AUSTRALIAN COMPETITION TRIBUNAL Application by Epic Energy South Australia Pty Ltd [2003] ACompT 5 TRADE PRACTICES - access to pipelines - review of decision of Australian Competition and Consumer Commission to draft access arrangement - whether decision reasonable - calculation of optimised replacement cost - whether covered pipeline - Moomba to Adelaide Pipeline System - Gas Pipelines Access (South Australia) Act 1997 (SA) Gas Pipelines Access (South Australia) Act 1977 (SA) Application by Epic Energy South Australia Pty Ltd [2002] AcompT 4 followed Associate Provincial Picture Houses Ltd v Wednesdbury Corporation [1948] 1 KB 223 applied House v The King (1936) 55 CLR 499 applied Duke Eastern Gas Pipeline Pty Ltd [2001] ACompT 2 followed APPLICATION BY EPIC ENERGY SOUTH AUSTRALIA PTY LTD File No 6 of 2002 JUSTICE COOPER, PROFESSOR D K ROUND AND MISS M M STARRS BRISBANE (HEARD IN ADELAIDE) 10 DECEMBER 2003 IN THE AUSTRALIAN COMPETITION TRIBUNAL Matter No 6 of 2002 RE: APPLICATION UNDER SECTION 38 AND/OR SECTION 39 OF THE GAS PIPELINES ACCESS (SOUTH AUSTRALIA) LAW FOR REVIEW OF THE DECISION BY THE AUSTRALIAN COMPETITION AND CONSUMER COMMISSION MADE ON 31 JULY 2002 TO DRAFT AND APPROVE AN ACCESS ARRANGEMENT FOR THE MOOMBA TO ADELAIDE PIPELINE SYSTEM UNDER SECTION 2.20(a) OF THE NATIONAL THIRD PARTY ACCESS CODE FOR NATURAL GAS PIPELINE SYSTEMS BY: EPIC ENERGY SOUTH AUSTRALIA PTY LTD TRIBUNAL: JUSTICE R E COOPER (DEPUTY PRESIDENT) PROFESSOR D K ROUND MISS M M STARRS DATE: 10 DECEMBER 2003 WHERE MADE: BRISBANE (HEARD IN ADELAIDE) THE TRIBUNAL ORDERS THAT: 1. The decision of the Australian Competition and Consumer Commission made on 31 July 2002 to approve an Access Arrangement pursuant to s 2.20(a) of the National Third Party Access Code for National Gas Pipeline Systems in respect of the Moomba to Adelaide Pipeline System ('the Decision') be varied. 2. The Decision be varied in the following manner: (a) An exchange rate adjusted cost of line pipe of $1 302.20 per tonne be used to calculate the Optimised Replacement Cost of the existing pipeline for the purpose of determining the initial Capital Base in valuing the Moomba to Adelaide Pipeline System ('MAPS') as a Covered Pipeline. (b) The expansion of capacity of the MAPS by 25 terajoules per day to satisfy the requirements of the operator of the Pelican Point Power Station be excluded from the System Primary Capacity ('the SPC') calculation and that the SPC of the MAPS for the purpose of the approved Access Arrangement be 323 terajoules per day. (c) All necessary and consequential amendments be made to give effect to such variations in terms of the amendments contained in 'Scenario 2' of the document 'Statement of Amendments Required to Access Arrangement'. IN THE AUSTRALIAN COMPETITION TRIBUNAL Matter No 6 of 2002 RE: APPLICATION UNDER SECTION 38 AND/OR SECTION 39 OF THE GAS PIPELINES ACCESS (SOUTH AUSTRALIA) LAW FOR REVIEW OF THE DECISION BY THE AUSTRALIAN COMPETITION AND CONSUMER COMMISSION MADE ON 31 JULY 2002 TO DRAFT AND APPROVE AN ACCESS ARRANGEMENT FOR THE MOOMBA TO ADELAIDE PIPELINE SYSTEM UNDER SECTION 2.20(a) OF THE NATIONAL THIRD PARTY ACCESS CODE FOR NATURAL GAS PIPELINE SYSTEMS BY: EPIC ENERGY SOUTH AUSTRALIA PTY LTD TRIBUNAL: JUSTICE R E COOPER (DEPUTY PRESIDENT) PROFESSOR D K ROUND MISS M M STARRS DATE: 10 DECEMBER 2003 PLACE: BRISBANE (HEARD IN ADELAIDE) the tribunal BACKGROUND TO THE APPLICATION and ISSUES ARISING 1 Epic Energy South Australia Pty Ltd ('Epic') is the owner and operator of a natural gas pipeline system known as the Moomba to Adelaide Pipeline System ('MAPS'). By virtue of the Gas Pipelines Access (South Australia) Act 1977 (SA) ('the Act'), MAPS is a pipeline that is a 'Covered Pipeline' under the National Third Party Access Code for Natural Gas Pipeline Systems ('the Code'). As a pipeline listed in Schedule A to the Code, it has been a Covered Pipeline since the commencement date of the Code. The Code is applied as a law of South Australia, it being part of the Gas Pipelines Access (South Australia) Law ('the Law'): s 7 of the Act. The Law is contained in Schedules 1 and 2 to the Act, the Code forming Schedule 2. Epic is a 'Service Provider' as defined in s 2 of the Law. 2 As the Service Provider, Epic was obliged by s 2.2 of the Code to submit to the Australian Competition and Consumer Commission ('the ACCC'), as the Relevant Regulator, a proposed Access Arrangement ('AA') together with applicable Access Arrangement Information ('AAI') for the MAPS. The ACCC was obliged to consider the AA and approve it as submitted or as amended in accordance with the requirements of the ACCC (s 2.16 of the Code) or not approve the AA and itself draft and approve an AA instead of the AA proposed under s 2.2 by Epic. On 1 April 1999, Epic lodged its proposed AA for the MAPS with the ACCC. 3 Section 3 of the Code requires that an AA must include a Reference Tariff and Reference Tariff Policy which must, in the opinion of the Relevant Regulator, comply with the Reference Tariff Principles described in s 8 of the Code: ss 3.3, 3.4, 3.5 of the Code. For the purposes of s 8 of the Code, it is necessary to calculate the Total Revenue which is the revenue to be generated from the sales of all services over the AA period. Under s 8.4 of the Code, Service Providers are given a choice of three alternative methodologies to calculate the Total Revenue. The methodology chosen by Epic required that it include an estimate of the initial Capital Base included in Epic's AAI. The methodology chosen by Epic required the application of the Depreciated Optimised Replacement Cost ('DORC') in valuing the Covered Pipeline: s 8.10(b) of the Code. 4 The ACCC refused to approve Epic's proposed AA. 5 Epic refused to make the revisions to its proposed AA required by the ACCC to secure its approval. As a consequence, the ACCC pursuant to s 2.20 of the Code, drafted and approved its own AA instead of the AA proposed by Epic. Final approval under s 2.20(a) of the Code was given by the ACCC on 31 July 2002. 6 On 14 August 2002, Epic filed in this Tribunal an application for review of the decision of the ACCC to draft and approve an AA for the MAPS pursuant to s 2.20 of the Code and of all decisions relating thereto. On 27 November 2002, the Tribunal ruled that the application for review was, and was limited to, an application for review under s 39(1) of the Law. 7 In its application, Epic relied upon, inter alia, the following ground of appeal (at [15]): '15 ... (c) made errors of fact and exercised its discretion incorrectly and/or unreasonably and/or when the occasion did not arise having regard to the matters in (a) and (b) above, and in any event having regard to all the circumstances, in relation to: (i) the initial Capital Base ("ICB"), Rate of Return ("ROR") and Total Revenue and the establishment of a Reference Tariff; (ii) the definition of the Covered Pipeline in relation to the Pelican Point Power expansion; (iii) the Extensions/Expansion policy ("EEP").' 8 Section 39(2)(a)(ii) of the Law contained in Schedule 1 to the Act, provides: '39(2) An application under this section- (a) may be made only on the grounds, to be established by the applicant- (ii) that the exercise of the relevant Regulator's discretion was incorrect or was unreasonable having regard to all the circumstances;' 9 In Application by Epic Energy South Australia Pty Ltd [2002] AcompT 4 ('Epic No 1'), the Tribunal said: '[30] Section 39(2)(a)(ii) is concerned with the correctness or unreasonableness of an exercise of discretion having regard to the circumstances relevant to the proper exercise of that discretion. Those circumstances are ones which are demonstrable from the matters to which the Tribunal may refer under s 39(5). For the purposes of the subsection, error is made out if it is demonstrated that the exercise of the discretion was so unreasonable on the basis of the matters available to the decision maker that no reasonable decision maker could ever come to it: Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223 at 223 - 234. It also deals with the situation where the decision is so far outside the range of decisions open to a reasonable decision maker that it bespeaks of error even though the particular error cannot be identified: House v The King (1936) 55 CLR 499 at 505. For the purposes of s 39(2)(a)(ii) of GPA Law, correctness and reasonableness are to be determined by reference to applicable criteria contained in the Code applied to the matters which were before the relevant Regulator before the decision under review was made.' 10 The concept of 'Wednesbury unreasonableness' involves a consideration of whether the exercise of decision-making was ultra vires because it was in excess of the statutory grant of power: Wednesbury at 234; Minister for Immigration and Multicultural Affairs v Eshetu (1999) 197 CLR 611 at 649. A decision made beyond power by a Regulator is reviewable under s 39(2)(a)(ii) of the Law, it being either incorrect, or unreasonable due to the absence of any statutory power to sustain it. However, s 39(2)(a)(ii) is not limited in its operation to decisions which are unreasonable in the Wednesbury sense. 11 A decision may be unreasonable in either of two ways. First, it may be the product of some error made during the decision-making process. This would include acting upon a wrong principle, taking into account extraneous or irrelevant considerations and the like: House v The King (1936) 55 CLR 499 at 505. Second, the decision itself may be unreasonable or plainly unjust in all the circumstances thereby enabling it to be inferred that there has been an error to properly exercise the decision-making power: House v The King at 505. 12 Absent evidence of a particular error in making a decision sufficient to establish unreasonableness in the first sense, what test is to apply to determine whether the decision made is an unreasonable one in the second sense? 13 The context in which s 39(2)(a)(ii) of the Law appears is one of administrative review for correction of error falling within the grounds specified in s 39(2). The section requires that the applicant establish such error to the satisfaction of the Tribunal on the basis of the limited matters specified in s 39(5) of the Law. It is not a rehearing on the merits at large; albeit one limited to the materials that were before the Regulator: Epic No 1 at [20]. It is nevertheless a review by a specialist Tribunal with the power to substitute its own decision on the merits to correct established reviewable error. There is nothing in the context of the Law, or in the statutory objects of the Law, to suggest that the legislature intended to prevent interference with the decision of the Regulator other than where the decision was wholly untenable in the public law Wednesbury sense. 14 The word 'unreasonable' does not stand alone in s 39(2)(a)(ii). It forms part of the cognate phrase 'unreasonable having regard to all the circumstances'. There is within the phrase a causative relationship between the character of a decision as 'unreasonable' and all of the relevant circumstances which cause the decision to acquire that characteristic. The ordinary meaning of s 39(2)(a)(ii) is that the exercise of the relevant Regulator's discretion is unreasonable because the totality of the relevant circumstances, viewed objectively, render it so. 15 The question for the Tribunal in each case must be whether, in the judgment of the Tribunal, having regard to all relevant circumstances, the decision in its effect and operation is unreasonable for the reasons particularised by the applicant in the grounds relied upon to give a right of review. It is a straightforward factual test based upon the Tribunal's own appraisal of all the relevant materials that were before the Regulator subject to the limitations imposed by s 39(2) and s 39(5) of the Law. It is not an examination at large but one constrained by the particular circumstances relied upon by the applicant which it contends render the decision unreasonable. Obviously, different observers may hold different views as to what is unreasonable having regard to all the relevant circumstances. However, before it will intervene and substitute its own view, the Tribunal must be satisfied that the applicant has established that the decision under review is wrong for the reasons particularised in the ground of review. The Tribunal may then intervene to correct the error. 16 If the Tribunal is satisfied that the applicant has established to its satisfaction that the decision is unreasonable and therefore it bespeaks error, the Tribunal may, in the exercise of its discretionary powers given under s 38(9) of the Law, make such consequential orders, if any, as it considers necessary to remedy the error: s 39(6) of the Law. 17 On the hearing of the application, Epic relied on two grounds only in respect of which it sought review of the decision of the ACCC to draw and finally approve its own AA on 31 July 2002. The first ground concerned the use by the ACCC of the lowest cost estimate contained in a report of Microalloying International Inc ('Microalloying') of 7 December 2000 ('the Microalloying Report') for line pipe to establish the initial Capital Base of the MAPS as required by s 8.10 of the Code. The second ground concerned the decision of the ACCC to increase the System Primary Capacity ('SPC') of the MAPS from 323 terajoules ('TJ') per day to 348 TJ per day. This increase was to take into account an expansion in the capacity of the system flowing from an expansion of capacity of 25 TJ per day to satisfy the requirements of the operator of the Pelican Point Power Station. 18 The first ground involves the ACCC's application of the DORC methodology. 19 The DORC methodology requires that one determines the Optimised Replacement Cost ('ORC') of the existing pipeline. It is in this respect that Epic submitted the ORC value determined by the ACCC was unreasonable. Having regard to all the circumstances, Epic submitted that the method by which the ACCC estimated the costs of materials with respect to the pipeline, gave rise to the circumstances which render the ACCC determination of the DORC, and ultimately its approval of the AA, unreasonable. In particular, Epic submitted that the use of the lowest indicative price of $1053 per tonne for line pipe contained in the Microalloying Report, was: '... erroneous and unreasonable, given the conceptual level and indicative pricing involved, for the ACCC to assume that the most probable cost of supply of mainline pipe for any replacement of the Moomba to Adelaide pipeline would be the lowest indicative price from overseas'. 20 The second ground involves the decision of the ACCC to include the additional pipeline capacity arising from the Pelican Point expansion in the SPC of the pipeline system available for the provision of a firm transportation service ('FT service') being the Reference Service in the AA Epic lodged on 1 April 1999. background to the aCcc decision ON THE ORC 21 The ACCC formed the view that the ORC estimate of Epic in its AA was too high. The ACCC commissioned Connell Wagner Engineers ('Connell Wagner') to review the Epic pipeline design and costs. Epic's ORC estimate was based upon estimations by Connell Wagner as well as an expert report of GPA Engineering Pty Ltd ('GPA') dated 1 April 1999. Connell Wagner, like GPA, recommended different unit costs for 10 MPa and 15 MPa pipelines. When Epic received the Connell Wagner report, it engaged Stephen Timms Consulting ('STC') to review the Connell Wagner report. In the result, the experts were divided and their reports covered a wide range of pipeline unit costs. 22 The ACCC decided to make its own estimates of the ORC for the MAPS. It did not use unit costs but rather 'built up' costs from the input components to pipeline construction, and actual construction conditions. The explanation of the method is contained in Annexure 2 of the Draft Decision (at 199): 'Pipeline costs are dependent upon a range of variables including size, pipeline location, terrain, rock, the tendering process, competition between pipeline suppliers, the exchange rate and project management, any of which may at any stage have a major influence on the total cost. Competition between suppliers of the major materials such as pipe or of pipeline construction services, improvements in industrial relations and changes in technology have, if anything, tended to ensure that Australian unit pipeline construction costs have declined over time. The Commission's analysis has attempted where possible to start from unit costs that reflect current conditions, and to take the preceding factors into account. As such, it relies on a more detailed methodology than that employed by Epic in its submission. The analysis addresses the effect of factors such as the price of the pipe and the difficulty of construction by breaking up the mainline and lateral pipelines into smaller, individual sections and takes into account factors such as pipe specification, design pressure, locality and terrain. Compressor pressure regulating, meter, scraper and valve stations and other facilities normally found on a pipeline system are also costed separately, again taking into account factors similar to those listed for the pipeline.' 23 The ACCC concluded that Option D, a 610 mm (or 24 inch) diameter pipeline operating at 10 MPa and one which was very similar to the Option D proposed by Epic, was the lowest cost option and that the mainline and loop line costs were $18 400 per inch.km and $24 500 per inch.km respectively at June 2000 price levels. Some unit costs adopted by the ACCC for other components of the pipeline system were higher than those proposed by Epic, and some costs were higher because of differences in the optimised design. A spreadsheet costing model was developed by the ACCC to apply its method. 24 Following the release of the Draft Decision on 25 August 2000, the ACCC provided Epic with a report of a review of the spreadsheet costing model by Sinclair Knight Merz ('SKM') dated 24 August 2000 and offered to provide a copy of the model to Epic subject to a confidentiality agreement. Of relevance to the present Tribunal review, the SKM report endorsed the line pipe cost of $1250 per tonne used by the ACCC in its model, preferred the ACCC method over that of Epic because it was more detailed, and agreed that it 'generate[d] a 50th percentile estimate of replacement costs'. 25 Epic sought expert opinion on the ACCC pipeline construction costs from Worley Limited on 27 September 2000 and from Venton and Associates ('Venton') on the costs and method more generally on 12 October 2000. Both considered that the unit pipeline construction cost was too low, however this is not relevant to the matters under review by the Tribunal. Venton queried the type of steel used by the ACCC in estimating its pipeline costs and made comments on the level of other cost components. 26 In November 2000, the ACCC commissioned Microalloying to investigate the current price of line pipe. The Microalloying Report provided a range of values for various types of steel pipe based on data 'from a variety of the company's international sources, including recent bid packages and are considered very reliable.' 27 The Final Decision of the ACCC dated 12 September 2001 retained Option D and listed changes made to the optimised pipeline design and unit costs. Of relevance to the Tribunal review, a line pipe cost of $1053 per tonne, the lowest price in a wide range presentedin the Microalloying Report, was adopted to estimate the ORC of the pipeline. A second review by SKM dated 29 August 2001 (that came to generally the same conclusions as the first report) reported, in particular, that the detailed method adopted by the ACCC was more robust than that adopted by Epic, and agreed again that the method 'generate[d] a 50th percentile estimate of replacement costs'. 28 The Final Decision included a table comparing the main cost components of the ACCC's Option D with Epic's Option B in June 2001 price levels. The ORC values were similar, although as the options had different pipeline capacities, pressures and diameters, the similarity is coincident. Comparison of the unit pipeline costs per inch.km in these circumstance is not appropriate. The ACCC mainline and loop line costs were $20 100 per inch.km and $25 800 per inch.km respectively in June 2001 price levels. 29 The Microalloying Report, the SKM report and the revised spreadsheet model were provided to Epic after the release of the Final Decision. In a submission to the ACCC on 22 January 2002, Epic advised that it did not make all the required revisions to the proposed AA because it did not agree with the method used to estimate the initial Capital Base. A further submission on 1 July 2002 raised concerns in detail about the method and unit costs used by the ACCC in its final decision. Of relevance to the Tribunal review are the cost of line pipe from the Microalloying Report, and the meaning of the SKM endorsement of the ACCC's ORC estimate. 30 As Epic did not make the changes to its proposed AA required by the ACCC, the ACCC drafted and approved its own AA. 31 The cost of line pipe, one of the major input costs, was obtained from the Microalloying Report, adjusted forexchange rate movements since Epic lodged its proposed access arrangement. In addition, the ACCC Statement of Facts and Contentions states that the reason for seeking line pipe costs was that Venton suggested that a different type of steel should be used (X70 not X80) than was used by the ACCC in the Draft Decision. For the quality of line pipe that would be required by Epic, the following estimates of the prices ($A per tonne at then-current exchange rates, landed in New South Wales), sourced from manufacturers in six countries/areas, were provided: Greek ($1053); Korean ($1190); Japanese ($1235); Australian ($1255); North American ($1270); and Brazilian/Argentinean ($1340). The ACCC used the lower bound cost of $1053 per tonne in the estimated ORC contained in the Final Decision. 32 The Final Decision contained no rationale for the selection of the lower bound price provided by Microalloying. The Final Approval continued to use the lower bound price, with adjustment for exchange rate movements so that it was then $1120 per tonne. The use of the lowest price from theMicroalloying Report was defended by the ACCC on the grounds that the prices were observed and not estimated, because the prices were observed there was a reasonable degree of certainty, and a reasonable and prudent Service Provider would choose the lowest priced line pipe. It was also noted that some other costs used by the ACCC were higher than those proposed by Epic. 33 Epic made two alternative submissions about the cost of line pipe used by the ACCC: either the ACCC was in error in selecting the lower bound price because of assumptions about the precision of the prices provided by Microalloying; or it was unreasonable in the circumstances. 34 Epic argued that the Microalloying Report provided broad indicative prices, based on the wording in the Report, letter of engagement and Microalloying's response, so that the use of the lower bound price was erroneous and unreasonable. The prices should be regarded as indicative for a number of reasons: the prices related to a limited number of pipeline criteria, did not purport to be firm prices that would be supplied on tender, and did not relate to a specific pipeline. As counsel for Epic submitted, the prices provided: '... an indicative pricing level which may never be guaranteed at the relevant time. It's only indicative pricing. To take the lowest of those prices doesn't do justice to what is reasonable and substantiated for the purposes of developing an initial capital base which provides the service provider with an appropriate return on capital for its investment. That's really the point.' 35 In these circumstances, Epic submitted that the lower bound price would not be selected but rather the mean or the mean excluding the highest and lowest prices. In its submissions in reply, Epic referred to the use of the median of the relevant prices provided in the Microalloying Report but did not pursue that option in the 'Statement of Amendments Required to Access Arrangement'. 36 In written submissions on behalf of Epic, it was also pointed out that there was no indication that any of the prices in the Microalloying Report came from bids/mills that had the capacity to supply the whole order, that it was usual to contract with a domestic supplier who may then sub-contract from overseas, and that there appeared to be inconsistencies in the prices of other line pipe sizes included in the Microalloying Report. These arguments were not pursued by Epic's further written submissions nor at the hearing before the Tribunal. background to the ACCC decision on the pelican point power station capacity 37 Section 3.1 of the Code requires Epic to include in its AA a policy on the service or services to be offered. The Reference Service offered by Epic was a FT service, together with a rebatable service, being an interruptible transportation service ('IT service'). The SPC of the pipeline system for the FT service provided in Epic's AA was 323 TJ per day. This figure represents the reliable capacity of the system. The notional maximum stated capacity of the system was 393 TJ per day. 38 After Epic lodged its AA for approval on 1 April 1999, it expanded the capacity of MAPS by looping thepipeline. This expansion was to provide additional services to the operator of the Pelican Point Power Station. 39 The treatment of an extension to, or expansion of, the capacity of a Covered Pipeline as part of the Covered Pipeline is dealt with by s 1.40 of the Code. That section provides: '1.40 An extension to, or expansion of the Capacity of, a Covered Pipeline shall be treated as part of the Covered Pipeline for all purposes under the Code if the Extensions/Expansions Policy contained in the Access Arrangement for that Covered Pipeline provides for that extension or expansion to be treated as part of the Covered Pipeline.' 40 The requirements to be included in an Extensions/Expansion Policy in the AA are set out in s 3.16 of the Code. 41 From the time that the AA was proposed until the Final Approval, there were several changes made to the Extensions/Expansion Policy ('E/EP') by both Epic and the ACCC. 42 The main changes to the E/EP during the development and approval process related to the treatment of extensions to the pipeline as opposed to expansions of capacity, and whether the Service Provider (Epic) or the Regulator (ACCC) had responsibility for deciding whether an extension or an expansion would become part of the MAPS and thereby become part of the Covered Pipeline. The AA drafted and approved by the ACCC contained an E/EP that provided for extensions to be included in the Covered Pipeline at the election of the Service Provider, for expansions constructed prior to the AA coming into operation to be covered, and for other expansions to be included unless the Regulator agreed that they should not be covered. This was given effect by cl 10.10(b) which provided: '(b) Any New Facility, except for an extension to the Pipeline, is to be considered part of the Covered Pipeline: (i) in the case of a New Facility which came into operation before this Access Arrangement became effective, upon this Access Arrangement becoming effective; (ii) in the case of any other New Facility at the time it becomes part of the Covered Pipeline, unless the Regulator, before the decision to construct the New Facility is made by the Service Provider, agrees that it should not be covered. Extensions to the Covered Pipeline will not be part of the Covered Pipeline, unless the Service Provider, by notice to the Regulator (given before the decision to construct the extension is made by the Service Provider) elects otherwise.' 43 The Pelican Point expansion, the subject of the decision under review, is an expansion of the first kind, that is it was constructed prior to the AA coming into operation so it automatically became part of the Covered Pipeline. 44 In its Final Decision, the ACCC argued for the inclusion of all future capacity expansions in the Covered Pipeline, including Pelican Point, for the reason that Epic may have market power. Section 2.24 of the Code lists several factors that the Regulator must take into account in assessing an AA. The AA made by the ACCC relied exclusively on the existence of excess demand for the gas transportation service. That in turn was used to suggest that a degree of market power may be able to be exercised by Epic in setting terms and conditions, with the potential to extract monopoly rents. The potential consequences of such actions were said to be lower investment and entry in downstream markets, reduced competitiveness of entrants in downstream markets, lower efficiency gains, and higher gas prices which may lower economic growth in South Australia. As a result, expansions should be covered to ensure the economically efficient operation of the Covered Pipeline (s 2.24(d) of the Code), that the public interest, including the public interest in having competition in markets (s 2.24(e) of the Code), and that the interests of users and prospective users (s 2.24(f) of the Code) are met. The ACCC also noted that coverage of expansions would enable prospective users access to the dispute resolution processes of the Code. 45 In response to the Final Decision, Epic said that the ACCC had not properly considered its commercial interests or the state of the market in South Australia as at 1 July 2002. The commercial interests discussed were the fact that the Reference Tariff would be used as the benchmark for the re-negotiation of the existing haulage contracts which all expire at the end of 2005, the expected costs of future capacity expansions which are expected to be higher than the costs of existing capacity as the pipeline is now fully compressed, and Epic's aim to maximise the capacity of the pipeline. Epic also claimed that the threat of coverage would be a constraint on its use of any market power and that one of the primary reasons expansions were included by the ACCC was that it wanted the Pelican Point expansion to be covered, that is, an expansion after lodging but before approval of the AA. 46 Epic provided a separate and lengthy confidential submission dated 18 January 2002 on the state of the market, rebutting the analysis of demand and stranding risk the ACCC undertook for the Final Decision. This submission presented different potential demand figures for the MAPS in the future, taking into account new infrastructure and revised bids for the use of the MAPS. It was as a result of this submission that the ACCC amended its position on the coverage of future capacity expansions in the Final Approval. 47 In the Final Approval, the ACCC repeated its view that expansions of capacity should be included where there was excess demand for transportation services. It agreed with Epic that since the Final Decision uncertainty had arisen about whether excess demand would continue, so the E/EP in the ACCC drafted and approved AA was amended to the effect that future capacity expansions would be considered on a case-by-case basis prior to their construction. The ACCC said the reason for coverage of expansions was concern about the exercise of market power, not coverage of the Pelican Point expansion per se, as claimed by Epic. The Pelican Point expansion remained covered 'because Epic would be in a position to exercise market power in respect of that capacity in the future', as argued in the Final Decision. In coming to the decision to include the Pelican Point expansion in the regulated coverage, the ACCC did not discuss Epic's commercial interests and the impact that the decision would have on those interests. 48 The discussion by the ACCC of market power in the Final Decision and Final Approval was in terms of potential rather than actual consequences. There was no mention of the s 2.24 factors other than those mentioned in par [44] of these Reasons, no attempt to assess the relative importance of the factors and no evidence of any analysis of the size and extent of the market power and its effects. The only reason given for why Epic may have market power was excess demand, and that had reduced between the Final Decision and Final Approval. That reduction led the ACCC to amend the E/EP for future capacity expansions but not for the already completed Pelican Point expansion. 49 Epic and the ACCC agreed that the capacity available for the FT service was 323 TJ per day, in the absence of the Pelican Point expansion. The ACCC added the Pelican Point capacity expansion of 25 TJ per day (giving a total capacity of 348 TJ per day for FT service) and added an additional amount of $11 million to the initial Capital Base. 50 Other than the view of the ACCC that Epic would be in a position to exercise market power in respect of the capacity represented by the Pelican Point expansion in the future, no other reason was given or relied upon by the ACCC for the decision it made to include that capacity for the purposes of the AA. 51 Epic submitted that the reasoning of the ACCC was erroneous, within the meaning of s 39(2)(a)(i), in that it incorrectly found that Epic was in a position then, or would be at any future time during the relevant period, to exercise market power in respect of the additional capacity created by the Pelican Point expansion. The error lies, Epic submitted, in the knowledge of the ACCC that the MAPS was fully contracted until 2006 and there was no available capacity for the MAPS in respect of which Epic would have an opportunity of exercising market power during the period of the AA. Further, the additional capacity of the Pelican Point expansion was contracted to 2019. Alternatively, Epic submitted that, having the knowledge which it did as to the lack of capacity available to Epic due to existing contractual arrangements, the decision of the ACCC to include additional capacity of 25 TJ per day in the SPC of the MAPS was incorrect or unreasonable within the meaning of s 39(2)(a)(ii) of the Law. 52 At the hearing, the ACCC sought to sustain its position on the market power associated with the Pelican Point expansion by reference, not to the Reference Service (FT service) but to the IT service. This was the first time that that position had been detailed as the basis of the ACCC's decision on the potential exercise of market power by Epic. 53 The ACCC argued that the Pelican Point expansion did not only create an FT service, but a commensurate IT service and market power could exist in respect of that service. It argued that an IT service would be available to other pipeline users if Pelican Point did not need to use its contracted FT service on a particular day, which IT service had a potential capacity of up to 25 TJ per day, depending on the extent to which Pelican Point did not take its contracted FT service. An incentive and risk-sharing mechanism was included in the AA such that the IT service was rebatable: the tariff for IT service was a fixed proportion (1.15) of the Reference Tariff and the revenues raised from IT service tariffs were shared between Epic and any user with an FT contract, in this case Pelican Point. The ACCC argued that Epic had the ability to exercise market power as a result of this commercial interest in the IT service, despite the FT service being fully contracted until 2019. To show error on the part of the ACCC, it was argued that Epic had to demonstrate that it had no market power at all and that had to be done with the factual material that was before the ACCC at the time of its decision. 54 In response, Epic submitted that Reference Services and Reference Tariffs apply to a significant part of the market by virtue of s 3.2(a)(i), s 3.3(a) and (b) of the Code. The IT service is not a Reference Service in the AA drawn up and approved by the ACCC. By implication the ACCC had determined that the IT service does not apply to a significant part of the market. The existing IT service arose due to the difference between the notional maximum capacity and FT service capacity of the MAPS and the potential use of FT service when it was not required by a contracted user. The difference between the notional and FT service capacity of 393 TJ per day and 323 TJ per day respectively without the Pelican Point expansion, or, 418 TJ per day and 348 TJ per day respectively with the Pelican Point expansion, was 70 TJ per day. 55 The ACCC accepted that only the FT service capacity (Reference Service) would be regulated as to terms and conditions of use, and be used to calculate the Reference Tariff. In other words, it did not use the potential market power associated with the 70 TJ per day or the unused FT service to require that the IT service be a Reference Service under the AA. The ACCC adopted a 'wait and see' position with respect to whether the IT service should be regulated. It also did so for future capacity expansions as a result of the change to the E/EP between the Final Decision and the Final Approval because of the uncertainty with respect to excess demand. 56 Given this, Epic submitted that it was not reasonable for the ACCC to determine that the Pelican Point FT service expansion of 25 TJ per day should be regulated because of the potential market power associated with any IT service that arises in respect of that capacity, and at the same time to determine not to regulate the potential 70 TJ per day presently available for the IT service. CONCLUSION ON THE ORC ISSUE 57 A key component of Epic's ORC was the cost of line pipe. Such high quality high strength pipe was available from a wide spectrum of both domestic and overseas suppliers. There appear to be no supply shortages, with capacity increasing internationally and with new pipe producers entering the market in Australia and overseas. 58 In its Report dated 7 December 2000 to the ACCC, Microalloying advised that international supply trends, together with producers in Korea and Greece showing a willingness to cut prices, led it to the conclusion that 'the highly competitive market is likely to continue for the foreseeable future', and that 'competitive pressures, rather than currency fluctuations will continue to dominate pricing practices and policies'. 59 Microalloying provided the ACCC with line pipe prices from six countries: see par [31]. 60 Microalloying did not disclose the identity of the manufacturers supplying line pipe cost quotations in each case, nor did it disclose how many estimates were obtained in each country, nor the range of prices obtained from within each country. It stated that the data, some of which was confidential and proprietary, had been collected from a variety of its international sources, including recent bid packages, and that they were considered to be 'very reliable', and 'would be expected to be achievable as a result of a competitive bidding process'. 61 However, in a situation of flexible prices, a price quoted on one day could be quite different to one quoted for the same job a month later. Indeed, Microalloying had warned the ACCC in a fax dated 17 November 2000 that 'Pipe prices can be volatile and may change within weeks or months depending on the "go-ahead" of major projects for construction'. It also said that its report would 'background our understanding of the world market situation at this point of time' but cautioned that it was 'not in a position to offer informed comment on market trends into the future except in a general sense'. 62 The Tribunal notes that quoted prices from suppliers are very much job- and time-specific, and thus indicative prices obtained in this manner may not provide a useful or accurate guide to the appropriate cost of line pipe to be included in Epic's ORC. An accurate costing of such line pipe can only finally be determined after a tendering process based on detailed specifications. Normally, in such circumstances, sound business practice would dictate accepting the lowest tender bid, taking all other factors into account. The ACCC has treated the Microalloying figures with a degree of specificity and certainty that was inappropriate given the qualifications Microalloying put on its findings. Such treatment would be more appropriate to the decision to be made by a pipeline operator when evaluating firm tender bids for the supply of specified line pipe. Here, a reasonable and prudent Service Provider would naturally select the lowest bid consistent with quality, delivery and other commercially important requirements, as a decision must be made with important implications for the firm's capital base and its rate of return. 63 For planning purposes, however, this price cannot be known with any certainty and a prudent operator would likely find it to be commercially unwise to plan a pipeline project based on the lowest known line pipe cost, or even the average line pipe cost of suppliers in the lowest-cost producing country. The risk here is highly asymmetric, all on the upside. Thus a prudent operator, in the absence of perfect information, would factor into its estimates the expected value of line pipe costs, based on its estimation of the range of likely future prices and the assessed probability of occurrence of each possible price. 64 In the absence of knowledge of such a probability distribution at the planning stage, an operator might therefore obtain some indicative estimates based on less-than full information being available, compared with a specific tender to job specifications, and take either a simple arithmetic average, a modified arithmetic average, or the median of these prices as the indicative planning parameter value. It would be a highly risky commercial action to take the lowest figure found in any such non-detailed price-seeking activity. 65 However, in the situation at hand, a forward estimate was needed, in an ex-ante or planning sense. Tenders had not and could not be called in the circumstances. A representative figure was needed that was as accurate as could be hoped for, given the inability to specify precisely the necessary parameters, the way in which the price data was collected, the relatively small number of prices that the Tribunal infers were obtained, and given the lack of any statement of statistical reliability that could be attached to the findings presented by Microalloying. 66 In such a situation the price of line pipe may be considered as a random variable. For a given grade of line pipe, delivery terms and so on, different prices may be quoted by different suppliers for a number of reasons. Each of these prices has associated with it some probability of occurrence. For planning purposes, in the absence of formal bids, a prudent Service Provider would be justified in following SKM's advice to select the 50th percentile or median value of the range of indicative prices, one that is just as likely to be exceeded as it is to overstate the relevant price. Alternatively, it could determine the expected value of the parameter, and this average would normally be close to the median if there were not too much skewness in the probability distribution of prices. 67 In such circumstances, taking an average of the prices recorded for each of the six countries, would be what a prudent operator might reasonably be expected to do, in order to determine a representative expected line pipe costing in the presence of incomplete and imperfect information. 68 For the figures provided by Microalloying for the six countries/regions, the simple average price was $1224. Omitting the highest and lowest figure from the calculation yields an average of $1238, while the median, or 50th percentile value, was $1245. Each of these amounts is considerably higher than the Greek price of $1053, the lowest country/region price that was presented by Microalloying, and the value finally selected by the ACCC. 69 The Tribunal observes that this specific price for Greek line pipe was not recommended to the ACCC by Microalloying as the one that should be used to calculate the ORC. 70 The ACCC itself made the decision to use the lowest indicative single country price provided to it in the Microalloying Report. This decision was taken in spite of advice provided to it by SKM in the report dated 24 August 2000, which stated: '... the cost estimate appropriate for this application should be made at the 50th percentile of estimating uncertainty ... This is different from a project budget cost estimate which would normally include extra contingency amounts to reduce the risk of the estimated cost being exceeded.' The Report concluded that it agreed: '... with the methodology adopted in the Commission's analysis to generate a 50th percentile estimate of the replacement costs'. 71 Subsequently, SKM was asked to provide the ACCC with an update of this desktop review, due to changes in key parameter values and in light of the Microalloying Report. This second report, dated 29 August 2001, appears to the Tribunal to consist largely of a reproduction of the earlier report, with a few additional paragraphs inserted to cover new developments. Herein lies a problem. 72 SKM still advocated the use of the 50th percentile cost estimate on the second page of its second report, using identical words to its previous report. However, on the third page, when discussing the ACCC's use of the low Greek price for line pipe, it cautions that: 'It is the case that the purchaser would logically accept the lowest price tendered from an acceptable supplier able to meet the quality and delivery requirements. It is of course an element of high uncertainty to use such data as a predictor as to what might be bid for piping material for replacement piping for the system as the specification circumstances may be different to that for which the Microalloy [sic] particular price was obtained'. The Tribunal agrees with these sentiments. 73 Yet the SKM report then goes on to indicate somewhat incongruously in the following sentence that 'There is nothing however to indicate that selection on this basis provides a biased estimator however one-way-or-the-other'. Epic argued that this conclusion was inconsistent with the preceding sentences. 74 The Tribunal agrees that in a situation where volatility in line pipe prices over time is quite evident, and where considerable price dispersion exists at any given time between suppliers in different countries (and quite possibly between suppliers within any given country, although no evidence was produced to that effect), there is a great risk that by using the lowest price obtained in indicative quotes from one country there could result a significant underestimation of the line pipe costs to be included in Epic's ORC. 75 This is especially so when it is considered that Microalloying provided no individual supplier data to the ACCC, nor did it reveal how many price quotes it received for each of the countries/regions, nor whether its figures were based on a truly random sample of representative prices. In addition, no estimate was provided of the error allowance that might be expected to accompany its indicative figures. 76 There is no dispute between the parties that line pipe can be sourced from many manufacturers around the world. The exercise of estimating the ORC for Epic's pipeline is of necessity a theoretical one done in the name of regulatory appraisal - no such pipeline was in fact being built. However, if it had sought tenders for line pipe, Epic (or any other prudent pipeline operator) would clearly have seen it as a commercial imperative to contract with the supplier offering the lowest price consistent with its engineering specifications and corporate priorities. 77 This is the logic that the ACCC appears to have used. But the exercise that confronted it was an ex-ante planning or indicative one, a conceptual estimating process divorced from precise tender specifications. It was common ground that s 8.10 requires a 'fairly high level of estimation', as described by counsel for Epic. In its submission to the Tribunal, the ACCC accepted that 'the ACCC's function under the Code is not to replicate the detailed assessment process undertaken by engineers prior to the construction of a pipeline'. It went on to say 'it is important that the estimates are reasonable and substantiated'. 78 Even though prices in the world market may have been competitive and trending downwards (but there is no evidence to this effect apart from the qualified and descriptive assessment made by Microalloying), the Tribunal finds that the selection of the Greek price by the ACCC assumed a degree of specificity appropriate to a final tender bid selection, where all operating parameters would have been completely specified and prices would be quoted as firm. There is no evidence that such precision in specification was used by Microalloying in getting its quoted prices. The ACCC was only told that they came from 'a variety of the company's international sources, including recent bid packages'. 79 In its 1 July 2002 submission to the ACCC, Epic argued that the ACCC's use of unit costs at the lower end of the range was 'an unreasonable approach, since there are often commercially sound reasons for not adopting the lowest cost offer. While this approach might be useful in establishing whether a proponent is deliberately overstating the cost, it is not a reasonable basis for establishing the probable cost of the project.' 80 Counsel for the ACCC argued that nothing more had been said by Epic on this matter - it had, for example, put in no contrary material by a competing expert. It should be noted that Epic did not dispute the accuracy of the prices produced by Microalloying in its submission to the ACCC - a fact correctly noted by counsel for the ACCC but then used as evidence that Epic had raised no problems with the estimation process. 81 It was put to the Tribunal by counsel for the ACCC that Epic had in fact endorsed what Microalloying had presented in Epic's submission to the ACCC. The Tribunal reads it differently. Epic's complaint has never been with the Microalloying Report as such. It accepted that Microalloying drew 'appropriate' conclusions in its report regarding costs, but pointed out that it had ignored quality differences between different pipe suppliers. 82 In its submission, Epic had noted that SKM had recommended use of a 50th percentile price, and agreed with this procedure, but went on to point out that 'none of the estimates have been undertaken at a level of detail that would allow this value to be reliably determined.' It also argued that: 'Industry experience is that estimators invariably underestimate project costs because they attempt to undertake accurate estimates, but in doing so do not provide sufficient contingency for those items that are omitted from the scope definition at the time, and those items that changed between the date of the estimate and the date of completion.' 83 Epic also submitted to the ACCC that the apparent endorsement by SKM of the lowest line pipe price provided by Microalloying conflicted with its earlier statement (both in time and within the second SKM report) that the proper estimate should be at the 50th percentile of indicative prices. Epic pointed out to the ACCC that 'this approach would require firstly validation of the quality and production capacity of the mills, the landed cost, coating costs, and then the application of a median unit cost to the estimate.' 84 The key issue here is that Microalloying did not recommend the Greek price - the ACCC made that decision, presumably because it was the lowest provided in Microalloying's survey. No evidence has been presented to the Tribunal that this choice was the result of a rigorous and systematic evaluation process. 85 It is with the process of the ACCC's reasoning that Epic took issue. It put the ACCC on notice that the median price was the appropriate one to use in the circumstances, as had been recommended to the ACCC by SKM. Yet in its Final Approval, the ACCC dismissed Epic's claim, albeit acknowledging the relative importance of line pipe costs (about thirty per cent during construction) in the calculation of Epic's ORC. 86 In the Final Approval, under a heading 'Reasonableness of lowest unit costs commercially offered', the ACCC stated without any detailed argument that on the basis of the figures provided in the Microalloying Report that it had 'adopted prices at the lower bound of the commercially offered prices for pipe as there was a reasonable degree of certainty involved'. It concluded that it 'considers that such a choice would also be the choice adopted by a reasonable and prudent service provider'. 87 In concluding thus, the ACCC has ignored the advice of SKM, or alternatively has followed the later, albeit inconsistent, advice provided by SKM in its second report, detailed above. It has provided no supporting argument as to why it believed there was 'a reasonable degree of certainty involved', or to what this certainty related - whether it was to the accuracy of Microalloying's reporting and calculations, or to accurately specifying the requirements for the Epic pipeline in its search for indicative prices, or to some statistical confidence in the estimating procedure followed by Microalloying. 88 In his oral submissions counsel for the ACCC stated that the issue for the Tribunal was whether the ACCC was 'wrong in accepting (a) the nature of this report and (b) that it was logical to calculate [the line pipe cost] by reference to the lower price', especially in light of what he claimed to be no joinder of factual issue by Epic with the ACCC on this matter - it was, he avowed, 'bare assertion'. The Tribunal does not accept these contentions. 89 The Tribunal finds that in raising its disquiet in its July 2002 submission, Epic had in fact put the ACCC on notice that it disagreed with its assessment. Epic clearly regarded the use by the ACCC of the lowest observed country price as unreasonable in practical terms, and as being inconsistent with what a prudent service operator would have done in similar circumstances. 90 The Tribunal agrees with the submissions of counsel for Epic, set out in paragraph [34] above. 91 The ACCC regarded s 8.1(a) and s 8.1(b) of the Code as requiring the use of the lowest line pipe cost such as the Greek price provided by Microalloying. Counsel for Epic argued that the ACCC had previously recognised in other decisions that 'the service provider has to be provided with an opportunity to earn a stream of revenue that recovers the efficient costs of delivering the reference service'. As a matter of principle, he put it that in the present circumstances there was no 'tendered quote which has all the certainty and firmness that one could then use to apply the reasoning in 8.1(a) [and (b)]'. 92 The Tribunal does not accept, as the ACCC asks it to do, that s 8.1(a) and s 8.1(b) of the Code require the ACCC to adopt the lowest stated figure in the Microalloying Report. Epic must be allowed the opportunity to earn a revenue stream that recovers the efficient costs of operating the Reference Service, and the need to replicate the outcomes of a competitive market does not demand the use of the lowest indicative price based on general, albeit informed, inquiries. Nor does the Tribunal accept, as was put by Counsel for the ACCC, that the ACCC's use of the Greek line pipe cost from the Microalloying Report could be justified because it was looking for 'greater accuracy, greater commerciality'. It is the process followed by the ACCC that was flawed and therefore was unreasonable, not the price chosen as such. 93 The ACCC has acted as if the Microalloying prices were actual firm prices tendered to precisely engineered specifications. It is true that the prices reported by Microalloying could be described as observed commercial prices, but they did not necessarily refer to line pipe with the same specifications, geographical context, delivery conditions et cetera as the Epic pipeline. If they had been actual firm prices resulting from a tender process, then a prudent Service Provider acting efficiently would of course select the lowest value. But the exercise was not for the purpose of building a pipeline, as counsel for Epic frequently pointed out. Rather, the ACCC's consideration occurred at a much earlier conceptual stage where caution was necessary and ex-ante expected values were both more commercially prudent and more reasonable in a regulatory sense. 94 In taking the approach it did, the ACCC exposed Epic to an asymmetric risk whereby the likelihood of underestimating the true actual line pipe cost was much greater than that of overestimating it. To take the lowest price from one source of supply, runs the risk of serious commercial understatement of the expected cost of line pipe. 95 The Tribunal finds that the ACCC made a choice in selecting the Greek price for line pipe for the determination of the ORC that in all the relevant circumstances is unreasonable. It is not an exercise of discretion that a reasonable person would have made in all the circumstances. It falls beyond the boundaries of what a prudent commercial operator would objectively be expected to do in these forward-looking planning circumstances. 96 Based on all the materials that were before the ACCC up to the time of its Final Approval, the Tribunal finds that the particular circumstances set out above rendered the ACCC's determination with respect to the allowable cost of line pipe unreasonable. As a consequence, the Final Approval has been affected by reviewable error and it becomes necessary for the Tribunal to make orders to remedy the ACCC's error. average of the microalloying figures 97 The final matter requiring resolution by the Tribunal is the method of calculating the average line pipe cost from the figures in the Microalloying Report. The alternatives are a mean of the six figures, a mean excluding the highest and lowest prices, and the median. As noted in par [68], these methods give $1224, $1238 and $1245 respectively in December 2000 price levels. No preference was stated by the parties, although the median was not pursued by Epic in the 'Statement of Amendments Required to Access Arrangement'. The choice for the Tribunal is therefore between the mean and the mean excluding the highest and lowest prices. 98 In selecting the appropriate method, some judgment is required as there is no theoretical or practical rationale to argue for one method over the other. The Tribunal has had recourse to the factors in s 2.24 of the Code, the factors that the Regulator must take into account in assessing an AA. Epic's commercial interest is accommodated and given substantial weight by the increases that will be made in the ORC by setting aside the decision of the ACCC and substituting either mean figure. Each figure produces a significantly higher figure for the line pipe cost. In these circumstances, consideration of the interests of users and prospective users, and the public interest must be given significant weight in deciding the best method of calculation of the line pipe cost. The method giving the lowest figure for the mean of the six figures in the Microalloying Report, having regard to the respective interests to be considered, is appropriate. That method is the unadjusted mean. 99 The selection of the unadjusted mean implies that Scenario 2 in the 'Statement of Amendments Required to Access Arrangement' be adopted for changes to the AA consequent upon the Tribunal's review. conclusion on the pelican point power station capacity 100 The matter subject to review is whether the ACCC erred by including the Pelican Point expansion as part of the Covered Pipeline because Epic was said to be able to exercise market power in respect of that expansion, or alternatively, whether its inclusion was unreasonable having regard to all the circumstances. 101 The ACCC gave as the basis for the inclusion of the Pelican Point capacity expansion its belief that excess demand for gas transportation could potentially lead to Epic being able to exercise 'a degree of market power in setting the terms and conditions, including tariffs, for an expansion ... because it is not [otherwise] constrained by competition or regulation'. 102 The existence of excess demand is only one indicator of whether market power exists. Its presence is neither necessary nor sufficient for the exercise of market power. Other factors include the structure and operation of the market for gas transportation, supplies of gas from upstream markets, demands for gas in downstream markets, contracts for services, substitution possibilities involving other forms of energy and, in the longer term, competition from new pipelines, commercial incentives to sell services, and the countervailing power of gas producers and gas users. None of these factors was considered by the ACCC. 103 Not only was a sole indicator of market power mentioned, namely excess demand, but the certainty associated with the presence of this factor was not high. The level of uncertainty was such that the E/EP was changed between the Final Decision and the Final Approval, with the consequence that future capacity expansions are not to be included as of right in the Covered Pipeline. 104 The purpose of coverage of a pipeline is to prevent the consequences of misuse of market power. If the market power is not significant then it cannot be used effectively to prevent an economically efficient pipeline operation or to deter competition in downstream or upstream markets or thwart the interests of users or prospective users. In Duke Eastern Gas Pipeline Pty Ltd [2001] AcompT 2 at [75], the Tribunal held that the increase in competition to result from coverage should be 'not trivial' before a pipeline was brought under the provisions of the Law for regulation as a Covered Pipeline. That is, given the costs of coverage, coverage should be pursued only if a 'not trivial' increment to competition can be expected to follow. 105 In response to Epic's submissions set out in par [51] above, the ACCC at the hearing, and for the first time, submitted that the Pelican Point expansion did not only create an FT service, but a commensurate IT service and that 'Epic could or would be in a position to exercise market power in relation to this extra 25 terajoules'. An IT service would be available to other pipeline users if Pelican Point did not need to use its contracted FT service on a given day. This capacity (up to 25 TJ per day) would then be available for other users or potential users. 106 An incentive and risk-sharing mechanism is included in the AA such that the IT service is rebatable. The tariff for IT service is a fixed proportion (1.15) of the Reference Tariff and approximately half of the revenues raised from IT service tariffs are retained by Epic with the balance rebated to the customer with an FT contract, in this case Pelican Point Power. 107 In the hearing the ACCC argued that the source of Epic's market power was this commercial interest in the IT service. 108 In the Tribunal's view, it is not reasonable to argue that the Pelican Point FT service expansion of 25 TJ per day should be now be regulated because of the potential market power associated with any IT service that Epic might be able to exercise. 109 The Tribunal does not accept the submission put by counsel on behalf of the ACCC at the hearing, that any level of market power justifies coverage. The ACCC, when it decided to regulate the IT service associated with the Pelican Point expansion, must also have formed the view that Epic had market power in respect of the IT service which was significant enough to justify coverage of the expansion and that coverage would increase competition in respect of the provision of IT services by a not trivial amount. Otherwise, the decision would appear to be an arbitrary one. 110 There is no evidence in the documents before the Tribunal, and nothing was put in oral argument, to suggest that Epic is able to exercise market power in any non-trivial way with respect to the Pelican Point expansion in the future. Epic and the ACCC agree that the expansion has been built and the expanded capacity is fully contracted on an FT service basis until 2019. The mechanism for the exercise of market power submitted by the ACCC relates to Epic's commercial interest in the revenues earned from rebatable IT services. No evidence was presented on the size of such revenues as they did not form part of the evidence before the ACCC when it made its decision. The Tribunal was presented with only a theoretical construct by the ACCC of its contention that Epic could exercise market power in the provision of IT services in the absence of coverage. 111 The IT service can only be a mechanism for the exercise of market power where the full capacity of the FT service is not used. Although no direct or explicit evidence was available on the use of IT services, there are several items suggesting that the potential to exercise market power is commercially remote. 112 First, the capacity expansion was financed by the operator of the Pelican Point Power Station for the transport of gas for use in its power station. This suggests that the 25 TJ might yield gas available to others as an IT service only at random, that is, on a non-systematic basis that makes it difficult to plan and execute any market power of a non-trivial nature. Second, the Pelican Point contract with Epic is for 25 TJ/day of FT service and 20 TJ/day of IT service, suggesting that little if any of the 25 TJ expansion will be available to market as an IT service. Third, the AAI submitted by Epic states that revenues from IT services were expected to be zero during the initial AA period. Fourth, as Epic noted in its response to the Final Decision, 'there is substantial uncertainty associated with the market for IT service, and with potential revenues from that service. This is the very reason why the MAPS Access Arrangement treats the IT service as a rebatable service, and not as a reference service'. Finally, the increase in the available IT service would be small if the Pelican Point expansion were covered. Without coverage the available IT service is 70 TJ per day, being the difference between the notional capacity of the system and the SPC of 323 TJ per day plus unused FT service of the SPC itself. With coverage it is 70 TJ per day plus unused FT service of the SPC of 348 TJ per day. 113 What needs to be addressed is whether the alleged ability of Epic to exercise market power is such as to have a commercially and socially significant impact, both in quantum and in sustainability. It must be non-trivial, and non-transient. It must have a significant effect on some aspect of competition in a relevant market. The Tribunal is of the opinion that the existence of market power, and the ability to exercise it systematically, have not been satisfactorily established by the ACCC and the ACCC does not point to any of the factual materials which were available to it which would sustain a conclusion that such market power exists. 114 The price of IT service is capped at 1.15 times the price of the FT service. So no market power can be exercised over price. The ACCC provided the Tribunal with no evidence as to how Epic might be able to exercise its market power in the price of supply of IT services. It is in Epic's commercial interest to have its IT service fully taken up, and this would not likely be the case if it were to try and impose conditions on the supply of the service that did not suit potential users. 115 The ACCC argued that the extra 25 TJ of capacity brought about by the Pelican Point expansion gives Epic the ability to exercise market power because of the likelihood that on occasions not all of the added capacity will be taken by Pelican Point and that some gas will accordingly then become available as an IT service. 116 With respect to the Pelican Point Power IT service, the amount that is available daily for customers is not known in advance. It can fluctuate daily. The daily quantity of gas not taken by Pelican Point can therefore be regarded as a random variable. Epic has argued that in such circumstances it simply has no market power with respect to the IT service. The ACCC has asserted that it does have such power, but has not provided any cogent reasons as to why this is the case. 117 This temporary spare capacity arising from the Pelican Point expansion is no different from what could happen to daily variations in demand for the original 323 TJ of capacity. Yet the ACCC was content to leave this IT service unregulated. Both conceptually and from an engineering capacity, the two types of spare capacity are as identical as is the extra gas put into the system as the result of the expansion. The additional 25 TJ associated with the Pelican Point expansion will likely represent only a marginal increment to what is available for sale as an IT service. 118 Even if market power exists, it is still necessary to ask whether the ACCC acted reasonably in exercising its discretion in requiring coverage of the Pelican Point expansion? One of the factors to be considered in assessing an AA is the legitimate business interests of the service provider (s 2.24(a) of the Code), and there is no mention of that factor in the ACCC's assessment of the inclusion of the Pelican Point expansion in the covered pipeline, even though Epic raised it in its submissions on the Final Decision. Coverage of the Pelican Point expansion capacity would entail a significant reduction in the Reference Tariff because the proportional increase in capacity is significantly greater than the proportional increase in costs. 119 The Reference Tariff does not affect the revenues earned by Epic in the period to which the AA applies because all gas transportation service capacity is fully contracted throughout the period. Where it will have its effect is in the negotiations when the existing contracts expire at the end of 2005. The effect of including the Pelican Point expansion will almost certainly be lower tariffs for all contracted users when the contracts expire, even though the capacity was fully paid for by Pelican Point which has a contract for the FT service capacity until 2019. 120 The Tribunal finds that there were no factual circumstances which would have entitled the ACCC to reasonably conclude that Epic would be able to exercise any market power of a significant nature in the market for IT service. The ACCC did not provide any cogent explanation of what the mechanism was, through which Epic could systematically act to exercise any market power, nor did it provide any evidence as to what the effects would be, or who would be affected, and for how long the effects would last. The ACCC could do no more than say that in the absence of proof that the Pelican Point expansion 'was fully contracted in every sense of the word' then it was correct for it to conclude that Epic could potentially exercise market power in the supply of IT service. That, in the view of the Tribunal, was not a sufficient basis upon which to conclude that the Pelican Point expansion of capacity ought to be covered and included in the SPC of the Pipeline System available for the provision of a FT service being the Reference Service in the AA Epic lodged on 1 April 1999. Nor was it open to the ACCC to come to the decision which it did without due consideration of the legitimate business interests of Epic which the ACCC failed to consider. 121 In the view of the Tribunal, the decision of the ACCC to include the Pelican Point expansion capacity in the SPC of the pipeline system was in error, and unreasonable in all the circumstances, and ought to be set aside. ORDERS 122 The Tribunal concludes that the AA approved by the ACCC on 31 July 2002 should be varied to substitute an exchange rate adjusted cost of line pipe of $1 302.20 in lieu of the figure used by the ACCC to calculate the ORC, and to exclude the additional 25 TJ per day from the SPC of the MAPS used by the ACCC for the purposes of its Approved AA. At the request of the Tribunal the parties prepared a document setting out the consequential variations and amendments which would be necessary if the Tribunal determined the matter as it has done. The agreed document is entitled 'Statement of Amendments Required to Access Arrangement'. The necessary changes are those set out under the heading 'Scenario 2'. I certify that the preceding one hundred and twenty-two (122) numbered paragraphs are a true copy of the Reasons for Decision of the Tribunal. Associate: Dated: 10 December 2003 Counsel for the Applicant: GH Murphy with GD Edmonds-Wilson Solicitor for the Applicant: Minter Ellison Lawyers Counsel for the Respondent: A Robertson SC with AI Tonking Solicitor for the Respondent: Australian Competition and Consumer Commission Date of Hearing: 1, 2 September 2003 Date of Judgment: 10 December 2003