Pilbara Infrastructure Pty Ltd v Economic Regulation Authority
[2014] WASC 346
•26 September 2014
JURISDICTION : SUPREME COURT OF WESTERN AUSTRALIA
IN CIVIL
CITATION: THE PILBARA INFRASTRUCTURE PTY LTD -v- ECONOMIC REGULATION AUTHORITY [2014] WASC 346
CORAM: EDELMAN J
HEARD: 21 - 22 AUGUST 2014
DELIVERED : 26 SEPTEMBER 2014
FILE NO/S: CIV 2511 of 2013
BETWEEN: THE PILBARA INFRASTRUCTURE PTY LTD
Applicant
AND
ECONOMIC REGULATION AUTHORITY
First RespondentBROCKMAN IRON PTY LTD
Second Respondent
Catchwords:
Administrative law - Need for an application to identify precisely a ground of review - Ultra vires theory of judicial review - Different grounds of jurisdictional error - Characterisation of general assertions of error of law - Relationship between jurisdictional error and error of law on the face of the record - Issues arising in relation to claims of error of law on the face of the record
Administrative law - Jurisdictional facts - Relationship between review of jurisdictional facts and merits review - Whether cost contingencies are a jurisdictional fact in determination of costs by Regulator
Administrative law - Relevant and irrelevant considerations - Requirement that a relevant consideration be mandatory - Whether cost contingencies are a mandatory relevant consideration in 'total costs' - Whether Regulator had regard to an irrelevant consideration of speculative future possibility of use
Administrative law - Unreasonable exercise of a discretion - Whether review based on unreasonableness applies to findings of fact - Distinction between a discretionary decision and a factual determination with a range of possible answers - Whether unreasonableness review of facts is possible in Australian law - Whether Australian law could accommodate the extension - Lack of argument on this issue - Whether issue best left to circumstances where the point is fully argued and where result depends upon it
Administrative law - Procedural fairness - Content of obligation of procedural fairness - Whether the Regulator was required to notify the railway owner of adverse parts of its determination concerning contingencies and economic life - Size of the effect of the approach taken by the Regulator to these two issues
Legislation:
Administrative Decisions (Judicial Review) Act 1977 (Cth)
Interpretation Act 1984 (WA)
Railway and Port (The Pilbara Infrastructure Pty Ltd) Agreement Act 2004 (WA)
Railways (Access) Act 1998 (WA)
Railways (Access) Code 2000 (WA)
Result:
Decision of the Economic Regulation Authority quashed
Category: A
Representation:
Counsel:
Applicant: Mr B Dharmananda SC, Mr M J Feutrill & Mr C P K Russell
First Respondent : Mr R M Mitchell SC & Ms R Young
Second Respondent : Mr K J Mony De Kerloy
Solicitors:
Applicant: Allen & Overy
First Respondent : State Solicitor for Western Australia
Second Respondent : Herbert Smith Freehills
Cases referred to in judgment:
Abebe v The Commonwealth [1999] HCA 14; 197 CLR 510
Anisminic v Foreign Compensation Commission [1969] 2 AC 147
Apache Northwest Pty Ltd v Agostini (No 2) [2009] WASCA 231
Australian Broadcasting Tribunal v Bond [1990] HCA 33; (1990) 170 CLR 321
Australian Heritage Commission v Mount Isa Mines Ltd [1997] HCA 10; (1997) 187 CLR 297
Azzopardi v Tasman UEB Industries Ltd (1985) 4 NSWLR 139
Banbury v Bank of Montreal [1918] AC 626
Baxter v New South Wales Clickers' Association [1909] HCA 90; (1909) 10 CLR 114
Begum v Tower Hamlets London Borough Council [2003] UKHL 5; [2003] 2 AC 430
Bressington v Commissioner for Railways (New South Wales) [1947] HCA 47; (1947) 75 CLR 339
Caccavo v Collins [2014] TASFC 7
Clambake Pty Ltd v Tipperary Projects Pty Ltd [No 2] [2007] WASC 244; (2007) 35 WAR 394
Collector of Customs v Agfa-Gevaert Ltd [1996] HCA 36; (1996) 186 CLR 389
Commissioner for Australian Capital Territory Revenue v Alphaone Pty Ltd [1994] FCA 1074; (1994) 49 FCR 576
Commissioner of Taxation v McCabe (1990) 26 FCR 431
Corporation of the City of Enfield v Development Assessment Commission [2000] HCA 5; (2000) 199 CLR 135
Craig v State of South Australia [1995] HCA 58; (1995) 184 CLR 163
Davis v Bunn [1936] HCA 44; (1936) 56 CLR 246
Detsongjarus v Minister for Immigration, Local Government and Ethnic Affairs (1990) 21 ALD 139
Electrolux Home Products Pty Ltd v The Australian Workers' Union [2004] HCA 40; (2004) 221 CLR 309
Giblin v McMullen (1868) LR 2 PC 317
Hanna v Australian Postal Corporation [1992] FCA 177
Hiddle v National Fire and Marine Insurance Company of New Zealand [1896] AC 372
Hockey v Yelland [1984] HCA 72; (1984) 157 CLR 124
Hope v Bathurst City Council [1980] HCA 16; (1980) 144 CLR 1
House v The King [1936] HCA 40; (1936) 55 CLR 499
Jewell v Parr (1853) 13 CB 909; 138 ER 1460
Kirk v Industrial Relations Court of New South Wales [2010] HCA 1; (2010) 239 CLR 531
Kostas v HIA Insurance Services Pty Ltd [2010] HCA 32; (2010) 241 CLR 390
Labocus Precious Metals Pty Ltd v Thomas (No 3) [2007] FCA 1346
McKay v Commissioner of Main Roads [2013] WASCA 135
McPhee v S Bennett Ltd (1985) 52 WN (NSW) 8
Metropolitan Railway Co v Jackson (1877) 3 App Cas 193
Minister for Aboriginal Affairs v Peko-Wallsend Ltd [1986] HCA 40; (1986) 162 CLR 24
Minister for Immigration and Citizenship v Li [2013] HCA 18; (2013) 249 CLR 332
Minister for Immigration and Citizenship v SZMDS [2010] HCA 16; (2010) 240 CLR 611
Minister for Immigration and Ethnic Affairs v Wu Shan Liang [1996] HCA 6; (1996) 185 CLR 259
Minister for Immigration and Multicultural Affairs v Bhardwaj [2002] HCA 11; (2002) 209 CLR 597
Minister for Immigration and Multicultural Affairs v Epeabaka [1999] FCA 1; (1999) 84 FCR 411
Minister for Immigration and Multicultural Affairs v Eshetu [1999] HCA 21; (1999) 197 CLR 611
Momcilovic v The Queen [2011] HCA 34; (2011) 245 CLR 1
Myers v Medical Practitioners' Board of Victoria [2007] VSCA 163; (2007) 18 VR 48
Naxakis v Western General Hospital [1999] HCA 22; (1999) 197 CLR 269
Nudrill Pty Ltd v La Rosa [2010] WASCA 158
OzEpulse Pty Ltd v Minister for Agriculture, Fisheries and Forestry [2007] FCA 1601; (2007) 163 FCR 562
Paridis v Settlement Agents Supervisory Board [2007] WASCA 97; (2007) 33 WAR 361
Pelechowski v Registrar, Court of Appeal (NSW) [1999] HCA 19; (1999) 198 CLR 435
Plaintiff M70/2011 v Minister of Immigration and Citizenship [2011] HCA 32; (2011) 244 CLR 144
Pownall v Conlan Management Pty Ltd (1995) 12 WAR 370
Puhlhofer v Hillingdon London Borough Council [1986] UKHL 1; [1986] AC 484
R (Cart) v Upper Tribunal [2011] UKSC 28; [2012] 1 AC 663
R (Prolife Alliance) v British Broadcasting Commission [2003] UKHL 23; [2004] 1 AC 185
R v Earl of Banbury (1694) Skinner 517; (1694) 90 ER 231
R v Hull University Visitor, Ex parte Page [1993] AC 682
Re McBain; Ex parte Australian Catholic Bishops Conference [2002] HCA 16; (2002) 209 CLR 372
Re Michael; Ex parte Epic Energy (WA) Nominees Pty Ltd [2002] WASCA 231; (2002) 25 WAR 511
Re Minister for Immigration and Multicultural Affairs; Ex parte Applicant S20/ 2002 [2003] HCA 30; (2003) 77 ALJR 1165
Re Minister for Immigration and Multicultural Affairs; Ex parte Lam [2003] HCA 6; (2003) 214 CLR 1
Re Minister for Immigration and Multicultural Affairs; Ex parte Miah [2001] HCA 22; 206 CLR 57
Re Minister for Immigration and Multicultural and Indigenous Affairs; Ex parte Palme [2003] HCA 56; (2003) 216 CLR 212
Refugee Review Tribunal; Ex parte Aala [2000] HCA 57; (2000) 204 CLR 82
Residues Treatment & Trading Co Ltd v Southern Resources Ltd (1989) 52 SASR 54
Rooke's Case (1598) 5 Co Rep 99b; (1598) 77 ER 209
Ryder v Wombwell (1868) LR 4 Ex 32
S v Crimes Compensation Tribunal [1998] 1 VR 83
Saeed v Minister for Immigration and Citizenship [2010] HCA 23; (2010) 241 CLR 252
Salemi v MacKellar [No 2] [1977] HCA 26; (1977) 137 CLR 396
Secretary, Department of Social Security v Hilton (1991) 22 ALD 746
Sharp v Wakefield [1891] AC 173
Soulemezis v Dudley (Holdings) Pty Ltd (1987) 10 NSWLR 247
Stead v State Government Insurance Commission [1986] HCA 54; (1986) 161 CLR 141
Television Capricornia Pty Ltd v Australian Broadcasting Tribunal [1986] FCA 458; (1986) 413 FCR 511
The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345
Tisdall v Webber [2011] FCAFC 76; (2011) 193 FCR 260
Toomey v London, Brighton and South Coast Railway Company (1857) 3 CB (NS) 146; (1857) 140 ER 694
WAJS v Minister for Immigration and Multicultural and Indigenous Affairs [2004] FCAFC 139
Warley Pty Ltd v Adco Constructions Pty Ltd (1988) 8 BCL 300
Wecker v Secretary, Department of Education, Science and Training [2008] FCAFC 108; (2008) 168 FCR 272
Wheelton v Hardistry (1857) 8 El & Bl 232; 120 ER 86
X v The Commonwealth [1999] HCA 63; (1999) 200 CLR 177
Texts cited:
Allan TRS 'Common Law Reason and the Limits of Judicial Deference' in Dyzenhaus D, The Unity of Public Law (2004)
Allan TRS, The Sovereignty of Law: Freedom, Constitution and Common Law (2013)
Allsop J, 'The Control of the Executive by the Courts' [2002] Federal Judicial Scholarship 17
Anderson S, 'Public Law' in The Oxford History of the Laws of England (vol XI, 2010)
Aronson M and Groves M, Judicial Review of Administrative Law (5th ed, 2013)
Aronson M, 'Jurisdictional Error without the Tears' in Groves M and Lee HP (eds), Australian Administrative Law: Fundamentals, Principles and Doctrines (2007)
Barak A, Proportionality: Constitutional Rights and their Limitations (2012)
Basten J, 'Jurisdictional error after Kirk: Has it a Future?' (2012) 23(2) Public Law Review 94
Craig P, Administrative Law (7th ed, 2012)
Galligan D, Discretionary Powers: A Legal Study of Official Discretion (1986)
Hamburger P, Is Administrative Law Unlawful? (2014)
Leeming M, Authority to Decide: The Law of Jurisdiction in Australia (2012)
Selway B, 'The Principle Behind Common Law Judicial Review of Administrative Action - The Search Continues' (2002) 30 Federal Law Review 217
Wade W and Forsyth C, Administrative Law (10th ed, 2009)
Wade W, 'Unlawful Administrative Action: Void or Voidable? Part 1' (1967) 83 Law Quarterly Review 499
Table of Contents
Introduction
The Code regime and the requirements for floor and ceiling prices
(i) The floor and ceiling prices
The floor price calculation
The ceiling price calculation
(ii) Establishing costing principles
(iii) The Regulator's requirement to make a determination
The Regulator's determination
(i) The manner in which the Regulator's determination was made
(ii) The Regulator's determination in relation to contingencies
The meaning and operation of contingencies
The Regulator's findings on contingencies
(iii) The Regulator's determination in relation to 'economic life'
The meanings of economic life and technical life
The calculation of economic life
The errors in the Costing Principles concerning economic life
TPI's request to change the Costing Principles
The Regulator's economic life findings
Issue 1: Existence of a jurisdictional fact: a proposal
Issue 2: Alleged errors in relation to the omission of contingencies
(i) TPI's grounds of judicial review concerning contingencies
(ii) TPI's first two grounds of review
The need to identify a category of jurisdictional error of law
The nature and operation of jurisdictional facts
The inclusion of contingencies is not a jurisdictional fact
The failure to include contingencies was not a misapprehension of power
(iii) Whether the exclusion of contingencies omitted a relevant consideration
The ground of review for failing to consider a relevant consideration
The Regulator did not fail to take into account a relevant consideration
(iv) Whether the exclusion of contingencies was unreasonable
The determination concerning contingencies was a decision of fact
The nature of judicial review for unreasonableness and judicial restraint
Obstacles to extending judicial review for unreasonableness to findings of fact
Can judicial review for unreasonableness be extended to findings of fact?
The reasons why I decline to express a concluded view on this issue
Issue 3: Alleged errors in relation to the use of an economic life
(i) The grounds of judicial review
(ii) TPI's first two grounds of review: misconstruing jurisdiction
Misconstruing jurisdiction in relation to economic life (singular)
Misconstruing jurisdiction in relation to possible life of commercial use
(iii) Possibilities of future use as an irrelevant consideration
(iv) Alleged unreasonableness of the Regulator's decision concerning economic life
Issue 4: Alleged failure to provide procedural fairness
(i) The rules of procedural fairness
(ii) The context of the statutory implication of procedural fairness
(iii) The manner of reaching the contingency and economic life decisions
(iv) The significance and size of the contingency and economic life issues
(v) Conclusion in relation to procedural fairness
Conclusion and relief
Appendix 1: Admissibility ruling in relation to Mr Davis' report
Mr Davis' expert opinion generally
(1) The relevance objection
(2) The non-expert opinion objection
(3) The hearsay objection
EDELMAN J:
Introduction
This application alleges errors potentially amounting to more than $2 billion by the Regulator, the Economic Regulation Authority. The errors are alleged to have been made in a ceiling price determination made under the Railways (Access) Code 2000 (WA). The Regulator purported to determine under the Code the ceiling price payable by Brockman Iron, an applicant for access to TPI's railway infrastructure. TPI seeks prerogative relief to quash the determination of the Regulator. One basis for that relief, namely that Brockman Iron had not made a valid proposal, must be rejected for reasons I gave in The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2].[1]
[1] The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345.
TPI's central claims in this proceeding are based upon allegations that the Regulator had erred in its calculation of the ceiling price in two ways: (i) by excluding contingencies from its calculation, and (ii) in its approach to economic life. TPI also says that it was denied procedural fairness in relation to the Regulator's determinations in these respects.
Although it is neither necessary nor appropriate for this Court to attempt to resolve the extent of either of the Regulator's alleged errors, it is relevant to the question of procedural fairness that the size of each alleged error is extremely significant. On TPI's case, the combined error alleged, assuming aggregation of the two classes of alleged error and an equal annuity over a 20‑year period, would be in excess of $2 billion for a 20‑year access agreement.[2] This includes the alleged annuity errors of $62 million and $39 million for each of the contingencies and economic life issues.
[2] See the alleged single annuity error in Exhibit B (Witness Statement of Mr Lee) [83].
This application was heard immediately after the trial in The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2].[3] My reasons in this application should be read together with my reasons for decision on that trial. They have been written separately only for clarity in separating the many issues that arise.
[3] The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345.
In this application, Brockman Iron's appearance was only to submit to the Court's jurisdiction. Although the Regulator made submissions, it was not a real contradictor of the application. The matters addressed by the Regulator included the operation of grounds of judicial review. This was a necessary exercise in circumstances in which some of the issues of judicial review raised by TPI are very unsettled.
One controversial issue that arose on this application is the extent to which, if at all, a Court can review factual findings for unreasonableness. Although I have canvassed the arguments in relation to that potential ground of review it was not ultimately necessary to reach a concluded view. As my review of those issues shows, this is a matter of considerable difficulty in Australian law and I do not consider that it should be determined in the absence of full argument, and where the determination of the case does not depend upon it. The reason why the determination of this application does not depend upon a concluded view on this issue is because TPI's application succeeds in relation to both issues on the established ground of lack of procedural fairness.
The Code regime and the requirements for floor and ceiling prices
As I explained in The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2],[4] the Code establishes a regime which involves the following key components.
[4] The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345 [26].
(1)The owner publishes information.[5]
[5] See Railways (Access) Code 2000 (WA) s 6, s 7A ‑ s 7E, sch 2.
(2)A proponent can request further information from the owner.[6]
[6] Railways (Access) Code 2000 (WA) s 7(1), s 7(2).
(3)A proponent can make a proposal.[7]
(4)The owner can require proof of various matters.[8]
(5)The proponent may withdraw the proposal at any time prior to arbitration or entry into an access agreement.[9]
(6)An arbitration can be held concerning whether the proponent has complied with the matters about which the railway owner required proof.[10]
(7)The railway owner must negotiate in good faith with the proponent subject to the proponent providing proof of the matters in (4) above[11] or determination by an arbitrator that the requirements of s 14 and s 15 of the Code have been met.[12]
(8)The proponent can refer the matter to arbitration if the railway owner refuses to negotiate[13] or negotiations fail.[14]
(9)The arbitrator can terminate the arbitration at any time if he or she thinks that the proponent's referral was vexatious, that the dispute is lacking in substance or the proponent has not engaged in negotiations in good faith.[15]
(10)The proponent may elect, within 14 days of being notified of the determination, not to give effect to the arbitrator's determination.[16]
(i) The floor and ceiling prices
[7] Railways (Access) Code 2000 (WA) s 8.
[8] Railways (Access) Code 2000 (WA) s 14, s 15.
[9] Railways (Access) Code 2000 (WA) s 9A.
[10] Railways (Access) Code 2000 (WA) s 18, s 25(2)(b), s 26(1).
[11] Railways (Access) Code 2000 (WA) s 13(2)(a).
[12] Railways (Access) Code 2000 (WA) s 32.
[13] Railways (Access) Code 2000 (WA) s 25(2)(a), s 26(1).
[14] Railways (Access) Code 2000 (WA) s 25(2)(c), s 26(1).
[15] Railways (Access) Code 2000 (WA) s 35.
[16] Railways (Access) Code 2000 (WA) s 34(2).
It is apparent from the structure of the Code that the purpose of the Regulator's determination of floor and ceiling prices is to establish a framework for negotiation by defining the bounds of the negotiations. In broad terms, those bounds also ensure that the railway owner does not charge monopoly prices and also does not pursue predatory pricing such as by cross‑subsidising a particular route.[17]
The floor price calculation
[17] Exhibit E (Expert Report of Ms Jaffer) page 6.
Schedule 4, cl 7 of the Code provides for the floor price and the manner in which the floor price is to be calculated.
(1)An operator that is provided with access to a route and associated railway infrastructure must pay for the access not less than the incremental costs resulting from its operations on that route and use of that infrastructure.
(2)The total of -
(a)the payments to the railway owner by -
(i)all operators; and
(ii)all other entities,
that are provided with access to a route, or part of a route, and associated railway infrastructure (the route); and
(b)the revenue that the railway owner’s accounts and financial statements show as being attributable to its own operations on the route,
must not be a sum that is less than the total of the incremental costs resulting from the combined operations on the route of all operators and other entities and the railway owner.
Ms Jaffer, an economic consultant in Australia and the United Kingdom for the last 30 years, gave expert evidence, called by TPI. She described the floor price as 'no less than the incremental costs of providing the service or services, where incremental cost is the additional cost incurred by the monopolist in providing just that [ie only that] service'.[18]
The ceiling price calculation
[18] Exhibit E (Expert Report of Ms Jaffer) page 13.
Schedule 4, cl 8 provides for the ceiling price and the manner in which the ceiling price is to be calculated.
(1) An operator that is provided with access to a route and associated railway infrastructure must pay for the access not more than the total costs attributable to that route and that infrastructure.
(2) For the avoidance of doubt it is declared that the calculation of total costs under subclause (1) ‑
(a) is for the whole of the route and associated railway infrastructure; and
(b) is to be the same for all operators,
regardless of the extent of the operations or use of the route and infrastructure by any particular operator.
The term 'total costs' is defined in sch 4, cl 1 as the total of all (a) operating costs; (b) capital costs; and (c) the overheads attributable to the performance of the railway owner’s access‑related functions whether by the railway owner or an associate.
This definition of 'total costs' therefore means that in order to calculate the total costs it is necessary to calculate 'capital costs'. It must not be forgotten that the capital costs being calculated are for the ultimate purpose of calculating 'the total costs attributable to [the proposed access] route and that [proposed access] infrastructure'. Capital costs are defined in sch 4, cl 2(1) as 'the costs comprising both the depreciation and risk‑adjusted return on the relevant railway infrastructure'. Schedule 4, cl 2 then provides as follows:
(3) Capital costs (other than capital costs under subclause (5)) are to be determined as the equivalent annual cost or annuity for the provision of the railway infrastructure calculated in accordance with subclause (4).
(4) The calculation is to be made by applying -
(a) the Gross Replacement Value (GRV) of the railway infrastructure as the principal;
(b) the Weighted Average Cost of Capital (WACC) as the interest rate; and
(c) the economic life which is consistent with the basis for the GRV of the railway infrastructure (expressed in years) as the number of periods,
where -
GRV is the gross replacement value of the railway infrastructure, calculated as the lowest current cost to replace existing assets with assets that -
(i) have the capacity to provide the level of service that meets the actual and reasonably projected demand; and
(ii) are, if appropriate, modern equivalent assets;
and
WACC is the target long term weighted average cost of capital appropriate to the railway infrastructure.
(5) Capital costs include amounts for the amortisation of -
(a) the costs incurred by the railway owner or an associate of the railway owner to acquire any interest in land; and
(b) any other costs incurred by the railway owner or an associate of the railway owner in relation to the acquisition of any interest in land (for example, costs in connection with Aboriginal heritage or native title issues or other transaction costs),
but only to the extent that the Regulator determines that those amounts relate to the acquisition after the commencement of this Code of an interest in land used for constructing, maintaining or operating the relevant railway.
Two points should be noted about the required calculation of capital costs.
First, although the Code requires the calculation of the annuity for the provision of the railway infrastructure to be made by 'applying' the GRV, WACC and economic life, it does not explain the manner in which that calculation must be made. The equation used by the Regulator was explained by Mr Watkinson as follows:[19]
[19] Exhibit F (Affidavit of Mr Watkinson) [56].
In this equation, n is the number of interest periods for the overall time period and P is the yearly payment or annuity.
Secondly, the Code does not define in detail any of the concepts of GRV, WACC and economic life.
The concept of GRV, or gross replacement value, as defined, is concerned with the lowest current cost to replace existing assets with hypothetical railway infrastructure. The replacement assets must (i) have the capacity to provide the level of service that meets the actual and reasonably projected demand; and, (ii) if appropriate, be modern equivalent assets. These matters, (i) and (ii), are concepts that are not defined in the Code.
The first undefined concept is the meaning of 'the capacity to provide the level of service that meets the actual or reasonably projected demand'. The literal meaning of this phrase requires consideration of the actual or reasonably projected demand for the hypothetical railway infrastructure and the capacity to provide the level of service that meets that demand.
As for the meaning of a modern equivalent asset, although this is not defined in the Code, Ms Jaffer explains that this is 'an asset which provides the same service capability of the existing asset, but takes into account any potential savings that could be gained from using an optimised asset configuration'.[20]
[20] Exhibit E (Expert Report of Ms Jaffer) page 17.
The Regulator observed in its determination that Brockman Iron had estimated that a railway would need a capacity of 120 million tonnes per annum. In contrast, TPI had estimated a railway with capacity of 155 million tonnes per annum.[21] It appears that TPI's estimate was based upon an estimate of TPI's demand for the hypothetical railway infrastructure.[22] It was not submitted that sch 4, cl 2(4)(c)(i) of the Code was concerned only with the capacity of hypothetical railway infrastructure for the actual and reasonably estimated demand for the proposed use (20 million tonnes per annum).[23] Nor were any submissions made on the meaning of the words 'if appropriate' in sch 4, cl 2(4)(c)(ii). All parties accepted that the hypothetical replacement railway infrastructure should be a modern equivalent asset and that the level of service required was concerned with an equivalent railway operating with existing demand which does not relate to the proposed use.[24] The Regulator said in its determination that the specification of the modern equivalent asset railway at the capacity of 155 million tonnes per annum, as estimated by TPI, was appropriate.[25]
[21] Exhibit 1/82, page 1568 [58].
[22] Exhibit 1/82, page 1568 [59].
[23] Exhibit 1/82, page 1571 [88].
[24] ts 139 (22 August 2014).
[25] Exhibit 1/82, pages 1569 [69], 1571 [92].
The concept of WACC, or weighted average cost of capital, is the market return required for a given investment.[26]
[26] Exhibit E (Expert Report of Ms Jaffer) page 17.
The concept of economic life is explained separately in detail below because it is the focus of one of the alleged errors.
Overall, the ceiling price was described by Ms Jaffer as 'no greater than the stand-alone cost of providing the service ... where stand‑alone cost is determined as the cost that an efficient competitor would incur in providing just that service'.[27] She explained that the effect of a ceiling price is to ensure that a person such as a railway owner cannot charge higher that the price that an 'efficient new entrant would charge if they entered the market, namely the stand alone cost of providing the infrastructure service'.[28]
(ii) Establishing costing principles
[27] Exhibit E (Expert Report of Ms Jaffer) page 13.
[28] Exhibit E (Expert Report of Ms Jaffer) page 15.
Section 46 of the Code requires each railway owner to prepare and submit to the Regulator a statement of 'principles, rules and practices', which the Code defines as 'costing principles'. Section 46(1) requires these costing principles to be applied and followed by the railway owner in the determination of the costs referred to in sch 4, cl 7 and cl 8 of the Code.
Section 46(2) permits the Regulator to approve the statement submitted by the railway owner either with or without amendments, or to determine the costing principles. Both the Regulator and the railway owner (with the approval of the Regulator) can amend the costing principles.
The Costing Principles relevant to this case were provided by TPI to Brockman Iron, as TPI was required to do,[29] on 22 May 2013.[30] Several points should be made about those Costing Principles.
[29] Railways (Access) Code 2000 (WA) s 9(1)(c)(iii).
[30] Exhibit 1/4.
(1)The Costing Principles were not definitive. As they explain, they 'need to be supported by databases and costing models containing considerable detail, which will change from time to time'.[31]
(2) In many places the Costing Principles replicated, verbatim, the requirements and definitions contained in the Code.
(3)In relation to calculating Gross Replacement Value for the assessment of capital costs, the assumption is made of a 'greenfields site' and hence costs related to constructing around rail traffic, surface restoration and other surface diversions are excluded from the GRV.[32]
(4)In relation to capital costs, the Costing Principles, like the Code, referred in the singular to the value of the infrastructure assets, the economic life of the assets and the rate of return.[33] The Costing Principles also referred in the singular to 'economic life' in the annuity calculation being 'expressed in years and based on the relevant economic life of the track sections'. [34]
(5)The Costing Principles explained that for 'the purposes of calculating the GRV, the replacement cost calculations are to assume a greenfields site and hence costs related to constructing around rail traffic, surface restoration and other surface diversions are excluded'.[35] A greenfields site is one without any development. It requires all the earthworks necessary for rail construction and the associated costs. But it also avoids any costs associated with building bridges over existing roads or re-routing existing roads.
(6) In a section dealing with design, construction and project management fees, the Costing Principles explained that 'TPI will apply design, construction and project management fees at a rate of 20% of the total cost of the infrastructure and based on an economic life of 50 years'.[36]
(7)A separate section of the Costing Principles was concerned with 'Economic life', section 3.2.2. That section is considered in detail later in these reasons.
(iii) The Regulator's requirement to make a determination
[31] Exhibit 1/4, page 228.
[32] Exhibit 1/4, page 232.
[33] Exhibit 1/4, page 231.
[34] Exhibit 1/4, page 234.
[35] Exhibit 1/4, page 232.
[36] Exhibit 1/4, page 233.
Schedule 4, cl 10 then provides as follows:
(1) Where -
(a) a proposal has been made; and
(b) clause 9 does not apply,
the railway owner is to determine the costs referred to in clauses 7 and 8 that are relevant to that proposal in accordance with the costing principles for the time being approved or determined by the Regulator under section 46.
...
(3) The Regulator is to either -
(a) approve the railway owner’s determination; or
(b) if he or she is not willing to do so, determine the relevant costs,
and the Regulator is to give that approval or make that determination not later than the 30th day after -
(c) the day on which he or she receives notice under subclause (2); or
(d) if -
(i) an application is made under section 11(2); and
(ii) an approval is given for the purposes of section 10(1),
the day on which that approval is so given.
(4) The costs so approved or determined by the Regulator in respect of a proposal are the costs that are to apply under clauses 7 and 8 for the purposes of the proposal.
The costs which the Regulator must approve or determine, described in sch 4, cl 7 and cl 8 are the 'ceiling costs' and 'floor costs' described above.
The Regulator's determination
This section of my reasons provides an outline of the facts, structure, and procedure surrounding the Regulator's determination.
(i) The manner in which the Regulator's determination was made
TPI and the Regulator provided substantial evidence concerning the background to the Regulator's determination and the manner in which that determination was made. Much of that evidence is discussed below in relation to the ground of judicial review concerned with procedural fairness. At this stage it is sufficient to describe the general procedure for the Regulator's determination.
Mr Watkinson, an economist and the Chief Executive Officer of the Regulator, explains the operation and procedure of the Regulator as follows.[37]
[37] Exhibit F (Affidavit of Mr Watkinson).
The Regulator is comprised of a Governing Body and a Secretariat. The Governing Body has one full-time member and two part-time members. The Secretariat has three operational divisions and two business support divisions.
One of the operational divisions of the Regulator's Secretariat is the Access Division. The Access Division is responsible for matters including floor and ceiling costs and weighted average cost of capital determinations under the Code.
On 15 May 2013, the Regulator received a letter from Brockman Iron which enclosed copies of Brockman Iron's proposal for access to TPI railway infrastructure.[38]
[38] Exhibit 1/2.
On 17 May 2013, in response to Brockman Iron's proposal, the Regulator directed TPI to amend its costing principles by 20 May 2013 to define six route sections of TPI's railway.[39]
[39] Exhibit 1/3.
On 22 May 2013, the Regulator received an email from Brockman Iron forwarding a letter from TPI.[40] In the letter, TPI made the following points.
(1) TPI alleged that the proposal for access was not a valid proposal under the Code for various reasons. Two of those reasons were Brockman Iron's lack of finance to enable access (such as for the cost of connecting infrastructure) and its failure to undertake a process of economic feasibility to put itself in a position to enter an access agreement.
(2)TPI also alleged that Brockman Iron's preferred solution was the East Pilbara Independent Railway, which was a proposal for an independent railway to be used by Brockman Iron and other companies to transport iron ore and other product to the Port Hedland port. TPI alleged that the consequence of this preferred solution was that Brockman Iron 'only seeks an access agreement which gives it an option to obtain access at some future time, if it so desires'. TPI also said that it required Brockman Iron to demonstrate under s 14 and s 15 of the Code that Brockman Iron had financial and managerial ability, and that there is capacity for the proposed access route.
(3) TPI also said that it was not possible for it to comply 'strictly' with s 9(1)(c) of the Code which required TPI to calculate floor and ceiling prices. Although TPI provided a calculation of those prices in an attachment to the letter, TPI said that the values of the assumptions contained in sch 4 of the Code and the Costing Principles for the date from which access is sought by Brockman were unknown. Further, TPI said that the Weighted Average Cost of Capital (WACC), on which the floor and ceiling prices rely, is determined by the Regulator on an annual basis and the most recent WACC is only current until 30 June 2013.
[40] Exhibit 1/4.
On 23 May 2013, TPI sought approval from the Regulator to enter into negotiations with Brockman Iron.[41] Approval was required because TPI considered, under s 10(1)(b) of the Code, that the access sought by Brockman Iron 'would involve the provision of access to railway infrastructure to an extent that may in effect preclude other entities from access to that infrastructure'. That approval was provided by the Regulator on 14 August 2013.[42]
[41] Exhibit 1/5.
[42] Exhibit 1/41.
On 23 May 2013, TPI also provided the Regulator with its determination of costs (floor prices and ceiling prices), qualified by reiterating TPI's objections that Brockman Iron had not made a valid proposal.[43] TPI also provided the costing model from which that determination had been made.[44] The ceiling price estimated by TPI was $575,642,663 for the two route sections.[45]
[43] Exhibit 1/6.
[44] Exhibit 1/7.
[45] Exhibit 1/4, page 221; Exhibit 1/7, page 255.
On 27 May 2013, the Regulator engaged AECOM Australia Pty Ltd (AECOM) as its technical advisor in relation to a determination of TPI's proposed floor and ceiling costs.
On 5 September 2013, the Regulator received AECOM's 52‑page final report on TPI's floor and ceiling costs determination.[46]
[46] Exhibit 1/75.
On 12 September 2013, the Regulator provided embargoed copies of the redacted decision on costs and a publication notice to TPI. A copy of the unredacted decision was also provided by email to TPI.[47] Several hours later, at 2:00 pm, the Regulator published a redacted copy of its decision on its website.
[47] Exhibit 1/82, page 1574 - 1575.
The Regulator imposed a total ceiling cost of $316,901,814 for the two route sections.[48]
(ii) The Regulator's determination in relation to contingencies
The meaning and operation of contingencies
[48] Exhibit 1/82, page 1610.
Neither the term 'contingencies' nor any definition of the term is contained in the Code or in the Costing Principles. The judicial review application by TPI defined 'contingencies' as 'amounts intended to cover unknown or unidentified costs or unknown amounts for identified costs that were nevertheless reasonably expected to be incurred'.[49]
[49] Amended application for judicial review [5].
Evidence concerning the meaning and operation of contingencies was given by Mr Gardiner (who was instructed by TPI) and Mr Davis (who was instructed by the State Solicitor's Office). For the reasons explained in Appendix 1 to these reasons, the evidence from Mr Davis was admissible.
Mr Gardiner is an experienced civil engineer who gave evidence concerning capital cost estimation in large infrastructure construction.
Mr Gardiner relied upon the Association for the Advancement of Cost Engineering's (AACE) International Recommended Practice No. 18R‑97: Cost Estimate Classification System - As Applied in the Engineering, Procurement, and Construction for the Process Industries. Mr Gardiner said that this recommended practice should be applied to cost estimate classification in relation to large transport infrastructure.[50]
[50] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.9.
Mr Gardiner provided the following graphic illustration of each of the components in a capital cost estimation. The diagram also illustrates how contingencies are applied to the base estimate in the estimation of outturn capital costs.[51] This diagram illustrates what the experts agreed at their joint conference, namely that '[t]he contingency costs included in a cost estimate are for unknown costs that are expected to be incurred but are not included in the Base Estimate'.[52]
[51] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.14.
[52] Exhibit D (Joint Memorandum of Conferral between Experts) [2.1].
Mr Gardiner explained that a contingency is a cost allowance for unforeseen items of work that experience suggests will eventuate. The allowance for contingency reduces as the level of project definition is improved during the evolution of a project from concept through to completion of commissioning.[53]
[53] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.18.
There are two types of contingency. The first type is 'known unknowns' (or 'inherent risk'). These are risks which usually related to measured items in the base estimate such as quantities of material required or costs of materials or labour. The second type of contingency is 'unknown unknowns'. These are project-specific risks concerning unmeasured items affecting the scope of the costs estimate such as cyclones, unforeseen ground conditions or industrial relations issues. [54]
[54] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.19.
As the diagram illustrates, the contingency is applied to the Base Estimate. Mr Gardiner explained that the contingency is applied to the Base Estimate in order to achieve a 50% probability of the outturn capital cost not being exceeded.[55] In other words, the base estimate of capital cost will be increased until there is an even chance that the estimate will not be exceeded. In that sense, the combined figure is the most likely capital cost. There was no dispute among the experts that this method of estimating contingencies represented a recognised approach.[56]
[55] Exhibit A (Expert Report of Mr Gardiner) 4.3.1.5.
[56] Exhibit D (Joint Memorandum of Conferral between Experts) [2.3].
It is easy to be misled by probability curves and probability estimates. They give an impression of accuracy and clarity. But a fair reading of the evidence before the Court was that although the process of estimating contingencies is performed in a very mathematical way, in many cases it will be a heavily subjective judgement, albeit one which will be based on experience. That was so in the approach taken by TPI in this case where the 20% contingency was assessed as a matter of judgement based on experience, without utilising statistical modelling, and without attempting to assess contingencies for individual items.[57]
[57] See, for example, Exhibit C (Witness Statement of Mr Murray) [33] - [46].
Mr Gardiner opined that it is necessary for all capital cost estimates for large infrastructure assets to include a contingency allowance in order to reflect the level of uncertainty associated with a Base Estimate which is based on limited or incomplete information.[58] On the evidence before the Court, and on the approach taken to cost estimation in this case, this is an inescapable conclusion.
[58] Exhibit A (Expert Report of Mr Gardiner) 4.3.1.7.
There may be many examples where a particular contingency allowance might not be included at all. Several can be given.
(1)Where there is no uncertainty in relation to a particular item such as where the design length of the rail line had been determined so there would be little or no uncertainty about the quantity of rail line required.[59]
(2)Where the contingency is for an item where there is close to zero risk of the cost contingency arising.[60] In relation to rail infrastructure, this might arise in relation to some items where the relevant asset has already been recently built, and the optimised replacement design is very similar. For example, as Mr Davis says, if the replacement railway infrastructure is being built on the same land and using the same materials as a recently built railway then there might be little contingency included for 'unidentified geotechnical conditions'.
(3)Another example is where the uncertainty about a particular item involved a symmetric risk.[61] An example not given by any expert but which might roughly illustrate the point is where there is an even risk of the ground being harder or softer than expected. If the result of harder ground will increase excavation costs just as much as the result of softer ground will decrease costs then no contingency will be allowed for that item.[62] Perhaps it would be more accurate to say that this item would not even be included in the assessment of contingencies.
[59] Exhibit D (Joint Memorandum of Conferral between Experts) [2.4].
[60] Exhibit G (Expert Report of Mr Davis) [100].
[61] Exhibit D (Joint Memorandum of Conferral between Experts) [2.4].
[62] Compare Exhibit G (Expert Report of Mr Davis) [98].
The important point, however, is that none of these examples would suggest the exclusion of contingencies from the entire estimate. Even example (2) might still require some contingency for 'geotechnical conditions' if those conditions were to include risks; for example, the possible occurrence of a cyclone in the Pilbara. Moreover, the examples are given from Mr Davis' perspective as a regulatory economist. Although they assist in the understanding of the regulatory and economic context of the assessment that the Code requires, I do not accept that they detract from Mr Gardiner's opinion as a civil engineer that in estimating costs for large infrastructure projects, it is necessary to add some contingency to the Base Estimate.
Mr Davis, giving evidence on instructions from the Regulator, purported to give evidence about the onus of proof of infrastructure providers in putting forward evidence of contingencies to a Regulator. He opined that it would be reasonable for a Regulator to reject claims for contingency where proof does not exist, is not put forward, or is highly speculative.[63] To the extent that this evidence attempts to invite a conclusion about any onus of proof in the Code then it is plainly inadmissible for that purpose. In any event, there is no onus of proof on TPI to adduce evidence of contingencies when the Regulator decides, under sch 4, cl 10(3)(b), to determine for itself the relevant costs.
[63] Exhibit G (Expert Report of Mr Davis) [94].
There is another concept which is separate from contingency. This is the concept of range of accuracy. The range of accuracy is a concept which is applied to the combined Base Estimate and contingency at P50 probability. The range of accuracy then permits an assessment to be made of the level of risk that the party bearing the costs is prepared to tolerate. If, for example, the party is risk averse then that party might build in an additional allowance for risk to bring the estimate from P50 to P90. The increase in costs to the level P90 would then mean that there is only a 10% chance that outturn costs would exceed the estimate.
This summary of the range of accuracy and the addition of a risk estimate to the contingency estimate can be explained in more detail as follows.
Mr Gardiner defined the expected accuracy range as[64]
[t]he level of accuracy (relative to the P50 contingency-adjusted estimate) which experience suggests can be achieved, based on the level of project definition deliverables and the corresponding estimating methodology.
[64] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.11.
The range of accuracy of an estimate will therefore depend upon the level of definition of the project. Projects with the highest level of definition (class 1) will have a small range of accuracy (ie -3% to -10% for the low range and +3 to +10% for the high range). Projects with the lowest range of definition will have a large range of accuracy (reflecting the inaccurate nature of the cost estimation for that level of definition). In this case, the level of project definition adopted by both TPI and by the Regulator was class 3.[65]
[65] Exhibit 1/82, page 1575 [109].
Class 3 is defined by AACE as follows:[66]
*Level of project definition deliverables: 10% to 40%
*End usage (purpose of estimate): Budget authorisation or control
*Methodology: Semi-detached unit costs with assembly level line items
*Expected accuracy range (Low and High ranges):
-10% to -20% (Low) +10 to +30% (High)
[66] Exhibit A (Expert Report of Mr Gardiner) 4.1.1.10.
Importantly, as AACE explain, the expected accuracy range of the cost estimate is a determination made after application of the contingency (which, as I have mentioned, is applied at the P50 level of confidence).[67]
[67] Exhibit A (Expert Report of Mr Gardiner) 4.3.1.28.
These matters were expressed diagrammatically by Mr Gardiner as follows below.
This diagram shows the following:
(1) The Base Estimate of capital cost is around $9.9 million.
(2)To that Base Estimate it is necessary to add a contingency of around $0.4 million to take the capital cost (P50 Mean) estimate to around $10.3 million.
(3)The expected accuracy range for the level of project definition is then applied at a confidence interval which is generally 80%. As an illustration only, in the case of class 3 if midpoints in the accuracy range were chosen then the range would be a low of ‑ 15% and a high of +20%.
(4) Based on that expected accuracy range, the cost estimate involving the Base Estimate plus contingency could then be adjusted for the risk that the party is prepared to tolerate. In the illustration above, only a 10% risk tolerance for the estimate to exceed outturn costs is allowed. So, at the P90 level the cost estimate becomes around $11.1 million.
The Regulator's findings on contingencies
The crucial paragraphs of the Regulator's decision on this issue are as follows.[68]
107. In general, the capital cost estimates prepared by TPI's consultants include a contingency allowance of 20 per cent applied to each asset category. TPI's costing principles do not propose the inclusion of contingency amounts.
108. AECOM has advised that the extent of contingencies reduce as the design development of a project progresses. That is, that the requirement for contingencies will diminish as the design unit quantities become more certain towards the completion of construction. The [Regulator] considers the [Modern Equivalent Assets] specification of a replacement railway to be a well‑developed design.
109. AECOM has advised that all per unit capital costs proposed by TPI are based on pre-feasibility standard estimates and that this class of estimate has an accuracy range of minus 10 per cent to plus 30 per cent. Therefore, the application of this class of cost estimate effectively provides for a contingency in cost estimates.
110. The [Regulator] considers that it is not appropriate to include a further contingency allowance on top of the contingencies afforded by the class of capital cost estimates used by TPI.
111. Accordingly, the [Regulator] has removed the 20 per cent contingency allowance from all capital cost items in order to make its determination.
112. The [Regulator] notes that AECOM's assessments of direct costs include a number of costs assessed to be in the upper end of the reasonable range.
[68] Exhibit 1/82, page 1574 - 1575.
Several points about the Regulator's determination emerge from the discussion above.
First, it will be clear from my discussion above that the Regulator's statements at [109] ‑ [110] conflates the accuracy range with the issue of inclusion of contingencies. There is a difference between the level of accuracy, or the expected accuracy range, and contingencies or contingent costs. The level of accuracy is determined only after the contingencies are calculated. The accuracy range does not provide for contingencies. It operates on contingencies.
Secondly, the Regulator was correct to observe in [107] that the Costing Principles do not propose the inclusion of contingency amounts. But nor do the Costing Principles exclude contingencies. The extent of contingencies, like other categories of costs, are matters that fall within the 'considerable detail' which the Costing Principles recognised would need to supplement the principles.
Thirdly, the observation in [108] about the development of the design is not a matter that affects contingency. The design was accepted by both TPI and the Regulator as a Class 3 design. The class of design affects the accuracy range rather than the amount of contingency. However, the accuracy range is relevant to the amount of additional cost that is included if it is desired to reduce the risk of outturn costs exceeding the estimate over and above the P50 mean. So, in the example above, this involved an increase in cost from $10.3 to $11.1 million.
It may be that the point that the Regulator was making in [108] was that the amount of contingency should be reduced because the new (hypothetical) construction would be based on an existing rail line. For instance, the experts all agreed in their joint conference that on a 'greenfield' approach to estimating the cost of a modern equivalent rail infrastructure asset, to the extent that the new construction would be similar in design to the existing rail line there would be less uncertainty regarding the quantities of materials required and this component of the contingency could be quite small. But, nevertheless, other contingencies would remain.[69]
[69] Exhibit D (Joint Memorandum of Conferral between Experts) [2.2].
Fourthly, it is not clear whether the comment at [112] informed the Regulator's conclusion in the previous paragraph that contingencies should be excluded. It is hard to see how this could be so. If the Regulator concluded that some of the direct costs were higher than should be allowed, and that overall this meant that the Base Estimate was higher than should be allowed, then a lower Base Estimate should be chosen. There would still be a contingency applied to that lower Base Estimate.
(iii) The Regulator's determination in relation to 'economic life'
The meanings of economic life and technical life
The concept of 'economic life' is not a term defined in the Code or in the Railways (Access) Act 1998 (WA).
The economic life is one of the three components in the calculation of capital costs which, in turn, are a component in the calculation of the total costs (for the ceiling price). The other two components are GRV (gross replacement value) of the railway infrastructure and WACC (weighted average cost of capital). The Code requires the economic life to be consistent with the basis for the GRV and to be expressed in years as the number of periods.
The concept of economic life is explained in the expert report of Ms Jaffer, who says the following:[70]
Economic life represents the period over which network assets are able to provide services and, hence, earn revenues from users. Economic life can differ for different parts of a network, for example if different track segments serve different mines with different remaining lives. However, any given track segment can have only one economic life, which is determined by the period over which it will be used to provide access services.
The purpose of economic life is to define the period over which costs can be recovered and as such is a key input to determining the level of annual revenues allowed under an access regime. All else equal, with a longer economic life the amount of the annual annuity cost will be lower in any given year. Conversely, a shorter economic life means that the annual annuity will be higher.
Similarly, economic life affects the depreciation allowance under a return on and of capital approach. Thus under a [Depreciated Optimised Replacement Cost] approach, the depreciation allowance will be lower in any given year given a longer economic life. This will be partially offset by a higher rate of return (since the slower depreciation rate will raise the value of the capital base on which a return is earned).
[70] Exhibit E (Expert Report of Ms Jaffer) page 18.
The concept of economic life is therefore an estimate of the period over which assets are productive, in the sense of delivering access services and earning access revenues.
Ms Jaffer also explains that the economic life about which the calculation of capital is made is a different concept from 'technical life'. Technical life involves an estimate of how long the assets will be physically capable of doing the task for which they were intended. Naturally, the economic life of assets cannot exceed their technical life because an asset that is physically incapable of doing its intended task will not be economically productive.
The effect of this approach is that an economic life can be shorter than a technical life. For instance, suppose railway infrastructure has a technical life of 50 years but the mine that its exclusive purpose is to service will only be economically productive for 20 years. Although the technical life of the railway infrastructure is 50 years, its economic life will only be 20 years. In summary, the economic life required to be used by the Code will never be more than the technical life of railway infrastructure, but it might be less.
The calculation of economic life
The Code does not provide for the principles by which economic life is to be calculated other than to provide that it must be consistent with the basis for the gross replacement value of the railway infrastructure (which involves the assessment of the lowest current cost to replace existing assets with specified assets).
There are various possible ways to calculate economic life of the 'railway infrastructure' (defined in s 3 of the Code as the 'facilities necessary for the operation of a railway, including ...') consistently with the Code. These are discussed later in these reasons. It suffices for current purposes to observe that one manner in which the economic life can be calculated for the purposes of the annuity is to assign an economic life to each of the facilities necessary for the operation of the railway. This was the approach taken by both the Regulator and TPI. Although TPI assigned an economic life of 19 years to every component of the railway infrastructure, it still did so by reference to each component (bridges, track materials, signalling etc).[71]
The errors in the Costing Principles concerning economic life
[71] See Exhibit B (Witness Statement of Mr Lee) GVL2; Exhibit 1/86, page 1658.
The discussion below sets out the best construction of the Costing Principles in relation to economic life. The Costing Principles are far from pellucid in this area. Although I conclude that the Regulator erred in its determination of the meaning of those paragraphs, the errors in expression in the paragraphs, and the difficulty in the wording of the paragraphs was undoubtedly a significant contributing factor when the Regulator came to interpret the paragraphs. The Regulator did not exercise a power to replace or amend the paragraphs.
In a section entitled 'Economic life', section 3.2.2, the Costing Principles explained the following:[72]
The assets['] lives assumed by TPI are based on economic life of the covered infrastructure (being the shorter of the economic life of the mines served by the railway infrastructure and the technical life of the railway infrastructure), or estimated lives of individual assets based on [Modern Equivalent Assets].
In calculating a ceiling (and if appropriate a floor) cost, the economic life assumption underpinning the annuity payment calculation for these types of capital costs will be based on the economic life of assets listed in [Appendix] A unless a shorter life is adopted due to the assets servicing a time limited project. In assessing the life of a project serviced by the assets, TPI may have regard to the term of contractual arrangements that are entered into by the parties. The [Regulator] will be advised as to the reasons for any shorter life assumption. (Emphasis added).
[72] Exhibit 1/4, page 233.
Several points can be noted about this discussion of economic life.
(1)The word economic that I have italicised in the second paragraph is an error. The point being made in the second paragraph must be that the 'economic life' (singular) assumption that is required by the Code will be based on the technical life of assets listed in Appendix A 'unless a shorter life is adopted due to the assets servicing a time limited project'. There are two reasons why the word 'economic' is an error and why the word should be 'technical'.
(i)The first reason is that unless 'economic' is replaced with 'technical', the Costing Principles would be inconsistent with the Code. The Code requires the use of an 'economic life'. It does not permit 'a shorter life [to be] adopted' than economic life.
(ii)The second reason is that unless the italicised word 'economic' is read to mean 'technical' then the paragraph makes no sense. The reference to a time limited project must be a reference to the economic life of the project. If the word 'economic' is read literally in the paragraph it would then describe a nonsensical notion of using an economic life of assets unless the economic life of the project which the assets service is shorter. But if there is a shorter economic life for the project which the assets service then that shorter life is the economic life of the assets. For the paragraph to have a sensible meaning it must involve a reference to using a technical life of assets as the assets' economic life unless the economic life of the project which they service is shorter (and hence the economic life of the assets is shorter). The reference to the lives in Appendix A must therefore be a reference to the technical lives in that Appendix.
(2)The point of the second paragraph is to therefore say that the economic life assumption (singular) underpinning the annuity payment will be either
(i) based on an assumption of the technical life being the economic life for each of the individual assets whose technical lives are listed in Appendix A to the Costing Principles, or
(ii) based on a shorter economic life for the assets listed in Appendix A than the technical life in (i) if the assets service a time limited project.
(3)The point of the first paragraph is far from clear. It refers to the 'assets lives' assumed by TPI but it does not suggest that those are the lives in Appendix A or that those are the lives to be used in the calculation of the annuity. The assumption appears to be that economic life for all the assets could be calculated in two ways.
(i)The first way would be based upon the shorter of the economic life of the whole mine or the technical life of the railway infrastructure. In other words, the economic life of the assets as a whole will be for as long as the mine is productive unless the assets are, broadly assessed as the whole covered infrastructure, expected to have a shorter life.
(ii)The second way is to assess the lives of the assets individually based on Modern Equivalent Assets.
In summary, the approach to determination of economic life that was adopted in the Costing Principles can be described as a necessity to determine an economic life for the railway infrastructure being the shorter of (a) an estimate based on the technical lives of individual assets described in Appendix A, or (b) the time period for a shorter project where the assets service a time limited project.
Appendix A to the Costing Principles was entitled 'Economic Life of Assets' and was as follows.
| Asset | Life Expectancy (Years) (Years) | |
| 1. | Earthworks for track | 100 |
| 2. | Bridges, tunnels and culverts | |
| a. Bridges (not footbridges) | 50 | |
| b. Culverts | 50 | |
| 3. | Level crossings | 20 |
| Access roads | 10 | |
| 4. | Fencing of track | 15 |
| 5. | Track materials | |
| a. Rail life | >20MGT | |
| Curve < 400m | 6 | |
| Curve 400-800m | 10 | |
| Curve > 800m & tangent | 25 | |
| b. Turnouts | >20MGT | |
| Bearers concrete | 30 | |
| Blades and stock rails | 4 | |
| Rail bound crossing | 10 | |
| Balance of turnout | 20 | |
| c. Sleepers | 50 | |
| d. Ballast | 25 | |
| e. Jewellery | 25 | |
| 6. | Track construction | 50 |
| 7. | Roads and shunter pathways | 10 |
| 8. | Signalling | |
| a. Track construction | 20 | |
| b. Flashlights | 20 | |
| c. Boomgates | 20 | |
| 9. | Communications | 20 |
| 10. | Maintenance | |
| a. Track signs | 10 |
The heading to Appendix A might be understood as a shorthand reference to the technical lives to be used in the calculation of the economic life of individual assets. Otherwise, if the heading were a reference to the economic lives of the assets in Appendix A it would be an error.
One reason why the lives in Appendix A could not be economic lives is because this would be inconsistent with the substantive discussion in section 3.2.2 of the Costing Principles which, as I have explained, assumes that the lives in Appendix A are technical lives.
Another reason why the lives in Appendix A could not be economic lives is because discussion elsewhere in the Costing Principles assumes an economic life for the railway infrastructure as a whole of 50 years. The technical life for earthworks (100 years) would be inconsistent with this assumption if it were an economic life. This is because the economic life of an asset within the mine could not exceed the economic life of the mine as a whole (50 years).
The assumption of the economic life of the railway infrastructure as a whole being 50 years is made for the purposes of calculating (i) design, construction, and project management fees, and also (ii) for financing charges during railway infrastructure construction.[73]
[73] Exhibit 1/4, page 233.
This 50‑year economic life of the mine is also consistent with the assertion by the Regulator in correspondence that TPI's annual report had indicated 'sufficient resources to sustain mine life for at least the remaining duration of the TPI State Agreement Act'.[74] The agreement between TPI and the State of Western Australia is contained in sch 1 to the Railway and Port (The Pilbara Infrastructure Pty Ltd) Agreement Act 2004 (WA). Clause 14(2) of that Agreement provides that the term of TPI's Special Railway Licence shall be a period of 50 years from the date of grant.
[74] Exhibit 1/67, page 1207.
Of course, many of the technical lives in Appendix A might also be economic lives if they are shorter than the economic period for which the asset would otherwise be productive. For example, on any view, the technical life for roads and shunter pathways of 10 years would also be an economic life.
TPI's request to change the Costing Principles
On 1 May 2013, TPI wrote to the Regulator seeking to change TPI's costing principles.[75] TPI explained that in Appendix A to its costing principles it had incorrectly referred to 'Economic Life'. TPI said that the reference should have been to 'Technical Life'.
[75] Exhibit 1/16, page 421.
On 15 May 2013, the Regulator refused to make this change.[76] The Regulator said that the change would not be consistent with the terms used in the Code. This answer is correct. The Code is ultimately concerned with the economic life of a railway, not its technical life. But, as I have explained above, the context of Appendix A was only for the technical lives to be used as part of the means of calculating the economic life of the particular assets which, in turn might have been what the single economic life required by the Code could have been based upon (unless the time limited project was for a shorter period). It was, and is, a matter for the Regulator but future confusion might be avoided by a clearer expression of the Costing Principles.
The Regulator's economic life findings
[76] Exhibit 1/16, page 420.
The crucial paragraphs of the Regulator's determination in relation to economic life are as follows.
279. TPI has nominated a uniform economic life of 19 years for all capital items. These lives are shorter than the economic lives nominated in Appendix A of TPI's Costing principles.
...
282. The [Regulator] notes that TPI has based the annuity calculations for capital costs of all asset classes on a uniform 19 years, and that this differs from the economic lives of assets nominated in Appendix A of TPI's Costing Principles. TPI has advised the [Regulator] that it has applied the economic life of FMG's mines because it is shorter than the economic lives nominated in its Costing Principles.
283. TPI has advised the [Regulator] that its proposed economic life of mines is based on FMG's 2011 Ore Reserve Statements for the Chichester hub (Cloudbreak and Christmas Creek Mines). [footnote: TPI have advised that the Ore Reserves reported in FMG's 2012 Annual Report are not materially different to those reported in the 2011 report. The [Regulator] notes that, nonetheless, the Mineral Resources shown in the 2012 report for the Chichester Hub are 32 per cent greater than those reported in the 2011 report, or 18 per cent greater, excluding Inferred Resources]. The 2011 Reserve Statement shows Ore Reserves of 1.7 billion wet tonnes at a production rate of 90 million tonnes per annum. This indicates an extraction period of 19 years. TPI has indicated that 1.7 billion wet tonnes equates to 1.5 billion dry tonnes for export, but has not indicated a per annum export rate.
284. In making public reports, ASX-listed mining companies are required to comply with the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (the JORC Code). The [Regulator] has referred to the 2012 edition of the JORC Code.
285. An Ore Reserve is defined in the JORC Code as the 'economically mineable part of a Measured or Indicated Mineral Resource'. A Mineral Resource is defined as a resource of 'such form, grade (or quality), and quantity that there are reasonable prospects for eventual economic extraction' [footnote: The guidelines for the JORC Code state that the term 'reasonable prospects for eventual economic extraction' implies an assessment in respect of all matters likely to influence the prospect of economic extraction, including mining parameters].
286. A Mineral Resource is classified as Measured, Indicated or Inferred, depending on the level of geological confidence, with an 'Inferred Mineral Resource' having the lowest level of confidence. The JORC Code states that an Inferred Mineral Resource estimate may not be used for planning in pre-feasibility or feasibility studies. Measured Mineral Resource is estimated with sufficient confidence to support detailed mine planning and evaluation of the economic viability of a deposit.
287. TPI has advised the [Regulator] of its view that the economic life of the mines served by TPI's railway is to be determined on the basis of Fortescue's Ore Reserves, and not its Mineral Resources. TPI has justified this on the basis that an Inferred Mineral Resource cannot be converted to an Ore Reserve for JORC-compliant reporting.
288. TPI has stated that the establishment of economic life on the basis of Ore Reserves is consistent with the approach adopted by the Australian Competition and Consumer Commission and the Independent Pricing and Regulatory Tribunal in respect of the Hunter Valley coal network.
289. The [Regulator] notes the most recent ACCC decision in relation to the ARTC Access Undertaking for the Hunter Valley Network [footnote: 11.pdf] requires that the 'useful life' of a route section be determined having regard to a number of measures of resources, specifically:
•the average remaining mine life of coal mines utilising the Pricing Zone of which that Segment or group of Segments forms part;
•average mine production levels anticipated during the Term having regard to Coal Chain Capacity at any time; and
•marketable coal reserves estimated for each mine existing at the time of the determination or expected to commence during the 5 year period following the time of the determination.
290. Marketable Coal Reserves are defined in the JORC Code as beneficiated or otherwise enhanced coal product where modifications due to mining, dilution and processing have been considered. Marketable Coal Reserves must be publicly reported in conjunction with, but not instead of, reports of Coal Reserves.
291. The [Regulator] does not consider that the approach to determining 'useful life' outlined in the ACCC Access Undertaking for the ARTC should apply to the establishment of economic life of railway assets for the purposes of this determination for the following reasons:
•The ARTC Access Undertaking is a guideline for the calculation of an access price on a Depreciated Optimised Replacement Cost basis, and is not a determination of recoverable costs on a Gross Replacement Value basis. The purpose of the 'useful life' calculation is to determine straight line depreciation in the DORC method, which is not the function of the economic life in the GRV method.
•The measure of Coal Reserves required to be considered is only one of a number of measures of mine life nominated for consideration.
292. The [Regulator] also notes that the ARTC does not operate above-rail on its own network and its below-rail business is not underpinned by the owners of the mines the railway serves. For both these reasons, TPI therefore faces a lower stranding risk in respect of the mines its railway serves, than does the ARTC.
293. Brockman Mining's submission (at page 2) provides a view that the economic life of the mines served by the TPI railway network is in excess of 50 years, and that this is based on a Mineral Resource indicated in FMG's 2012 Annual Report of 6 626 million tonnes, exported at 80.8 million tonnes per annum [footnote: 6 626 million tonnes exported at 80.8 million tonnes per annum implies an 80 year export task]. TPI has advised the [Regulator] of its view that this approach should be rejected, as it based on FMG's total Mineral Resources, including its Inferred Mineral Resources.
294. The [Regulator] has noted that the 6 626 million tonnes referred to by Brockman Mining is the total Mineral Resources for the Chichester and Solomon Hub, and includes Inferred Resources. The [Regulator] notes that by excluding Inferred Resources, the Mineral Resources of the Chichester and Solomon Hubs reported in the 2012 FMG Annual Report are 3 210 million tonnes.
295. The [Regulator] notes that TPI's Costing Principles refers to the economic life of the infrastructure as the shorter of the economic life of the mines served by the railway infrastructure and the technical life of the railway infrastructure. The [Regulator] does not consider that the mines served by the railway infrastructure should necessarily be confined to mines owned by FMG.
296. The [Regulator] is mindful that Brockman's Access Proposal is for access over 25 years commencing in 2016. The [Regulator] does not consider that it is appropriate for TPI to make a determination of costs relevant to that proposal on the basis of an economic life of the railway limited to 19 years from 2013. In this regard, the [Regulator] is mindful of the objectives of the Code and the Act, in particular the requirements of section 20(4) of the Act relating to the economically efficient use of the railway, and the objective to facilitate a contestable market for rail operations.
297. In addition to the lives of mines potentially served by the below-rail infrastructure, the [Regulator] is aware that TPI is negotiating to provide above-rail (haulage) services to miners in the Pilbara. The [Regulator] does not have any information regarding the impact of any above-rail customers on the economic life of TPI's railway assets.
298. The [Regulator] has previously approved provisions in TPI's Costing Principles which allow TPI to truncate the economic life of assets to remaining mine life in order to accommodate the stranding of assets where mines are towards the ends of their lives.
299. The Chichester Hub Reserve Statement of September 2011 includes statements that there is considerable upside potential to mine life and that Inferred Resources provide significant potential for Reserve upside.
300. The [Regulator] notes that FMG corporate presentations available in the public domain, and in particular the ASX announcement of 19 March 2013 indicate current gross mineral resources [footnote: The 'Gross Mineral Resources' referred to by FMG include the Inferred Mineral Resources shown in its 2012 Annual Report] in excess of 15 billion tonnes with an annualised growth of 1.5 billion tonnes per annum.
301. This data is replicated in a chart included in an Investor Presentation in Hong Kong, dated March 2003 entitled 'Resource Portfolio Sets Expansion Platform'. The title of this chart implies that expansion plans are based on the Inferred Mineral Resources shown in FMG's 2012 Annual Report.
302. The [Regulator] does not accept the reasons provided by TPI for limiting the life of the mines served by the two route sections relevant to Brockman's Access Proposal to 19 years. The [Regulator] has used the economic lives of assets shown in Appendix C of TPI's Costing principles for the purposes of calculating annualised capital costs [footnote: The [Regulator] has applied an economic life of 25 years to all track rail for the purposes of this determination. The [Regulator] notes that TPI's Costing Principles allow shorter economic lives for track rail curves. TPI 's Costing Model provides only a total rail GRV, and does not provided a breakdown by rail type. On the basis of the track rail quantities provided for the 2010 determination of costs for the Cloudbreak to Port Hedland section, and as the two route sections subject to this determination are wholly contained within that section, the [Regulator] has assumed the proportion of curved rail for the two route sections subject to this determination to be less than 10 per cent. The [Regulator] considers making an adjustment on this basis to the economic life of track rail to allow for a shorter economic life of curved rail is not warranted].
The Regulator's determination in these paragraphs is discussed in more detail below, but for present purposes two general points, arising from the immediately preceding sections of these reasons, can be made about the Regulator's determination in relation to economic life.
First, in relation to [279] and [282] of the Regulator's reasons, as I have explained in the preceding sections of my reasons above, TPI's assessments in Appendix A of the Costing Principles was not an assessment of economic lives. It was an assessment of technical lives.
Secondly, although the Regulator spoke on a number of occasions about 'economic lives' this was not necessarily an error. The Code requires the application of an economic life as part of the annuity calculation but it does not specify the process or manner by which the economic life for use in the annuity calculation must be determined. As I have explained above, that process can occur by using different economic lives for each component of the railway infrastructure.
Issue 1: Existence of a jurisdictional fact: a proposal
It was common ground between TPI, Brockman Iron, and the Regulator that the existence of a valid proposal was a jurisdictional fact which was a prerequisite for the exercise of the Regulator's power to set floor and ceiling prices.
The first issue is whether that jurisdictional fact was satisfied. This issue can be disposed of quickly. It was the subject of my decision in The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2].[77] For the reasons I explained in that decision, Brockman Iron made a valid proposal. There was no jurisdictional error.
[77] The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345.
Issue 2: Alleged errors in relation to the omission of contingencies
(i) TPI's grounds of judicial review concerning contingencies
Various grounds of judicial review brought by TPI were based on allegations of errors by the Regulator in refusing to allow for contingencies in its estimate of costs. Those grounds can be summarised as follows:
(1) The Regulator made an error of law by misconstruing TPI's costing principles.[78]
(2)The Regulator made an error of law by misconstruing or misapplying the provisions in sch 4 of the Code.[79]
(3)The Regulator made an error of law by excluding the relevant consideration of contingencies, which it was required to take into account under sch 4 of the Code or the costing principles.[80]
(4)The Regulator made an error of law by making a determination that was unreasonable and irrational by not including contingencies in its estimate and conflating contingencies with a range of accuracy.[81]
[78] Amended application for judicial review [6].
[79] Amended application for judicial review [7].
[80] Amended application for judicial review [8].
[81] Amended application for judicial review [9].
Much of the submissions by senior counsel for the Regulator were concerned with these basic propositions of administrative law. Nevertheless, as I explain below, the boundaries and the application of these principles can often be very difficult.
(ii) TPI's first two grounds of review
The need to identify a category of jurisdictional error of law
TPI's submissions in relation to the first two grounds, although otherwise extremely thorough, did not identify with precision the reason why a writ of certiorari should issue to quash the decision if the alleged errors were found to have been made. No detailed submissions were made concerning why the Regulator's alleged error would be properly conceptualised as an error of law rather than an error of fact or why the error of law was one which should be recognised as a ground for judicial review. To illustrate the importance of these distinctions it is helpful to contrast the English approach to judicial review with the Australian approach.
English administrative law now differs in radical respects from Australian administrative law. English law no longer draws a distinction between jurisdictional errors of law and non-jurisdictional errors of law.[82] Professors Aronson and Groves have argued that the English experience with abolishing a distinction between jurisdictional and non-jurisdictional errors of law has been to create a 'seriously dysfunctional' system of what is 'in effect untrammelled appeal rights for all legal errors'.[83] Further, as Basten JA has observed extra-judicially,[84] the English system of judicial review is today further complicated by the leave requirement for judicial review as well as statutory 'gateway' criteria for appeals from the Upper Tribunal to the Court of Appeal.[85]
[82] Anisminic v Foreign Compensation Commission [1969] 2 AC 147; R v Hull University Visitor, Ex parte Page [1993] AC 682.
[83] Aronson M and Groves M, Judicial Review of Administrative Law (5th ed, 2013) 220.
[84] Basten J, 'Jurisdictional error after Kirk: Has it a Future?' (2012) 23(2) Public Law Review 94.
[85] Tribunals, Courts and Enforcement Act 2007 (UK); R (Cart) v Upper Tribunal [2011] UKSC 28; [2012] 1 AC 663.
I therefore reject this ground of review.
(iii) Possibilities of future use as an irrelevant consideration
The legal principles governing the ground of judicial review for irrelevant considerations are set out above at [126] - [130]. This ground, based on the taking into account of irrelevant considerations, must be dismissed for the reasons expressed immediately above. On a proper construction of the Regulator's determination, it did not take into account the speculative possibilities as to the future use of the infrastructure and estimated mineral resources.
(iv) Alleged unreasonableness of the Regulator's decision concerning economic life
The ground of review concerning unreasonableness of the Regulator's decision in relation to economic life relied upon both the alleged errors concerning (i) the alleged use of multiple lives and technical lives, and (ii) the use of speculative possibilities for future use.
I have already explained that the only error that I consider was made was the use of technical lives as economic lives. In relation to the unreasonableness ground of review this raises again the difficult question which I considered above in relation to contingencies. The question is whether the erroneous use of technical lives as economic lives is (i) an error of fact or (ii) an error in the exercise of a discretion. Putting aside a question of law that might be involved in the construction of the Costing Principles, the erroneous use by the Regulator of technical lives as economic lives was an error of fact. The Regulator did not misunderstand the meaning or the concept of an economic life. The error was in using the wrong factual data as the input (that is, the wrong factual data to the extent that the technical lives in Appendix A to the Costing Principles do not correspond with technical lives).
For the same four reasons that I have expressed in relation to contingencies at [190] - [194], I do not consider that it is appropriate to decide whether it is possible to judicially review an error of fact for unreasonableness.
Nor is it appropriate to decide whether, in this case this error of fact was unreasonable within such a ground of review. The most important reason why I do not do so is because it is not necessary. In the absence of necessity, I do not venture a concluded view on this issue. The reason why it is not necessary to do so is because TPI succeeds in its ground of review based on procedural fairness. It is to that ground that I now turn.
Issue 4: Alleged failure to provide procedural fairness
(i) The rules of procedural fairness
The rules of procedural fairness are derived by implication from the statute. As five members of the High Court explained in Saeed v Minister for Immigration and Citizenship[225] the implication proceeds upon 'the assumption that the legislature, being aware of the common law principles, would have intended that they apply to the exercise of a power [to destroy or prejudice a person's rights or interests]'. Their Honours also cited with approval[226] a passage from reasons of Barwick CJ, where the Chief Justice had explained that the construction of a statute will have regard, in particular, to 'the terms in which the power is granted, the nature of the power of decision or action, the identity of the donee of the power and its subject matter'.[227]
[225] Saeed v Minister for Immigration and Citizenship [2010] HCA 23; (2010) 241 CLR 252, 258 - 259 [12] (French CJ, Gummow, Hayne, Crennan & Kiefel JJ).
[226] Saeed v Minister for Immigration and Citizenship [2010] HCA 23; (2010) 241 CLR 252, 258 - 259 [12] (French CJ, Gummow, Hayne, Crennan & Kiefel JJ).
[227] Salemi v MacKellar [No 2] [1977] HCA 26; (1977) 137 CLR 396, 401.
The assumption concerning the legislature's intention in relation to procedural fairness is derived from the principle of legality.[228]
[228] Electrolux Home Products Pty Ltd v The Australian Workers' Union [2004] HCA 40; (2004) 221 CLR 309, 329 [21] (Gleeson CJ); Momcilovic v The Queen [2011] HCA 34; (2011) 245 CLR 1, 46 - 47 [43] (French CJ).
Senior counsel for the Regulator submitted that TPI had made a submission to the Regulator that contingencies should be allowed and the rules of procedural fairness would not ordinarily require the Regulator to signal its decision.[229] As a general proposition this can be accepted. Procedural fairness will not ordinarily require a decision maker to disclose mental processes, provisional views or proposed conclusions before a final decision is made.[230] But two points should be emphasised.
[229] ts 112 (21 August 2014).
[230] Re Minister for Immigration and Multicultural and Indigenous Affairs; Ex parte Palme [2003] HCA 56; (2003) 216 CLR 212 [22] (Gleeson CJ, Gummow & Heydon JJ); Commissioner for Australian Capital Territory Revenue v Alphaone Pty Ltd [1994] FCA 1074; (1994) 49 FCR 576, 590 ‑ 591 (the Court); OzEpulse Pty Ltd v Minister for Agriculture, Fisheries and Forestry [2007] FCA 1601; (2007) 163 FCR 562, 576 [55] (Emmett J).
First, whether procedural fairness will require a decision maker to 'signal' its decision will always depend on the statutory and factual context of the decision. As Murphy JA (Martin CJ and Buss JA agreeing) explained in McKay v Commissioner of Main Roads,[231] the principles of procedural fairness 'ultimately involve matters of degree and judgment' and these principles are not 'susceptible to hard and fast rules'.
[231] McKay v Commissioner of Main Roads [2013] WASCA 135 [158].
Secondly, the well‑known principle from which senior counsel cited is subject to equally well-known exceptions. Three related exceptions are as follows:[232]
[(i)] procedural fairness may require the [decision maker] to hear the parties further if certain matters emerge in the [decision maker's] consideration of the case after trial which the [decision maker] regards as potentially dispositive but in relation to which, in all the circumstances, it is to be inferred that the parties did not have a proper opportunity to address at trial.
...
[(ii)] The subject of a decision is entitled to have his or her mind directed to the critical issues or factors on which the decision is likely to turn in order to have an opportunity of dealing with it ...
[(iii)] The subject is entitled to respond to any adverse conclusion drawn by the decision-maker on material supplied by or known to the subject which is not an obvious and natural evaluation of that material.
[232] McKay v Commissioner of Main Roads [2013] WASCA 135 [156] - [157] (Murphy JA, Martin CJ & Buss JA agreeing); Commissioner for Australian Capital Territory Revenue v Alphaone Pty Ltd [1994] FCA 1074; (1994) 49 FCR 576, 591 - 592 (the Court). See also Apache Northwest Pty Ltd v Agostini [No 2] [2009] WASCA 231 [215], [217] (Buss JA, Wheeler JA & Newnes JA agreeing).
Although these remarks were made in the context of findings by a trial judge, they also apply generally in the context of administrative decision makers. As senior counsel for the Regulator accepted, a useful (although not definitive) approach is to ask whether the adverse conclusion is one that TPI could reasonably have anticipated.[233] If TPI could reasonably have anticipated the adverse conclusion and if it had the opportunity of dealing with it, then its failure to make submissions or to provide evidence on the point could not permit a conclusion of denial of procedural fairness.
(ii) The context of the statutory implication of procedural fairness
[233] ts 113 (21 August 2014).
The content of the procedural fairness that the Regulator should afford to a railway owner must be assessed in the context that an express possibility recognised by the Code is that the Regulator might simply adopt the determination of the railway owner. In other words, a railway owner might expect (as TPI did) that there was a real possibility that the Regulator might adopt the railway owner's determination. This possibility does not mean that the Regulator must consult a railway owner about all the terms of its determination. Nor does it mean that the Regulator is required to provide a railway owner with an advance copy of its determination (which it is not). It is, however, an important context for the expectation that a railway owner would have about being given the opportunity to respond to aspects of a determination that would have a significant adverse effect.
It is also possible that the Regulator might adopt the determination by the railway owner subject to changes and modifications, or even formulate its own determination afresh. The factors that increase the likelihood that the railway owner would expect to be given the opportunity to respond to adverse changes or modifications or substantial differences in a new determination include the following, all of which point towards a denial of procedural fairness in this case:
(1)the significance of the modifications to, or departures from, the railway owner's determination,
(2) the extent to which the modifications involved departures from, or differences with, the Costing Principles, and
(3) the period of time over which the modifications or departures were considered.
(iii) The manner of reaching the contingency and economic life decisions
The manner in which the Regulator reached its determination enhanced the need for procedural fairness. The essence of this point is that from 19 August 2013, the Regulator had assured TPI that the Regulator would give TPI notice of any 'credible, relevant and significant adverse information' that might affect the Regulator's decision. And yet, despite repeated requests, and despite the Regulator contemplating the removal of contingencies, TPI was not given notice that this would occur. Nor was TPI given notice of the approach that the Regulator would take to economic lives.
The requests by TPI and the responses of the Regulator must be assessed in context. I put to one side the suggestion made in correspondence sent by TPI to the Regulator that natural justice required the Regulator to provide it with a draft of the Regulator's reasons.[234] That was not a submission that TPI repeated in this Court. Nor could it have been.
[234] Exhibit 1/66, page 1204.
TPI had written to the Regulator numerous times asking for an opportunity to respond to potentially adverse issues. TPI wrote to the Regulator on 18 June 2013 and said that it wanted a reasonable opportunity to make submissions in response to any matters raised by third parties.[235] TPI wrote to the Regulator again on 21 June 2013 reiterating its desire to respond to matters raised by third party submissions.[236] TPI again wrote to the Regulator on 16 August 2013 emphasising its desire to respond to any adverse matters if the Regulator decided to make its own determination.[237]
[235] Exhibit 1/54, page 1128.
[236] Exhibit 1/55, page 1129.
[237] Exhibit 1/64, page 1201.
On 19 August 2013, the Regulator responded to TPI's requests with the following assurance:[238]
The [Regulator] acknowledges that the requirements of procedural fairness apply to its decision under [clause] 10(3) of Schedule 4 of the Railways (Access) Code 2000 (Code).
Consistent with these requirements, the [Regulator] will give notice to [TPI] of any credible, relevant and significant adverse information which may affect the [Regulator's] decision under [sch 4, cl] 10(3) of the Code and will allow TPI a reasonable opportunity to respond to any such information before making its decision.
[238] Exhibit 1/65, page 1203.
TPI's expectations of the opportunity to make submissions on any significant adverse finding would also have been enhanced by another letter sent by the Regulator on 26 August 2013 in which the Regulator apparently enunciated the issues on which it might make adverse findings:[239]
Before finalising its decision under [sch 4] clause 10(3), the [Regulator] would like to provide TPI with the opportunity to comment on a number of matters which have arisen in the [Regulator]’s review.
These matters relate to the cost of the flash butt welding facility that forms part of the Track Construction, the time period over which the capital annuity is calculated and the cost of the fuel facility which forms part of the Buildings Cost.
[239] Exhibit 1/67, page 1205.
This 26 August 2013 letter contained no mention of any possible adverse conclusion in relation to the exclusion of contingencies.
There was mention of a possible adverse conclusion in relation to the 'determination of economic life' (singular).[240] As the Regulator explained in its determination, its 26 August 2013 letter had sought information from TPI 'to provide further explanation of [TPI's] claim that the economic life of the mines served by the route sections is limited to 19 years'.[241] TPI was invited to comment on the discrepancy between the estimated economic life of 19 years that TPI provided and the economic life of the mine which was said to be in excess of 50 years. TPI had notice that the Regulator might increase the economic life, perhaps even very substantially.
[240] Exhibit 1/67, page 1207.
[241] Exhibit 1/82, page 1570 [80].
But TPI did not have any notice that an approach would be taken to economic life which, as I have explained above, incorrectly, but directly, applied the technical lives used in the Costing Principles. As I have also explained above, the Regulator was aware that TPI had applied a single economic life and that TPI had based that on 'the technical lives specified in Appendix A to the Costing Principles [which] were applied in the Cost Model'.[242]
[242] Exhibit 1/6, page 251.
On 2 September 2013, TPI wrote to the Regulator, emphasising that TPI had not received a copy of AECOM's report and confirming that:[243]
Based on the ERA's letter received on 20 August 2013 and the terms of the Notice, we understand that the three matters contained in the Notice to be the only issues affecting TPI's interests in relation to the ERA's review of TPI's floor and ceiling cost determination.
[243] Exhibit 1/70, page 1397.
Although the basis for this assumption is clear, the Regulator never corrected this incorrect assumption.
In contrast with the Regulator's 19 August 2013 assurance, and TPI's assumption on 2 September 2013, the exclusion of contingencies was a matter considered by the Secretariat of the Regulator on 18 June 2013[244] and 24 June 2013.[245]
[244] Exhibit 1/53, page 1126.
[245] Exhibit 1/56, page 1132.
Finally, in relation to both the economic life and the contingency issue, the Regulator had received draft reports from AECOM on 19 June 2013,[246] 5 July 2013,[247] and a final report on 5 September 2013.[248] None of those reports was provided to TPI. Although the failure to provide those reports to TPI could not constitute a breach of procedural fairness, it is relevant that those reports contained some of the basis upon which the Regulator relied for its conclusions on both of the relevant issues. The Regulator was aware of these matters from 19 June 2013 during the course of correspondence I have described.
(iv) The significance and size of the contingency and economic life issues
[246] Exhibit 1/38A.
[247] Exhibit 1/60.
[248] Exhibit 1/75.
The significance of the exclusion of contingencies and the use of technical lives rather than economic lives is a further matter which reinforces the need for the Regulator to have given TPI an opportunity to respond to each of these issues.
On TPI's calculation, the exclusion of contingencies had the effect of reducing the ceiling costs by around $62 million from the equivalent annual cost. Over the 20‑year duration of the proposed access agreement this is an alleged error of $1.24 billion.
On TPI's calculation, the use of technical lives rather than economic lives had the effect of reducing the ceiling costs by around $39 million from the equivalent annual cost. Over the 20‑year duration of the proposed access agreement this is an alleged error of $780 million.
Nothing in these reasons should be construed as accepting the calculations of TPI. Indeed, for the reasons I have explained, the uniform application of a 19‑year economic life by TPI might not be an accurate approach. The purpose of making reference to the numerical difference between the approaches of TPI and the approach of the Regulator is to illustrate the significance of these two matters which enhances the reasonable expectation that TPI would have concerning whether it would be given a proper opportunity to respond to the approach that the Regulator proposed to take.
(v) Conclusion in relation to procedural fairness
These three matters of context (points (ii), (iii), and (iv) above) combine to support the conclusion that TPI was entitled to have the opportunity of responding to both of the conclusions that (i) contingencies would be excluded, and (ii) economic lives would be used for each of the assets based on the technical lives in Appendix A to the Costing Principles.
Conclusion and relief
My ultimate conclusion is that TPI was denied procedural fairness in relation to the decision reached by the Regulator. In two respects, contingencies and economic life, the decision of the Regulator had consequences for the ceiling price which were massive. In the particular factual circumstances of this case, including the correspondence exchanged and the opportunity that the Regulator had to inform TPI of possible adverse findings on these issues, procedural fairness required that TPI be given the opportunity of making submissions on these matters. I will hear from the parties concerning the orders that should be made consequent upon these reasons, but my preliminary views are as follows.
The first claim for relief was for a declaration that Brockman Iron did not make a valid proposal under the Code. That relief is rejected.
The second claim for relief is for declarations that each of the Regulator's decision of 14 August 2013 (approval to enter into negotiations) and its determination of 12 September 2013 (the determination concerning costs) is invalid and of no lawful force or effect.
The third claim for relief is for a writ of certiorari quashing each of the Regulator's decision of 14 August 2013 and its determination of 12 September 2013.
The fourth claim for relief arises if, as I have concluded, the Regulator's power to make a determination was enlivened, for a writ of mandamus compelling the Regulator to make its decision under s 10 of the Code and to determine TPI's costs, under sch 4, cl 10 of the Code, according to law.
TPI's submissions concerning the 14 August 2013 approval by the Regulator were based upon its argument that Brockman Iron had not made a valid proposal. I rejected that claim in The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2].[249] But, subject to one point, it was uncontroversial that a finding of denial of procedural fairness meant that the second, third and fourth claims for relief concerning the Regulator's determination on 12 September 2013 were appropriate. The one possible exception is whether affording procedural fairness to TPI would have made (a 'backward looking test') or could have made (a 'forward looking test') a difference.
[249] The Pilbara Infrastructure Pty Ltd v Brockman Iron Pty Ltd [No 2] [2014] WASC 345.
It is not necessary to enter the debate concerning whether the appropriate test, deriving from Stead v State Government Insurance Commission,[250] is a backward looking test, or a forward looking test, or either, or both. It is sufficient to say that in this case both tests are satisfied. In Re Refugee Review Tribunal; Ex parte Aala[251] even McHugh J, who was the only judge to conclude that the denial of procedural fairness could not have made a difference to the outcome in that case, said that '[i]t is no easy task for a court ... to satisfy itself that what appears on its face to have been a denial of natural justice could have had no bearing on the outcome'. In circumstances where I have concluded, based on meticulous submissions from TPI, that the Regulator made errors which were matters of considerable significance it is impossible to conclude that the denial of procedural fairness could not have made a difference to the result on either test. It is appropriate that each of the second, third and fourth claims for relief concerning the Regulator's 12 September 2013 determination be made.
[250] Stead v State Government Insurance Commission [1986] HCA 54; (1986) 161 CLR 141, 145 (the Court).
[251] Refugee Review Tribunal; Ex parte Aala [2000] HCA 57; (2000) 204 CLR 82, 122 [104], quoting Stead v State Government Insurance Commission [1986] HCA 54; (1986) 161 CLR 141, 145 (the Court).
Appendix 1: Admissibility ruling in relation to Mr Davis' report
TPI objected to the admissibility of the whole of Mr Davis' expert report on the grounds that it is (1) irrelevant, (2) comment, argumentative, and not expert opinion, and (3) hearsay.
Neither TPI nor the Regulator sought a ruling prior to trial; both preferred the course of this matter being ruled on in my reasons for decision.[252]
Mr Davis' expert opinion generally
[252] ts 5 (21 August 2014).
Mr Davis was asked by the State Solicitor's Office to provide an expert opinion on three questions.
The first question was
[i]n the field of regulatory economics is it a usual or invariable practice for an assessment of Gross Replacement Value [GRV] of an asset, for the purposes of an access regime to an infrastructure facility, to include an allowance for contingencies (either as the term 'contingencies' is defined in the judicial review application or otherwise)?
The second question was
[i]s the approach described in part 4.3 of Mr Gardiner's report, of adding a contingency component to a base estimate of cost in order to achieve a 50% probability of the outturn costs not being exceeded, a usual or invariable practice for an assessment of the gross replacement value of an asset, for the purposes of the application of an access regime to an infrastructure facility?
The third question was
[i]f the answer to question 1 or 2 is that incorporation of an allowance for contingencies, or the adoption of that approach, is a usual but not invariable practice then in what circumstances might a person determining the gross replacement value of an asset for that purpose reasonably decide not to incorporate an allowance for contingencies or adopt that approach?
(1) The relevance objection
As to relevance, TPI submitted that Mr Davis' evidence concerned whether contingencies had been taken into account in different contexts and under different regulatory schemes. TPI submitted that this expert evidence is irrelevant to questions concerning whether the Regulator acted rationally, or reasonably, in excluding contingencies from the calculation of gross replacement value under the Code.[253]
[253] TPI's objections to the Regulator's evidence, 4 July 2014 [21].
I reject this submission. TPI's submissions concerned whether contingencies must be included in the estimation of costs in the calculation of GRV (as part of 'capital costs', which in turn is part of 'total costs'). Although the Code defines GRV it does not prescribe how the calculation is to be made. The evidence of Mr Davis involved asking:
(1) what it means to estimate GRV;[254]
(2) the use of contingencies in regulatory asset valuation (including how GRV is usually estimated;[255] what a contingency is;[256] approaches taken to contingencies in other sectors;[257] comments on the approach of Mr Gardiner);[258] and
(3)when the inclusion of contingencies might be rejected.[259]
[254] Exhibit G (Expert Report of Mr Davis) [16] - [22].
[255] Exhibit G (Expert Report of Mr Davis) [23] - [30].
[256] Exhibit G (Expert Report of Mr Davis) [31] - [38].
[257] Exhibit G (Expert Report of Mr Davis) [39] - [83].
[258] Exhibit G (Expert Report of Mr Davis) [84] - [87].
[259] Exhibit G (Expert Report of Mr Davis) [88] - [100].
Although Mr Davis acknowledged that his expertise lies in the area of regulatory economics rather than infrastructure construction,[260] he explains that GRV is the same as other regulatory asset valuations in other industries,[261] and then considers the operation of contingencies across industries generally.
[260] Exhibit G (Expert Report of Mr Davis) [85].
[261] Exhibit G (Expert Report of Mr Davis) [15], [20] - [21].
In a context in which there is no reference to, or definition of, contingencies in the Code, each of these matters is relevant to the issues raised by TPI's submissions. Although the general regulatory nature of Mr Davis' evidence contrasts with the specific nature of Mr Gardiner's evidence about capital cost estimation for construction of infrastructure assets, admissibility is a separate question from the weight to be given to Mr Davis' extrapolation of general principles concerning contingencies that apply across all industries.
(2) The non-expert opinion objection
As to expertise, TPI submitted that Mr Davis' opinion is in the nature of comment and argument, asserts conclusions, and involves speculation based on a survey of regulatory decisions taken from different regulatory contexts, none of which requires the application of expertise.[262]
[262] TPI's objections to the Regulator's evidence, 4 July 2014 [22].
The Regulator relied on the evidence of Mr Davis as expert evidence including for the purpose of the comparisons below in various infrastructure areas (telecommunications, gas pipelines, rail, and airports):[263]
(1) the gross replacement value approach, and an Optimised Replacement Costs approach, and
(2) the gross replacement value approach, and a Depreciated Optimised Replacement Cost approach.
[263] The Regulator's submissions on evidence, Schedule 1 to the Regulator's written submissions, 14 August 2014 [11].
As to the question of lack of expertise, there was very little in TPI's submissions which clarified this point. It might be thought surprising to suggest that no expertise is required to explain and compare concepts such as Gross Replacement Value, Optimised Replacement Cost, and Depreciated Optimised Replacement Cost. These are not terms which are in common everyday use, at least not as far as I am aware.
In Re Michael; Ex parte Epic Energy (WA) Nominees Pty Ltd,[264] Parker J (with whom Malcolm CJ and Anderson J agreed) explained that expert evidence could be given about the meaning of words such as 'efficient', 'efficiency', 'competition', 'monopoly', 'market', and 'economically' in relation to the use of those words in the National Third Party Access Code for Natural Gas Pipeline Systems 1997.[265] That conclusion applies with even greater force in this case where the relevant concepts of contingencies, Optimised Replacement Cost, and Depreciated Optimised Replacement Cost are not terms contained in the Code.
[264] Re Michael; Ex parte Epic Energy (WA) Nominees Pty Ltd [2002] WASCA 231; (2002) 25 WAR 511.
[265] Re Michael; Ex parte Epic Energy (WA) Nominees Pty Ltd [2002] WASCA 231; (2002) 25 WAR 511, 541 ‑ 544 [101] - [107].
Finally, as the Regulator pointed out, Ms Jaffer had also been called by TPI to give expert evidence which included a comparison between the GRV and DORC approaches.
As to the submission that Mr Davis' evidence involved impermissible argument, assertion, and speculation, this was simply a bald assertion. No particular passages were identified as involving impermissible argument, assertion or speculation. Rather, the submissions simply sought wholesale exclusion of Mr Davis' evidence or large parts of his evidence. The closest points to speculation in Mr Davis' report are his statements of possible circumstances in which contingencies might be excluded. This is not impermissible speculation or argument. It is a relevant opinion which is an expert assessment that contingencies are not always required. This is directly relevant to the question of whether contingencies are a mandatory relevant consideration.
(3) The hearsay objection
This was an objection that Mr Davis had based his report on inadmissible hearsay evidence concerning the reasons of other regulators.[266] TPI effectively submitted that the hearsay evidence relied on by Mr Davis was not the 'non‑specific' hearsay of the type permitted in Pownall v Conlan Management Pty Ltd.[267] In general terms, all expert opinion requires some hearsay since specialised knowledge is rarely the product of pure personal experience. Almost every expert stands on the shoulders of a giant before him or her. But a line is drawn between matters that inform the general experience of an expert and specific matters upon which the expert relies in the particular. So, in Pownall, the difference was between the permissible use by a valuer of particular transactions in order to make general conclusions (a general exposition, market trends, an assessment of outliers) and the impermissible use of particular transactions to infer directly the value of the property in issue.
[266] TPI's objections to the Regulator's evidence, 4 July 2014 [23].
[267] Pownall v Conlan Management Pty Ltd (1995) 12 WAR 370, 374 - 376 (Ipp J, Malcolm CJ agreeing) 387 ‑ 388 (Anderson J).
The Regulator relied upon the decision of E M Heenan J in Clambake Pty Ltd v Tipperary Projects Pty Ltd [No 2].[268] In that case, his Honour explained reasons why the application of the distinction drawn in Pownall must be applied 'with discrimination'.[269]
[268] Clambake Pty Ltd v Tipperary Projects Pty Ltd [No 2] [2007] WASC 244; (2007) 35 WAR 394, 405 ‑ 407 [38] - [44].
[269] Clambake Pty Ltd v Tipperary Projects Pty Ltd [No 2] [2007] WASC 244; (2007) 35 WAR 394, 407 [44].
No party in this case questioned the decision in Clambake. With respect to E M Heenan J, it might not be an approach which would have commended itself to a common lawyer prior to the Common Law Procedure Act 1852, but today it is an approach which is 'the proper way to run litigation'.[270]
[270] Labocus Precious Metals Pty Ltd v Thomas (No 3) [2007] FCA 1346 [15] (Allsop J).
It would have been an waste of time, and of resources, if the Regulator were required to call each of the authors of the reports relied upon by Mr Davis in order to prove the truth of the contents of those reports as the foundation for his ultimate conclusion. Those reports included reports from the Australian Competition and Consumer Commission, decisions of the Australian Competition Tribunal, and publications on economic models commissioned by infrastructure companies and consultants. Mr Davis' reference to the reports is admissible in this case because he relied on those reports as particular contexts in order to draw a general conclusion that it is neither an invariable nor a usual practice for contingencies to be taken into account in the field of regulatory economics.[271]
[271] Exhibit G (Expert Report of Mr Davis) [83].
45
21
5