance." In other words, the facilitation of any competition closer to effective competition will satisfy criterion (a), provided the movement is not trivial. 1146 Here access to the lines will enable FMG, and possibly other firms if they choose, to provide a haulage service. At least mining companies will have the choice of paying the provider's asking price. Under the status quo, they have no such choice. We regard the choice as significant. It constitutes an increase in competition even if the supplier of haulage (eg FMG) is a monopolist. Though it may be a monopolist, it may not be able to exercise monopoly power. Once haulage is provided by a third party, the incumbent may be encouraged also to provide haulage services to junior miners. Even if the incumbent does not provide a service to junior miners, it would constrain the third party haulier because of its potential to provide haulage services. 1147 Is this any different from the position that would obtain if there were no access to the lines? The answer is "yes" in respect of three lines, the Goldsworthy, Hamersley and Robe lines, and "no" in respect of one line, the Mt Newman line. 1148 It is necessary to explain why we have reached these answers. Once again they are dictated by the effect on the rail haulage market if the likelihood is that a substitute facility will be constructed if there is no access to the service. In the case of the Mt Newman service, the Chichester line, extended by the Marillana spur, will offer a very similar service to the Mt Newman service. The Marillana spur will likely not be built if the Mt Newman service is declared. Thus, the situation in the access factual and no access counterfactual is the same: in either event there will be a single line which provides the service. In the case of the Goldsworthy line there will be no substitute facility. It follows that the future with access will promote competition whereas the future without access will not. 1149 The same is true as regards the Hamersley line. If there is no access the new facility will be only a partial substitute for the Hamersley line. We say that it will be only a partial substitute because the new Dixon line (from the Solomon area) would not enable haulage services to be provided to significant parts of the Hamersley haulage market, particularly to the tenements in areas to the south and east of the Solomon area. 1150 The most difficult haulage market to assess, from a competition perspective, is the Robe market. It is worth restating some of our earlier observations about this line. 1151 The first observation is that some tenements in the vicinity of the Robe may have alternative rail options. Some of these options will exist regardless of whether the Robe service is accessed. For example, Aquila will build a line from its projects in the West Pilbara to at least near Mesa J (in order to at least connect with the Robe line). Likewise, some tenements in the vicinity of the Robe line might also be able to access the Hamersley line or the Dixon line (as the case may be). The only alternative rail option the existence of which might depend on whether there is access to the Robe service, is the section of the Aquila line continuing beyond the Mesa J area to Anketell Point (the Anketell section). If there is no access to the Robe service, that section is likely to be built. Indeed, the Anketell section might well be built even if there is access to the Robe service. 1152 The second observation is that the proposed course of the Anketell section diverges considerably from the course of the Robe line. Much of the section is 50km or more to the north of the Robe line. This means that for tenements close to the Robe line, it may not be viable to truck longer distances to connect with the Anketell section, and vice versa. 1153 How, then, does this somewhat complicated factual situation translate into the language of criterion (a)? 1154 Let us begin with the definition of the market. It is clear that some (but certainly not all) potential consumers of the Robe service may have alternative rail options (eg the Aquila line or the Hamersley/Dixon line). The question is whether those services are within the same market. The way to test for this is to apply the hypothetical monopolist test: Would a hypothetical monopolist be able to profitably implement a price increase on the Robe service? Whether this is possible will depend on the extent to which consumers can substitute to other lines. We do not have enough data to determine whether the product market includes the alternative lines. Nonetheless, we think that, regardless of whether the market definition incorporates several rail alternatives or is confined to the Robe service, it is possible to conclude that access to the Robe service would promote a material increase in competition in the market. 1155 It is convenient to consider first the position if the product market includes both the Robe service and alternatives. It is, we think, clear that access would promote competition in this market for various reasons. First, the only identified source of demand for the Robe service, at this stage, is Aquila. Access to the Robe line creates choice for Aquila. It may choose haulage on the Robe line, or it may choose to have its ore hauled on its own line, if it elects to build the Anketell section of the Aquila line. It can be no worse off if it has access to the Robe line, but it may be better off. 1156 Second, if there are other, as yet unidentified, consumers they may also benefit from access. Some, who could access haulage on both the Robe line and other lines, would benefit from the choice between two or more suppliers. Others who could not access rail would clearly benefit from having a rail option available. 1157 The only consumers who might be worse off if there were access to the Robe service are those close to the Anketell section of the Aquila line. If there is access, the Anketell section may not be built. However, we think that those consumers are likely to be limited in number. They are relatively close to the coast, so trucking direct to port may be the preferable option in any event. Many would also be close to Cape Preston and therefore not interested in the Robe service. Further, a significant number of tenements in the vicinity of the proposed Anketell section are large magnetite projects, which have already been ruled out as potential sources of demand for the Robe service: see Chapter 15.6.2.2. And ultimately, the Anketell section may be built even if the Robe service is declared, in which case no consumers would be worse off. 1158 All in all, we are satisfied that access to the Robe service would promote a material increase in competition in the market for haulage broadly defined on the Robe service. We think the majority of potential consumers of the service would benefit from access, thereby increasing the likelihood of a vibrant, competitive haulage market emerging in the Robe area. 1159 An alternative scenario is that the market is defined narrowly to only include the Robe service (and not encompass alternative rail facilities). Here too, it is clear that access to the Robe service would promote a material increase in competition. By definition, in this scenario the vast majority of consumers would not have an effective substitute to the Robe service (because otherwise, they could defeat a price increase). In the access scenario, there would be a competitive market for the service; in the no access scenario, there would be no market transactions. In saying this, we recognise that a small proportion of consumers (ie too small to defeat a price increase) would have alternative rail options in different markets. They might or might not take up access on the Robe line. But for most consumers of the Robe service, their only transport option would be the Robe service. We are satisfied that creating a market for at least those miners (and, indeed, potentially for other miners who elect to use the Robe service rather than some alternative) would promote a material increase in competition.

1160 Criterion (f) requires the Tribunal to be satisfied that "access (or increased access) is not contrary to the public interest". The criterion is expressed in the negative, ie it is not necessary for the Tribunal to positively be satisfied that access is in the public interest. 1161 The TPA provides no definition of the expression "public interest". Dr Fitzgerald says that what must be considered is the welfare, particularly the economic welfare, of the Australian community as a whole. This is in line with the views of the NCC. Professor Fels put it that the Tribunal should consider whether the costs of access do not outweigh the benefits, though it must be borne in mind that, apart from the fact that the assessment may at best be robust, the object is to achieve not a state of perfection but a better outcome. We are in broad agreement with these views, but with one important qualification. On close analysis it may be that access will be manifestly unjust to a section of the community while, at the same time, benefiting the community as a whole. In that circumstance access may nevertheless be contrary to the public interest. 1162 In the past the Tribunal has emphasised that criterion (f) should not be used to call into question the results obtained by the application of the earlier criteria. Rather, criterion (f) is concerned with other considerations, not caught by earlier criteria, which bear upon the public interest: see eg Virgin Blue at [587]-[588]. While the results of earlier criteria cannot be questioned, they are not to be ignored. The satisfaction of criteria (a) and (b) indicates that benefits will occur from access. Those benefits, and other benefits not considered under earlier criteria, as well as the costs of access, must be taken into account under criterion (f). 1163 The Tribunal also has a discretion whether or not to declare a service. In the past the Tribunal has said that, in light of the detailed considerations required to be undertaken to determine whether the statutory criteria have been satisfied, the discretion is extremely limited. This is not consistent with Sydney Airport (No 2). There the Full Court observed the discretion "may be affected by a wide range of considerations of a commercial, economic or other character not squarely raised by, nor falling within, the necessary preconditions in s 44H(4)": 155 FCR at 137. That is to say, the discretion is a very broad one. 1164 The relationship between criterion (f) and the discretion is important. Criterion (f) requires the Tribunal to consider whether "access" is contrary to the public interest. The NCC points out that if the Full Court's decision in Sydney Airport (No 2) applies to this criterion (and it suggests that, for consistency of approach, it must apply), the only issue to be considered is the impact on public welfare of the mere fact that a third party has access to the service on reasonable terms and conditions. The NCC says (and no party disagrees) that it is necessary for there to be a more detailed inquiry to appreciate the effect on public welfare of both a declaration and access under Part IIIA, but that inquiry is only relevant to the Tribunal's discretion. 1165 It is, we think, impossible to disagree with the NCC. The result, however, is problematic. Mere access to the service may, in theory, have some consequences, albeit their measurement is impossible. In the real world many consequences will only arise if access is actually taken up. And the nature and extent of those consequences will depend upon the extent to which access is taken up. Moreover, whether, and the extent to which, access is taken up will be a reflection of the terms upon which access is obtained. 1166 The Sydney Airport (No 2) approach does not permit a detailed consideration of the likely terms of access that might be imposed under Part IIIA. It does allow for the assumption that access will be on reasonable terms. So, we will apply the Sydney Airport (No 2) approach to criterion (f) by making some broad assumptions about the nature of access "on reasonable terms and conditions". Issues of substitutability may also be taken into account. We will also take into account the effect of access being taken up: see Chapter 17.4.3. Issues which arise because of the specific operation of Part IIIA will be considered under the discretion. 1167 Our approach could have this consequence. There may be a finding that access simpliciter is in the public interest (or, more correctly, that access is not contrary to the public interest), but taking into account the effects of both a declaration and access under Part IIIA, a different conclusion might be reached. The limitations of the Sydney Airport (No 2) approach highlight the importance of the discretion to withhold making a declaration if, in the Tribunal's opinion, a declaration is not in the community's best interests. Indeed, this is consistent with the Full Federal Court's view about the wide scope of the discretion. 1168 The factors which the Tribunal proposes to take into account will not be confined to strict cost-benefit issues. The Tribunal will also have regard to broader issues concerning social welfare and equity, and the interests of consumers. Where possible the Tribunal will distinguish between criterion (f) issues and discretionary factors, although the distinction is not always easy to draw. 1169 While many factors for and against a declaration and access will be discussed, their impact will, in most cases, be difficult, if not impossible, to quantify. A good example is dynamic efficiency considerations which involve speculation about the future path of technology in the iron ore industry or rail industry. In part the difficulty of quantification arises because many of the alleged costs and benefits of access are esoteric or qualitative in nature. Another reason is that many of the alleged costs and benefits depend upon the occurrence of future events which are necessarily uncertain. Hence, the cost-benefit analysis that the Tribunal performs will not be purely quantitative, and will have significant qualitative aspects. 1170 In weighing up the costs and benefits, and in the exercise of its discretion, the Tribunal will keep in mind that Pt IIIA was intended by the Hilmer Committee to protect "the legitimate interests of the owner of the facility": Hilmer Report at 252. Thus, Part IIIA specifically protects the incumbent's reasonably anticipated requirements for the facility (s 44W(1)(a)) and requires the ACCC during an access arbitration to take into account the legitimate business interests of the provider (s 44X(1)(a)). Clause 4(i)(ii) of the Competition Principles Agreement also requires the ACCC to take into account the costs to the owner of providing access other than costs associated with losses arising from increased competition in upstream or downstream markets. But when giving consideration to the owner's interests, it is necessary to distinguish between its legitimate interests and other, potentially anti-competitive, interests. 1171 This brings us to a point already mentioned in these reasons, namely, what issues may properly be taken into account at the declaration stage of the two-stage process. FMG and the NCC suggest that many of the alleged costs raised by the incumbents should only be considered at stage two, the negotiation (and if necessary) arbitration of an access dispute. There are two premises underlying this suggestion. The first is that Part IIIA makes provision for the protection of the incumbent's legitimate business interests so it cannot be assumed that those interests will be prejudiced as a result of access. The second premise is that whether there is a factual basis for many of the incumbents' concerns will only become apparent at stage two. This is for the reason that the impact of access on the incumbent's operations will depend on the nature and extent of the access sought. Thus, according to the NCC, it is necessary to distinguish between those costs caused by access per se and those caused by the type of access sought at stage two. The NCC does not, however, make clear what costs, if any, are relevant to criterion (f) under this approach. 1172 We think that the approach suggested by FMG and the NCC gives too limited a role to the decision-maker at the declaration stage. It is true that, when it comes to the kind of access which might be sought in the event of a declaration, nothing is certain. But criterion (f) requires the Tribunal to consider the consequences of access (even if only in the narrow sense described in Sydney Airport (No 2)). This necessarily involves some speculation. In any event, there will be some consequences which, while not certain, are likely to occur. It is inappropriate to simply ignore those consequences. The better view, we think, is that the Tribunal should consider consequences that are likely to arise as a result of access, giving them a weight that pays regard to their degree of likelihood. 1173 As regards the proposition that Part IIIA provides some protection for the legitimate interests of the incumbent, the Tribunal makes two observations. First, there may be costs which will be borne by an access seeker in order to protect the incumbent's legitimate interests. That the incumbent does not bear those costs does not make them irrelevant to the declaration decision. We are, after all, concerned with costs from society's perspective. The second observation is that Part IIIA does not provide the incumbent with certainty that its legitimate business interests will be protected: see generally Chapter 10. We think this is relevant both to the public interest criterion and to discretion. 1174 The final introductory comment we wish to make is that, as regards criterion (f) and the discretion, the Tribunal considers that its role at the declaration stage is to be concerned with the "big picture". At stage two the minutiae of terms of access will be dealt with. When considering the public interest and exercising its discretion at the declaration stage, the Tribunal is, to a significant extent, concerned with broad issues of policy that are unlikely to be dealt with at stage two.

1175 There is a great deal of literature regarding the regulation of railways. The discussion tends to focus on the traditional "monolithic model" of organising a railway, where the State or a single firm owns the infrastructure and controls all operating and marketing functions. Professor Willig, in an article he co-authored with Professor Ioannis Kessides, wrote: "[This structure is] production-oriented, is unresponsive to market demands for services, and is hierarchical (if not bloated) in organizational architecture": Ioannis Kessides and Robert Willig, "Restructuring Regulation of the Rail Industry for the Public Interest" (Policy Research Working Paper No 1506, World Bank Private Sector Development Department, 2005) 21. The authors went on to say of the monolithic model: "No-one would deliberately choose it for the public interest". 1176 Unsurprisingly, the response by governments around the world to the monolithic model has been to attempt to "unbundle" above and below rail services, to separate the naturally monopolistic segment from the competitive ones, and to allow access to the natural monopoly infrastructure. Whether those attempts have been successful is debatable. Clearly problems have arisen. But even a cursory examination of the voluminous literature on the regulation of railways shows it is quite wrong to immediately assume that regulation has been a failure. Louis Thompson, a railway adviser to the World Bank and one of the world's leading experts on railway regulation, makes the very good point that one should be wary of criticism of regulation because the comparison often made is between the nature of a particular reform and perfection rather than with what went before: see eg Louis Thompson and Karim-Kacques Budin, "Final AICCF Directions of Railway Reform" (Paper presented at the OECD Presentation, Beijing, January 2002) 12. 1177 In this case the Tribunal has been provided with considerable evidence about the experience of rail access regimes in other jurisdictions, especially Europe, the UK, North America and the east coast of Australia. The evidence has been tendered to show the problems which have arisen under those regimes, and to invite the Tribunal to conclude that the same, or similar, problems will arise if the lines are declared. 1178 The Tribunal has considered all of the evidence regarding other access regimes but found it generally to be of little assistance. For one thing, while the rail industry has been one of the most heavily regulated sectors, the models of regulation differ markedly from country to country and, within Australia, from state to state. There are differences in forms of ownership: some models have public ownership, some have private ownership, and some have either public or private ownership with a concession or franchise awarded to users. There are also structural differences. The structure might be integral (where the infrastructure and all operations are unified under a single control), dominant (where a dominant operator controls the infrastructure with separate minority users as tenants) or where there is complete separation between ownership and use. There are differences in the purpose to which rail is being put (eg for haulage of minerals, coal freight or passengers). The existence of these differences makes very difficult, if not impossible, the task of drawing valid conclusions from the experience of other regimes. 1179 Second, we note that, in recent times, inquiries have been conducted into a number of the Australian regimes, recommending the introduction of new measures. It is perhaps too early to determine whether these measures, if implemented, will be successful. If successful, they could subvert the utility of reliance on the regime to show that access will necessarily be contrary to the public interest. They might also be a useful precedent to follow in stage two. 1180 Third, and most importantly, we think that the probative value to be obtained from an in-depth inquiry into the failings of other access regimes is far outweighed by the time and effort required to undertake the inquiry. What does appear is that we have been presented with a very one sided picture of other access regimes. Our own necessarily limited inquiries show there are other views. Even some of the experts called by the incumbents have, in published papers, favoured reform of the railway sector through regimes not far removed from that created by Part IIIA: eg Janusz Ordover and Russell Pittman, "Restructuring the Railway for Competition" (Paper presented at the OECD/World Bank Conference on Competition and Regulation in Network Infrastructure Industries, Budapest, 1994). To deal fully with the success or failure of other regimes is a task that, in the experience of those who have attempted to undertake it, takes an inordinate time to complete. It would be a significant diversion from the path contemplated by Part IIIA. 1181 Finally on this point, we are satisfied that, having regard to the voluminous evidence, only some of which is directly relevant to the situation in the Pilbara, we are able to draw appropriate, if very general, conclusions on the risks as well as the benefits of access. 1182 There is one important exception to the Tribunal's position regarding evidence of other access regimes. The Tribunal will pay some regard to the State access regime which applies to the Chichester line

1183 Ordinarily an inutility argument would fall under discretion. Here aspects of the argument merge with a consideration of the public interest. 1184 The argument the incumbents make is there is no "real" demand for access on any line, so there is no point in making a declaration. 1185 First, BHPB says that FMG is the only potential source of demand for access to BHPB's lines but FMG will not take it up. We do not agree with the first part of the argument. We have identified various potential sources of demand for BHPB's lines and for RTIO's lines: see generally Chapter 15.5. Even if FMG were not to take up access following a declaration (and we accept it has no formal business plans to do so), other junior miners, either independently or as a group, might. 1186 Second, reliance is placed on the fact that there are no port facilities for third parties to use. But this is only a temporary situation. The evidence shows that, with the involvement of the Western Australian government, new port facilities will be developed. Even if that takes time, smaller junior miners are likely to obtain access to the port facilities that will be built by or on behalf of the large miners. For example, Atlas has some of its ore processed through FMG's facilities at Anderson Point. 1187 Third, it is argued that (1) capital expenditure associated with rail is so high as to make it infeasible for small junior miners to take up access; and (2) larger miners can afford to build their own rail facilities. Even if these premises were correct, it does not follow that there is no utility in declaration. First, it might be more efficient for miners who can afford their own rail facilities to use the incumbents' lines. Second, the evidence suggests that rail may be affordable by smaller junior miners. By way of example, the BRS study for [c-i-c] suggests that rail infrastructure is affordable for small operations if the junior miner considers measures such as leasing trains and using bottom-dumping ore cars. Admittedly, the study excluded certain infrastructure costs but, as Dr Fitzgerald says, relatively low volume mining operations in the Mid-West use rail. 1188 The fourth argument, which is the most difficult, is that the terms of access are likely to be so unattractive as to deter access. The terms which will create the problems are those that will be imposed to protect the incumbents' reasonably anticipated requirements for the use of the lines. 1189 No doubt conditions will be imposed upon a third party to avoid any displacement of the incumbent's throughput. The restrictions or conditions would likely involve either (1) paying for expansions to increase the capacity of the line and/or purchasing additional consists for the incumbent; or (2) requiring the third party to operate flexibly; or (3) a combination of (1) and (2). 1190 Whether these conditions are required, and the cost to the third party of meeting them, will depend on the circumstances in which access is sought. Nonetheless, it is possible to make some broad observations based on the modelling undertaken by TSG. It will be recalled that TSG modelled a number of scenarios involving a third party operating on a highly flexible basis with its trains having lower priority than the incumbent's trains in order to fit in with the incumbent's "run when ready" approach. Even then, the modelling results suggest that the Mt Newman, Hamersley and Robe lines, without further expansions, could not handle the potential demand (other than relatively small levels of demand) without displacing the incumbent's throughput. Hence, a rigid form of third party operation could not realistically be accommodated. And, to meet total demand, it is likely that those lines will need to be expanded, in some cases significantly. 1191 As to expansion costs, it is not possible at this stage to say that it is unrealistic to expect that a third party will be able to pay the costs. The evidence indicates that the cost of expanding a line is likely to be somewhere around $3-4m per km of new track required (although the cost may vary depending on terrain). It is plausible that a third party could meet those costs, particularly if iron ore prices continue to increase over the long run. 1192 Whether it is realistic for a third party to operate as flexibly as may be required is questionable. For example, a third party may be required to queue for extended periods before getting on the incumbent's line. This will likely increase the number of consists a third party requires to move its desired output. As Dr Dallimore said, a third party will require fewer consists on its own dedicated line than when sharing a line with the incumbent. It does not necessarily follow, however, that the cost of extra consists would be so prohibitive so as to render third party operations unrealistic. Once again, the Beyond Rail study suggests otherwise. 1193 One possible approach to deal with the difficulties is for the third party to adopt flexible operating practices. Mr Tapp and Mr Flanagan said that FMG's and Atlas' operations might be made viable by operating larger stockpiles at mine and port to cater for the unpredictability of train movements. But such stockpiling may be prohibitively expensive, particularly given that junior miners producing ore with low demand characteristics will be subject to greater shipping order variability and, hence, would require larger stockpiles. Perhaps a junior miner selling its ore into the spot market might tolerate more uncertainty regarding train availability than a miner selling under supply contracts, given that it may be able to build up stockpiles before selling into the spot market. (This would depend, of course, on the terms of the spot contract and delivery conditions). In this connection, it is significant that there has been substantial growth in spot market activity in recent years. 1194 Mr Hay, manager of rail development within RTIO's rail division, said that because of the constant re-routing and rescheduling of trains on the RTIO network, it is unlikely that RTIO could confidently give more than an hour's notice to a third party that a train path is available. FMG's response that it could operate with such limited notice is, we think, somewhat simplistic. Mr Thomas, General Manager of FMG's rail department, said that it was possible for FMG to operate with 24 hours notice of a train slot window of 4 to 6 hours with 30 minutes notice of the final running time. He acknowledged that such an arrangement would be unfairly onerous. 1195 The overall impression we have is that flexible operations may technically be possible, but in many instances would not be commercially attractive. Still, the evidence is not so strong as to enable us to conclude that no-one would seek access due to the flexibility of operations that would be required. It must be understood that the degree of flexibility required of a third party will vary from line to line and between different points of a line. For example, there is ample capacity on sections of the Goldsworthy line and operational requirements will likely not be onerous.

1196 The benefits which might flow from access must be understood in the context of our conclusion that, if the lines are not declared, there will likely be other railways which could be accessed by third parties. Apart from accessing the Chichester line, railways likely to be built include the Kennedy line, Dixon line and Aquila line. 1197 In Chapter 13.4, we considered the extent to which potential access seekers to the relevant BHPB/RTIO line could access those alternative railways. We observed that in the case of the Hamersley line and, to a lesser extent, the Robe and Goldsworthy lines, there is a risk of tenements being "stranded" (ie without viable transport options) in the absence of access to those lines. Clearly, a benefit of access to the lines is the potential for "unlocking" a stranded deposit. Moreover, if access causes a net increase in the quantity of iron ore exported, there will be the creation of incremental employment, wages and salaries, export revenue and, through taxes, an increase in government revenue which may be redistributed through government expenditure or lower tax rates. The increased economic activity will have flow-on benefits for other parts of the economy. 1198 As we also observed, there might well be tenements which, while not technically stranded without access, would nonetheless benefit from access. For example, without access some projects may be required to rely on trucking, an option which is potentially more expensive than rail haulage and subject to volume limitations. This may mean that a larger project is unable to move its desired throughput due to the limited volumes it can truck. Some projects might be further away from the alternative railway than a BHPB/RTIO line, so accessing the alternative railway may be more expensive. For the miners for whom access is preferable there are benefits from access associated with reducing their costs and potentially increasing their throughput. 1199 In some cases, there may be benefits from avoiding the construction of alternative railways in the first place. In Chapter 15, we observed that for all the lines except the Mt Newman line, there would be significant savings in capital costs by accommodating third party demand on the incumbents' lines rather than building alternative facilities. These savings are an important benefit which needs to be taken into account. 1200 When considering the application of criterion (a) we concluded that access to some lines would promote competition in dependent markets. The extent of this benefit should not be overstated. Access will not promote a material increase in competition in either the global iron ore market or the Pilbara tenements market. It will affect the rail haulage market, but this only (directly) benefits projects in the vicinity of the relevant line. To a large extent, the benefit overlaps with the benefits already discussed about access being a necessity, or at least preferable, for some junior miners. 1201 Another potential benefit is that junior miners using the incumbents' lines may be able to bring forward production earlier than they would if required to wait for the construction of an alternative facility. This is a limited benefit. Many junior miners are not in a position to take advantage of the earlier availability of the incumbent's line. Those wishing to transport ore on the Robe and Hamersley lines must wait until the Anketell Point port facilities are constructed. Those who would transport ore on the Mt Newman and Goldsworthy lines require port capacity to become available at Port Hedland. 1202 A suggested related benefit is that juniors might use access to the incumbent's line temporarily to "ramp up" production in anticipation of building its own line. This seems to us to be an expensive option that will require the junior to incur significant sunk costs, eg the cost of building a spur to connect to the incumbent's line and payment of the costs of expansions to the incumbent's line. We doubt very much whether a junior miner would think it worthwhile to obtain temporary access. 1203 Finally, as the NCC says, there will be social and environmental benefits from minimising (1) the amount of below rail infrastructure in the Pilbara; and (2) the use of public roads to truck ore. In respect of (1), FMG has raised environmental concerns about the impact of a spur through the Fortescue Marsh area but it is fairly clear these environmental problems can be avoided. As regards (2), because trucking is already regulated, the benefit is not likely to be great.

18.5.1 Loss of throughput 1205 The incumbents argue that the mere presence of a third party on a line will lead to a loss of throughput. Many of the reasons are to be found in the evidence of Mr Hoare and Dr Dallimore. The first reason is based on the hypothesis that variability reduces the extent to which a line can practically be used. The contention is that the additional variability which the third party brings will reduce system throughput. An example cited is a phenomenon known as the "bull-whip effect", where an incident in one part of the system can have flow-on effects (such as congestion) for other parts of the system. The bull-whip effect, it is argued, will be more pronounced with third parties present. The second reason is that the lines form part of an integrated production system which requires "buffer capacity" to manage variability within that system. The presence of a third party will, so it is said, reduce the buffer. A third, and related, reason is that a third party introduces rigidity into the system and this will hinder the incumbent's ability to operate on a run when ready basis. 1206 Subject to the limitations of the requirements safeguard (as to which see Chapter 10), there are two reasons for rejecting these contentions. First, we think that reasonable terms of access would give an incumbent priority for the use of its line. Second, by reason of the requirements safeguard, a third party's operations will be required to be structured in a way that will not result in displacement of an incumbents' throughput. 1207 RTIO also argues that, no matter how flexibly a third party is required to operate, once its train is on the line, the inability of RTIO, when it would otherwise be opportune to do so for operational reasons, to divert the third party train will lead to loss of throughput. Mr Hay suggests that it is also impractical to "park" third party trains on the line to minimise their impact. 1208 While FMG gives several answers to this argument, we think that the true answer is to be found in the results of the dynamic modelling requested by the Tribunal. The modellers were asked to assume that third party trains, once on the line, would be given equal priority to the incumbent's trains. We expected that any major problem with being unable to divert a third party train would have been apparent in the results. Yet, the results show that, at least for relatively low volumes, the impact of a third party on the line is minimal. The Tribunal does not, therefore, accept that the inability to divert a third party train will necessarily, or likely, lead to major displacement of the incumbent's output and thinks this is a matter that, if it is a problem, can be considered at stage two.

18.5.2 Compatibility and safety 1209 It seemed to have been accepted that criterion (d) (health and safety) was satisfied. Nonetheless, the incumbents have suggested that access, or at least the type of access contemplated by FMG, might pose significant safety risks. 1210 The starting point for the analysis is to appreciate that below and above rail infrastructure for a heavy haulage railway involves highly sophisticated and intricate technology. There are tremendous forces generated by moving heavy volumes of iron ore. Managing those forces pushes technology to the limit, and sometimes beyond, as the not uncommon instances of derailments show. BHPB and RTIO both devote significant resources to developing, maintaining and monitoring below and above rail infrastructure to prevent and detect problems. 1211 Expert witnesses largely agreed that heavy haulage railways require compliance with strict technical standards to avoid damage to the lines and to maintain safety. Of particular importance for the safe running of trains is the wheel/rail interface. Trains are engineered to operate effectively on a given track profile, and are sensitive to small changes in operating conditions. 1212 As slight deviations in the wheel/rail interface can create safety issues, it can be problematic to run a train on multiple tracks with different profiles. This creates a potential obstacle for an access seeker who contemplates using one of the incumbent's lines in conjunction with its own line. Take FMG as the example. It contemplates that its trains will run on both the Mt Newman line and the Chichester line which has a different profile. 1213 Mr Clark, an engineer called by BHPB, said that if rolling stock with BHPB's wheel profile ran on the Chichester line it would cause damage to the train's wheels which could lead to damage to the line and even potential derailments. Mr Clark observed that because FMG's trains would spend most of their time on the Mt Newman line, the odds were that the damage will manifest itself on that line. 1214 The most straightforward way of dealing with the compatibility problem is to impose a requirement that the third party use dedicated rolling stock on the incumbents' lines and "quarantine" those trains on the incumbents' lines. This would prevent FMG using its trains on a Pilbara-wide network, imposing costs which are likely to be significant. The dynamic efficiency losses may also be significant. So also would be losses resulting from the restriction on the third party's ability to experiment with new, more efficient, above rail technology. 1215 There might be an alternative way to manage the problem. Mr Tew, a rail engineer, suggested that interface problems caused by running on two different tracks can be mitigated by operating at reduced speeds. (This would, of course, have an impact on the turnaround time for trains) This is how BHPB manages its trains which operate on a section of the Goldsworthy track which is poorly maintained. Another possible safety measure which might be considered is to establish a regime for detecting wheel failures for trains operating on two lines. 1216 How the issue is to be managed, and its costs worked out, is for stage two. For present purposes all we need note is that there will be a cost, and the cost may be significant, possibly to the point, at least in the case of some junior miners, of making access prohibitive. 1217 Passing to another topic, RTIO suggested that a third party with smaller volumes of iron ore (10mtpa or less), will use (lower-cost) front-end loaders to load its trains. It claims that because front-end loading risks off-centre loading, the third party will be restricted to using partly loaded ore cars. Assuming this to be so, we fail to see how this is a cost (lower throughput) attributable to access. Presumably, without access, the third party will either (1) use rail on an alternative facility (in which case the off-centre loading issue would still arise); (2) use trucking, which will probably be even more inefficient in many cases; or (3) not transport its product at all. It is not an issue which is of concern to the Tribunal at the declaration stage.

18.5.4 Scheduling issues 1222 In addition to short notice of time slots, several other scheduling concerns were raised. The first relates to the disclosure of confidential information. The incumbents' train scheduling operations are integrated with other central planning functions to ensure a coordinated approach to mine, rail and port operations. Rail schedulers regularly meet with other planners and discuss confidential information about production activity. The incumbents fear that this confidential information will be at risk of being disclosed if third party schedulers become involved in the incumbents' train scheduling activities. 1223 It is not clear why a third party should be involved with the incumbents' internal scheduling discussions. In dealing with a third party, it should be sufficient for the incumbent to inform the third party of the outcomes of its internal discussions in so far as they relate to train slots available for third parties. Undertakings of confidentiality could be given to deal with any residual problems. 1224 The second concern is the possibility that a third party will "overbook" train slots and ultimately not use them all. A related concern is that the third party may give insufficient notice of non-use to enable the incumbent to make its own use of the slot. The Tribunal considers that appropriate disincentives could be put in place to prevent overbooking. 1225 The third concern is that the incumbent may be required to segregate or "ring-fence" its train scheduling function from other functions in order to avoid conflicts of interest. The incumbents note that ring-fencing is contemplated for TPI's operation of the Chichester line under the Western Australian Rail Access Code. In the Tribunal's view, so long as a third party's capacity entitlement is clearly defined and transparent scheduling protocols are imposed, conflicts of interests can be managed. In any event, the Tribunal doubts that the incumbents can be forced to ring-fence their operations.

18.5.5 Train failures and investigations 1226 Several issues have been raised about the possibility of third party train failures on the incumbents' lines. We proceed on the basis that a train failure has the potential to cause serious disruption to operations. 1227 First, the incumbents argue they have extensive resources which can be mustered to quickly respond to train failures, whereas third parties are unlikely to be able to respond as quickly. In the Tribunal's view, it is not a point which needs to be dealt with at the declaration stage. For one thing, FMG says that train failures are likely to be relatively uncommon. Even if this were not so, it would be in all parties' interests to clear the track as quickly as possible. We expect that the terms of access will deal adequately with what should be done. For example, the terms could provide for the incumbent to take remedial action, with appropriate penalties on the third party for costs and delays caused. It will be possible for the incumbent to take action provided, as seems almost inevitable, the third party is required to use trains which are compatible with the incumbent's lines. 1228 BHPB makes the point that in the event of a train failure, the third party may not have an incentive to deal urgently with the situation. The Tribunal does not accept that a third party would dawdle in response to a breakdown, especially when it is likely to be liable for direct and consequential loss to the incumbent by reason of the breakdown. 1229 On the other hand, the Tribunal does accept that identifying the cause of a train failure and any alleged resultant damage for purposes of assessing compensation may prove contentious. For example, a third party may argue that a derailment was caused by poor maintenance of the incumbent's track. And there will inevitably be disputes about the quantum of any claim. The Tribunal is, however, mindful of the fact that sophisticated monitoring equipment and stringent auditing records are likely to be required under the terms of access, and these will, to a considerable extent, reduce the scope for argument. Still, there will be the cost of any dispute, which may be large.

18.5.6 Constraints on third party operations and dynamic efficiency considerations 1230 The Tribunal has repeatedly made the point that a third party operating on an incumbent's line will be subject to operational constraints. For example, (1) the third party will be required to run its trains with minimal notice in less than ideal conditions; and (2) the third party will likely be required to operate trains which are compatible with the incumbent's trains, potentially restricting or precluding their use on lines owned by the third party. 1231 This situation is to be contrasted with the regime that exists, or is likely to exist, if the third party is to use an alternative rail facility. Use of the Chichester line is a good example. We have outlined some of the features of the WA access regime which applies to the Chichester line: see Chapter 13.2.3. It is likely that new railway lines in the Pilbara will also be subject to the Access Code and a similar regime developed under it. It must be conceded that the WA regime provides greater benefits to third party users than does Part IIIA. The benefits include scheduled timetables which give greater certainty in the use of the line and a more equitable sharing arrangement between the third parties. 1232 This is not to criticise Part IIIA. A significant difference between the two regimes is that Part IIIA applies to existing infrastructure and so, quite rightly, makes provision to ensure the owner is not prejudiced. Railway owners who fall under the WA regime know in advance of construction of a line that third parties can apply for access: see the Railways (Access) Act, s 7. This enables the owner to forward plan to accommodate line sharing. 1233 To the extent that access to the incumbents' line is more constrained than would be access to a new line that would be built if there is no access, the resultant cost needs to be taken into account. Of course, this would not be a cost for juniors that could not, in any event, obtain access to the alternative rail facility. For them, constrained access to an existing line might still be preferable to the unpalatable alternatives of trucking, mine-gate sales or not mining at all. 1234 There are other social benefits from the construction and use of a new railway line compared with expanding and gaining access to an existing line. While it may be difficult to assign a precise dollar value to those benefits, they are nevertheless real and should be considered. 1235 First, a new line which potentially uses more efficient technology (both above and below rail) may lead to opex savings compared with use of an (expanded or not) older line and older operating technologies. 1236 Second, the construction of a new line, while likely to be more expensive than mere access to, or an expansion of, an existing line, should lead to greater external benefits (both in the short run and the long run through the multiplier effect) such as economic growth and employment both in the region and from wherever equipment and raw material supplies emanate. 1237 Third, in some cases a new line will create greater capacity than is immediately needed to cater for third party demand (in contrast to an expansion). In one sense, this is inefficient, particularly from a productive efficiency perspective. In another sense, this could promote dynamic efficiency in the long run. Having spare capacity on an alternative line would give greater future flexibility and room for currently unanticipated growth in the demand for above or below rail services, either by currently identified or future access seekers. A relatively large increment in line capacity now could turn out to be more socially efficient in the long run than a series of short-term increments in capacity by the incumbent, and could lead to greater certainty for future planning purposes. In saying this we are mindful that one of the goals of Part IIIA is to promote efficient investment in infrastructure facilities over the long run.

18.5.7 Inefficiencies associated with delayed or sub-optimal new operating practices and technology 1238 Inefficiencies will arise when there is a need to implement new operational practices or new technology that might require the participation of a third party or otherwise impact the third party's use of the line. Assume, for example, an incumbent is considering implementing automatic train operations. This is a change that is considered from time to time. The technology involves the remote control of trains from a central location. Such an advance may be impossible if a third party train continues to operate in the traditional manner. 1239 Not only will certain changes be impossible to introduce, others will be delayed. Each incumbent regularly experiments with new practices and technology. The experiments include changes to train configurations, mine hub strategies, blending practices, port strategies and use of "sprinting" and new technologies. All could have an impact on a third party using the line. Hence a rational incumbent will first consult with the third party about the proposed changes. 1240 The need to consult with, or even obtain the consent of, a third party will inevitably involve delays, particularly where disputes arise in relation to changes that might have a detrimental impact on the third party (eg by increasing access charges). 1241 Quite apart from the problem of delay, if the incumbent is required to compensate the third party for detriment caused by a change in operating practices or technology, the incumbent may be dissuaded from pursuing the change. 1242 The problems that may arise cannot, as FMG suggests, be dealt with adequately by terms of access. Whether a change will have a material, negative impact on a third party will be the subject of dispute and, if not resolved, will lead to arbitration. This is likely to be a common occurrence. 1243 The Tribunal is of the view that any delayed or sub-optimal changes to technology or operating practices may cause significant harm to the incumbents and to the public interest generally. The evidence shows that keeping at the forefront of technological developments is an important part of RTIO's and BHPB's strategy in the highly competitive global iron ore market. The loss of any advantage may have serious long-term consequences.

18.5.9.1 Is there a need to consult? 1247 The Tribunal is of the clear view that, having regard to the structure of Part IIIA, a rational incumbent contemplating a line expansion will consult with a third party user to reach agreement or consensus about the works to be undertaken. 1248 Before explaining why we have arrived at this view, it is necessary to make a few observations about the nature of expansions to railway lines. An expansion often involves large capital expenditure. It can cost hundreds of millions of dollars, if not more. Planning for an expansion is a drawn out process, progressing from an order of magnitude study to a pre-feasibility study and so on. During the process, plans can be substantially revised and fine-tuned. When an expansion proposal is finalised, the board of directors (or relevant management committee) may be unwilling to approve such large expenditure if there is significant uncertainty or serious risks regarding the project. The approach taken by RTIO and BHPB to approving large capital expenditure is conservative and relatively risk-averse. 1249 How does Part IIIA impact on this process? We must begin with the requirements safeguard. We have observed that the requirements safeguard is forward looking: see the discussion in Chapter 10. If, at the time a dispute is notified, an incumbent reasonably anticipates that it will expand its line and use the capacity created by that expansion, its requirements will be protected. 1250 In such a case there is no legal requirement for the incumbent to consult with a third party user about the proposed expansion. As a practical matter, however, it would be sensible for the incumbent to consult with the third party to determine whether the third party also has plans for expanding its operations which may be incorporated into the incumbent's plans. If the incumbent does not take that course it may later face a request that it expand the line to accommodate the third party's needs. The result could lead to a series of expansion projects, most likely an inefficient way to proceed. 1251 Another factor that bears on this issue is that the requirements safeguard protects the incumbent's reasonably anticipated requirements measured at the time when a dispute is notified. The requirements safeguard does not address what happens if those requirements later change. In the context of an expansion, the timing limitation is important. Expansion plans change regularly, and new expansion plans may emerge, all of which may not have been reasonably anticipated when the dispute was notified. It may be asking too much of an incumbent to state at an early stage of expansion planning precisely how it expects to expand, and what its requirements will be in relation to that expansion. This is particularly so if the term of access is for an extended period; access can be granted for 20 years. 1252 What is to happen if there is an unanticipated need for expansion? There are several complications which can arise. The first is that the incumbent could be prevented from carrying out the expansion works if the works disrupt third party use of the service. By s 44ZZ the incumbent is prohibited from hindering a third party's access under a determination. This complication could be overcome if the terms of access provide for a temporary disruption to a third party's use to permit construction works to take place. Presumably the incumbent would be required to use reasonable endeavours to minimise disruption and provide an alternative rail slot where possible. 1253 The second complication concerns so-called "cheap capacity". We have noted that investment in below rail infrastructure can be "lumpy". Lumpy increments in capacity can be particularly expensive; other increments may be relatively cheap. Assume that when a dispute is notified, there is cheap capacity available for third party use. Assume also that following notification of a dispute the incumbent's requirements are greater than originally anticipated, so that it becomes necessary to undertake an unanticipated expansion. The incumbent will no doubt contend that the cost of the expansion should be borne by the third party, given that the incumbent would not have undertaken the expansion but for the third party's presence. The incumbent would obviously prefer to have this issue resolved, either by negotiation or arbitration, before going ahead with the works. 1254 The third complication is that the incumbent appears to have no guarantee that it will be entitled to use the capacity created by an unanticipated expansion. This reflects the limitation of the requirements safeguard. Assume that a third party user wishes to use some of the capacity created by a proposed expansion. Assume also there is a dispute about this which is referred to arbitration. The dispute could, in some circumstances, be a dispute "in relation to an earlier determination", in which case s 44W(2) would be triggered and the requirements safeguard would cease to apply. Alternatively, if s 44W(2) is not triggered each of the third party and the incumbent will be existing users whose reasonably anticipated requirements at the time of notification of the dispute must be accommodated, in effect giving them something akin to equal priority: see Chapter 10. 1255 It is, of course, possible that in an arbitration the ACCC will give the incumbent's requirements priority. But there is no guarantee this will happen. An incumbent, mindful that it faces a risk its requirements will not be satisfied, will consult with the third party to see if some agreement on use of capacity can be reached. 1256 All in all, the Tribunal is satisfied that in the case of most, if not all, expansions a rational incumbent will negotiate with a third party user in an attempt to reach agreement on the extent and cost of an expansion and the allocation of any additional capacity that is created.

18.5.9.2 Will negotiating lead to delay? 1257 The incumbents' case is that the participation of a third party in expansion planning will necessarily involve considerable delays, even in an ideal world where there are no disputes about the expansion. The mere fact that multiple parties are involved creates complexities which take time to resolve. The third party will need time to liaise with others (eg customers, port authorities) to work out a final position. If a third party is to contribute to the cost of an expansion, it will require time to arrange finance. Mr Leahey, a banking expert, suggests that it would take at least three to six months to obtain finance, but the time may be much longer. 1258 To test this argument, it is necessary to consider how the expansion process is likely to play out in practice. Mr Ranson is responsible for directing mine development and infrastructure studies at RTIO. Mr Ranson explained the likely sequence of events involved in an expansion which may involve a third party. (BHPB's investment approach is broadly similar). RTIO would not notify a third party of a possible expansion project before the order of magnitude study had been completed. RTIO would, however, need to ascertain the extent and terms of the third party's participation before finalisation of the pre-feasibility study. The third party's involvement at the pre-feasibility study stage would involve the following steps: (1) RTIO would inform the third party of the proposed expansion and ascertain how much extra capacity the third party sought; (2) The third party would carry out its own analysis of its expansion options and the commercial viability of those options; (3) Modelling and engineering work would have to be carried out to determine the optimum expansion project which would provide the additional capacity required by RTIO and the third party; (4) RTIO would notify the third party of the details of the optimum expansion project including modelling and engineering calculations and the likely cost (within a range) to the third party; (5) RTIO would seek confirmation from the third party of its participation; and (6) If and when terms and conditions of the third party's involvement have been finalised by agreement, RTIO would complete the pre-feasibility study. 1259 Mr Ranson said that at least the following issues are likely to be the subject of disagreement with the third party: (1) The need for an expansion to meet the additional capacity requirements of RTIO and the third party; (2) The scope of works required to meet the additional capacity requirements; (3) The timing of the expansion works; (4) The cost of the works. 1260 Mr Ranson said it is unlikely that a third party could be compelled to provide a binding commitment to participate in an expansion project until the rail components of the feasibility study had been completed. Those components would include engineering studies, geotechnical surveys and results, cost estimates, approvals under the applicable Aboriginal Heritage Acts (if relevant) and tenure arrangements where necessary. Only then would the third party be in a position to have a "bankable" feasibility study with which to seek finance. This takes time. And, ultimately, Mr Ranson suggested that RTIO is likely to be unwilling to progress the project past the feasibility stage until it has a clear indication from the third party one way or the other, because the removal of a third party may significantly alter the scope of the project. 1261 Mr Taylor, who has experience with the east coast coal supply chains, said that negotiating with third parties at an earlier stage would not necessarily avoid delays. As he put it, "ultimately, the major decisions will be made in the final year leading up to commitment, regardless of the start date for planning, because market information and the imperative to make final commitments both increase as this deadline approaches. It is in respect of these decisions during the final year that the delay will be experienced". Mr Taylor emphasised that, in his experience, miners plan expansions by "working backwards". The miner identifies a "gap in the market" to be supplied, and then works out how long it would take to construct the infrastructure necessary to expand production to supply that gap. The critical date, in Mr Taylor's view, is the date by which construction must commence to meet the gap. Mr Taylor said all parties would be aware of the critical date, and for tactical reasons will very often delay making a binding commitment until the very last moment. For this reason, according to Mr Taylor, consulting with a third party earlier will not necessarily bring forward the date by which negotiations are resolved if they are resolved at all. 1262 FMG points out that steps could be taken to minimise the likely delays. Mr Morton, an expert with experience of various access regimes, cites the Queensland Rail's Access Undertakings, which provide a formal framework for negotiations giving parties fixed time limits to reach agreement. In addition, FMG suggests that the incumbent could regularly publish a "master plan" for expansion which would establish a transparent means for third parties to consider their position regarding proposed expansions, thereby avoiding delays. 1263 The Tribunal accepts FMG's point - which Messrs Ranson and Taylor willingly concede - that it would be advantageous for an incumbent to engage as early as possible with third parties regarding expansions. For the same reason, establishing a conceptual "master plan" may be desirable. It is clear, however (and, for that matter, not contested by FMG), that the third party's position as regards an expansion would only be finalised at a relatively late feasibility stage of planning. 1264 We have considered whether the feasibility stage could be brought forward. Mr Taylor said that it cannot, because parties, gaming the system, will delay making commitments until the "true" deadline date is reached. We think that, in any event, bringing the process forward artificially is not likely to work. There is evidence which shows that, acting prudently, firms often delay making infrastructure investments for as long as possible to reduce the uncertainty of future events. That is, it is rational for a firm to delay committing to major expenditure for as long as possible. 1265 There will, we think, be at least two problems associated with trying to bring the decision-making process forward. First, it will force an incumbent to make decisions regarding expansions earlier than it would otherwise, increasing the risk to the incumbent with no means of compensation if losses do occur. Second, a rational access seeker will be just as unwilling to commit earlier than it would like, and will seek to defer making any commitment for as long as possible. 1266 This brings us to the next point, which is whether time limits could operate effectively to speed up the resolution of a dispute about an expansion. Assume the third party is interested in participating in an expansion, but disputes the cost, or says it has been provided with incomplete information about the proposal. Time limits might operate unfairly on the third party. Some restrictions would be called for. More importantly, imposing time limits would not necessarily save time - it might simply prompt the third party to seek to have any dispute referred to arbitration. 1267 Overall, the Tribunal doubts that time limits, early engagement or the use of master plans will substantially reduce the delays associated with engaging with third parties about an expansion. 1268 So far we have been considering delays caused by attempts to reach an agreement about expansion works. A delay to expansion works might be caused even after consensus (short of agreement) or agreement is reached. First, there is a risk that the third party may, in breach of agreement, change its mind. If so, the incumbent may need to "go back to the drawing board" and reformulate its plan. The Tribunal acknowledges this risk, but notes that a third party who has previously agreed to an expansion would be liable for breach of contract if it withdrew. Still, any claim would not be straightforward. Damages caused by a delay are likely to be inherently difficult to quantify, in part because the extent of any delay will often be uncertain. 1269 Second, there is a risk that if the third party withdraws from a proposal before there is any agreement (which is likely to happen on occasion), there will then be no foundation for a claim in damages.

18.5.11 Real options 1272 There is a large body of literature which holds that when an investment is irreversible and future demand and cost conditions are uncertain, the option of being able to wait and choose whether or not to invest in the future is valuable. The theory is that the option to invest is much like a call option on a stock that gives its holder the right to buy at a certain future date and at an agreed price. At the expiration of the period, the option holder will only exercise the option if the price is greater than the exercise price. The value of the option is in part due to the ability to delay exercising the option until future uncertainty is resolved. 1273 The incumbents' businesses have real options that are of value. Professor Triantis, an expert on real options, describes them. Holding tenements gives an incumbent real options to invest in mine, rail and port infrastructure in order to expand iron ore production to meet changes in market conditions. Its infrastructure is developed in a way which gives the incumbent a variety of options for investing in further infrastructure and/or new technology as market demand or technological changes dictate. The flexible and integrated nature of production gives the incumbent real options to respond quickly to variations in demand, for example by shutting down production where necessary. 1274 Access is said to lead to a loss or reduction in the value of these real options in several ways. First, the ability to exercise options will be impeded. It will take longer to exercise options to introduce new technology or expansions, for example, if engagement with third parties is required. It will be harder to quickly respond to changes in market demand if third parties reduce "flex" in the rail system. 1275 Second, the value of investments will be eroded, for example because of uncertainty regarding entitlements to capacity created by new expansions. As the expected return from exercising the option is smaller, this will (1) reduce the value of the real option and (2) reduce the likelihood of the option being exercised (because the returns are not high enough to justify the risk of exercising the option early). 1276 Third, access prices will not adequately compensate the incumbent for the loss or diminution in value of real options. Professor Triantis makes the point that while regulators are beginning to realise the importance of taking into account real options in appropriate cases (and the incumbents would contend that the price of access to a railway would be an appropriate case), the task is extremely challenging. He also makes the point that the high degree of complexity involved in properly valuing real options would, at a minimum, result in protracted negotiations or a complex arbitration. This is no doubt correct. 1277 Mr Houston has explained why he thinks Professor Triantis overstates the value of BHPB's options. First, he says that the principal investment decision (ie to build the line in the first place) has already been taken. We are not persuaded by this argument. Although the principal decision to build the line has been taken, the evidence suggests that there will nonetheless be significant and expensive investments to be made to the lines over the next decade or two. Second, Mr Houston notes that rail systems cannot be adjusted (eg to gauge) or moved without great difficulty, and the lines can likely only be used for transporting iron ore. This is clearly right but no one has suggested otherwise. Finally, Mr Houston also notes that BHPB has "free optionality" for expansion plans, because it can increase the capacity of the railway through relatively small incremental measures such as passing loops. We do not understand how this argument leads to the conclusion that the value of real options regarding expansions is likely to be relatively small. If anything, the fact that BHPB has the option to pursue investments at relatively low cost (instead of high cost alternatives) would suggest that that option is more valuable. 1278 Professor King said that the loss of real options for a monopolist is an inevitable, proper and intended result of regulation. Professor King observes that declaration removes certain options from the access provider. The provider no longer has the option of deciding whether or not to bargain with an access seeker. To the degree that this option is of value to the regulated firm, declaration will remove its value. That, Professor King suggests, is the aim of the declaration. 1279 The Tribunal agrees with Professor King, but only to a point. A distinction needs to be made between affecting a monopolist's anti-competitive power and affecting a monopolist's legitimate business interests. Removing a real option that enables an incumbent to exercise anti-competitive market power is not a relevant "cost" for public interest analysis. Removing other real options which the incumbent might legitimately exercise is a different matter. 1280 Professor King also argues that the value of real options may be increased as a result of declaration, once real options created for third parties are taken into account. One is the option of being able to seek access. This is acknowledged by Professor Triantis. But he makes the point that, unlike most options, the third party is not required to make an up-front payment to secure its options. This, Professor Triantis says, will skew the incumbents' incentives to invest. 1281 Ultimately, we think that issues regarding real options and their value overlap to a large extent with the topics already discussed concerning the costs and benefits of access. Whether there is a net loss or gain in the value of real options will, we think, correspond with the broader assessment of whether there is a net benefit or loss from access. To avoid "double-counting", we do not propose to focus specifically on real options, apart from making the obvious point that the value is likely to be a significant topic in any dispute on price.

18.5.12 Costs of arbitration 1282 It is highly likely that regulatory intervention (ie arbitration) will be required if a line is declared. First, the incumbents and any third party that gains access will be competitors in the global iron ore market, which increases the risk of disputes arising. Second, negotiations to date regarding haulage have not been amicable, and we do not expect this situation to improve. Third, large amounts of money are involved. Fourth, some access issues (eg capacity) are likely to be highly contentious. Fifth, if there are multiple access seekers, then the prospects for reaching collective agreement are relatively low. Sixth, it is likely that the lines will undergo regular expansions over the next decade or so, and many of those expansions have the potential to result in disputes. 1283 What are the costs of arbitration? The first broad group of costs are a consequence of the time it would take for a dispute to be resolved. Mr Phillips, a partner in Allens Arthur Robinson, provided a summary of a number of disputes that have been determined by the ACCC under Part IIIA and Part XIC. Only one dispute was resolved within seven months. Most took between one and (approximately) three years to resolve. 1284 We think the majority of access disputes that will arise will take a substantial amount of time to resolve. An application for access will require, at a minimum, a consideration of the impact of the access seeker's proposed usage on the incumbent. That would need to be modelled and the results will be strenuously debated. If there is a potential for displacement of the incumbent's throughput, the ACCC would need to determine what measures are needed to remedy that displacement. There will be significant debate about the appropriate costs to be taken into account to determine the access charge. 1285 Any proposed expansion will also likely lead to disputes. It will likely be necessary to conduct engineering studies, geotechnical surveys and undertake cost estimates of the proposed works. Once the works are identified (by agreement or arbitration), the question of the contracting of the works will arise. Arrangements will have to be made to ensure contractors are paid what is due. Complications may arise during the construction phase, eg about whether the works have been properly undertaken. At each stage there is the potential for many disputes, most of which could go to arbitration. 1286 This is by no means a complete list of the kinds of disputes that could arise. Operational or technological changes which might require an increase in access charges or which might cause detriment to a third party user, will lead to dispute and, likely, arbitration. Arguments about whether a third party is seeking to consume "cheap" capacity are a potential source of debate. 1287 Perhaps an initial "test case" arbitration could lay down general rules and guidelines for later arbitrations, which might then be of much shorter duration. But no "precedent" can be worked out for the three key issues which are likely to arise in most arbitrations: (1) capacity; (2) the extent and nature of expansions; and (3) the cost of the works. Nor can there be many rules for disputes arising from a breach of an access declaration. The issues that will arise are fact-specific and, more often than not, will be very difficult to resolve. 1288 The necessarily lengthy arbitrations will have at least two presently relevant cost consequences. First, there are the direct costs of the arbitration (eg legal fees etc). Second, and more importantly, there are the costs from yet another source of delay, especially for expansions and other developments. 1289 Apart from timing and the associated direct and indirect costs, the incumbents have a more fundamental concern about arbitration. They fear that the ACCC might get things wrong. A theme underlying the evidence of a number of the economists is that regulators at stage two will fall into error. There appears to us to be something of an article of faith held by many economists that regulation is more likely to lead to a wrong (ie inefficient) result than would be the situation if parties are left to their own devices. We know that a good deal has been written on this topic - much of it one way: ie in support of the view that regulation is prone to error. 1290 For our part we see little likelihood of error. Legislation which authorises regulatory intervention for the most has the following features: (a) the decision-maker has relevant expertise; (b) the decision-maker has the means of obtaining the material facts; and (c) the parties have a right to be heard. While the decision-maker does not have a stake in the outcome, other than a desire to arrive at the correct decision, he/she has more information upon which to base his/her decision than the parties. We rather think that the regulator is often better placed than the parties to arrive at the most efficient decision, and is likely to do so. But importantly, for a regulator to do its job properly will take time, and that time cannot be ignored as a potential source of delay. 1291 In any event, we do not think we are entitled to make an assumption about wrong decision-making under Part IIIA. Arbitral decisions are subject to merits review. Decisions by the reviewing body are reviewable for error of law. These are adequate safeguards.

1292 Mr Taylor has estimated the NPVs of several junior miners' projects in the Pilbara. The NPVs of those projects vary depending on factors such as the size and quality of the deposit, transport options and so on. The value of each project considered by Mr Taylor is significant, ranging from $77m for a relatively low volume project to around $2.6bn for a large project. We think there is little doubt that the value of the projects of miners who might benefit from access is likely to be substantial. We will, in a moment, discuss how those benefits might be assessed. 1293 On the costs side, there is evidence that attempts to quantify the incumbents' costs arising from (1) delays to expansions and (2) lost throughput due to more congested, less flexible multi-user operations. As regards (2), we have explained why we do not accept that operational inefficiencies which might arise from access will cause lost throughput. We will therefore focus on the costs arising from delays. 1294 Dr Fitzgerald and Mr Taylor quantify the costs of delay to BHPB and RTIO respectively. Their approaches are broadly similar. Both establish a base case representing the relevant incumbent's planned expansions over the next fifteen years or so, in the absence of a third party user. They then develop alternative scenarios assuming various average delays caused by access. Dr Fitzgerald considers scenarios involving delays of twelve and twenty-four months; Mr Taylor considers scenarios involving delays ranging from three to eighteen months. They identify the value of "lost" throughput as the additional throughput from expansion which is delayed. They then make assumptions about whether any of this lost throughput is diverted to other Australian producers or recovered by the incumbent. Dr Fitzgerald assumes that lost throughput is diverted to overseas producers, and that BHPB catches up on lost capacity. Mr Taylor assumes that 20% of lost revenue is diverted to other Australian producers, and 20% is subsequently recovered by RTIO. Both then calculate the net loss in export revenues, and how this will impact on Australia's GDP and Australia's economic welfare. It should be noted that both experts' analyses consider the effects of delays to expansions to the incumbents' rail network as a whole, rather than on an individual line basis. 1295 Dr Fitzgerald concludes that lost export revenues from an average delay of twelve months for BHPB expansions would be $8.7bn in NPV terms. He notes the impact on Australia's GDP would be smaller, as some resources which would have been used for expansions would be diverted elsewhere. As a rough rule of thumb, he estimates that GDP loss resulting from a fall in export revenue is typically between one-half and two-thirds the size of the fall in export revenue. This equates to around $4.4bn to $5.8bn based on the assumed twelve month average delay. Taking into account losses in royalty, taxation income and distributions to Australian shareholders, Dr Fitzgerald calculates that the community's loss of economic welfare would be in the order of $3.2bn in NPV terms. The loss would be greater if longer delays or larger demand (and higher iron ore prices) are assumed. 1296 Mr Taylor approaches the issue in a different way. He estimates that the NPV of lost export revenues to Australia ranges from around $10bn for a three month average delay to around $40bn for an eighteen month average delay. Mr Taylor estimates that the impact on GDP would be around 40-60% of the NPV of revenue lost to Australia. This would equate to around $4-6bn for a three month average delay. 1297 In relation to the impact on GDP, RTIO puts forward more detailed analysis from Mr Brown, an economist from Access Economics. Adopting Mr Taylor's estimates of lost exports, Mr Brown models the impact on GDP using a computable general equilibrium model developed by Access Economics. Mr Brown's estimate is that, in current market conditions with a relatively tight labour market, the impact on GDP would be around 60-70% of lost export revenues. This is consistent with other export industries Mr Brown has previously modelled. If this estimate were adopted, a three month average delay to RTIO expansions would cause losses in GDP of $6-7bn. 1298 FMG is critical of these witnesses. First, it argues that their modelling misunderstands the nature of the enquiry under criterion (f): it says the criterion should be limited to considering factors not already covered by other criteria. We have already expressed our disagreement with this approach. Second, FMG attacks the factual assumptions the modelling makes about delays and inefficiency. We, on the other hand, think delays are a real possibility for reasons which we have explained. Third, FMG points out that the modelling does not take into account potential gainsin export revenue from granting access, eg, by allowing otherwise stranded deposits to be developed. We agree that the modelling results estimate the cost associated with delays to expansions; the results are not a comprehensive assessment of the impacts generally on export revenue resulting from access. Finally, FMG points out the modelling (and particularly the macroeconomic modelling undertaken by Mr Brown) requires speculation about a variety of economic factors and is sensitive to the assumptions made. We agree that the modelling is imperfect. However, we think that the calculations are sufficiently robust to provide a broad indication, at an order of magnitude level, of the kind of costs which expansion delays may cause to the incumbents and, in turn, to the Australian community. Mr Taylor extensively tested the sensitivity of his findings to various key assumptions, and in all cases the lost export revenues were still in the billions of dollars. We give particular weight to Mr Taylor's estimates regarding an assumed three month average delay to expansions. The costs of delays for this short period are significant. The evidence suggests that the average delay may well be longer than three months. 1299 It was not only the economists who gave evidence about losses arising from delays. Senior personnel also had their say. Mr Walsh AO, RTIO's chief executive officer, said that any delay in expansion in the current economic environment (ie the growing demand by China for iron ore) will delay the revenue stream from extra sales that could be achieved from each expansion. He said a delay of twelve months to implement a 20mtpa expansion (not a large expansion in the scheme of things) would cost in excess of $2bn at current prices and exchange rates.

1300 We can now weigh up the benefits and costs of access to each service. Although this will be done on a line by line basis, some general observations can be made. 1301 It is particularly important that, if the services are not declared, alternative rail facilities are likely to be available for many access seekers. The situation in the Pilbara is unusual in that, notwithstanding the presence of facilities with natural monopoly characteristics, alternative facilities can be - and are highly likely to be - built if a declaration is refused. Other benefits which might ordinarily flow from access to a natural monopoly facility do not necessarily arise here. In particular, although we have concluded that criterion (a) is satisfied in respect of all of the services except the Mt Newman service, we doubt that access will result in large gains for competition. 1302 Sifting through all of the benefits and costs which have been raised by the parties, in the end there are only a few which have a major bearing on the public interest. 1303 On the benefits side of the ledger, there are two key potential benefits. First, access to some lines might potentially "unlock" some stranded deposits, and even for those who are not stranded, access may provide a significantly cheaper form of transport and allow greater volumes of iron ore to be transported. Second, there may be capital savings from accommodating third party demand on an incumbent's existing line rather than on a new railway. 1304 On the other hand, there are two key likely costs. First, there is the likelihood that access will discourage the development of alternative lines. Given that access to alternative lines may well be less constrained, and provide far more certainty of use for third parties than access to the existing lines, this is an important cost. The second critical cost is that for some services access will most likely lead to delays in the incumbent making changes on its lines, whether by way of expansions, adopting new operating practices or introducing new technology. These delays (and the resultant costs) are a likely consequence of having a shared facility. The delays (and costs) would probably arise in any event if third parties were to share an alternative facility. Hence, the third parties may be indifferent to the effect. But it is of great significance for BHPB and RTIO. 1305 We have said that criterion (f) and the discretion do not require a precise quantifiable cost/benefit analysis. Nonetheless, in what follows we have attempted to compare the benefits and costs of access, where possible giving them some order of magnitude value.

18.7.1 Mt Newman service 1306 To begin with, in the absence of a declaration, the Chichester line would likely be extended to Marillana. The result is that there would be no competition benefits: ie no tenements will be stranded or even inconvenienced by using the extended Chichester line. Indeed, some might benefit from being able to access the extended Chichester line at any point, rather than being required to truck to Mindy Mindy on the point to point Mt Newman service. 1307 We have elsewhere concluded that there are unlikely to be any capital savings associated with expanding the Mt Newman line rather than building an alternative line: see Chapter 15. Indeed, taking into account the ability to link with the Chichester line, there may be greater costs associated with expanding the Mt Newman line. 1308 The only real benefit we see from declaring the Mt Newman line is that it may allow FMG to accelerate its development of the Mindy Mindy mine (avoiding waiting for a rail spur to be built to the Chichester line). Other junior miners would likely need to wait for port facilities at South West Creek or Utah Point to become available. Given that Mindy Mindy is a relatively small mine, we see the benefit from accelerating its development as marginal. 1309 As to the costs of access, it must be remembered that the vast majority of BHPB's production is transported down the Mt Newman line. BHPB's demand for use of the line - and correspondingly, the constraints on third party use - are most acute. Several of those junior miners identified as potential access seekers are contemplating relatively large iron ore operations: ie operations of greater than 10mtpa. The effect of the operating constraints on those juniors would be keenly felt. 1310 The constrained nature of access is to be compared with the position that would exist if the junior miners obtain access to the Chichester line. We have outlined some of the advantages of access under the regulatory regime applying to the Chichester line from the perspective of an access seeker. Further, according to FMG, the Chichester line is more technologically advanced than the Mt Newman line and can handle trains which are more efficient. A third party which obtains access to the Chichester line would use more efficient trains. Alternatively FMG could provide the juniors with haulage services. FMG has expressed a keen interest in hauling for third parties. Junior miners are likely to be better off accessing the Chichester line. However, if the Marillana spur were not built, this might prevent many junior miners from accessing the Chichester line. 1311 Another key cost we have identified is the cost of delays to expansions or changes in operating practices or technology. Any delay will inevitably affect the Mt Newman line. And, as the evidence regarding the cost of delays to expansions shows, delayed throughput (whether due to delays in expansions, in changes to technology or for some other reason) is potentially a very large cost. 1312 Even if we were to ignore costs from delays to expansions, access to the Mt Newman service has limited benefits and significant costs, and would not be in the public interest. This conclusion is only strengthened once delays to expansions under Part IIIA are taken into account. Hence, access is not in the public interest. In any event, as a matter of discretion we would not declare the service.

18.7.2 Goldsworthy service 1313 Access to the Goldsworthy line will have two major benefits. The first is that it will provide access seekers with a rail option which otherwise would not be available. Unlike other lines, an alternative rail facility to the Goldsworthy service is unlikely to be developed in the absence of access. Some potential access seekers may admittedly have other transport options (eg trucking near Port Hedland or slurry pipe from the Ridley project). But we do not discount the possibility of other potential access seekers requiring access to the line to transport their iron ore, and in any event, rail may be cheaper even if trucking is possible. 1314 Access will also clearly achieve some capital savings. A conservative estimate of the minimum capital savings to be had is something in the vicinity of $550m. This is because limited, if any, expansion is needed to the eastern section of the line to accommodate third party demand. 1315 Turning to costs of access, the impact of operating constraints on third parties when accessing the Goldsworthy line are likely to be significantly less than for accessing the Mt Newman line. BHPB's demand for the eastern section is limited, and potential third party demand for that part of the line is not great. This suggests that a third party could operate with relative freedom on that section. The Finucane section is obviously more problematic. But this section is relatively short and a third party has a variety of options for bypassing this section altogether (eg using the Chichester line or the spur to South West Creek that is being investigated by the NWIOA). 1316 The risk that access to the Goldsworthy service will lead to delays to changes in operating practices or technology is, we think, limited. Technological development and maintenance of the eastern section of the line has not kept pace with changes to the remainder of BHPB's Pilbara network. Given BHPB's relatively limited anticipated use of the eastern section, changes to operating practices are not likely to be significantly impacted. Access to the Finucane section will have a greater impact, although the potential impact should not be overstated. The Finucane section is short. To the extent that there are technological or operating changes, those changes may well be manageable over a short distance (in the way, for example, that BHPB manages trains coming off the eastern section of the Goldsworthy line by running them at lower speeds). 1317 As regards expansions, access to the eastern section will have little, if any, impact given that no expansions are likely. Access to the Finucane section will have a greater impact. But we think the impact is unlikely to be large on an ongoing basis. Being a relatively short section, there is only so much that can be done to expand the line. Putting in numerous sidings, for example, is an unlikely option. We suspect that what will be required for any expansion will be relatively clear-cut. Because the distances are short, the cost of any expansion is likely to be relatively small and less prone to protracted dispute. 1318 We are satisfied that access to the Goldsworthy service is not contrary to the public interest. This is a line which, for most of its length, can comfortably accommodate third party demand. The problems associated with the Finucane section can, we think, be largely mitigated or avoided. We see no reason to exercise our discretion against declaration. We think it is appropriate to declare this service for the 20 year period sought.

18.7.3 Hamersley service 1319 Of all of the lines, demand for the Hamersley line - from both RTIO and potential access seekers - is likely to be the most intense. This means that the benefits and costs of access are likely to be of greater magnitude than for other lines. Although we have concluded that the benefits may be worth billions, there is the very real possibility - indeed probability - that these benefits could be dwarfed by the costs. 1320 To understand how we come to this conclusion, it is necessary to have regard to the way in which demand will arise for the Hamersley line. On the one hand, RTIO requires the line to carry the vast majority of its iron ore. Most of the mines which RTIO plans to bring online going forward will require use of the Hamersley line. On the other hand, we have identified 167mtpa of potential third party demand. To put this figure in context, this will be more than half of RTIO's output as at 2014. Of the potential third party demand we have identified, a large proportion is from tenements located near the Solomon area (many of whom could access the Dixon line if it were built), with a comparatively small group of tenements to the south and east who would perhaps be unable to access the Dixon line. 1321 Viewed in this context, there are two principal benefits from access. One is providing a rail service for tenements who would be "stranded" or severely inconvenienced without access and unable to access the Dixon line. As to the risk of tenements being stranded or severely inconvenienced, see Chapter 13.4. 1322 If it difficult to know the extent of the benefit. When we assessed potential third party demand for the Hamersley line, we considered that most would come from large projects in or near the Solomon area. But it must be remembered that we identified potential demand by applying some fairly strict rules which did not take into account other deposits (eg non JORC-certified deposits) which might also benefit from access. Other projects might emerge in the vicinity of the Hamersley line. This makes it impossible to quantify the benefits of access for those tenements, other than to suggest that the benefits will be substantial. It is also important to note that a number of those tenements, even if they cannot access the Dixon line, might be able to access other rail services (eg the Kennedy line or the Marillana spur to the Chichester line). 1323 The other principal benefit is the capital savings associated with avoiding duplication of the Hamersley line. There are potentially large capital savings if the Hamersley service is not duplicated. We have assessed the maximum amount that could be saved to be around $2.4-2.75bn, although we did not, and could not, purport to calculate accurately the actual savings. The actual savings might be significantly lower, but, even then, the savings will be substantial. It is important to understand who will benefit from these capital savings. The largest capital savings arise from avoiding duplication of the line south and east of Rosella. The capital savings from Rosella north - which, for a large proportion of identified potential third party demand, is the only section which they will use - are less clear-cut. The cost of expanding this section to cater for third party demand may not be that much less than building the Dixon line. 1324 There is no denying that savings of this order (ie savings incurred by avoiding the construction of an alternative facility) have a significant bearing on the public interest. But to focus on construction costs is to take too blinkered a view. It is just as important - and perhaps more important - to consider whether access to the Hamersley line will encourage the optimal development of rail infrastructure in the Pilbara. The Hamersley line services parts of the Pilbara where significant mining activity already takes place and where activity will likely grow. A vital question is whether access will encourage rail infrastructure which best services these regions. There are three possibilities presented before us. The first is that there is access to the Hamersley line and the Dixon line is not built. The second possibility is that there is access to the Hamersley line and the construction of the Dixon line is delayed (but it is eventually built). The third possibility is that there is no access to the Hamersley line and (as is most likely) the Dixon line will be built. 1325 Let us consider the implications if the first possibility were to eventuate. This would involve potentially cramming a very large amount of traffic (from both RTIO and third parties) on to one rail network. We have potential third party demand to be 167mtpa; in the long run this likely understates potential demand. Apart from anything, this will increase the potential for complexity in rail operation and for disputes. 1326 From the perspective of third parties using the line, the constraints associated with fitting in with RTIO's requirements may be very significant, if not prohibitive. Many of the juniors who might benefit from access have relatively large projects contemplating target production rates of 10mtpa or more. In FMG's case, its project contemplates 100mtpa of demand for the Hamersley line. These projects will be particularly affected by constrained operations. The reality is that the proposed Dixon Island line would likely offer a much more attractive arrangement for many potential access seekers. In this regard, we note that if FMG were to build the Dixon Island line, it would be able to use it as part of an integrated network with the Chichester line (and any other lines it is contemplating building). Obviously, the Hamersley line could not be used in that same way. 1327 From the perspective of RTIO, the consequences of third party access would likely be grave. The Hamersley line is regularly expanded to enable RTIO to take advantage of growing iron ore demand. The expansions are significant, and take years to plan and implement. If access is granted it would be necessary for expansion to cater not only for RTIO's future needs but also those of the third parties. The resultant delays caused by the need to consult with third parties on any expansion proposed (RTIO's or the third parties) will produce significant loss to RTIO. Likewise, the costs associated with delaying or preventing the introduction of innovations on the line (eg through changed operating practices or new technology) could be significant. 1328 On Mr Taylor's analysis and his assumption of a three month average delay to an expansion - which we think is a very conservative assumption - lost export revenue would be in the order of $10bn. While Mr Taylor's calculation was for the combined RTIO rail network, most of this loss would be attributable to the Hamersley line. Mr Taylor did not take the next step, as Dr Fitzgerald did in BHPB's case, of deriving the likely impact of this loss on Australian economic welfare. If he had done so, we have little doubt that the cost would be extremely large, if Dr Fitzgerald's findings are anything to go by. 1329 Turning to the second possibility (access to the Hamersley line will delay but not prevent the construction of the Dixon line), we do not think that much could be said to commend this result. The notion of third parties paying for expansions to the Hamersley line and then "ramping up" production while building the Dixon line is, we think, uncommercial. Whatever capital savings there are from access would be greatly reduced. And for the period the Dixon line is under construction, the costs that will be incurred by both third parties and RTIO will be large indeed. 1330 If we consider the final possibility (no access to the Hamersley service), then we think the consequences are as follows: · A relatively small proportion of access seekers who could not otherwise access the Dixon line or some other line would be prejudiced. · For a relatively large proportion of access seekers who could access the Dixon line, the Dixon service will offer a much more commercially practical rail solution, although they will incur the additional costs of constructing the Dixon service either directly or indirectly (through access charges to the Dixon service). · RTIO will avoid the potentially enormous costs associated with delays to implementing expansions and innovations on the Hamersley line. 1331 In none of the above scenarios would we be satisfied that access (even if one ignored delays to expansion) would not be contrary to the public interest. Indeed, when expansion delays are taken into account (and given the enormous potential demand for the Hamersley line, delays are likely to be most prominent), we think access would be contrary to the public interest. We would, in any event, for the reasons outlined, exercise our discretion not to declare the Hamersley service.

18.7.4 Robe service 1332 Here there are parallels with the Goldsworthy service. One parallel is the need to consider two separate sections of the line. 1333 Access to the south-west section has significant benefits and relatively minor costs. There are large savings from avoiding duplication of this part of the line, since it appears that any expansions required to meet the demand for this section is limited. Correspondingly, issues such as constraints to third party operations, delays to changes in technology or operating practices, or delays to expansions, are not particularly troubling. 1334 Access to the northern section of the line (ie north of Western Creek) gives rise to greater complexities. The starting point is that much of the iron ore which RTIO plans to bring online will be directed to Cape Lambert. Equally, however, it appears that RTIO's current expansion plans would not fully utilise capacity of the northern section until at least 2018. 1335 We note also that the northern section may have some bearing on the proposed joint venture between BHPB and RTIO. It is far from clear whether this joint venture will proceed. At the present time, it is too speculative a prospect to be given significant weight in assessing the public interest. 1336 If we were declaring the Hamersley service, there might have been large potential third party demand for the northern section. But we are not declaring the Hamersley service, and the third party demand for the northern section is relatively limited. It appears that for at least some time, that demand can be comfortably accommodated with relatively minor, if any, expansions. 1337 Access to the Robe service will result in large capital savings and, because there is significant capacity, problems associated with constrained third party operations and delays to changes to operating practices or technology are not as pronounced as on other lines. Importantly, Mr Ranson has suggested that in the absence of a joint venture with BHPB, RTIO does not have any intention of fully utilising the capacity of the northern section of the Robe line before 2018. We are concerned about the likelihood of extensive use by RTIO of the line after that time. Even taking into account potential expansion delays, we think that access would not be contrary to the public interest if the period for declaration is limited to 2018. However we are not satisfied that, for any period beyond 2018, access would not be contrary to the public interest.

MISCELLANEOUS ISSUES 1338 There are a few remaining legal issues which, for want of a better term, are miscellaneous. They may be disposed of relatively quickly. 1339 The first was raised by the Tribunal itself. The question is whether the application for a declaration covers a service provided by a facility which, at the time of hearing, is, in many respects, different from what it was at the time of application. The Tribunal raised this question because several of the railway lines, particularly the Hamersley line and the Mt Newman line, have, during the period, undergone significant expansion, including having sections altered from single track to double track. 1340 We think the answer lies in drawing a distinction between the service for which the application is made and the infrastructure pursuant to which the service is provided. While the infrastructure might be different, the service applied for has not changed: see Chapter 12. 1341 The second issue is confined to the application for declaration of the service on the Mt Newman line. According to the application the service sought is described, relevantly, as "the use of the facility being part of the Mt Newman railway line which runs from a rail siding that would be constructed near Mindy Mindy in the Pilbara to port facilities at Nelson Point in Port Hedland." In fact, FMG does not wish to exit the Mt Newman line at Port Hedland. Instead, if access is granted, it will exit the line somewhere near the Goldsworthy junction, probably at a point 10 to 12km south of the port at Port Hedland. 1342 BHPB says that this is a material variation to the service that has been applied for. It says that the section of the line that FMG wishes to access is materially shorter in length than that stated in the application. And it is says it has a different character because it does not involve transport to port facilities. 1343 We accept that FMG has changed the service to which it seeks access. We do not agree that the change is in any way material. In Rio Tinto Ltd v Australian Competition Tribunal, the Full Federal Court said (at [59]) that there need not be a "slavish attachment to the words of an application". A "literal or pedantic adherence" to the description of the service is not required, provided that the substance and essential natural of the service is not altered. That is the position here. 1344 The last point arises out of s 44W(1)(c). This subsection provides that the ACCC must not make a determination in an arbitration of an access dispute that would have the effect of depriving any person of a protected contractual right. Each of BHPB and RTIO contend that were access to be granted each would be deprived of protected contractual rights. The contractual rights are those created pursuant to the State agreements, pursuant to special leases of the rail corridors granted by the Western Australian government and pursuant to several joint venture agreements. 1345 The Tribunal will not enter upon this issue. It is an issue appropriately considered at stage two when a particular request for access is considered: Hamersley Iron Pty Ltd v National Competition Council (1999) 164 ALR 203 at [80]. Indeed, there is little utility in dealing with the point at this stage. Even if it is a good one as a matter of theory, in appropriate cases consents may be forthcoming from counterparties to the relevant agreements so as to undermine the argument.

Concluding Remarks 1346 There are four observations we wish to make. 1347 First, while we set aside two ministerial decisions, it does not follow that we disagree with those decisions. The iron ore industry is a dynamic industry. Iron ore operations change rapidly and dramatically. They change to meet ever changing market demand, and to enable firms to keep up with, or get ahead of, competitors in the iron ore market. Little of what we see today will exist in a few years time. The nature of the industry that was before the Minister when the decisions were first taken is significantly different from the industry that we have here. 1348 Second, the mechanism to promote community welfare identified by the Hilmer Committee - competition - is as vital today as it was in 1993 when the Committee published its report. Part IIIA is an important means of achieving community welfare. Yet the Part does require reform, and we think the following issues should be investigated:

The possibility of compelling inefficient investment
The desirability, and mechanics, of an expansion power
Whether access on a first-come, first-served basis is inequitable 1349 These are just some of the important issues which the legislation does not deal with adequately. There are other, less significant, problems which we have identified along the way that might also be worth looking at. 1350 Third, one matter of procedure should be considered, namely whether it is appropriate to collapse a rather complicated process. The multitude of steps included in a Part IIIA application for access are: (1) NCC recommendation; (2) Ministerial declaration; (3) Tribunal review; (4) Appeals to the court; (5) Possible remitter; (6) Negotiations for access; (7) Arbitration; (8) Further Tribunal review; (9) Possibly more appeals to court. Getting through this process will inevitably take years: If a complex case was run in the fast lane, the earliest it will still take is 4 to 5 years to complete. The imposition of time limits on administrative decision-makers will reduce the time a little, but will not address the core problem. 1351 Finally, we cannot leave these reviews without acknowledging the significant effort that the parties and the NCC went to in order to prepare for, and present, what is a most difficult case. Without the assistance of the parties and their lawyers it would not have been possible for us to come to grips with so many issues in a (relatively) short period of time. We recognise that we have placed significant burdens on the lawyers by imposing limits, and requiring prompt responses to Tribunal requests and directions. Still, the cooperation we received from them was complete. We wish to thank everyone for their assistance. I certify that the preceding one thousand, three hundred and fifty one (1351) numbered paragraphs are a true copy of the Reasons for Determination herein of the Tribunal.

In the matter of Fortescue Metals Group Limited [2010] ACompT 2

At a glance

Court

Decision date

Before

Catchwords

Source

Catchwords

Judgment (287 paragraphs)

Legislation Cited (14)

Cases Cited (12)

Cited By (8)