AUSTRALIAN COMPETITION TRIBUNAL
Application by EnergyAustralia and Others [2009] ACompT 8
EVIDENCE – Tribunal limited to information, documents, material and matters before the Regulator – extent to which the Tribunal may have regard to non ‘review related matter’ – matters to which the Tribunal may have regard
NATIONAL ELECTRICITY LAW – National electricity pricing and revenue regime – application for review of distribution and transmission determinations – distribution and transmission network service providers – assessment of appropriate regulatory control period – whether Regulator reasonably withheld agreement to proposed averaging period – whether decision was made by reference to national electricity objectives and pricing principles
Gas Pipelines Access (South Australia) Act 1997 (SA)
National Electricity Law
National Electricity Rules
Application by East Australian Pipeline Ltd (2005) ATPR 42-047
Application by EnergyAustralia (2009) ACompT 7
Application by Epic Energy South Australia Pty Ltd (2003) ATPR 41-932
Australian Competition and Consumer Commission (ACCC) v Australian Competition Tribunal (2006) 152 FCR 33
East Australian Pipeline Pty Ltd v Australian Competition and Consumer Commission (2007) 233 CLR 229
Minister for Immigration and Ethnic Affairs v Wu Shan Liang (1996) 185 CLR 259
File No 2 of 2009
BY: ENERGYAUSTRALIA
Applicant
AND: SOUTHERN SYDNEY REGIONAL ORGANISATION OF COUNCILS
Intervener
File No 3 of 2009
RE: APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO TRANSGRID PURSUANT TO CLAUSE 6A.13.1 OF CHAPTER 6A OF THE NATIONAL ELECTRICITY RULES
BY: TRANSGRID
Applicant
File No 4 of 2009
RE: APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A DISTRIBUTION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO INTERGRAL ENERGY PURSUANT TO CLAUSE 6.11.1 OF APPENDIX 1 OF CHAPTER 11 OF THE NATIONAL ELECTRICITY RULES
BY: INTERGRAL ENERGY
Applicant
File No 5 of 2009
RE: APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO TRANSEND PURSUANT TO CLAUSE 6A.13.1 OF CHAPTER 6A OF THE NATIONAL ELECTRICITY RULES
BY: TRANSEND
Applicant
AND: Nyrstar Australia Pty Ltd
Intervener
File No 6 of 2009
RE: APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A DISTRIBUTION/TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO COUNTRY ENERGY PURSUANT TO CLAUSE 6.11.1 OF APPENDIX 1 OF CHAPTER 11 OF THE NATIONAL ELECTRICITY RULES
BY: COUNTRY ENERGY
Applicant
JUSTICE MIDDLETON (DEPUTY PRESIDENT), MR R DAVEY AND MR R Shogren
12 november 2009
MELBOURNE (HEARD IN SYDNEY)
| IN THE AUSTRALIAN COMPETITION TRIBUNAL |
|
|
| File No 2 of 2009 |
| RE: | APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A DISTRIBUTION/TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO ENERGYAUSTRALIA PURSUANT TO CLAUSE 6.11.1 OF APPENDIX 1 OF CHAPTER 11 OF THE NATIONAL ELECTRICITY RULES
| |
| BY: | ENERGYAUSTRALIA | |
|
| Applicant | |
| AND | SOUTHERN SYDNEY REGIONAL ORGANISATION OF COUNCILS | |
|
| Intervener | |
|
|
| |
|
| File No 3 of 2009 | |
| RE: | APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO TRANSGRID PURSUANT TO CLAUSE 6A.13.1 OF CHAPTER 6A OF THE NATIONAL ELECTRICITY RULES
| |
| By: | TRANSGRID | |
|
| Applicant | |
|
|
| |
|
| File No 4 of 2009 | |
| RE: | APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A DISTRIBUTION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO INTERGRAL ENERGY PURSUANT TO CLAUSE 6.11.1 OF APPENDIX 1 OF CHAPTER 11 OF THE NATIONAL ELECTRICITY RULES
| |
| BY: | INTERGRAL ENERGY | |
|
| Applicant | |
|
|
| |
|
| File No 5 of 2009 | |
| RE: | APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO TRANSEND PURSUANT TO CLAUSE 6A.13.1 OF CHAPTER 6A OF THE NATIONAL ELECTRICITY RULES
| |
| By: | TRANSEND | |
|
| Applicant | |
| AND | NYRSTAR AUSTRALIA PTY LTD | |
|
| Intervener | |
|
|
| |
|
| File No 6 of 2009 | |
| RE: | APPLICATION UNDER SECTION 71B OF THE NATIONAL ELECTRICITY LAW FOR A REVIEW OF A DISTRIBUTION/TRANSMISSION DETERMINATION MADE BY THE AUSTRALIAN ENERGY REGULATOR IN RELATION TO COUNTRY ENERGY PURSUANT TO CLAUSE 6.11.1 OF APPENDIX 1 OF CHAPTER 11 OF THE NATIONAL ELECTRICITY RULES
|
| By: | COUNTRY ENERGY |
|
| Applicant |
| the Tribunal: | JUSTICE MIDDLETON (DEPUTY PRESIDENT), MR R DAVEY AND MR R Shogren |
| DATE: | 12 November 2009 |
| PLACE: | MELBOURNE (HEARD IN SYDNEY) |
| APPLICANTS AND APPLICATIONS......................................................................... | [1] |
| GROUNDS OF REVIEW.............................................................................................. | [6] |
| MATTERS, DOCUMENTS, INFORMATION, OR MATERIAL BEFORE THE TRIBUNAL.................................................................................................................... | [8] |
| THE REGULATORY FRAMEWORK........................................................................ | [10] |
| DISTRIBUTION AND TRANSMISSION SERVICES.............................................. | [15] |
| Building block determinations.................................................................................... | [23] |
| Requirements relating to draft and final determinations.......................................... | [41] |
| The Regulatory Proposals.......................................................................................... | [45] |
| TransGrid’s and Transend’s Revenue Proposals..................................................... | [51] |
| FUNCTION OF TRIBUNAL........................................................................................ | [56] |
| COST OF CAPITAL: WITHHOLDING AGREEMENT............................................ | [68] |
| WHAT AVERAGING PERIOD SHOULD BE APPLIED?........................................ | [104] |
| COST OF DEBT............................................................................................................. | [114] |
| INFLATION FORECASTS........................................................................................... | [124] |
| ENERGYAUSTRALIA SPECIFIC MATTERS........................................................... | [126] |
| OPERATING EXPENSES............................................................................................. | [126] |
| Step Changes.............................................................................................................. | [128] |
| Clauses 6.5.6........................................................................................................... | [133] |
| The first Wilson Cook report................................................................................. | [136] |
| Proposed Step changes from Base Year................................................................. | [140] |
| The AER’s draft decision on EA’s step changes.................................................. | [142] |
| EA’s response to the draft decision....................................................................... | [143] |
| The PwC report....................................................................................................... | [144] |
| The Concept Economics report.............................................................................. | [145] |
| The second NERA report....................................................................................... | [149] |
| The Huegin report.................................................................................................. | [157] |
| EnergyAustralia’s revised regulatory proposal..................................................... | [165] |
| The AER’s Final Decision on EA’s step changes................................................ | [166] |
| The second Wilson Cook report............................................................................. | [169] |
| Wilson Cook’s response to the PwC report........................................................... | [170] |
| Wilson Cook’s response to the Concept Economics report.................................. | [171] |
| Wilson Cook’s response to the second NERA report............................................ | [172] |
| Wilson Cook’s response to the Huegin report....................................................... | [175] |
| The Tribunal’s consideration of EA’s step changes............................................. | [186] |
| The Tribunal’s conclusion on EA’s step changes................................................ | [199] |
| Maintenance Costs..................................................................................................... | [200] |
| EnergyAustralia’s June 2008 regulatory proposal............................................... | [202] |
| The first Wilson Cook report................................................................................. | [205] |
| The AER’s draft decision....................................................................................... | [211] |
| EnergyAustralia’s response to the draft decision................................................. | [212] |
| The Sinclair Knight Merz report........................................................................... | [213] |
| The Huegin report.................................................................................................. | [220] |
| EnergyAustralia’s revised regulatory proposal..................................................... | [224] |
| The second Wilson Cook report............................................................................. | [226] |
| Wilson Cook’s response to the SKM report........................................................... | [229] |
| Wilson Cook’s response to the Huegin report....................................................... | [240] |
| Wilson Cook’s use of a mid-point.......................................................................... | [245] |
| The Tribunal’s consideration of EA’s maintenance costs................................... | [246] |
| The Tribunal’s conclusion on EA’s maintenance costs....................................... | [254] |
| PASS ThROUGH.......................................................................................................... | [255] |
| PUBLIC LIGHTING...................................................................................................... | [262] |
| TRANSGRID SPECIFIC MATTER: DEFECT MAINTENANCE........................... | [263] |
| The meaning of ‘defect maintenance’........................................................................ | [265] |
| The meaning of ‘new growth asset’........................................................................... | [266] |
| TransGrid’s May 2008 regulatory proposal............................................................. | [267] |
| The first PB report...................................................................................................... | [272] |
| TransGrid’s response to a draft of the first PB report............................................. | [274] |
| The AER’s October 2008 draft decision................................................................... | [278] |
| The SKM report......................................................................................................... | [281] |
| TransGrid’s January 2009 revised regulatory proposal........................................... | [283] |
| The second PB report................................................................................................. | [288] |
| The AER’s April 2009 Final Decision....................................................................... | [293] |
| The Tribunal’s consideration of TransGrid’s defect maintenance........................... | [295] |
| The Tribunal’s conclusion on TransGrid’s defect maintenance costs..................... | [310] |
| GENERAL OBSERVATIONS...................................................................................... | [311] |
| CONCLUSION............................................................................................................... | [313] |
REASONS FOR DETERMINATIONS
APPLICANTS AND APPLICATIONS
1 The applicants are as follows:
(a) EnergyAustralia (‘EA’), the owner and operator of an electricity transmission and distribution network located in New South Wales;
(b) TransGrid, the owner and operator of an electricity transmission network located primarily in New South Wales;
(c) Integral Energy, the owner and operator of an electricity distribution network located in New South Wales;
(d) Country Energy, the owner and operator of an electricity distribution network located primarily in New South Wales; and
(e) Transend, the owner and operator of an electricity transmission network located in Tasmania,
referred to collectively as ‘the Applicants’.
2 In addition, South Sydney Regional Organisation of Councils (‘SSROC’) was granted leave to intervene in the review in File 2 of 2009 in relation to EA’s application. Nyrstar Australia Pty Ltd (ACN 124 535 468) (‘Nyrstar’) was granted leave to intervene in file 5 of 2009 in relation to Transend’s application. Nyrstar relied upon written submissions, which have been considered by the Tribunal in the reaching of its conclusions.
3 The applications were heard by the Tribunal together.
4 On 19 June 2009, pursuant to s 71B(1) of the National Electricity Law (‘the NEL’) the Tribunal granted the Applicants leave to review final determinations of the Australian Energy Regulator (‘AER’) respectively entitled:
(a) EA distribution determination 2009-10 to 2013-14 dated 28 April 2009, (‘EA Final Determination’). The AER’s reasons for the EA Final Determination are set out in its Final Decision, New South Wales distribution determination 2009-10 to 2013-14 (‘Final Decision’) also dated 28 April 2009 and published on 30 April 2009;
(b) TransGrid transmission determination 2009-10 to 2013-14 dated 28 April 2009 (‘TransGrid Final Determination’). The AER’s reasons for the TransGrid Final Determination are set out in its Final Decision, TransGrid transmission determination 2009-10 to 2013-14 (‘TransGrid Final Decision’) dated 28 April 2009 and published on 30 April 2009;
(c) Integral Energy distribution determination 2009-10 to 2013-14 dated 28 April 2009 (‘Integral Energy Final Determination’). The AER’s reasons for the Integral Energy Final Determination are set out in the Final Decision;
(d) Country Energy distribution determination 2009-10 to 2013-14 dated 28 April 2009 (‘Country Energy Final Determination’). The AER’s reasons for the Country Energy Final Determination are set out in the Final Decision; and
(e) Transend transmission determination 2009-10 to 2013-14 dated 28 April 2009 (‘Transend Final Determination’). The AER’s reasons for the Final Determination are set out in its Final Decision, Transend transmission determination 2009-10 to 2013-14 (‘Transend Final Decision’) also dated 28 April 2009 and published on 30 April 2009;
(referred to collectively as the ‘Final Determinations’).
5 On 21 August 2009, the Tribunal reserved its decisions in these matters and on 4 September 2009, pursuant to s 71Q of the NEL, the Tribunal extended the standard period for making its determinations until 27 November 2009.
GROUNDS OF REVIEW
6 A review of the decision of the AER is conducted under Pt 6 Div 3A of the NEL. The available grounds of review are specified in s 71C(1). They are that the AER made an error or errors of fact in its findings of facts which error or errors, in itself or combination, were material to the making of the decision under review, or that the AER’s exercise of discretion was incorrect having regard to all the circumstances, or that its decision was unreasonable, having regard to all the circumstances.
7 Section 71C(2) of the NEL provides that it is for the Applicants to establish the ground of review which is pursued.
MATTERS, DOCUMENTS, INFORMATION, OR MATERIAL BEFORE THE TRIBUNAL
8 Section 71R identifies the matters, documents, information or material the Tribunal may consider or have regard to in making its determination. It is convenient to set it out in full. It provides:
(1) Subject to this section, the Tribunal, in reviewing a reviewable regulatory decision, must not consider any matter other than review related matter.
(2) The Tribunal, in reviewing a reviewable regulatory decision, must have regard to any document –
(a) prepared, and used, by the AER for the purpose of making the reviewable regulatory decision; and
(b) that the AER has made publicly available.
(3) In addition, if in a review, the Tribunal is of the view that a ground of review has been established, the Tribunal may allow new information or material to be submitted if the new information or material –
(a) would assist it on any aspect of the determination to be made; and
(b) was not unreasonably withheld from the AER when it was making the reviewable regulatory decision.
(4) Subject to this Law, for the purpose of subsection (3)(b), information or material not provided to the AER following a request for that information or material by it under this Law or the Rules is to be taken to have been unreasonably withheld.
(5) Subsection (4) does not limit what may constitute an unreasonable withholding of information or material.
(6) In this section –
review related matter means –
(a) the application for review and submissions in support of the application; and
(b) the reviewable regulatory decision and the written record of it and any written reasons for it; and
(c) in the case of a reviewable regulatory decision that is a network revenue or pricing determination – any document, proposal or information required or allowed under the Rules to be submitted as part of the process for the making of the determination; and
(d) any written submissions made to the AER before the reviewable regulatory decision was made; and
(e) any reports and materials relied upon by the AER in making the reviewable regulatory decision; and
(f) any draft of the reviewable regulatory decision; and
(g) any submissions on the draft of the reviewable regulatory decision or the reviewable regulatory decision itself considered by the AER; and
(h) the transcript (if any) of any hearing conducted by the AER for the purpose of making the reviewable regulatory decision.
9 The Tribunal was informed by the AER that all documents prepared, and used, by the AER for the purpose of making the Final Determinations, and that the AER had made publicly available, were before the Tribunal. The Tribunal has had regard to such documentation.
THE REGULATORY FRAMEWORK
10 The NEL and the National Electricity Rules (‘the Rules’) provide the economic and legal framework for the regulation of the revenues of the Applicants operating in the national electricity market.
11 The NEL requires that in performing or exercising its economic regulatory functions or powers the AER must:
· do so in a manner that will, or is likely to, contribute to the achievement of the national electricity objective (s 16(1)); and
· take into account the revenue and pricing principles (s 16(2)).
12 The national electricity objective, found in s 7 of the NEL, is:
... to promote efficient investment in, and efficient operation and use of, electricity services for the long term interest of consumers of electricity with respect to:
(a) price, quality, safety, reliability and security of supply of electricity; and
(b) the reliability, safety and security of the national electricity system.
13 The revenue and pricing principles set out in s 7A of the NEL are:
(2) A regulated network service provider should be provided with a reasonable opportunity to recover at least the efficient costs the operator incurs in –
(a) providing direct control network services; and
(b) complying with a regulatory obligation or requirement or making a regulatory payment.
(3) A regulated network service provider should be provided with effective incentives in order to promote economic efficiency with respect to direct control network services the operator provides. The economic efficiency that should be promoted includes –
(a) efficient investment in a distribution system or transmission system with which the operator provides direct control network services; and
(b) the efficient provision of electricity network services; and
(c) the efficient use of the distribution system or transmission system with which the operator provides direct control network services.
(4) Regard should be had to the regulatory asset base with respect to a distribution system or transmission system adopted –
(a) in any previous –
(i) as the case requires, distribution determination or transmission determination; or
(ii) determination or decision under the National Electricity Code or jurisdictional electricity legislation regulating the revenue earned, or prices charged, by a person providing services by means of that distribution system or transmission system; or
(b) in the Rules.
(5) A price or charge for the provision of a direct control network service should allow for a return commensurate with the regulatory and commercial risks involved in providing the direct control network service to which that price or charge relates.
(6) Regard should be had to the economic costs and risks of the potential for under and over investment by a regulated network service provider in, as the case requires, a distribution system or transmission system with which a regulated network service provider provides direct control network services.
(7) Regard should be had to the economic costs and risks of the potential for under and over utilisation of a distribution system or transmission system with which a regulated network service provider provides direct control network services.
14 The national electricity objective provides the overarching economic objective for regulation under the NEL: the promotion of efficient investment and efficient operation and use of, electricity services for the long term interests of consumers. Consumers will benefit in the long run if resources are used efficiently, that is if resources are allocated to the delivery of goods and services in accordance with consumer preferences at least cost. As reflected in the revenue and pricing principles, this in turn requires prices to reflect the long run cost of supply and to support efficient investment, providing investors with a return which covers the opportunity cost of capital required to deliver the services.
DISTRIBUTION AND TRANSMISSION SERVICES
15 It is necessary to say something about the economic regulation of distribution and transmission services and the process of making a determination by the AER.
16 Pursuant to cl 11.15.2 of the Rules, Ch 6 as it is set out in Appendix 1 to Ch 11 (‘the Transitional Rules’), applies to the Distribution Network Service Providers (DNSPs) EA, Country Energy and Integral Energy
17 Clause 11.15.1 of the Rules provides that there is to be a regulatory control period of five years of the DNSPs commencing on 1 July 2009, which is referred to as the regulatory control period 2009 – 2014.
18 Clause 6.2.4 of the Transitional Rules provides that the AER must make a distribution determination for each DNSP. A distribution determination is to impose controls over the prices of direct control services, the revenue to be derived from direct control services or both (cl 2.5.5(a) of the Transitional Rules).
19 The definitions in Ch 10 of the Rules provide that a direct control service is a service that:
(a) the Rules specify as a service the price for which, or the revenue to be earned from which, must be regulated under a distribution determination or transmission determination; or
(b) if the Rules do not do so, the AER specifies, in a distribution determination or transmission determination, as a service the price for which, or the revenue to be earned from which, must be regulated under the distribution determination or transmission determination.
20 The Transitional Rules (cl 6.2.3A(b)) further divide direct control services into two subclasses: standard control services and alternative control services. A standard control service is a direct control service that is subject to control mechanisms based on a DNSP’s total revenue requirement (Ch 10 of the Rules). An alternative control service is a distribution service that is a direct control service but not a standard control service (Ch 10 of the Rules).
21 The control mechanisms for direct control services and alternative control services are set out in cl 6.2.5 of the Transitional Rules. For standard control services the control mechanism must be of the prospective CPI minus X form, or some incentive-based variant of the prospective CPI minus X form, in accordance with Pt C of the Transitional Rules (which relates to building block determinations for standard control services) – Transitional Rules cl 6.2.6(a). The control mechanism for alternative control services may (but need not) utilise elements of Pt C – Transitional Rules cl 6.2.6(c).
22 TransGrid and Transend, which are Transmission Network Service Providers (TNSPs), are also subject to a regulatory control period of 2009 – 2014 which is, for the purposes of this review, governed by Ch 6A of the Rules. Clause 6A.2.1 of the Rules provides that the AER must make transmission determinations for TNSPs in accordance with Ch 6A in respect of prescribed transmission services and negotiated transmission services.
Building block determinations
23 A building block determination is the component of a distribution or transmission determination.
24 The procedure for making a building block determination involves the submission of a building block proposal to the AER by each DNSP and TNSP.
25 A DNSP’s building block proposal is required to be prepared in accordance with the post-tax revenue model prepared and published by the AER and must comply with the requirements of, and must contain or be accompanied by the information required by, any relevant regulatory information instrument (cl 6.3.1(c)(1) and (2) and cl 6.4.1(a) of the Transitional Rules). The contents of a DNSP’s post-tax revenue model must include, amongst other things, a method that the AER determines is likely to result in the best estimates of expected inflation (cl 6.4.2(b) of the Transitional Rules). Like requirements which apply to a TNSP appear in Pt C of Ch 6 of the Rules.
26 The building blocks relevantly include:
(a) indexation of the regulatory asset base;
(b) a return on capital for that year;
(c) the depreciation for that year;
(d) the estimated cost of corporate income tax of the provider for that year; and
(e) the forecast operating expenditure for that year.
(see cl 6.3.1(c)(1) and (2) and cl 6.4.1(a) of the Transitional Rules and cl 6A.5.3 of the Rules.)
27 The indexation of the regulatory asset base is calculated in accordance with cl 6.5.1 and Sch 6.2 of the Transitional Rules for a DNSP and cl 6A.6.1 and Sch 6A.2 of Ch 6A for a TNSP, which includes the rolling-forward of the regulatory asset base from one regulatory year in a regulatory control period to another regulatory year in a regulatory control period.
28 The return on capital is calculated in accordance with cl 6.5.2 of the Transitional Rules for a DNSP and cl 6A.6.2 of Ch 6A of the Rules for a TNSP.
29 The forecast operating expenditure for the year is the forecast operating expenditure as accepted or substituted by the AER in accordance with cl 6.5.6 of the Transitional Rules for a DNSP and cl 6A.6.5 of Ch 6A of the Rules for a TNSP.
30 The return on capital for each regulatory year is calculated by applying a rate of return for each DNSP and TNSP for that regulatory control period to the value of the regulatory asset base for the relevant distribution or transmission system as at the beginning of that regulatory year (cl 6.5.2(a) of the Transitional Rules and cl 6A.6.2(a) of the Rules, respectively).
31 The rate of return for a DNSP or TNSP for a regulatory control period is the cost of capital as measured by the return required by investors in a commercial enterprise with a similar nature and degree of non-diversifiable risk as that faced by the business of the provider and must be calculated as a nominal post-tax weighted average cost of capital (WACC) in accordance with the formula set out in cl 6.5.2(b) of the Transitional Rules and cl 6A.6.2(b) of the Rules.
32 Two important parameters in the WACC formula are the nominal risk free rate and the debt risk premium (DRP).
33 In respect of the nominal risk free rate, cl 6.5.2(c) of the Transitional Rules and cl 6A.6.2(c) of the Rules provides that the nominal risk free rate for a regulatory control period is the rate determined for that regulatory control period by the AER on a moving average basis from the annualised yield on Commonwealth Government bonds with a maturity of 10 years using:
(a) the indicative mid rates published by the Reserve Bank of Australia; and
(b) a period of time which is either:
(i) a period (‘the agreed period’) proposed by the relevant DNSP, and agreed by the AER (such agreement not to be unreasonably withheld); or
(ii) a period specified by the AER, and notified to the provider within a reasonable time period prior to the commencement of that period, if the period proposed by the provider is not agreed by the AER under (i),
and for the purposes of (i);
(iii) the start date and end date for the agreed period may be kept confidential, but only until the expiration of the agreed period; and
(iv) the AER must notify the DNSP or TNSP whether or not it agrees with the proposed period within 30 business days of the submission of the building block proposal.
34 In respect of the DRP, cl 6.5.2(e) of the Transitional Rules and cl 6A.6.2(e) of the Rules provides that the DRP for a regulatory control period is the premium determined for that regulatory control period as the margin between the 10 year Commonwealth annualised bond rate and the observed annualised Australian benchmark corporate bond rate for corporate bonds which have a maturity of 10 years and a credit rating of BBB+ from Standard and Poors.
35 A building block proposal must include the total forecast operating expenditure for the relevant regulatory control period which the DNSP or TNSP considers is required in order to achieve the operating expenditure objectives, which are:
(a) meet or manage the expected demand for standard control services over that period;
(b) comply with all applicable regulatory obligations or requirements associated with the provision of standard control services;
(c) maintain the quality, reliability and security of supply of standard control services;
(d) maintain the reliability, safety and security of the distribution system through the supply of standard control services
(see cl 6.5.6(a) of the Transitional Rules and cl 6A.6.6(a) of the Rules).
36 The forecast of required operating expenditure of a DNSP or TNSP that is included in a building block proposal must:
(a) comply with the requirements of any relevant regulatory information instrument;
(b) be for expenditure that is properly allocated to standard control services in accordance with the principles and policies set out in the Cost Allocation Methodology for the DNSP or TNSP; and
(c) include both:
(i) the total of the forecast operating expenditure for the relevant regulatory control period; and
(ii) the forecast of the operating expenditure for each regulatory year of the relevant regulatory control period.
(see cl 6.5.7(a) of the Transitional Rules and cl 6A.6.6(b) of the Rules)
37 The AER is required to accept the forecast of required operating expenditure of a DNSP or a TNSP that is included in a building block proposal if the AER is satisfied that the total of the forecast operating expenditure for the regulatory control period reasonably reflects the operating expenditure criteria, which are:
(a) the efficient costs of achieving the operating expenditure objectives;
(b) the costs that a prudent operator in the circumstances of the relevant DNSP would require to achieve the operating expenditure objectives; and
(c) a realistic expectation of the demand forecast and cost inputs required to achieve the operating expenditure objectives.
(see cl 6.5.6(c) of the Transitional Rules and cl 6A.6.6(c) of the Rules)
38 A building block proposal must include the total forecast capital expenditure for the relevant regulatory control period which the DNSP or TNSP considers is required in order to achieve the capital expenditure objectives, which are:
(a) meet or manage the expected demand for standard control services over that period;
(b) comply with all applicable regulatory obligations or requirements associated with the provision of standard control services;
(c) maintain the quality, reliability and security of supply of standard control services;
(d) maintain the reliability, safety and security of the distribution system through the supply of standard control services.
(see cl 6.5.7(a) of the Transitional Rules and cl 6A.6.7(a) of the Rules)
39 The forecast of required capital expenditure of a DNSP or a TNSP that is included in a building block proposal must:
(a) comply with the requirements of any relevant regulatory information instrument; and
(b) be for expenditure that is properly allocated to standard control services in accordance with the principles and policies set out in the Cost Allocation Method for the DNSP;
(c) include both:
(i) the total of the forecast capital expenditure for the relevant regulatory control period; and
(ii) the forecast of the capital expenditure for each regulatory year of the relevant regulatory control period; and
(d) identify any forecast capital expenditure that is for an option that has satisfied the regulatory test.
(see cl 6.5.7(b) of the Transitional Rules and cl 6A.6.7(b) of the Rules)
40 The AER is required to accept the forecast of required capital expenditure of a DNSP or TNSP that is included in a building block proposal if the AER is satisfied that the total of the forecast capital expenditure for the regulatory control period reasonably reflects the capital expenditure criteria, which are:
(a) the efficient costs of achieving the capital expenditure objectives;
(b) the costs that a prudent operator in the circumstances of the relevant DNSP would require to achieve the capital expenditure objectives; and
(c) a realistic expectation of the demand forecast and cost inputs required to achieve the capital expenditure objectives.
(see cl 6.5.7(c) of the Transitional Rules and cl 6A.6.7(c) of the Rules)
Requirements relating to draft and final determinations
41 A distribution or transmission determination is predicated on a number of decisions by the AER which include:
(a) a decision on the DNSP’s or TNSP’s current building block proposal in which the AER either approves or refuses to approve the annual revenue requirement for the provider, as set out in the building block proposal, for each regulatory year of the regulatory control period;
(b) a decision in which the AER either:
(i) accepts the total of the forecast capital expenditure for the regulatory control period that is included in the DNSP’s or TNSP’s proposal; or
(ii) does not accept the total of the forecast capital expenditure for the regulatory control period that is included in the DNSP’s or TNSP’s proposal, in which case the AER must set out its reasons for that decision and an estimate of the total of the DNSP’s or TNSP’s required capital expenditure for the regulatory control period that the AER is satisfied reasonably reflects the capital expenditure criteria, taking into account the capital expenditure factors;
(c) a decision in which the AER either:
(i) accepts the total of the forecast operating expenditure for the regulatory control period that is included in the current proposal; or
(ii) does not accept the total of the forecast operating expenditure for the regulatory control period that is included in the proposal, in which case the AER must set out its reasons for that decision and an estimate of the total of the DNSP’s or TNSP’s required operating expenditure for the regulatory control period that the AER is satisfied reasonably reflects the operating expenditure criteria, taking into account the operating expenditure factors.
(see cl 6.12.1(2), (3) and (4) of the Transitional Rules and cl 6A.14.1(1), (2) and (3) of the Rules)
A distribution determination must also include:
(a) a decision in relation to the rate of return;
(b) a decision on the control mechanism for alternative control services; and
(c) a decision on the additional pass through events that are to apply for the regulatory control period.
(see cl 6.12.1(5), (12) and (14) of the Transitional Rules).
42 The reasons given by the AER for a draft distribution determination or a final distribution determination must set out the basis and rationale for the determination including:
(a) details of the qualitative and quantitative methods applied in any calculations and formulae made or used by the AER;
(b) the values adopted by the AER for each of the input variables in any calculations and formulae, including:
(i) whether those values have been taken or derived from the provider’s current building block proposal; and
(ii) if not, the rationale for the adoption of those values; and
(c) details of any assumptions made by the AER in undertaking any material qualitative and quantitative analyses; and
(d) reasons for the making of any decisions, the giving or withholding of any approvals, and the exercise of any discretions, for the purposes of the determination.
(see cl 6.12.2 of the Transitional Rules and cl 6A.14.2 of the Rules)
43 The AER has limited discretion in making a distribution determination with respect to the total revenue requirement for a DNSP and its annual revenue requirement or with respect to a TNSP’s total revenue cap and maximum allowed revenue. The AER is required to approve the total revenue requirement for a DNSP or TNSP for a regulatory control period, and the annual revenue requirement for each regulatory year of the regulatory control period, as set out in the provider’s current building block proposal, if the AER is satisfied that those amounts have been properly calculated using the post-tax revenue model on the basis of amounts calculated, determined or forecast in accordance with the requirements of Pt C of the Transitional Rules or Pt C of Ch 6A (cl 6.12.3(d) of the Transitional Rules and cl 6A.14.3(b) of the Rules).
44 The AER also has limited discretion in substituting an amount, value or methodology in respect of certain constituent decisions. If the AER refuses to approve an amount, value or methodology referred to in cl 6.12.1 of the Transitional Rules or cl 6A.13.2 of the Rules, the substitute amount, value or methodology on which the determination is based must be:
(a) determined on the basis of the current regulatory proposal; and
(b) amended from that basis only to the extent necessary to enable it to be approved in accordance with the Rules.
(see cl 6.12.3(f) of the Transitional Rules and cl 6A.13.2 of the Rules)
The Regulatory Proposals
45 EA, Integral Energy and Country Energy were required to submit a regulatory proposal to the AER for distribution services provided by means of, or in connection with, its distribution system, on or before 2 June 2008. Each submitted its regulatory proposal to the AER on 2 June 2008.
46 Relevantly, the regulatory proposal submitted was required to include, amongst other things:
(a) for direct control services classified as standard control services – a building block proposal;
(b) for direct control services classified as alternative control services:
(i) the proposed control mechanism, a demonstration of the application of the proposed control mechanism, and the necessary supporting information; and
(ii) in the case of a departure from the AER’s likely approach to the relevant control mechanisms for alternative control services a statement of the reasons justifying the departure.
47 The regulatory proposal was also required to comply with the requirements of and to contain or be accompanied by the information required by, any relevant regulatory information instrument.
48 The AER released its draft decision and determination on the DNSPs’ respective regulatory proposals in November 2008 and in March 2009 released a supplementary draft decision for alternative control (public lighting) services for the DNSPs.
49 In response to the draft determination, each DNSP was entitled to submit a revised regulatory proposal to the AER not more than 30 business days after the publication of the AER’s draft decision. Each DNSP submitted a revised revenue proposal on 14 January 2009.
50 In submitting a revised revenue proposal, a DNSP may only make revisions so as to incorporate the substance of any changes required to address matters raised by the draft distribution determination or the AER’s reasons for it (cl 6.10.3(b) of the Transitional Rules).
TransGrid’s and Transend’s Revenue Proposals
51 Both TransGrid and Transend were required to submit a revenue proposal to the AER for their transmission services on or before 2 June 2008. Transend submitted its revenue proposal to the AER on 30 May 2008 and TransGrid submitted its on 31 May 2008.
52 Relevantly, the revenue proposal submitted by each TNSP was required to include, amongst other things:
(a) a forecast of the required capital expenditure that complies with the requirements of the Rules;
(b) a forecast of the required operating expenditure that complies with the requirements of the Rules;
(c) the estimated total revenue cap for each TNSP for the relevant regulatory control period and the maximum allowed revenue for each TNSP for each regulatory year of the regulatory control period using the post-tax revenue model together with details of all amounts, values and other inputs used by the TNSP for that purpose.
53 The AER released its draft decisions on the TNSP’s transmission determination in November 2008.
54 In response to the draft determination, the TNSPs were entitled to submit a revised revenue proposal to the AER not more than 30 business days after the publication of the AER’s draft decision. Both TNSPs submitted their revised revenue proposals on 13 January 2009.
55 In submitting a revised revenue proposal, a TNSP may only make revisions so as to incorporate the substance of any changes required to address matters raised by the draft transmission determination or the AER’s reasons for it.
FUNCTION OF TRIBUNAL
56 In view of the submissions made by the parties to the Tribunal, it is necessary to briefly make some observations on the function of the Tribunal.
57 The NEL limits to four the grounds upon which it is open to the Applicants to challenge before the Tribunal the Final Determinations.
58 Attention has been focussed on all grounds, but some comment needs to be made upon the grounds specified in s 71C(1)(c) and (d) – incorrect exercise of discretion and unreasonableness.
59 It is to be observed that the ‘unreasonable’ ground is a separate ground of review. It is not, as in the Gas Pipelines Access (South Australia) Act 1997 (SA) (‘Gas Law’), related to the error of an incorrect exercise of discretion: see East Australian Pipeline Pty Ltd v Australian Competition and Consumer Commission (2007) 233 CLR 229, 250. The term ‘unreasonable’ does not just provide a basis for informing the presence of one or more of the established grounds which render a decision ‘incorrect’, in the sense of the incorrect exercise of discretion. It provides a separate and distinct ground of review.
60 The question arises then as to when it may be held that the decision under review is ‘unreasonable in all the circumstances’. For instance, is it limited to so called Wednesbury unreasonableness?
61 On this question, the comments of the Full Court in Australian Competition and Consumer Commission (ACCC) v Australian Competition Tribunal (2006) 152 FCR 33 are relevant, accepting that for the purposes of the NEL (as distinct from the Gas Law), unreasonableness is a separate ground of review.
62 In ACCC v ACT 152 FCR 33, it was stated:
176 The Tribunal has not been given a purely substitutive function in relation to the review of the ACCC’s discretion. That is to say, if the ACCC has exercised its discretion on correct principles and if the particular exercise of the discretion was open to it within the framework of the Code, the Tribunal is not empowered to set aside that decision simply because it thinks another decision would have been preferable. This is emphasised by the provision in s 39(2)(a)(ii) of the ground of review based on unreasonableness. The exercise of a discretion is not unreasonable simply because another decision-maker would have come to a different view. On the other hand unreasonableness in s 39(2)(a)(ii) is not limited to cases in which the exercise of the discretion was so unreasonable that no reasonable person could have so exercised it.
177 In Application by Epic Energy the Tribunal (Cooper J presiding) said (at [30]):
Section 39(2)(a)(ii) is concerned with the correctness or unreasonableness of an exercise of discretion having regard to the circumstances relevant to the proper exercise of that discretion. Those circumstances are ones which are demonstrable from the matters to which the Tribunal may refer under s 39(5). For the purposes of the subsection, error is made out if it is demonstrated that the exercise of the discretion was so unreasonable on the basis of the matters available to the decision maker that no reasonable decision maker could ever come to it: Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223 at 223–234. It also deals with the situation where the decision is so far outside the range of decisions open to a reasonable decision maker that it bespeaks of error even though the particular error cannot be identified: House v R (1936) 55 CLR 499 at 505. For the purposes of s 39(2)(a)(ii) of GPA Law, correctness and reasonableness are to be determined by reference to applicable criteria contained in the Code applied to the matters which were before the relevant Regulator before the decision under review was made.
That passage does not limit the ground of unreasonableness to so called Wednesbury unreasonableness. It is compatible with the wider view of ‘unreasonableness’ which would pick up logical error or irrationality in the decision. The ACCC’s submission which would limit the unreasonableness ground to so called Wednesbury unreasonableness is not accepted.
178 The concept of ‘unreasonableness’ imports want of reason. That is to say the particular discretion exercised by the ACCC is not justified by reference to its stated reasons. There may be an error in logic or some discontinuity or non sequitur in the reasoning. It may be that the decision has an element of arbitrariness about it because there is an absence of reason to explain the discretionary choices made by the ACCC in arriving at its conclusion.
63 The Tribunal considers it clear that the scope of the separate ground of review of ‘unreasonableness’ set out in the NEL goes somewhat beyond the so called Wednesbury unreasonableness ground. To a certain extent, there is an overlap between the exercise of a discretion which is ‘incorrect’, and a decision which is unreasonable having regard to all the circumstances. If the reasons for a decision contain an element arbitrariness, in the sense of an unexplained discretionary choice made in reaching a conclusion, then it may readily be concluded that the decision itself is unreasonable, and that the exercise of discretion miscarried or was in error.
64 If a decision is not determined by reference to the applicable criteria in the NEL and the Rules, then it will readily lead to a conclusion that the exercise of any discretion in reaching the decision was incorrect, and the decision was unreasonable in all the circumstances.
65 In considering whether the Applicants have established any ground of review, s 71R limits the matters which the Tribunal may consider on its review to ‘review related matter’ as defined in s 71R(6). It is only if a ground of review is made out that the Tribunal may allow new information or material to be submitted, and then only if it would assist on any aspect of the determination to be made and was not earlier unreasonably withheld from the AER: see s 71R(3). Also, s 71O(2) prevents a party to a review, other than the AER, from raising any matter that was not raised in submissions to the AER before the reviewable regulatory decision was made.
66 Therefore, the Tribunal’s review is not at large, but is a review of the AER’s decision on the factual and legal grounds available, but only on the material provided to or before the AER. Nevertheless, the Tribunal must consider the merits of whether the material provided to or before the AER leads to a finding or findings of material fact different from those made by the AER, or that it exercised its discretion incorrectly, or that its decision in all the circumstances was unreasonable.
67 Once the Tribunal is satisfied that a ground of review is established, the Tribunal must consider the various options available under the NEL. One option is to remit the matter to the AER. The Tribunal has already indicated its approach to the appropriateness or otherwise of remitting the matter to the AER: see Application by EnergyAustralia (2009) ACompT 7 at [30-38].
COST OF CAPITAL: WITHHOLDING AGREEMENT
68 The first principal issue for determination is whether the AER unreasonably withheld agreement to the original proposed averaging period within the meaning of cl 6.5.2c(2)(i) of Ch 11 of the Transitional Rules. In addressing the Transitional Rules the Tribunal should also be taken to address the equivalent Rule.
69 A number of principles should be stated which assist in a determination of this issue:
(a) an interpretation of the Transitional Rules that will best achieve the objective or purpose of the NEL is to be preferred to any other interpretation;
(b) in determining the averaging period an exercise of discretion by the AER is involved, but such is limited by and subject to express prescriptions otherwise in the Transitional Rules;
(c) the Applicants have the initial responsibility to propose the averaging period, which the AER is to consider, and then to determine whether there is a reasonable basis upon which the proposal should not be agreed to;
(d) relevant considerations bearing on the decision to withhold agreement include:
(i) whether the period proposed is likely to result in an unbiased risk free rate, given that the equity beta and market risk premium are deemed to be 1.0% and 6.0% respectively; and
(ii) the achievement of the national electricity objective revenue and pricing principles;
(e) the question of whether the decision of the AER withholding agreement was unreasonable can only be determined by reference to the circumstances under which the decision was made, taking into account the proposal, the information before the AER at the time of its decision, and the purpose, scope and provisions of the NEL and the Transitional Rules;
(f) there is no legal or economic presumption in favour of any particular averaging period; in deciding whether or not to agree to the proposed period, the AER must consider the period proposed by the Applicants, and determine whether it should withhold agreement to that period based on the period proposed, the information before the AER at the time of its decision, and the purpose, scope and provisions of the NEL and Transitional Rules;
(g) in considering its decision, the AER may take into account other averaging periods not proposed for the purpose of considering whether to withhold agreement to the proposed period. However, if the averaging period proposed by the Applicants is otherwise appropriate, the AER cannot withhold consent only on the basis that there is another averaging period which it prefers over that proposed by the Applicants;
(h) the reasons given by the AER at the time of its decision are to be examined to determine whether the AER has unreasonably withheld its agreement;
(i) any reasons relied upon by the AER after its decision may also to be examined to determine whether the AER has unreasonably withheld its agreement;
(j) the focus of the enquiry is upon the decision of the AER to withhold agreement to the proposed averaging period, not whether the proposed averaging period is itself reasonable although that consideration may impact upon an assessment of the decision of the AER to withhold agreement; and
(k) if the Tribunal came to the view that the AER did unreasonably withhold its agreement, this could amount to either the exercise of the AER’s discretion as being incorrect in all the circumstances (see s 71C(1)(a)) or the AER’s decision being unreasonable in all the circumstances (see s 71C(1)(d)).
70 It is convenient to consider the AER’s response to EA’s original proposed averaging period. On 8 July 2008, the AER purported to withhold agreement to EA’s original proposed averaging period, proposed a new averaging period for EA of 15 business days commencing on 2 March 2009 and ending on 20 March 2009 (the AER’s proposed averaging period), and invited EA to nominate an alternative averaging period between 1 February 2009 and 20 March 2009. The critical reasons for which the AER withheld agreement were that:
(a) the period was too early and thus too far removed from either the commencement of the five year regulatory period or the likely date of the AER’s final determination;
(b) it was contrary to accepted regulatory practice to adopt a period so early;
(c) the Capital Asset Pricing Model (‘CAPM’) theory, as interpreted by accepted experts, suggested that the risk-free rate should be calculated on or as close as possible to the day of final determination using the most up to date information; thereby giving an unbiased rate of return consistent with market conditions at the date of the final determination;
(d) the purpose of an averaging period is to address possible daily volatility in financial markets but within the constraint that ideally one would be measuring the risk free rate at the date of the final determination; and
(e) to the extent that EA required certainty for purposes of raising capital, it was enough to receive the ultimate regulatory determination.
71 For the purposes of this enquiry, the Tribunal takes the response of the AER dated 8 July 2008 to EA’s original proposed averaging period as representative of the response to the other Applicants.
72 Omitting formal parts, the AER’s response was as follows:
I refer to EnergyAustralia’s confidential attachment 8.1 to its regulatory proposal dated 2 June 2008 which proposes the averaging period for the nominal risk free rate for the regulatory control period 2009-14. EnergyAustralia proposed that the averaging period be the 15 business days starting on 2 June 2008. This proposal was restated in a letter from Geoff Lillis to the AER dated 2 July 2008.
Clause 6.5.2(c)(2)(iv) of the transitional chapter 6 in the National Electricity Rules (NER) requires the AER to notify a distribution network service provider (DNSP) whether or not it agrees with the proposed averaging period within 30 business days of the date of submission of its regulatory proposal.
The AER does not agree with the averaging period proposed by EnergyAustralia as the starting date of EnergyAustralia’s proposed averaging period is almost 12 months prior to the commencement of the regulatory control period. In this regard, the AER considers that the starting date of the proposed period is too far removed from the date by which the AER is likely to publish EnergyAustralia’s final determination, which is expected to be in April 2009, and the commencement of the 2009-14 regulatory control period. The averaging period proposed by EnergyAustralia is contrary to accepted regulatory practice as reflected in previous AER and ACCC determinations; the ACCC’s Statement of Regulatory Principles and previous jurisdictional regulators’ determinations, all of which apply a nominal risk free rate averaging period considerably closer to the final determination date.
The AER’s regulatory practice is supported by accepted expert views in economic and finance literature[1]. Capital Asset Pricing Model (CAPM) theory suggests that, ideally, the nominal risk free rate will be calculated on the day of the final determination. The CAPM is an ex ante model and therefore the most up to date information should be used if available.
Further, applying an averaging period which is closely aligned to the date of the final determination provides an unbiased rate of return that is consistent with market conditions at the time of the final determination. In this regard the AER notes EnergyAustralia’s concern about the need for certainty in order to manage its commercial risks and the CEG report which expresses the premise that ‘a particular business may wish for greater early certainty about its allowed rate of return’. 2 The AER does not agree that this premise goes to the issue of applying an averaging period, the purpose of which is to address the possible daily volatility in financial markets. This is because the regulatory determination itself provides a DNSP with certainty in relation to the rate of return that will apply during the forthcoming regulatory control period. Further, this information is provided to the DNSP prior to the commencement of the regulatory control period. Accordingly, the AER does not consider that early certainty of EnergyAustralia’s allowed rate of return is necessary for EnergyAustralia to deliver on its proposed capital expenditure program for the upcoming regulatory control period.
For the above reasons, the AER does not agree that EnergyAustralia’s proposed averaging period is appropriate. Instead, the AER proposes a new averaging period for EnergyAustralia of 15 business days (the length of EnergyAustralia’s proposed averaging period has not changed) starting on 2 March 2009 and ending on 20 March 2009. The AER accepts EnergyAustralia’s request that the averaging period be kept confidential. It also intends to use the new averaging period to establish the commercial debt margin. The AER’s proposed averaging period takes into account the time required for it to conduct financial modelling using the risk free rate, final determination drafting and the decision making process.
If EnergyAustralia does not agree with the dates proposed by the AER, EnergyAustralia has until 14 August 2008 to write to the AER and nominate an alternative averaging period between 1 February 2009 and 20 March 2009.
If you have any questions in relation to this matter please contact Scott Haig on (02)6243 1207.
[1] See, e.g., Martin Lally, The cost of capital for regulated entities, report prepared for the Queensland Competition Authority, 26 February 2004, p.63; Martin Lally, Determining the risk free rate for regulated companies, report prepared for the ACCC, August 2002, p.14; Kevin Davis, Report on risk free interest rate and equity and debt beta determination in the WACC, report prepared for the ACCC, 28 August 2003, p.16.
2 Competition Economists Group, Nominal Risk Free Rate, debt risk premium and debt and equity raising costs for EnergyAustralia, Attachment 8.2 of EnergyAustralia’s regulatory proposal, 2 June 2008.
73 The place to start in making any assessment of the AER’s decision is the Transitional Rules themselves and the NEL, under which the Transitional Rules were made.
74 The Transitional Rules provide the context for the proposing of an averaging period, but the proposal must be in accordance with the NEL, and more specifically with the national electricity objective and the revenue and pricing principles set out in s 7 and s 7A, respectively.
75 The principles in s 7A can be taken to be consistent with and to promote the objectives in s 7. The principles are themselves stated normatively in the form of what is intended to be achieved. They state that the price charged by a Network Service Provider (‘NSP’) for its service should allow a return commensurate with the regulatory and commercial risks involved in providing the service in the context that the NSP should be provided with a reasonable opportunity to recover at least the efficient costs it incurs and with effective incentives in order to promote economic efficiency with respect to the services it provides. Economic efficiency includes efficient investment in the system with which it provides services, efficient provision of services, and efficient use of the system.
76 It is well accepted in the literature of regulatory economics and in regulatory practice that all these efficiency objectives are in principle met by setting prices for services that allow the recovery of efficient costs, including the cost of capital commensurate with the riskiness of the investment in the assets (infrastructure or ‘system’, as the term is used in the NEL) used to provide services.
77 It might be asked why the NEL principles require that the regulated NSP be provided with the opportunity to recover at least its efficient costs. Why ‘at least’? The issue of opportunity is critical to the answer. The regulatory framework does not guarantee recovery of costs, efficient or otherwise. Many events and circumstances, all characterised by various uncertainties, intervene between the ex ante regulatory setting of prices and the ex post assessment of whether costs were recovered. But if, as it were, the dice are loaded against the NSP at the outset by the regulator not providing the opportunity for it to recover its efficient costs (eg, by making insufficient provision for its operating costs or its cost of capital), then the NSP will not have the incentives to achieve the efficiency objectives, the achievement of which is the purpose of the regulatory regime.
78 Thus, given that the regulatory setting of prices is determined prior to ascertaining the actual operating environment that will prevail during the regulatory control period, the regulatory framework may be said to err on the side of allowing at least the recovery of efficient costs. This is in the context of no adjustment generally being made after the event for changed circumstances.
79 The Transitional Rules, consistently with these principles, set out procedures for determining efficient costs, including the cost of capital, ie, the return on capital.
80 It is important to note that under the Transitional Rules the cost of capital is not the actual cost to the NSP of raising finance (debt and equity). The cost of capital is assessed by use of an economic theory, the CAPM, which in broad terms states that the cost of capital of a project is determined by the riskiness of the project relative to investment in the market as a whole, or equivalently, on average. In fact, there are two levels of abstraction from the actual NSP in making the calculation. The first is to measure the cost of capital ‘by the return required by investors in a commercial enterprise with a similar nature and degree of non-diversifiable risk as that faced by the distribution business of the provider’ (cl 6.5.2(b)). The second is to use the weighted average cost of capital (WACC) formula derived from the CAPM to estimate that required rate of return.
81 This approach is not driven merely by a practical need to find a procedure that can generate an estimate of the firm's cost of capital, but is based on an important principle. The principle being that the rate of return required by investors (and hence consistent with efficient investment) depends on the risk of the investment. The principle is embodied in s 7A(5), as mentioned above.
82 The rate of return, or WACC, is applied to the value of the regulatory asset base of the NSP as at the beginning of a regulatory year to produce the return on capital (in dollar terms) for that regulatory year (cl 6.5.2(a)). (The regulatory asset base is updated each year (cl 6.5.1(e)(2).) Thus the WACC is applied in each of the five regulatory years within the regulatory control period. It follows that the WACC to be applied each year should in principle be the rate of return required by investors at the beginning of that year. This rate of return would naturally be expected to differ from year to year.
83 That is not, however, the scheme set out in cl 6.5.2. Rather it provides for a single value of the WACC to be calculated and applied to each year's starting regulatory asset base.
84 The WACC formula has five inputs, viz. the equity beta, the market risk premium, the debt to equity ratio, the risk free rate, and the debt risk premium. The first three of these are given deemed values in cl 6.5.2(b). The choice of those values implies that the equity of NSPs is deemed to have the risk characteristics of the market as a whole. The risk free rate is defined in cl 6.5.2(c) but is to be averaged over a period, as described above. The purpose of averaging is to smooth day-to-day volatility that might make the choice of any particular day's rate in some way unrepresentative or inappropriate. The debt risk premium is defined in cl 6.5.2(e) such that the cost of debt of NSPs is deemed to be the benchmark corporate bond rate for corporate bonds which have a maturity of 10 years and a credit rating of BBB+ from Standard and Poors. Deciding this benchmark rate is a matter for determination by the AER.
85 On the face of it, this set of inputs and formula generates a rather strange rate of return estimate. The risk free rate, whether agreed or specified, is, it seems to be agreed by all parties, that which prevails at some time (the averaging period) prior to the start of the regulatory control period; similarly with the benchmark corporate bond rate. Those inputs might generate a rate of return value reasonably close to that actually required by investors at the start of the regulatory control period, and applied to the first year's starting regulatory base. But with changes in market conditions over the regulatory control period, it is hard to see why the rate of return value would represent the return required by investors at, say, the start of the final year of the regulatory control period.
86 In the meantime, the risk free rate and corporate bond rates would almost certainly have varied from their initial values. Consequently, there appears to be no virtue in setting those rates at values that prevailed close to the start of the regulatory control period, or to the publication of a final determination.
87 It may be accepted that, as the AER stated in its letter of 8 July 2008, its own regulatory practice and the practice of regulators more generally has been to apply a nominal risk free rate averaging period closer to the start of the regulatory control period. This practice has been supported by economic experts. The Tribunal observes, however, that this is not a universal practice. In market conditions that are not wildly out of the norm, this may be expected to provide a figure that is fairly close to being an unbiased estimate of the risk free rate consistent with market conditions at the time of the final determination; and may consequently be expected to provide a reasonable estimate of the rate of return on capital that would be required by investors at the time of the final determination.
88 But as explained above, there is no proper basis for seeking such an estimate. The views of economic experts appear to be based on a model where the regulatory control period is considered to be a single period (of five years), not five consecutive one-year periods. In the scheme set out in the Transitional Rules, the nexus is broken between the period to which the rate of return applies and the period for which that rate of return is estimated. Once that is realised, the basis for withholding agreement to an averaging period proposed by EA falls away.
89 It is useful to consider on what basis the AER might reasonably withhold agreement to an averaging period proposed by an NSP. The only clear ground is that the period proposed would be likely to generate a rate of return that was inappropriate, ie, too high or too low having regard to the period in which it was to be applied. No doubt that is the ground that the AER ultimately had in mind when it did withhold agreement. Bolstered by regulatory practice and economic expert opinion, it considered that setting the risk free rate and hence the rate of return so far ahead of the period to which it would apply was less likely to provide an appropriate return than was setting the rate of return at a period closer to when it would apply.
90 Putting to one side the Tribunal’s approach as set out above, the AER otherwise had no basis upon which to reject the averaging period proposed by the EA without further enquiry. Rather than assume that the rate at a closer date would give a better estimate, the AER should have examined the evidence regarding expected future rates. Such evidence of forward interest rates, ie, rates that will apply at some future time for a prospective period, is available from market data. Comparisons could be made between rates expected to prevail during the averaging period proposed by the NSP and rates expected at later periods. But it follows from the Tribunal's reasoning that it would be insufficient and inappropriate to only compare with rates expected to prevail close to the time of the final determination.
91 Given the uncertainties involved in forecasting future interest rates, withholding agreement to an NSP's proposed averaging period may be reasonable where market data indicated an expectation that rates would be considerably lower during the regulatory period than at the proposed period.
92 There is no suggestion that the AER made such an assessment prior to making its decision to withhold agreement to the proposed averaging periods. However, in its Final Decision the AER did appeal to the fact that in June 2008 the yield curve for Commonwealth Government Securities (‘CGS’) was downward sloping. This, the AER said, reflected prevailing market expectations that interest rates would be lower in the future. Therefore, setting the risk free rate based on the proposed averaging periods (in June, July and August 2008) would lead to systematic ex ante overcompensation of firms relative to the efficient cost of capital. This argument was repeated in the AER’s submissions to the Tribunal.
93 It may be noted that the NSPs’ regulatory proposals were made in May 2008, before the June yield curve came into existence. However, nothing crucial has been said to hinge on whether a yield curve for June or (say) May is examined. The AER was silent as to precisely what can be deduced from the downward-sloping yield curve beyond that it reflected market expectations of lower interest rates in the future. Which interest rates? How much lower? When in the future?
94 The yield curve at a given time does indeed provide the basis for estimating market expectations about interest rates that will apply in future periods. But it does not do so without further analysis. By itself, it shows only what rates apply at the time the curve is calculated – June 2008 in this case – for securities with varying maturity. Thus the downward sloping yield curve as at June 2008 showed that in June 2008 CGS maturing in two years’ time provided a lower yield than CGS maturing in one year’s time; a security maturing in three years’ time provided a still lower yield; and so on. From examination of the chart in the Final Decision, it appears that yields were falling from just over 7 per cent for a security maturing in June 2009 to a little less than 6.6 per cent for a security maturing in June 2018. That is not a large difference.
95 But the Rules prescribe that the risk free rate is a ten-year rate for CGS. What the AER needed to do, and apparently did not do, was inquire as to what the yield curve implied about how the ten-year risk free rate would change in the future.
96 The method for doing so is well established. Consider an investor who has $1 to invest and wants his capital returned in two years’ time (say). He knows the interest rate at which he can invest for two years, and therefore what total return (capital and interest) he could receive in two year’s time. He also knows the interest rate he could obtain now for investing the $1 for only one year. If he invests now for only one year, what interest rate will need to be available to him in a year’s time so that he can then reinvest and achieve the same total return as he would if he invested for two years now? This is the rate that is expected to prevail in the market in a year’s time. For if that expected future rate were higher than needed for the investor to be indifferent between his two choices, he would do better by investing now for one year and then reinvesting. Other investors would see things the same way, and the expected future interest rate would be driven down until it was such as to exactly equalise the expected returns.
97 By a generalisation of this logic, the so-called forward interest rates expected to apply at future dates can be calculated. One consequence is that for the yield curve to slope downwards it is sufficient for the market to expect a fall in short-term interest rates.
98 Counsel for EA submitted calculations prepared by its expert, Dr Hird, of the ten-year CGS expected, as at June (and alternatively May) 2008, to apply on 25 February 2009. The precise significance of this date will become clear later, but it is closer to the date of the Final Determinations than were the proposed averaging periods. The Tribunal accepts that the methodology used is the well established one outlined above. It notes that the methodology relies on assumptions beyond the factual information embedded in the yield curve. Any methodology must.
99 Without accepting as facts the specific results of the calculations provided by counsel for EA, the Tribunal finds that the yield curve provided by the AER was not a sufficient basis to lead it to the conclusions stated above, viz. that using the proposed averaging periods would lead to systematic ex ante overcompensation of firms relative to their efficient cost of capital.
100 For these reasons, we consider that the AER’s decision, as expressed in its letter of 8 July 2008, to withhold agreement with the averaging periods proposed to it was unreasonable. The AER’s submissions in support of its decision all fail because it did not have sufficient reason to believe that the proposed averaging periods were unlikely to produce an unbiased estimate of CGS rates in the regulatory control period, once it is properly understood how those rates are applied under the Transitional Rules.
101 In its written submission in reply, the AER resiled substantially from the position with respect to the yield curve that it had expressed in its Final Decision and its earlier submissions. It stated that it pointed to the trend in the yield curve, not in defence of its decisions to withhold agreement to the proposed averaging periods, but as illustrating why, in April 2009, the AER discounted the certainty argument advanced by the NSPs for their choice of those periods. In this context the Tribunal notes that, in reaching its decision, it has in any case placed no weight on the Applicants’ expressed desire for certainty regarding the WACC for the purpose of securing capital for its investment program. An NSP is entitled to take whatever considerations it wishes into account in proposing an averaging period. However, neither the NEL nor the Rules provide any basis for the regulator taking account of a desire for certainty of that kind. As the AER correctly pointed out in its letter of 8 July 2008, the very process of determining an NSP's allowable revenue provides certainty with respect to the rate of return that will apply. In any case, there is evidence that NSPs may or may not arrange the raising of finance to coincide with the averaging period, and that hedging arrangements are available to deal with risks associated with financing.
102 Moreover, the NEL and Rules seek to ensure that an NSP operates and invests efficiently in the manner of a firm in a competitive environment. Such a firm would never have the luxury of knowing its revenues years in advance.
103 In light of the above, the Applicants have demonstrated that the AER exercised its discretion incorrectly, or its decision in this respect was unreasonable, for the purposes of establishing a ground of review under s 71C(1)(c) and (d). The Tribunal is thus empowered to act under s 71P(2) of the NEL.
WHAT AVERAGING PERIOD SHOULD BE APPLIED?
104 Having decided that the AER unreasonably withheld agreement with the averaging periods proposed by the Applicants in their original regulatory proposals, the Tribunal must decide what averaging period should apply.
105 On the basis of the submissions put by the parties, three categories of possible averaging periods arise for consideration:
• Those originally proposed by the Applicants, from which, the Tribunal has found, the AER unreasonably withheld agreement. These periods differ from one NSP to another but all fall within the months of June, July and August 2008.
• Those set out in the Applicants’ Revised Proposals, which again differ, but all fall within the period ending 5 September 2008.
• The period specified by the AER: 2-20 February 2009.
106 The Applicants sought a decision by the Tribunal that adopted their revised proposed averaging periods or, as a second preference, their originally proposed averaging periods.
107 The Tribunal has already explained above that it sees no special virtue in an averaging period close to the date of the AER’s final decision (if indeed February be considered close to the end of April). In addition, it accepts the Applicants’ submissions that by late April 2009 it was apparent that yields on CGS in February and March were at an unusually low level. The only relevant facts agreed among the parties are that on 15 January 2009 the yield had fallen to 3.89 per cent, its lowest level in the 40 years since the Reserve Bank of Australia commenced publishing data in July 1969; and that, by comparison, between 1998 and mid 2008 the yield tended to range between 5 and 7 per cent and exceeded 6 per cent for most of the year immediately preceding the AER’s final decisions. The average rate during the AER’s averaging periods was around 4.3 per cent.
108 The Applicants submitted that these facts demonstrated that basing a risk free rate on the AER’s specified averaging periods would not achieve the objective of an unbiased rate of return consistent with market conditions at the date of the final decision. They appealed to expert opinion that the market risk premium was far higher than its deemed value while the risk free rate was abnormally low, so that the return required by investors was much higher than the AER’s specified averaging period would generate.
109 The AER argued that the decrease in official (ie, short term) interest rates – a policy response to the softening economic outlook – should not be ignored and that the AER could not assume that a low risk free rate was a short term phenomenon. It considered that the February 2009 averaging periods did not represent an abnormal period in relation to the observed CGS yields. The AER submitted that any argument that the actual market risk premium was out of kilter with the deemed value could not justify choosing an averaging period so as to consciously generate a higher risk free rate in compensation, nor would such a method of choice be permissible under the Rules. However, this does not help the Tribunal in now deciding what averaging period to apply.
110 In fact, the Tribunal must simply have regard to what the NEL and the Rules provide. This has been discussed above. The Tribunal considers that an averaging period during which interest rates were at historically low levels is unlikely to produce a rate of return appropriate for the regulatory period. While the economic outlook is still uncertain, and it is not impossible that rates during the regulatory period could be at levels somewhere near those prevailing in February/March 2009, the Tribunal considers that to be an unwise assumption. Even if economic conditions were to deteriorate again, there would be no basis at this stage for assuming that historically low interest rates will be representative of each of the five years commencing on 1 July 2009.
111 However, the Tribunal does not consider it appropriate to try to fine-tune the averaging period so that it produces a risk free rate most likely to prevail during the regulatory period. It is one thing to reject a particular period as aberrant. It is quite another to seek an averaging period that produces what is thought to be the best result. The Rules do not provide for the specification of a particular risk free rate. (They could simply have deemed one.) Rather, they provide for the choosing, either by agreement or by specification by the AER, of an averaging period.
112 In these circumstances, where the Tribunal has found that the AER erred in withholding agreement with the averaging periods originally proposed by the Applicants, the Tribunal considers that the objective and principles in the NEL are best met, and can be met under the Transitional Rules (and their equivalents), by agreeing to the Applicants’ revised proposals, as they have sought in these proceedings. Doing so is in accordance with the scheme by which an NSP proposes an averaging period, which cannot be rejected unreasonably. There is no suggestion that the revised averaging periods were chosen opportunistically. Indeed, they produce lower values for the risk free rate than do the originally proposed periods. Nor is there now reason to consider that they may lead to inappropriate outcomes. Nor is there material that was before the AER that suggests a superior choice of averaging periods under the NEL and Rules. The AER has not in these proceedings opposed such an outcome in circumstances where the withholding decision has been found to be in error.
113 Therefore the Tribunal considers it would be appropriate to vary the Final Determinations by agreeing to the NSPs’ revised proposed averaging period, and to remit the matter to the AER with a direction that the AER apply that averaging period in the calculation of the rate of return on capital for the NSP as if the rate of return had applied from the beginning of the regulatory control period (with the additional revenue for the first year of that period to be recovered in the second year).
COST OF DEBT
114 As mentioned above, in addition to the risk free rate, another input to the WACC formula is the observed annualised Australian benchmark corporate bond rate for corporate bonds which have a maturity of 10 years and a credit rating of BBB+ from Standard and Poors, as determined by the AER. The Tribunal considers this rate should be calculated based on the same averaging period as used for the risk free rate, as all the parties agree.
115 As to the remaining issue relevant to cost of debt, we are thus concerned with the period of two, three or four weeks immediately prior to 5 September 2009.
116 The Applicants submitted that the AER erred in its determination of the benchmark corporate bond rate. The AER derived the rate using Bloomberg estimates of fair yields. An alternative data source is provided by CBASpectrum. The two sources generated very similar series prior to the financial crisis, but then diverged. The precise time and magnitude of divergence is not altogether clear. It is agreed only that the divergence had reached historically unprecedented levels on or around October 2008.
117 The Applicants’ position is that the AER should have used an average of the Bloomberg and CBASpectrum-based estimates rather than rely solely on Bloomberg. For the three-week period leading up to 5 September 2009, the average is 22 basis points higher than the Bloomberg-based figure (9.04 per cent vs. 8.82 per cent). For the February 2009 averaging periods specified by the AER, the difference would have been of the order of 100 basis points, some five times as much.
118 By way of background, the AER had used the Bloomberg data rather than the CBASpectrum data in its draft decision based on reviews it had undertaken (using external consultants) during previous revenue determinations made in 2006. This does not seem to have been controversial, and the AER continued to use the same approach in revenue determinations in 2008. In their original proposals, the NSPs had not proposed the use of the CBASpectrum data. They adopted the idea of averaging the two series in their revised revenue proposals, supported by a consultant’s report by CEG, dated 25 March 2009.
119 The Applicants’ argument is lengthy and complex. Much of it was couched in terms of the impact on the estimated cost of debt of the divergence subsequent to the averaging periods in the revised proposals, and in terms of changes in the bond market and in the data series in that later period. (For example, there was dispute about the exclusion from the AER’s estimates of certain later data, and even of a particular bond that was issued only on 24 March 2009.) Indeed, most of the Applicants’ submissions consisted of an attack on a further review that the AER undertook in early 2009 in the light of the revised proposals and the new divergence between the series.
120 One fundamental element of the Applicants’ submissions is that the AER erred in approaching the issues by seeking to determine which of the series was a better predictor of observed yields. It should, they said, have sought to determine whether there was a strong reason to prefer one series to the other; and if not, use an average. An average of two series, each as likely to be an unbiased estimate, would be a better estimate than either taken alone.
121 However, the Tribunal considers that the Final Decisions disclose that the AER carefully considered the arguments that had been put to it. It remained persuaded that the Bloomberg series provided more accurate estimates. Once that conclusion had been reached, it would have been an error to average the two series. The Tribunal is not persuaded that the Applicants have established any reviewable error.
122 No doubt in future revenue determinations the AER will need to consider again the data sources and methodology. That will certainly be the case if there are competing series and continuing divergence between them. The materials provided in these proceedings may well be of assistance to it.
123 But for the averaging periods now determined by the Tribunal, viz. those in the revised proposals leading up to 5 September 2009, the Tribunal sees no compelling case for departing from the AER’s methodology and the Bloomberg series.
INFLATION FORECASTS
124 Transend at one time agitated, as a ground for review, the inflation forecasts determined by the AER. Transend specifically stated that the inflation ground only arises if it does not establish the averaging period ground of review. In the event, Transend has been successful in establishing the averaging period ground of review.
125 The AER’s methodology is to use the inflation forecasts of the Reserve Bank of Australia, which extend for two years, and beyond that the mid-point of the Bank’s target inflation band, ie 2.5 per cent. Transend argued that in the market conditions at the time it submitted its revised revenue proposal (January 2009) this methodology would lead to perverse results when combined with the prevailing risk free rate. Since the conditions giving rise to the concern with the AER’s approach did not apply in the August/September 2008 period, the Tribunal accepts that the inflation ground does not arise.
ENERGYAUSTRALIA SPECIFIC MATTERS
OPERATING EXPENSES
126 EnergyAustralia seeks a review of the AER’s decision not to accept its forecast operating expenditure (opex), in particular the AER’s decision:
(a) not to accept ‘step changes’ totalling $151.3 million in EA’s forecast opex; and
(b) to reduce EA’s forecast maintenance opex costs by $22.4 million.
127 The following deals with the AER’s decision relating to the step changes and maintenance costs.
Step Changes
128 EnergyAustralia’s opex forecasts included a number of what it describes as ‘step changes’ used by it to denote a step up or step down in the operating costs of an activity from a known change in the cost drivers. A step up occurring when a cost not incurred in the base year is forecast and a step down occurring when a cost incurred in the base year is anomalous or is not expected to be incurred in the future. For example, EA forecast:
(a) a step up of $20 million as a result of an increase in land tax payable by it during the regulatory period; and
(b) a step down of $10 million to account for unusual storm costs in the base year.
129 Using its 2006-2007 opex as the base year, EA’s opex forecast for the regulatory control period in its June 2008 regulatory proposal was $2.970 billion. The AER’s November 2008 draft decision did not accept the forecast and substituted $2.64 billion. EnergyAustralia’s January 2009 revised regulatory proposal opex forecast was $2.941 billion. The AER did not accept the revised forecast and its April 2009 Final Decision substituted $2.6281 billion.
130 The AER arrived at the opex figure of $2.6281 billion in its Final Decision by, amongst other things, reducing to zero the value of the opex claimed by EA in respect of the step changes. The Final Decision notes (page 168) that EA advised the AER that its decision resulted in a reduction of $177 million in its forecast opex. A footnote to the agreed Statement of Facts notes, however, that:
“EnergyAustralia calculated ... [the $177m] ..., using its Opex Forecast Model, to include the effect of real cost escalators, including real labour cost, which were a component of EnergyAustralia’s forecast operating expenditure. In its Final Decision, the AER adopted a different cost escalator (Final Decision: p181), resulting in a reduction in EnergyAustralia’s total allowable controllable opex (Final Decision: p204). Once the effect of the adjustment of the labour escalator is taken into account, the step changes decision results in a reduction of $166 million ($2008-09) to EnergyAustralia’s forecast operating expenditure.”
131 The agreed Statement of Facts:
(a) states that EA’s application for review relates to $151.30 million ($2008-09) of opex which was disallowed as a result of the AER’s decision; and
(b) identifies the following step changes and values in issue before the Tribunal:
No Disallowed Step Change Value $m1
1. Incremental iAMS2 / Field Computing 44.86
2. Intelligent Network System and Automation Comms 23.92
3. Incremental IT Capex( Corporate Systems) 18.19
4. Incremental Data Centre 18.19
5. Incremental IT Capex Network Systems 14.57
6. NVD3 Shared Telco Infrastructure 9.74
7. NVD – Demand Management Initiatives 7.86
8. Finance and Commercial – Business Systems4 4.47
9. NVD Telecommunications Support 2.18
10. Information Services 2.18
11. IT Applications – Network System Services 1.91
12. Incremental iAMS Pilot 1.76
13. Business Improvement Team 1.09
14. NVD – Business Systems Operations (-0.2)
15. Metering Systems (-0.22)
_________________________________________________________________
1. Values as they appear in paragraphs [119] to [351] of Chapter 2 of Part 3 of the agreed Statement of Facts.
Note: The total value ($150.5m) is not reconciled to the amount ($151.3m) in paragraph [37] of the agreed
Statement of Facts. [The values also differ from values appearing at Tab 2 of in EnergyAustralia’s Operating
Expenditure Supplementary Materials Volume referred to at Tp665.]
2. iAMS: integrated Asset management System.
3. NVD: Network Venture Development.
4. The AER’s Outline of Submission concedes that: (a) step change 8 should normally be part of the base year expenditure; (b) that it ought to be allowed; and (c) the value of the step change substituted for zero.
132 EnergyAustralia submits that the AER made a number of errors in rejecting the step changes listed and that those errors relate in various ways to the grounds in s 71C of the NEL.
Clauses 6.5.6
133 It is convenient at this point to summarise cl 6.5.6(c) of the Transitional Rules which governs the AER’s acceptance or rejection of an opex forecast by a DNSP. In essence, the clause provides that the AER must accept a DNSP’s forecast opex if it is satisfied that the total of the forecast reasonably reflects the following criteria:
(1) the efficient costs of achieving the cl 6.5.6(a) opex objectives; and
(2) the costs that a prudent operator in the circumstances of the relevant DNSP would require to achieve the cl 6.5.6(a) opex objectives; and
(3) a realistic expectation of the demand forecast and cost inputs required to achieve the cl 6.5.6(a) opex objectives.
134 Clause 6.5.6(a) of the Transitional Rules provides, in effect, that a DNSP’s building block proposal must include total forecast opex which it considers is required to meet each of four opex objectives, namely:
(1) meet or manage the expected demand for standard control services over that period;
(2) comply with all applicable regulatory obligations or requirements associated with the provision of standard control services;
(3) maintain the quality, reliability and security of supply of standard control services;
(4) maintain the reliability, safety and security of the distribution system through the supply of standard control services.
135 It is also convenient at this point to chronicle the several expert reports relied on by:
(a) EA in preparing its June 2008 regulatory proposal;
(b) the AER in preparing its November 2008 draft decision;
(c) EA in preparing its January 2009 revised regulatory proposal; and
(d) the AER in preparing its April 2009 Final Decision.
EnergyAustralia and the AER also relied on those reports in advancing their step change submissions before the Tribunal and they are central to the Tribunal’s consideration of the step change issue.
The first Wilson Cook report
136 In the course of preparing its November 2008 draft determination, the AER received from Wilson Cook & Co Limited (engineers and management consultants advisers and valuers), a report on, amongst other things, a review of EA’s forecast opex. The October 2008 report (the first Wilson Cook report) is in five volumes. Volume 1 covers general matters common to a number of DNSPs that Wilson Cook were reviewing for the AER and Volume 2 addresses issues specific to EA.
137 Wilson Cook was retained by the AER to, amongst other things, make recommendations it considered necessary for the fixing of appropriate levels of expenditure in the context of cl 6.5.6 of the Transitional Rules. Volume 1 of the first Wilson Cook report sets out the opex objectives and criteria in cl 6.5.6(a) and (b) and, noting that its terms of reference did not define ‘prudence’ or ‘efficiency’, states that it adopted the following approach:
We first noted that the objective of the review was in essence to assess the DNSPs’ expenditure proposals and report to the AER on whether in our opinion the forecast expenditure reasonably reflected the efficient costs of a prudent DNSP working in the circumstances of the DNSP concerned.
We noted that to ensure adequacy or effectiveness, a prudent operator might undertake more work than otherwise considered necessary but to ensure efficiency it might undertake less and thus a balance between the two is required.
We noted that prudence has connotations of exercising sound judgement especially concerning one’s own interests, being careful to avoid undesired consequences, being cautious or circumspect in one’s conduct, managing carefully and with economy.
Prudence is often best judged by the absence of evidence suggesting a lack of it. In the case of electricity networks, imprudence might be most discernible if there was evidence of failure to invest adequately, accompanied by identified adverse consequences, and is thus best assessed retrospectively.
Where we considered that there was an appropriate balance between these factors, prudence and efficiency, we have said in the text that the expenditure is “reasonable”. Where we found identifiable instances of imprudent expenditure, an imprudent failure to make expenditure or of what appeared to be inadequate provision for future expenditure, we have identified them.
We considered efficiency in terms of the nature or timing of expenditure and looked for evidence that as far as practicable the expenditure reflected optimal planning and design and competitive costs taking account of local factors, ‘good electricity industry practice’ and the defined security of supply and service standards of the DNSP concerned.
We interpreted good electricity industry practice to be the exercise of that degree of skill, diligence, prudence and foresight reasonably to be expected of a distribution business working under the prevailing conditions consistent with applicable regulatory, service, safety and environmental objectives.
138 The approach to the cl 6.5.6(c) concepts of ‘efficient’ and ‘prudent’ adopted by Wilson Cook in Volume 1 of its first report is non-controversial. Indeed, as may be seen from the following extract from a May 2008 NERA report (the first NERA report), Wilson Cook’s approach to prudent and efficient is consistent with the approach taken by EA’s own economic consultant when advising it on its preparation of its June 2008 regulatory proposal):
In principle, a distinction could be drawn between the ‘efficient costs’ required by the first criteria and ‘the costs that a prudent operator [..] would require’, as set out in the second criteria,... . However, the structure of clauses 6.5.6(c) and 6.5.7(c) effectively rules this out for the purposes of the AER’s assessment of the expenditure forecasts. If such a distinction were to be drawn (ie, if the AER considered that the forecasts reflected the costs a prudent operator would require, but not the efficient costs) then the forecasts could not simultaneously satisfy the first two expenditure criteria.
This conclusion begs the question of the relationship between efficiency and prudence. ... the efficient timing of investment is conditional on the view taken as to the probability distribution of asset failure, which is inherently uncertain. A prudent DNSP may take a more pessimistic view of the probability of asset failure, and decide to replace assets earlier. Such replacement is both prudent (since it reflects a degree of risk- aversion) and efficient (based on the DNSP’s view of the probability distribution).
... efficiency is a dynamic process. Whether or not a firm is operating on the efficiency frontier is also something that cannot be objectively verified. The reference to a ‘prudent operator’ in the expenditure criteria provides some guidance as to how efficiency may be identified in practice. We have already identified that a key aspect of prudence is the process followed by the DNSP. An important dimension of the prudence of a process is the degree to which it is motivated by (or reflects) improvements in efficiency. A process that is motivated by efficiency will in turn ensure that the DNSP moves closer to the efficiency frontier, even though that frontier will itself be moving.
For example, a prudent process is likely to be one that considers alternative options for undertaking an augmentation. The motivation behind that process is to select the least cost option for that augmentation (all other factors being equal), ie, it is an efficient option.
A prudent process can therefore be expected to result in the DNSP moving towards maximum cost efficiency, even as that efficiency benchmark is itself moving. In other words, an assessment of prudence, ie, satisfaction of criterion (2), can be expected to also lead to satisfaction of criterion (1), over time.
139 What is controversial is the tests outlined in the following extract from Section 9.3 of Volume 2 of Wilson Cook’s first report and how the tests were relied on and applied by the AER in its November 2008 draft determination:
Proposed Step changes from Base Year
EnergyAustralia has factored a large number of step changes into its forecast level of opex. Most of these occur between the base year and the start of the next period. Excluding adjustments for abnormal items in FY 200773, the step changes in this period total $64 m. The effect of the step changes is to add approximately 15% to average opex in the next period compared to the base year. The step changes are mostly the result of business decisions made by EnergyAustralia, not decisions made in response to outside factors. Some are proposed on the ground that the base year for some activities was not “normal”: others, particularly in IT, arise from incremental opex related to capital investments.
The proposed step changes are reviewed in section 9.5 but we note the following general points. First, in a competitive market, businesses do not normally add to their own costs unless they are satisfied that there is a benefit to customers in terms of the product delivered or to the business in terms of efficiency. Regulation presumably ought to incentivise natural monopolies in a similar way. Second, businesses are dynamic, with variations occurring from year to year. Such variations ought not to form the basis of a claim for a step change, as the effect of that would be to allow costs to be passed on readily in contravention of the efficiency objective implicit in the regulatory framework.
We consider that a methodology such as that used by EnergyAustralia that starts with a base year and then applies cost escalators, workload escalators and step changes (which apart from some adjustments for abnormal items in the base year are almost all additional costs) without any explicit consideration of business efficiency improvements or potential cost savings is likely to lead to a forecast of future costs that is above an efficient level.
We therefore consider that for acceptance as a step change, a cost ought to relate to a fundamental change in the business environment arising from outside factors or be offset by cost efficiencies in other areas.
________________________
73 The effects of the storm in 2007, the superannuation rebate and an accounting change relating to fleet and logistics recoveries.
140 Noting that:
(a) EA had applied a large number of step changes to its base-year forecast based on expected business cost changes;
(b) the majority related to increases in operating costs resulting from capital investment in IT systems and property; and
(c) the other major step change was for incremental apprenticeships,
Wilson Cook concluded:
Due to their large number, we did not review each change individually except to consider whether it met the test of a valid step change as set out in section 9.3. We did not consider any of the step changes listed in the table met the test of being necessitated by a fundamental change in business activity due to factors outside the control of the business. However, we accepted the step change for incremental apprenticeships on the basis that this is fundamental to the delivery of the proposed capital and maintenance programme in the next period. An adjustment is proposed to remove the other step changes.
141 The adjustment recommended by Wilson Cook was $316 million reflecting its proposal to remove the majority of EA’s step changes.
The AER’s draft decision on EA’s step changes
142 The manner in which the AER relied on and applied the tests in Wilson Cook’s first report and accepted its recommendation may be seen from the following extract from its November 2008 draft decision:
The AER has considered Wilson Cook’s review of EnergyAustralia’s controllable opex forecast for the next regulatory control period and accepts the recommendation that an adjustment of $316 million should be made. This adjustment reflects Wilson Cook’s recommendation to remove the majority of step changes included in EnergyAustralia’ s controllable opex forecast for the next regulatory control period.
Wilson Cook removed these proposed step changes on the basis that they did not meet its criteria for an acceptable step change. Wilson Cook noted that, in general, a step change should:
• deliver a benefit to customers in terms of the product delivered or to the business in terms of efficiency
• be non-recurring in nature or relate to a fundamental change in the business environment arising from outside factors.
Wilson Cook also considered that the application by EnergyAustralia of workload escalators as well as step changes did not include any consideration of business efficiency improvements and, therefore, has the potential to over-estimate the level of future costs.
The AER considers that the step change criteria adopted by Wilson Cook to assess EnergyAustralia’ s proposed step changes accord with the opex criteria in that they ensure any step changes reflect the efficient costs a prudent operator would require to achieve the opex objectives.
EA’s response to the draft decision
143 In preparing its January 2009 revised regulatory proposal in response to the AER’s November 2008 draft decision, EA received advice from four experts whose reports were submitted to the AER and advanced in argument before the Tribunal in support of EA’s step change-specific submissions. The following paragraphs summarise the salient points of the reports necessary to an understanding of the parties’ step change-specific submissions
The PwC report
144 The first of those reports, produced by PricewaterhouseCoopers (PwC), is a brief (in substance but three pages) case study of how EA’s opex forecasts for its integrated Asset Management System (iAMS) factored in future efficiency improvements. Based on information provided to it by EA that if iAMS were not rolled out, ‘significantly more’ support staff would be required than the two additional staff EA proposed with the rollout, PwC concludes that the extent of efficiency gains attributable to the iAMS is not readily quantifiable due to the iAMS being implemented concurrently with a number of other efficiency reform programs.
The Concept Economics report
145 The second report, produced by Concept Economics, provides a theoretical foundation for EA’s step change-specific submissions. The report addresses the question of how and when EA’s ‘non-system’ information technology opex may deliver efficiency gains in an environment of intensive capex and deployment of new IT systems throughout EA’s business and distribution network.
146 Concept Economics dismisses an approach, based on a narrow application of the concepts of economies of scale and scope, that postulates high levels of capex and the introduction of new technology would (all else being equal) justify lower opex forecasts and a more aggressive assumption of future productivity gains.
147 Relying on what it describes as a significant body of economic evidence, Concept Economics advances the proposition that during high capex investment phases, ‘complementary’ opex is also likely to be high and efficient because this type of ‘non-system’ opex is, by its nature, ‘enabling’ technology giving rise to firm wide adaptation costs. Key examples of such technologies include information and communication technologies. According to Concept Economics, economic literature and empirical evidence suggests that enabling technologies require significant complementary investments to produce gains at the firm level. Indeed, a failure to appropriately support capex with related opex risks ‘stranding’ the capex and promoting poor customer outcomes.
148 Concept Economics also observe that a characteristic of operational efficiencies produced by technological innovation is that gains are typically lagged at both a firm and an economy level and that evidence suggests it would be unsafe to assume that efficiency benefits are realised instantaneously at the point of initial investment, or flow evenly through time. Instead, it says, it is well recognised that efficiency gains from technology innovations and the firm-level changes they inspire are significantly lagged.
The second NERA report
149 The third report, produced by NERA in February 2009 (the second NERA report), addresses squarely the Wilson Cook criteria that EA submits are inconsistent with the cl 6.5.6(c) opex criteria. This report observes that:
(a) the cl 6.5.6(a) opex objectives effectively determine the parameters of allocative efficiency for a DNSP; and
(b) the cl 6.5.6(c) criteria determine:
(i) a DNSP’s productive efficiency (whether the DNSP’s opex forecast is efficient and the costs that a prudent operator would require to achieve the opex objectives); and
(ii) because the evaluation of costs in cl 6.5.6(c) is not limited to current costs, it also encompasses dynamic efficiency, a longer-term view of efficiency over time.
150 Noting that Wilson Cook’s characterisation of behaviour in a competitive market refers to:
(a) benefits to the business in terms of efficiency relates to productive efficiency; and
(b) benefits to customers in terms of the product delivered, reflect allocative efficiency, the second NERA report observes that:
“... the focus of Wilson Cook’s subsequent analysis and the AER’s consideration appears to be solely on productive efficiency. The question of whether the proposed expenditure provides benefits to customers in terms of the product delivered (ie, whether the expenditure is allocatively efficient) does not appear to have been directly considered.
In particular, Wilson Cook’s characterisation of step changes refers only to them being ‘offset by cost efficiencies in other areas’, and not to them being justified through providing benefits to customers and meeting the expenditure objectives in clause 6.5.6(a). This leads Wilson Cook to focus only on productive efficiency gains. This is evident in the following comments:
‘EnergyAustralia has not demonstrated that its proposed expenditure takes into account specific improvements in organisational efficiency or productivity’
And:
‘we consider that the large investment proposed in IT systems and property should lead to improvements in business efficiency and reductions in opex’.
Likewise, the AER refers to a lack of evidence in relation to ‘efficiency gains and other savings to offset the proposed expenditure associated with the step changes.’
Given the AER’s acceptance that step changes may deliver either cost efficiencies or benefits to customers, we would have expected the AER to consider the link between inputs and outputs in the case of EnergyAustralia’s proposed step change expenditure, and in particular what improvement in outputs the step change in expenditure may provide.
151 Expanding on the link between inputs and outputs, the second NERA report concluded:
In giving effect to its obligations, we would have expected the AER to consider the link between inputs and outputs in the case of EnergyAustralia’s proposed step change expenditure, and in particular what improvement in outputs that expenditure may provide. This would be consistent with the AER’s statement that a step change should include a benefit to the business in terms of efficiency or to customers in terms of the product delivered.
Expenditure that provides benefits to customers may be considered prudent and efficient (and therefore consistent with the expenditure criteria in the National Electricity Rules), even in the absence of off-setting cost efficiencies. In this context, the ‘benefit to customers’ need not represent an improvement from the current quality of service, but could be the maintenance of the current quality of service, if the alternative in the absence of the expenditure is a deterioration in service quality.
152 Commenting on the appropriate counterfactual to assess ‘offsetting’ productive efficiency gains associated with the proposed step changes, the second NERA report observes that:
(a) the implicit counterfactual upon which the AER and Wilson Cook relies is maintaining the status quo; and
(b) in fact, the relevant counterfactual need not reflect the status quo.
NERA observes in this regard that:
If expenditure was not subject to a step-change, output may stay the same, or it may increase or decrease, depending on the specific circumstances. For example, if demand is increasing, this may require an increase in expenditure, simply in order to maintain the specified level of reliability. Similarly, in a situation with increasing capex, there may also be a requirement to increase supporting opex, in order to maintain the same output levels.27
It is therefore important to consider explicitly how outputs will change if expenditure remains the same. This then becomes the relevant counterfactual against which the achievement of productive efficiencies should be assessed. Consideration of the appropriate counterfactual also informs a view as to what change in expenditure would be needed to maintain output as it is (or to increase or decrease it, as the case may be).
_____________________
27 We note that there is no a priori relationship between these two forms of expenditure: more capital expenditure may mean more or less operating expenditure, for the same level of outputs.
153 Concluding its observations on the relevance of an appropriate counterfactual NERA notes:
In our opinion, the AER’ s conclusion on whether the proposed step change expenditure is prudent and efficient therefore needs to consider both what service outcomes would be delivered in the absence of the expenditure [..] and the evidence presented on what overall operating expenditure levels would be in the absence of the proposed expenditure.
154 In the context of advocating a longer-term perspective, the second NERA report observes that the first Wilson Cook report states that:
(a) ‘We consider that the large investment proposed in IT systems and property should lead to improvements in business efficiency and reductions in opex.’; and
(b) such efficiencies should ‘offset’ the proposed step change in operating expenditure.
155 Based on those observations, NERA expresses a view that these statements by Wilson Cook infer that the associated efficiencies should be achievable within the same regulatory period as the planned expenditure.
156 Contrary to its view of Wilson Cook’s statements and consistent with the view held by Concept Economic that it is well recognised that efficiency gains from technology innovations and the firm-level changes they inspire are significantly lagged, the second NERA report concludes that:
It is incorrect to assume that all expenses will be offset by efficiencies and that these efficiencies will occur within the same regulatory period as those expenses are incurred. If forecast expenditure is disallowed or reduced on this basis, expenditure that is otherwise consistent with the operating expenditure criteria set out in the National Electricity Rules, may be disallowed.
The Huegin report
157 The fourth expert report obtained by EA in preparing its revised regulatory proposal was from the Huegin Consulting Group (Huegin). The Huegin report identifies three concerns with the recommendations contained in the first Wilson Cook report to remove certain step changes advanced by EA. First, that Wilson Cook’s definition of criteria for a valid step change has no reference in the Transitional Rules. Second, a concern with what Huegin identifies as the assumption by Wilson Cook that any cost increase should be offset by equivalent and immediate cost savings. Third, a concern about Wilson Cook’s use of the Composite Scale Variable as a comparative determinant of efficient levels of opex.
158 Expanding on its first concern, Huegin concludes that the following four step change criteria have no basis in the Transitional Rules:
Criteria Description
1. External Factors: relates to a fundamental change in the business
environment arising from outside factors.
2. Cost Efficiencies: delivers a benefit to the business in terms of
efficiency.
3. Customer Benefits: delivers a benefit to customers in terms of the
product delivered.
4. Transitional Costs: is non-recurring in nature.
159 Notwithstanding its view of the four criteria, each step change was reviewed by Huegin against those criteria. Huegin concluded that:
Once broken down to a level of detail sufficient to make meaningful assessments, only 12 of the step changes (representing 11 % of the overall cost increase) do not meet any of proposed criteria of a ‘valid’ step change. One common driver of expenditure which is not addressed by Wilson Cook & Co’s criteria is that of risk mitigation - generally an avoided cost. When risk mitigation is included as a valid driver of expenditure, only 7 of the proposed step change cases (representing 4% of the value) do not meet any of the criteria.
Given the tightly regulated environment that EnergyAustralia operates in, and the existence of penalties associated with the realisation of certain risks, we believe that risk mitigation is an appropriate criterion for inclusion in a step change validation test. Unlike a competitive business that can influence profit through the acceptance of higher risk, a DNSP must allow for the appropriate level of expenditure to operate at acceptable levels of risk.
160 In relation to its second concern, Huegin observes that while it is reasonable to expect benefits from particular cost increases, the assumptions of how and when these efficiency gains are made are overly simplified in the AER’s draft decision. Specifically, it says, they ignore the fact that:
1. Not all expenditure will result in efficiency or productivity gains - there are a number of other benefits that may be realised, e.g. regulatory compliance, risk avoidance and customer service levels; and
2. Where cost savings are possible as a result of investment, the effect is not always evident in that particular cost centre and is very rarely immediately realised - but may materialise later and/or in other parts of the business.
161 In expanding on its third concern, (what Huegin perceived as Wilson Cook’s use of the Composite Scale Variable (CSV) as a comparative determinant of efficient levels of opex) Huegin:
(a) sets out some history of the CSV, noting that it was first used for determining representative opex in 1999 by the Office of Gas and Electricity Markets (Ofgem, the body which regulates the power supply industries in the UK);
(b) notes limitations attaching to the use of CSV and regression analysis generally where the sample size is small;
(c) attacks Wilson Cook’s application of the CSV, in particular, for forcing the regression line through the origin which suggests that as the scale of a DNSP decreases, its fixed costs approach zero; and
(d) concludes that the first Wilson Cook report places undue reliance on the incorrect application of regression analysis using the CSV.
162 Huegin’s criticism of Wilson Cook’s use of and application of the CSV flows over to its criticism of Wilson Cook’s ‘top-down and bottom-up’ assessment of EA’s opex where, it says, the CSV was used as a determinative of cost efficiency. Huegin is of the view that ‘top-down and bottom-up’ assessment in the first Wilson Cook report is effectively the same activity conducted at the total level (labelled top-down in the review) and repeated at the slightly more granular level of opex, maintenance and other (labelled bottom-up in the review). Huegin claim that:
(a) both analyses rely on simple calculations of a small percentage change to the same base year costs; and
(b) Wilson Cook’s assumption that its top-down analysis validates its adjusted bottom up analysis is fundamentally flawed.
163 Huegin proposes an alternative method for assessing step changes against a modified version of an ISSR Framework (a concept developed by Booz Allen Hamilton in the 1980’s to describe the inverse relationship between the ability to impact cost and the ability to influence the cost drivers). As described by Huegin, the Framework aims to describe the relationship between the impact and controllability of four categories of costs, being:
• Inherent - Costs which have a significant impact, yet are beyond the ability of the business to change (without major strategic decisions) due to some third party or environmental influence (such as geography) or design constraint.
• Structural - Costs borne due to economic influence, the type of production materials or technology or as a legacy of historical events.
• Systemic - Costs resulting from business rules and policies.
• Realised - Cost resulting from work and labour force management practices and decisions.
164 Applying its modified ISSR Framework, Huegin found that:
(a) EA is constrained in the step changes that it can avoid;
(b) a large percentage of the step changes rejected by Wilson Cook will occur whether or not they are accepted by the AER; and
(c) should the AER accept the recommendations in the first Wilson Cook report, EA would be forced to find an unlikely cost reduction in its core business activities.
EnergyAustralia’s revised regulatory proposal
165 While EA’s January 2009 revised regulatory proposal made some adjustments to the step changes contained in its June 2008 proposal, based on the four expert reports summarised above, the revised proposal:
(a) criticised the AER’s draft determination for its reliance on, and application of, the first Wilson Cook report; and
(b) pressed its claim for the step changes to be accepted.
The AER’s Final Decision on EA’s step changes
166 In reaching its decision whether to accept EA’s forecast opex and in particular its step changes, the AER summarised three broad arguments advanced in EA’s revised regulatory proposal against the adjustment made to its opex in the draft decision as follows:
(a) the AER and Wilson Cook did not consider all of the material in EA’s regulatory proposal;
(b) the AER uncritically relied on Wilson Cook’s analysis rather than supplementing it with its own analysis; and
(c) much of Wilson Cook’s analysis was flawed.
167 EnergyAustralia’s attack on the AER’s reliance on and application of the first Wilson Cook report was particularised by the AER as follows:
(a) Wilson Cook’s criteria for accepting step changes were inconsistent with the Transitional Rules, too narrow (for example, they do not include risk mitigation) and were not applied consistently, resulting in the rejection of prudent expenditure;
(b) Wilson Cook’s bottom up analysis includes simplifying assumptions (specifically, that opex step changes would be off-set by efficiencies) to avoid a detailed review of the step changes; and
(c) Wilson Cook’s top down benchmarking analysis has significant methodological errors in the application of the cost scale variable analysis.
168 To assist it in addressing those arguments the AER sought a second report from Wilson Cook (the second Wilson Cook report).
The second Wilson Cook report
169 Wilson Cook’s consideration of the second NERA report was time constrained. It is, however, clear from a reading of the second Wilson Cook report that it addressed the salient points in the second NERA report and, concentrating on the arguments EA identified as crucial, each of the other expert reports submitted in support of EA’s revised regulatory proposal. The following review of its consideration of those reports is in the sequence that they are referenced above.
Wilson Cook’s response to the PwC report
170 The second Wilson Cook report observes, correctly, that the PwC report appears to be overly dependant on advice from EA, made no attempt to indicate the scale and timing of the benefits of the iAMS rollout and stopped short of investigating the broader implications of the rollout.
Wilson Cook’s response to the Concept Economics report
171 Wilson Cook also correctly observes of the Concept Economics report that the high-level conceptual picture painted in the report ought not to excuse EA from the requirement of identifying the expected benefits and savings or quantifying (or at least attempting to quantify) the benefits and savings that it expects will be realised from its opex. Most relevantly, Wilson Cook observes that:
Were these requirements to be excused, it would imply that:
(a) even the vaguest assurances of future materialisation of cost savings and benefits ought to be accepted by regulators or advisers assessing the efficiency of expenditure, without any requirement for their quantification or possibly even their identification; and
(b) customers would bear the full risk of the un-identified or non-quantified potential efficiency gains or benefits being realised in the future.
We considered it necessary, for the purpose of our work, to ask for this information but as already reported, we found that no such detailed information had been supplied.
Wilson Cook’s response to the second NERA report
172 Responding to the extract from the second NERA report to the effect that in considering whether a step change is prudent and efficient the AER needs to consider both:
(a) what service outcomes would be delivered in the absence of the expenditure; and
(b) any evidence on what overall operating expenditure levels might be in the absence of the expenditure,
Wilson Cook observes it did not find any explicit and/or quantified evidence upon which it might determine whether the outcomes or expenditure in the absence of the proposed expenditure might be prudent or efficient.
173 Wilson Cook rejects NERA’s conclusion to the effect that statements by Wilson Cook imply that associated efficiencies should be achievable within the same regulatory period as the planned expenditure. This, it says, is NERA’s inference. Wilson Cook’s position is that there should be future benefits, some of which would be expected in the current regulatory period because if the majority were realised far in the future, their present value might not justify the expenditure.
174 Concluding its observations on the second NERA report, Wilson Cook noted that as it was received only a week before its reporting deadline, it did not have time to examine it in detail.
Wilson Cook’s response to the Huegin report
175 Responding to Huegin’s first concern about the criteria used to assess the validity of step changes in Wilson Cook’s first report, Wilson Cook concluded that the criteria stated in its first report remained appropriate but that they ought to be expanded to clarify what it intended as their interpretation and to deal with certain situations explicitly. Accordingly, Wilson Cook revised the criteria in its first report as follows:
Revised Criteria
The revised criteria that we propose to address EnergyAustralia’s concerns in relation to its step changes are set out below. They are for application after the business has demonstrated: (a) that it has adjusted its base-year expenditure to remove items that were abnormal or will clearly not recur28 and to add items that would normally be present;29 and (b) that the step changes do not duplicate any allowances for workload escalation or inflation in the next period that have been applied separately.
For a step change to be accepted, the business should then be able to demonstrate that:
(a) it is related to a fundamental change in the business environment arising from outside factors or offset by cost efficiencies in other areas (the original criterion);
(b) it is attributable to the imposition of new or changed obligations due to external factors including, if relevant, mandated improvements in service levels (an extension of the interpretation of (a) above);
(c) it is of a type that will improve service levels voluntarily as opposed to being mandated - in respect of which customers’ willingness-to-pay for the improved service should be demonstrated (a further extension of the first criterion);
(d) it will bring cost savings or benefits to customers - in respect of which, the business should be able to demonstrate that: (i) it is continually looking for better ways of using its resources and improving its processes and systems to improve service levels or achieve cost efficiencies; (ii) it has defined the savings and benefits in terms of their nature and the expected time if their realisation; and (iii) where the savings and benefits are quantifiable, they have been quantified in sufficient detail for cost-benefit analyses to be prepared and that the cost-benefit analyses justify the investment; or
(e) alternatively, if it does not meet any of these criteria, the business has demonstrated that it will continue to operate efficiently as a whole, despite the cost increase.
__________________________
28 Done in our Final Report as a precursor to the step change analysis.
29 None identified in our Final Report but some identified in this review.
176 In relation to its criterion (d), Wilson Cook agreed with statements of EA’s experts that there may be a lag in the realisation of some benefits, but added two qualifications. First, it is reasonable to expect that the benefits from investments planned for implementation at or near the beginning of the period under review should accrue, or at least begin to accrue, before the end of the period. Second, that there may be a lag in the realisation of benefits, should not apply to all types of benefit. Wilson Cook also repeated its previously made point that the present value of benefits that are delayed substantially might not be sufficient to justify the investment.
177 In relation to its criterion (e), Wilson Cook stated the criterion would be satisfied only by a ‘... robust benchmarking of the business’ future costs over the regulatory period, sufficient to demonstrate its continued efficiency in comparison with industry norms.’
178 Finally, in relation to all its criteria, Wilson Cook considered it necessary to identify evidence of compliance with the criteria, including evidence of the quantification of benefits as envisaged in its criterion (d).
179 In revising its criteria, Wilson Cook considered Huegin’s belief that risk should be added to the factors used to determine the validity of a step change and concluded that:
(a) risk cannot be considered unless cost, benefits and potential adverse impacts are quantified;
(b) little or no quantification of benefits had been attempted by EA; and
(c) therefore no quantification of risk by Wilson Cook was possible.
180 In explicitly recognising improved service levels, Wilson Cook’s revised criteria also address concerns raised by NERA to the effect that the focus of Wilson Cook’s subsequent analysis and the AER’s consideration appears to be solely on productive efficiency.
181 Because the CSV was used only in a ‘top down’ analysis as a check of reasonableness, the quantum of its adjustment being determined from a different (‘bottom-up’) analysis, Wilson Cook rejects Huegin’s third concern, to the effect Wilson Cook used the CSV as a determinative of cost efficiency, as an incorrect interpretation that the CSV determined its findings. Other points raised by Huegin about Wilson Cook’s use of the CSV prompted Wilson Cook to re-examine its use and to undertake a fresh analysis. This analysis, relying solely on Australian data for a group of predominantly urban utilities, addresses particular concerns in the Huegin report about the sample size used by Wilson Cook and its regression analysis.
182 Addressing concerns in the Huegin report about its benchmarking, Wilson Cook undertook a completely revised and robust benchmarking of EA’s opex which confirmed its opinion that while EA’s base year opex should be accepted as efficient, its forecasts are not and should therefore be adjusted.
183 Applying its revised criteria in combination with new information presented in EA’s submissions, Wilson Cook accepted some additional step changes. It was not, however, satisfied that EA had quantified, or attempted to quantify, the cost savings and benefits it claimed to be associated with its step changes, nor shown where it had allowed for cost savings and benefits in its opex forecasts. Accordingly, Wilson Cook recommended against acceptance of the remaining step changes. These step changes relate almost exclusively to IT.
184 In assessing Huegin’s modified version of an ISSR Framework, Wilson Cook found that Huegin appeared to consider only the claimed necessity or unavoidability of the step changes and not the efficiency or cost effectiveness of the step changes. Wilson Cook notes in this regard that a necessary condition for its endorsement of opex was efficiency, or where benefits are said to be in the form of improved service, a statement of the improved service levels resulting from the opex.
185 Wilson Cook also notes that Huegin had not, as far as it was able to tell, quantified the claimed benefits in relation to step changes in terms of amount or time of occurrence. It considered this lack of quantification to be a material weakness in Huegin’s argument that the step changes were justifiable.
The Tribunal’s consideration of EA’s step changes
186 Clause 6.5.6(c) of the Transitional Rules is clear – the role of a DNSP is to provide the AER with an opex forecast that reasonably reflects the three opex criteria and the AER must accept the forecast if it is satisfied that the total of the forecast reasonably reflects the three criteria. EnergyAustralia is correct to submit that it is not the AER’s role to simply make a decision it considers best. It is also correct for it to say that the AER should be very slow to reject a DNSP’s proposal backed by detailed, relevant independent expert advice because the AER, on an uninformed basis, takes a different view. Nor, as EA submits, may the AER reject such a proposal merely because it has an expert opinion. The AER, based upon any expert advice, needs to make its own evaluation, an evaluation that is reviewable by this Tribunal.
187 The opinions of EA’s experts have been outlined. The Tribunal is satisfied that in its second report the AER’s expert (Wilson Cook), in the words of EA, ‘squarely engaged’ with each of the EA’s experts. To the extent EA’s experts may have impugned the first Wilson Cook report, it revised its approach and properly adjusted its recommendations, to accommodate those experts’ opinions. That it has done so is illustrated by reference to Wilson Cook’s revised criteria and the fact that Huegin’s criticism prompted Wilson Cook to re-examine its use of the CSV and to undertake a fresh analysis, using fresh data and addressing Huegin’s concerns about its regression analysis.
188 The AER’s terms of reference for each review conducted by Wilson Cook are apposite. Wilson Cook adhered to the terms of reference given to it by the AER. An exhaustive review of the material before the Tribunal satisfies it that Wilson Cook did have regard to the evidence advanced by EA in relation to its step changes. In reaching its final decision on EA’s step changes, the AER was entitled to rely on, and apply, the recommendations in the second Wilson Cook report. The AER evaluated those recommendations and decided accordingly.
189 EnergyAustralia submits:
(a) that the AER applied criteria developed by Wilson Cook that were inconsistent with the cl 6.5.6(c) opex criteria; and
(b) even if the Wilson Cook criteria were consistent with the cl 6.5.6(c), the AER misapplied them or applied them inconsistently, illogically and irrationally.
190 In developing its criteria Wilson Cook had proper regard to the cl 6.5.6(c) concepts of efficient and prudent. Its views on the acceptability of a step change quoted in the extract from its first report, do not stray from the cl 6.5.6(c) concepts of ‘efficient’ and ‘prudent’ such as to cause disquiet to the Tribunal. Wilson Cook is correct in its observation in that extract to the effect that EA’s methodology of applying step changes to a base year sans consideration of efficiency or potential cost savings is likely to lead to an opex forecast that is above an efficient level. The AER is likewise correct in its submissions that:
(a) if a step change is to reasonably reflect the opex criteria, a cost saving arising from efficiencies within EA’s business attributable to the planned capex and opex should be reflected in the forecast opex;
(b) alternatively, if a cost saving is not expected, a step change should result in a benefit to customers that warrant the forecast opex;
(c) identifying the expected benefit and giving a value to it is relevant to evaluating whether the expenditure is ‘efficient’ and ‘prudent’; and
(d) alternatively, if neither a cost saving nor a customer benefit is expected, the step change should be the consequence of an unavoidable change in activity due to an external obligation.
191 The AER also correctly submits that Wilson Cook’s revised criteria are but a reiteration of the criteria in its first report expanded to clarify what is intended.
192 EnergyAustralia takes issue with the AER’s reliance on the second Wilson Cook report, submitting that as the report was not disclosed to EA until after the AER’s final decision, the AER denied it a reasonable opportunity to make a submission in respect of matters under consideration by it, contrary to s 16(1)(b)(ii) of the NEL. In particular, EA submits that it was denied the opportunity to address an emphasis in the second Wilson Cook report on:
(a) the need for quantification of efficiencies;
(b) benefits to customers; and
(c) risk mitigation.
193 The first Wilson Cook report and the AER’s draft decision should have put EA on notice that the AER was seeking quantification of cost savings or benefits. Indeed, the final paragraph in the extract from Wilson Cook’s second report suggest that Wilson Cook sought from EA information relevant to quantification of efficiencies. That the AER was seeking quantification was in the minds of EA’s experts (and by implication EA itself) is also apparent from the PwC report which concludes that the efficiency gains in respect of one project were not readily quantifiable. Likewise, the extract from the second NERA report indicates that another of EA’s experts was alert to the AER’s concern with a lack of evidence in relation to efficiency gains and other savings to offset the proposed expenditure associated with step changes. The first NERA report commissioned by EA raises the issue of asset failure and, it was EA’s own expert, Huegin, which noted that Wilson Cook’s first report had not addressed the issue of risk mitigation. The issue of risk mitigation is addressed by Wilson Cook in the context of responding to Huegin. None of the matters above is a fresh issue out of the blue capable of taking EA by surprise and sufficient to make its case under s 16(1)(b)(ii) of the NEL, even if such a case was relevant to the review currently before the Tribunal.
194 Clauses 6.12.2 and 6.12.3(f) mandate what must follow the AER’s reliance on, and application of, Wilson Cook’s recommendation: cl 6.12.2 requiring that the AER must set out in its reasons and cl 6.12.3(f) by providing, in effect, that if the AER refuses to approve the total of EA’s forecast opex, the opex it substitutes must be determined:
(a) on the basis of the DNSP’s regulatory proposal: (cl 6.12.3(f)(1)); and
(b) amended from that basis only to the extent necessary to enable it to be approved in accordance with the Rules: (cl 6.12.3(f)(2)).
195 The Tribunal is satisfied that the AER’s final decision which references and attaches a copy of the second Wilson Cook report complies with the requirements of cl 6.12.2.
196 EnergyAustralia also takes issue with the AER’s compliance with cl 6.12.3(f). It submits:
(a) that the AER’s decision to reduce to zero the disallowed step changes was ‘arbitrary and illogical’ and made without regard to the requirement in cl 6.12.3(f)(2); and
(b) the AER should have:
(i) sought quantification from EA; or
(ii) applied a percentage reduction for inferred productivity savings as was done by another DNSP (Integral Energy) and by the Australian Competition and Consumer Commission (the ACCC) which had regulatory responsibility in previous periods.
197 It is apparent from what the Tribunal has already said that the Tribunal is of the opinion that the AER did seek quantification from EA. EnergyAustralia had the opportunity to provide the quantification sought. There is a suggestion in the PwC report that efficiency gains may not be readily identifiable by EA. If that be so, it may have adopted the course it submits the AER should, namely, apply a percentage reduction, as Integral Energy did. As submitted by EA, unlike other regulatory regimes, this regime gives considerable weight to the business experience, calculations and judgments of the regulated entity. EnergyAustralia is far better placed than the AER to undertake the exercise required to quantify the efficiency gains or to arrive at judgements about any percentage reduction for inferred savings. Because EA failed to undertake that exercise, the AER was simply unable to determine a substitute amount on the basis of a current regulatory proposal. The Tribunal is satisfied that the AER complied with cl 6.12.3(f)(2), assuming it applies to this decision making process.
198 EnergyAustralia submits that because the cl 6.5.6(c) opex criteria are the same as the cl 6.5.7(c) capex criteria it is unreasonable and illogical to approve the capex in relation to a project but to reject forecast opex in relation to it. Acceptance of this submission would render cl 6.5.6(c) otiose. It is rejected. There is a two stage process. The first is the DNSP’s decision to undertake capex which must be assessed by the AER in terms of cl 6.5.7(c). If the capex is accepted by the AER, it may have opex consequences, which are separately assessed by the AER against the cl 6.5.6(c) criteria. If, as is the case here, the AER identifies a deficiency in the DNSP’s opex forecast, the fact that it is linked to approved capex is irrelevant.
The Tribunal’s conclusion on EA’s step changes
199 In light of the above reasons, the Tribunal affirms the AER’s decision to reduce to zero each of the step changes other than step change No 8 (Finance and Commercial – Business Systems). The AER concedes step change No 8 should normally be part of EA’s base year expenditure and the Tribunal accepts this concession. The determination of the Tribunal will need to reflect this conclusion.
Maintenance Costs
200 Relying on the second Wilson Cook report, the AER’s final decision:
(a) rejected EA’s forecast maintenance opex because the forecast was based, in part, on an exponential relationship between asset age and maintenance costs;
(b) took a mid-point between:
(i) EA’s average annual estimated increase of 11% in its maintenance costs; and
(ii) Wilson Cook’s projected increase of 5.5% in those costs; and
(c) reduced EA’s forecast maintenance opex by $22.4 million.
201 Before turning to consider EA’s submissions why the AER should not have rejected its maintenance costs forecast, it is helpful to trace how the AER arrived at its decision. To that end, the following paragraphs traverse:
(a) EA’s June 2008 regulatory proposal;
(b) criticism of the proposal in the first Wilson Cook report upon which the AER based its November 2008 draft decision;
(c) two expert reports upon which EA based its January 2009 revised regulatory proposal (the Huegin report and a report by Sinclair Knight Merz (SKM)); and
(d) the second Wilson Cook report upon which the AER based its April 2009 final decision.
As with the step change issue, the parties relied on their experts’ reports in advancing their submissions before the Tribunal and an understanding of them is critical to the Tribunal’s consideration of EA’s forecast maintenance costs.
EnergyAustralia’s June 2008 regulatory proposal
202 Underpinning EA’s maintenance costs forecast in its June 2008 regulatory proposal is a model developed by it, building on earlier advice from SKM, to quantify the trade off between replacement and maintenance expenditure. In developing the model it used a top-down approach and, with the object of confirming the outcomes from that approach, it also undertook a bottom-up assessment of its maintenance requirements.
203 In brief, EA’s top-down approach uses weighted average age as a proxy for expected asset condition to predict maintenance costs with the resultant age curve displaying an exponential relationship between maintenance costs and asset age. EA’s bottom-up assessment involves an analysis of historic numbers of completed planned inspection tasks and calculation of the associated costs per task, which were then compared to actual recorded costs. Future planned inspection costs were adjusted to reflect the replacement of older assets with newer technology that would reduce the ongoing total maintenance costs.
204 In preparing its regulatory proposal EA engaged SAHA International to bench mark its asset management performance with particular focus on maintenance. Based on a number of factors peculiar to EA, SAHA concluded that:
(a) EA meets or exceeds best practice thresholds for asset management practices;
(b) its current asset management regime ensures that its maintenance programs are optimised for both cost and asset performance; and
(c) its maintenance practices were relatively efficient.
The first Wilson Cook report
205 Wilson Cook was not convinced that EA’s regulatory proposal demonstrated that the relationship between maintenance costs and asset age is exponential. Nor was it convinced that a valid exponential curve could be derived from the two data points used by EA which Wilson Cook described as follows:
... the actual current level of opex (based on FY 2006 actuals) and the current age of the assets, and an estimate of the level of opex applicable to new assets (based on a percentage of replacement costs, the percentage having been determined by SKM in 2002)
206 The first Wilson Cook report observes that quantitatively EA’s approach begs the following questions:
(a) whether the cost of maintaining a new asset is comparable with the costs of maintaining an old asset, as this affects the calculation of the ‘new asset’ point and its relationship with the 2007 cost point on the cost v age relationship curves;
(b) whether the relative costs of maintaining a new asset as a percentage of replacement cost, calculated in 2002, is relevant, given the significant change in asset replacement costs over the intervening years;
(c) whether the maintenance costs calculated for 2007 are efficient, given that EA was at that time in the process of catching up on a maintenance backlog from 2006; and
(d) whether the cost v age relationship curves should be exponential.
207 In relation to (d), Wilson Cook noted that exponential growth in expenditure of any type seldom occurs in reality.
208 Noting that although intuitively a relationship between maintenance costs and asset age would appear to exist, Wilson Cook states that evidence available to it from a New Zealand study it undertook suggests that costs may not increase exponentially with the average age of network components, although they may be related to age in another way. It notes that it:
(a) tested the assumption that there is an exponential relationship between maintenance cost and the average age of network components by looking at New Zealand data for 2005 and 2006;
(b) found that network type was a much stronger driver of cost; and
(c) even within networks of the same general type, found no obvious regression and, if anything, a direct linear relationship between direct costs and age seemed to have stronger trends.
209 Expressing doubt about the robustness of EA’s approach, the first Wilson Cook report observes that:
(a) EA may have overstated the relationship between maintenance costs and asset age;
(b) the relationship results in a forecast increase relative to its normalised base year of approximately 11% in average maintenance costs over the next regulatory period; and
(c) if the escalation were based on Wilson Cook’s CSV, an increase of 7% would result.
210 Wilson Cook notes, however, that some increase above that attributable to size alone can be expected. This, it says, is due to the level of EA’s replacement capex being largely directed at transmission, sub-transmission and zone substation assets, not at distribution assets where Wilson Cook expects many maintenance costs would lie. Taking this into consideration, and in the absence of better information, Wilson Cook took, as a reasonable estimate, an increase half way between the upper and lower bounds: that is, an increase of 9%.
The AER’s draft decision
211 While the AER’s November 2008 draft decision accepts EA’s proposition that other things being equal, the level of maintenance expenditure will increase as its network ages, it:
(a) notes Wilson Cook’s concerns regarding the determination of the relationship between maintenance costs and asset age and the application of that relationship to determine future maintenance workloads; and
(b) reduced EA’s maintenance costs forecast by, among other things, adopting the mid-point between its forecast of an 11% increase and Wilson Cook’s 7%.
EnergyAustralia’s response to the draft decision
212 In preparing its January 2009 revised regulatory proposal in response to the AER’s November 2008 draft decision, EA received advice on its maintenance costs from SKM and Huegin. It submitted their reports to the AER and used them to advance its submissions to the Tribunal. The following paragraphs summarise the salient points of the reports necessary to an understanding of the parties’ submissions.
The Sinclair Knight Merz report
213 SKM, which was responsible for developing the model used by EA to forecast its maintenance costs, observes that it is an oversimplification on the part of Wilson Cook to say that EA’s model is based on:
(a) an assumption of an exponential relationship between opex and age (it is based on multiple exponential relationships); and
(b) two known points (it is based on the two points identified by Wilson Cook and the expected profile of planned maintenance and failure rates over the life of the asset class).
214 SKM also observes that:
(a) there are a number of highly respected technical sources which make reference to failure rates of electrical equipment being exponential in character; and
(b) the relationship most commonly used to describe wear-out failures is exponential, or some variation (such as a Weibull curve with a shape factor close to that describing an exponential curve).
215 Responding to concerns in the first Wilson Cook report to the effect that adopting 2002 costs for maintaining new assets and 2007 costs for maintaining older ones was incorrect, SKM expresses the view that:
(a) if the costs used in the model were inefficient, this would not affect the model materially as the inefficiency would apply over the whole opex-age curve and while this may raise the curve it would be unlikely to materially distort it ; and
(b) changes in replacement cost are ‘largely irrelevant’ because:
(i) the replacement cost does not affect calibration of the curve; and
(ii) if the curves were recalculated using current-day replacement costs, the same annual percentage increases would be derived.
216 The SKM report states that:
(a) the purpose of the opex-age curve is not to definitively calculate the exact maintenance costs for a particular asset at a given age, but rather to characterise and approximate the average cost pressures on a DNSP as assets age;
(b) the ‘top end’ of the curve is largely irrelevant, as there are generally few assets in this part of the curve;
(c) the average annual percentage increase in costs applies to the network as a whole and, if the average age is maintained, will show no net increase in average maintenance costs;
(d) it is only as the network as a whole ages slightly (typically by retaining more of the oldest and most expensive assets) that the cost v age relationship is applied - and in practice this will be at the centre of the curve around the point at which the curve was calibrated.
217 Referring to the statement by Wilson Cook to the effect that based on New Zealand evidence, direct costs may not increase exponentially with the average age of the network components, SKM observes that:
(a) extreme care needs to be taken in comparing costs between Australian and New Zealand DNSPs;
(b) there are significant size differences between DNSPs in each country and significant regulatory and operational differences;
(c) it is unclear whether the differences are recognised or taken into account; and
(d) because the true relationship between maintenance costs and age may only be determined by considering the trend of maintenance costs over the lifetime of an asset class, not the two consecutive years used by Wilson Cook, no credibility attaches to Wilson Cook’s conclusion from its New Zealand study.
218 SKM also lists what it considers to be a number of flaws in Wilson Cook’s analysis which lead SKM to conclude it does not have the level of rigor necessary:
(a) for the AER to reject EA’s network and asset specific analysis as unreasonable; or
(b) to support a linear relationship between direct costs and age.
219 In addressing the first Wilson Cook report, SKM also observes that:
While it is difficult to find sufficiently detailed data in the right format to allow analysis that would definitely prove either a linear or exponential relationship (or other), in SKM’s experience an exponential relationship is considered more likely. In practice this will not be an exact exponential, but will be “lumpy” due to different technologies (typically with much higher inherent maintenance costs for older technologies, exacerbated by their age) and other factors.
... ... ...
New assets tend to be of a design and materials requiring less maintenance, and this will be reflected in the maintenance costs for these assets. It is therefore likely that the opex-age curve exhibits both age related cost increases (likely to be exponential) and also technology related increases (likely to exacerbate the age relationship further). SKM considers this is likely to move the overall relationship further away from a linear relationship.
The Huegin report
220 Noting that:
(a) the first Wilson Cook report based its recommendation for adjustment of EA’s maintenance costs on doubts about its age curve displaying an exponential relationship between maintenance costs and asset age; and
(b) while the evidence for the doubts is referenced in the report but not available to it, Huegin limited its assessment of the first Wilson Cook report to the validity of:
(i) EA’ use of the exponential curve; and
(ii) Wilson Cook’s alternative use of its CSV.
221 Consistent with the AER’s acceptance of EA’s proposition that other things being equal, the level of maintenance expenditure will increase as its network ages, Huegin observes that the increasing age of assets is widely accepted as a factor in the increase in maintenance costs. It then observes that:
(a) whether the increase exhibits exponential growth will be determined, in part, by the failure rate of the equipment, which in turn is directly related to its failure mode; and
(b) the failure mode will depend on the physical characteristics of the equipment, installed environment and operating cycles;
(c) failure rates are not the only factor in an exponential maintenance cost increase with age;
(d) the combination of long-life legacy assets and the increasing speed of technological change precipitates obsolescence of spares and diminishing availability of technical skill in repair;
(e) whether overall maintenance cost increase is linear or exponential will depend upon the asset, market and economy;
(f) studies in the aviation industry have shown that the costs compound with increasing fleet age; and
(g) in developing models for optimal asset retirement point decisions for several industries it has noted that many individual items of equipment exhibit exponentially growing maintenance costs beyond a particular age.
222 Conceding that much of the information upon which it bases its observations set out in the preceding paragraph is general, and some anecdotal, Huegin concludes that:
(a) the information demonstrates that EA’s use of an exponential curve is not without basis; and
(b) the countenance it provides to Wilson Cook’s doubts about an exponential curve also highlights the uncertainty in the basis for Wilson Cook’s recommendation for, and the AER’s acceptance of, an adjustment.
223 Based on an examination of Wilson Cook’s use of its CSV, Huegin observes that because the CSV is the combination of three closely correlated variables related to customer numbers and ignores many of the drivers of maintenance costs (such as asset age, network type, capital asset changes), it is considerably less robust than EA’s escalation methodology.
EnergyAustralia’s revised regulatory proposal
224 Drawing on the SKM and Huegin reports, EA submitted in support of its January 2009 revised regulatory proposal that the AER should not give weight to the first Wilson Cook report because, amongst other things:
(a) the SKM and Huegin reports support the assumption that maintenance costs increase exponentially with the age of an asset; and
(b) Wilson Cook did not provide sufficient evidence to support its conclusions.
225 More particularly, EA criticised Wilson Cook’s:
(a) reliance on New Zealand evidence;
(b) use of a mid point to determine a substitute amount for EA’s maintenance opex forecast; and
(c) CSV analysis as ignoring factors peculiar to EA and identified in the SAHA report submitted with its regulatory proposal.
The second Wilson Cook report
226 In order that it might respond to EA’s criticism, the AER obtained a further report from Wilson Cook. To lay a foundation for its response to the criticism, the second Wilson Cook report restates its understanding of EA’s model as proposing that its maintenance costs be adjusted in accordance with a set of exponential curves determined by relating maintenance costs to the average age of assets – separate exponential maintenance cost v age curves being determined for each of six asset groups, based on work undertaken for EA by SKM, with the maintenance costs for new assets (expressed as a percentage of replacement cost) being calculated by SKM using 2002 data and the cost of maintaining older assets using 2007 data. The method applied by EA, as understood by Wilson Cook, is to determine exponential curves, one for each of the chosen categories. Each fitted to two points: the first point being the maintenance cost of new assets (of age zero) and the second point being derived from the combination of:
(a) its records of maintenance expenditure on each of the six groups in 2007; and
(b) the average weighted age of the assets in each group at that time.
227 EA then projects its future maintenance costs by applying, to the developed curves, the changed average weighted ages of each of the asset groups in each future year to take account of the planned level of asset replacement in each asset group during the period.
228 Having restated its understanding of EA’s model and the questions it raised, Wilson Cook concludes that:
(a) because the rate of planned replacement in most asset categories is insufficient to arrest the increase in average weighted asset age, resulting in a net movement ‘up’ the exponential cost v age curve, its effect is to increase forecast maintenance costs each year; and
(b) EA’s maintenance costs forecasts were likely to be overstated.
Wilson Cook’s response to the SKM report
229 The following paragraphs summarize points raised in the second Wilson Cook report, responding to the SKM report, that are pertinent to the parties’ submissions.
230 Addressing SKM’s observation in the first paragraph of its report, Wilson Cook observes that because EA’s average network age is projected to increase over the next regulatory period, application of the exponential relationship results in the derivation of a greater increase in average maintenance cost compared to that which would be derived if, for example, a linear relationship had been applied with the same starting point and initial slope. Refraining from commenting on SKM’s experience on which SKM says it bases its observation, Wilson Cook reiterates its own view that exponentially increasing costs are seldom observed in practice.
231 Commenting on the SKM’s observation in the second paragraph of its report, Wilson Cook agrees with the proposition that a new asset costs less to maintain than an older equivalent asset and notes that:
(a) it never contended otherwise; and
(b) SKM appears to have misinterpreted its questioning whether the cost of maintaining a new asset is comparable with the costs of maintaining an old asset.
232 Rather, Wilson Cook states, what it contends is:
(a) because there is a difference in maintenance cost between new and old assets of the same type, it is not safe to apply a cost model using asset age alone when the gradient should be derived from mixed causes (namely, increasing costs with age and changing costs through replacement with new technology);
(b) the key point is that over time an asset will increase in age but not change in technology;
(c) because the gradient derives from mixed causes, EA’s model is likely to overstate the rise in maintenance cost with increasing average age and therefore overstate the effects of ageing alone; and
(d) a more tenable model would identify and apply, separately:
(i) the effects of ageing for assets of the same technological group; and
(ii) the effects of new technology under the asset turnover mechanism, not the asset ageing mechanism.
233 Drawing on SKM’s own observation that as there are generally few assets in the ‘top end’ of the curve it is largely irrelevant, Wilson Cook notes that while it may be tempting to fit an exponential relationship to a cost-age characteristic that is flat or linearly rising in the initial years but which exhibits steep increases near the end of life, because such a fit over-emphasises the end-of-life characteristic that applies to only a small proportion of the asset population it is not generally safe to do so. As Wilson Cook observes:
Fitting an exponential relationship applies an increasing gradient to the whole age span but the majority of the assets are unlikely to be described correctly by such a relationship.
This, Wilson Cook says, is evident in case studies SKM set out in its report
234 Responding to SKM’s observation that highly respected technical sources make reference to failure rates of electrical equipment being exponential in character and the relationship most commonly used to describe wear-out failures is exponential, or some variation (such as a Weibull curve with a shape factor close to that describing an exponential curve), Wilson Cook acknowledges that highly respected technical literature often characterise failure times as being exponential, but notes that a distinction needs to be drawn between failure times on the one hand and rates on the other. It observes in this regard that:
Failure times refer to the time elapsed before the failure of an asset occurs and are thus related to age, whereas the failure rate is the ratio of the number failing in a given time interval to the number present at the beginning of that interval. SKM appear to have confused the two.
Various technical sources and our own analysis support the conclusion of exponentially distributed failure times for electrical equipment such as circuit breakers but that is not the same as concluding that the failure rates of the assets increase exponentially. Exponentially distributed failure times are a special case of the Weibull distribution where the shape factor (beta) is unity.39 With exponentially distributed failure times, the failure rate, is constant over time. To illustrate this with an example: a fixed population of assets will decrease in number over time as its members fail. The number failing in any given time interval over time will show as the distribution of failure times.
__________________
39. See: “AppIied Life Data Analysis”, Nelson, John Wiley & Sons, 1982, ISBN 0-471-09458-7, p 40.
235 Wilson Cook concludes in this regard that:
If SKM accepts that the relationship most commonly used to describe wear-out failure is exponential or is a Weibull curve with a shape factor close to that describing an exponential curve, ... then SKM ought to have concluded that the relationship describing the failure rate against age, and therefore the implied cost of failures against age, is flat or linear. 40
____________________
40. Costs arising from failures will be determined by both the failure rate with age and the age profile of the asset population. For clarity, when discussing the cost effects of failure rate alone, we assume that the age profile of the asset population is flat.
236 Noting that it has experience in both Australia and New Zealand and that in its view the degree of similarity in maintenance work and work practices in both countries is greater than the degree of difference, the second Wilson Cook report rejects SKM’s observation that no credibility attaches to the conclusions from its New Zealand study. Wilson Cook also address SKM’s criticism of its use of two consecutive years of New Zealand data by explaining that:
(a) the 2005 and 2006 years data referred to in its analysis are data sets collected in those years, each comprising dispersed costs and ages upon which trends were examined; and
(b) two data sets were examined to provide a level of assurance that the observations were consistent between data sets collected in different years.
237 Also noting that it referred to the New Zealand data only to illustrate that it supported its general contentions and thus added weight to its concerns that SKM’s analysis was flawed, Wilson Cook concludes that SKM is incorrect in implying that it relied on its New Zealand study to reject EA’s approach and substitute a linear relationship.
238 Referencing concerns expressed in its first report about EA’s model using 2002 costs for maintaining new assets and 2007 costs for maintaining older ones being incorrect, the second Wilson Cook report:
(a) observes SKM’s responses fails to address Wilson Cook’s concern that the new asset point and the 2007 point have been calculated dissimilarly due to shifts that have taken place in replacement costs between the years used and therefore the gradient of the cost-age curve is incorrect; and
(b) states it is obvious that escalation of SKM’s 2002 costs to 2007 prices would lift that point and alter the gradient in its analysis.
239 Wilson Cook is of the opinion that SKM’s statement to the effect that the average annual percentage increase in costs applies to the network as a whole and, if the average age is maintained, will show no net increase in average maintenance costs, is not true. The second Wilson Cook report states:
Even if the average age were to be maintained constant, requiring a level of asset replacement matched to the ageing rate, old assets are still being replaced with new assets of modern technology with lower maintenance costs. Over time, the average maintenance cost will thus decrease as the asset population attains ever-higher proportions of modern assets.
In essence, the SKM model does not properly separate ageing effects and technology effects and is therefore not robust.
Wilson Cook’s response to the Huegin report
240 Some of the issues raised in the Huegin report (eg support in technical literature for the application of exponential maintenance cost growth), were also raised by SKM and, insofar as Wilson Cook response to the SKM report deals with them, they are not revisited.
241 Responding to points made by Huegin, Wilson Cook expresses a view that:
(a) in making those points Huegin appears to have relied on experience and data in the aviation industry; and
(b) that industry differs sufficiently from the electricity distribution industry such as to make comparisons of maintenance cost behaviour erroneous.
242 Drawing on Huegin’s views, Wilson Cook suggests that Huegin agrees with one of Wilson Cook’s principal points, namely, that an exponential relationship over the full life of assets as applied by EA is not necessarily correct.
243 As noted above:
(a) Huegin’s criticism of Wilson Cook’s use of the CSV in its first report prompted Wilson Cook to re-examine its use and to undertake a fresh CSV analysis; and
(b) the fresh analysis, relying solely on Australian data for a group of predominantly urban utilities, addresses particular concerns in the Huegin report about the sample size used by Wilson Cook and its regression analysis.
244 Wilson Cook’s response to the Huegin report also notes that Huegin does not address specific issues raised in the first Wilson Cook report, in particular, the incompatibility of using 2002 costs for maintaining new assets and 2007 costs for maintaining older ones.
Wilson Cook’s use of a mid-point
245 Both SKM and Huegin criticised Wilson Cook’s use of a mid-point to arrive at its proposed adjustment of EA’s forecast maintenance costs. Noting that it is common to accept a mid-point (or some other point) between the upper and lower bounds of calculation when there is reason to believe that neither bound is suitable for use without adjustment and where there is no better basis for determination, Wilson Cook rejects the criticism that its use of a mid-point between the two available estimates (EA’s and its) was wrong.
The Tribunal’s consideration of EA’s maintenance costs
246 EA submitted that the AER made three errors in rejecting EA’s forecast maintenance costs:
(a) the AER wrongly rejected EA’s assumption of an exponential relationship between maintenance costs and the average age of EA’s assets;
(b) even if the rejection were well founded, the ‘mid-point’ adjustment made by the AER was not consistent with the Transitional Rules; and
(c) consistently with the position taken by EA with respect to the AER’s reliance on the second Wilson Cook report on the step change issue, the AER denied EA a reasonable opportunity to make submissions in respect of matters raised in the report.
247 In light of the material in the foregoing paragraphs, in particular those summarising Wilson Cook’s response to the SKM and Huegin reports, the Tribunal is satisfied that the AER was correct in not accepting EA’s forecast of maintenance costs based on an exponential relationship between maintenance costs and the average age of EA’s assets. Again, the Tribunal is satisfied that:
(a) the AER’s expert was properly briefed;
(b) has ‘squarely engaged’ with each of EA’s experts; and
(c) to the extent that EA’s experts may have impugned the first Wilson Cook report, Wilson Cook addressed those expert’s concern in its second report and, in particular, any concern that Wilson Cook may not have understood EA’s model.
248 The Tribunal is also satisfied that the selection of a mid-point in the two assessments of EA’s maintenance costs is consistent with cl 6.12.2 and 6.12.3(f). In response to EA’s submission that the mid-point is ‘arbitrary’, the AER submits, correctly, that
(a) it is more accurate to describe the mid-point as an ‘approximation’ which is common to the outcome of all models; and
(b) it is a reasonable approximation because it draws on the outcomes of both models to achieve a reasoned outcome.
249 EA submits that the AER is not permitted to reject EA’s entire methodological approach and adopt some other approach. That is, the AER is only permitted to amend EA’s methodology, not depart from it. To do otherwise, it submitted, is not to approve an amount, value or methodology based on EA’s regulatory proposal amended only to the extent necessary to enable it to be approved in accordance with the Transitional Rules as required by cl 6.12.3(f)(2).
250 In this circumstance, the approach taken by the AER was in accordance with the Transitional Rules.
251 The primary discretion given to the AER by cl 6.12.3(a) is to refuse to accept or approve any element of a regulatory proposal. The AER’s power to substitute an amount or value or methodology exists so that it may properly perform its obligation under cl 6.12.1(4)(ii) to set an estimate of the total opex that the AER is satisfied reasonably reflects the opex criteria.
252 Once the basis of EA’s approach to the assessment of maintenance costs is rejected as above, then the approach undertaken by the AER is an appropriate way to proceed. No other ‘amendment’ to the ‘extent necessary’ to be approved in accordance with the Transitional Rules is appropriate or possible in keeping with the primary purpose of the Transitional Rules.
253 Finally, nothing in the second Wilson Cook report on the issue of EA’s maintenance costs forecast should have come as a surprise to EA. The second Wilson Cook report covers the same ground as its first report and, as may be seen by the references above, does not stray beyond providing appropriate responses to issues raised in the SKM and Huegin reports relied on by EA in preparing its revised regulatory proposal. The NEL and the Rules mandate a sequence of, and timetable for, a DNSP’s regulatory proposal, the AER’s draft decision, the DNSP’s revised regulatory proposal and the AER’s final determination. To avoid gaming of the sequence, the NEL and the Rules are quite detailed about what is to occur in each sequence, when it is to occur and about the rights and obligations of a DNSP and the AER. It is apparent from the foregoing paragraphs that EA was given a reasonable opportunity to make submissions and did in fact make submissions. A line must be drawn by the AER in its engagement with a DNSP, else it fails to meet the deadlines imposed on it. Certainly, nothing in the NEL or the Rules obliged the AER in this matter to, in effect, give EA a ‘second bite of the cherry’.
The Tribunal’s conclusion on EA’s maintenance costs
254 In light of the above reasons, the Tribunal will affirm the AER’s decision to reduce EA’s forecast maintenance costs by $22.4 million.
PASS ThROUGH
255 As part of EA’s Final Determination, the AER made a constituent decision on the additional pass through events that are to apply for the regulatory control period under the ambit of cl 6.12.1(14) of the Transitional Rules (the pass through decision). In it, the AER rejected a number of EA’s proposed specific pass through events, and nominated a ‘general nominated pass through event’ which it defined at 15.6.2 of the Final Decision.
256 EA and the AER now contend in common that the pass through decision was affected by error and that the Tribunal should vary EA’s Final Determination pursuant to s 71P(2)(a) of the NEL in the form set out in proposed orders. Specifically, amendments are proposed to the general nominated pass through event:
(a) to remove the qualification that the event ‘falls outside of the normal operations of the business’;
(b) to remove the qualification attached to the materiality threshold; and
(c) to change the time at which foreseeability is determined.
257 If those variations are made, in the circumstances of the present application, EA does not press for the pass through decision to be further varied to include its specific pass through events.
258 The Tribunal cannot make orders by consent. However, upon the basis the material before the Tribunal, including any concessions or admissions of the AER, the Tribunal may be satisfied as to the existence of reviewable error, and may be satisfied as to the appropriate relief to grant.
259 In relation to the pass through decision, EA and AER submitted the following common submissions (omitting footnotes):
18 There are three errors in the AER’s definition of the “general nominated pass through event” at 15.6.2 of the final decision.
19 First, as accepted by the AER in its submissions at paragraph [19], an error lay in confining the general pass through events to those that fall outside of the normal operations of the business. Thus framed, the general nominated pass through event:
(a) was uncertain in its operation; and
(b) excluded some events which ought to qualify for pass through.
20 Secondly, the formulation of the materiality threshold to require the change in costs to be not only material but also “likely to significantly affect the DNSP’s ability to achieve the operating expenditure objectives and/or the capital expenditure objectives…” introduced a rigidity into the pass through event that EnergyAustralia contends is contrary to the apparent intention behind the general nominated pass through event. The AER has confirmed that those words are not indeed to impose a second or higher threshold to the materiality requirement provided for in the Transitional Rules. Accordingly, those words are unnecessary and should be deleted as proposed.
21 Thirdly, the AER erred in deciding that the question whether an event is relevantly “unforeseeable” should be determined as at the time that the AER makes its distribution determination. This would produce the anomaly that an event that was unforeseeable, and could not be included in the regulatory proposals, would nevertheless be excluded from the pass through process if it became foreseeable (or eventuated) by the time of the AER’s distribution determination. Accordingly the parties propose that, to correct this anomaly, foreseeability should be determined as at the time when EnergyAustralia lodged its regulatory proposal.
22 The AER’s definition of the “general nominated pass through event” was therefore affected by reviewable errors answering the description of an incorrect exercise of discretion, or an unreasonable decision, and the Tribunal should be so satisfied.
260 Having read the submissions by the parties on the pass through decision, including the common submission of EA and AER, the Tribunal is satisfied that EA has demonstrated reviewable error, and considers it appropriate to vary the pass through decision in the manner proposed by EA and AER (see s 71P(2)(a) and (3) of the NEL).
261 Therefore, subject to any further submissions as to the form of the determination, the Tribunal proposes to make a determination in relation to the pass through decision in accordance with the form of order that the parties provided to the Tribunal on 20 August 2009.
PUBLIC LIGHTING
262 The Tribunal has already indicated its approach to the issues of public lighting: see Application by EnergyAustralia (2009) ACompT 7. The Tribunal will need to consider the appropriate determination to make in respect of public lighting.
TRANSGRID SPECIFIC MATTER: DEFECT MAINTENANCE
263 TransGrid seeks a review of the AER’s final decision to reduce its forecast opex by $13.5 million, a figure arrived at by reducing its forecast defect maintenance for new growth assets by $15 million and then allowing $1.5 million for non-recoverable and non-routine costs.
264 If found in favour of TransGrid, the first and foremost issue agitated by the parties under this heading (whether, in circumstances where the average age of TransGrid’s asset base will remain relatively stable over the next regulatory period (2009-2014), its acquisition of new growth assets skew TransGrid’s forecast opex for defect maintenance) is determinative of the question of error. TransGrid submitted that other issues addressed by the parties (such as the extent of defects in new growth assets, the scope of warranty cover for them and the reasonableness of the AER’s decision based on advice from an expert which raised issues requiring further investigation) would not need to be decided. While the Tribunal will need to return to that submission, the focus in the following paragraphs, which chronicle how the AER arrived at its decision, is on the first and foremost issue.
The meaning of ‘defect maintenance’
265 TransGrid defines ‘routine maintenance’ as scheduled inspection and preventive maintenance, and ‘defect maintenance’ as maintenance addressing out-of-specification conditions that may affect the performance or reliability of the transmission network. This definition of defect maintenance captures all costs of a non-routine nature, some of which are required on a regular basis based upon condition analysis.
The meaning of ‘new growth asset’
266 The term ‘new growth asset’ is short hand for an asset added to TransGrid’s system which results in the growth of the system, as distinct from an new asset that merely replaces an existing asset. Expenditure on new assets in the next regulatory period is estimated to be approximately $2.5 billion, of which $1.58 billion will be for new growth assets and the balance on replacement assets.
TransGrid’s May 2008 regulatory proposal
267 TransGrid’s opex model adopts 2006-2007 (adjusted for scope changes that will not be reflected in the forecast years) as the base year for forecasting its opex. The key inputs to the model are:
(a) routine maintenance forecasts;
(b) defect maintenance ratios;
(c) major operating projects forecasts;
(d) labour cost escalators;
(e) asset growth factors;
(f) economy of scale factors; and
(g) base year costs and adjustments.
268 TransGrid’s May 2008 regulatory proposal arrived at its opex forecasts by:
(a) building up from a ‘zero base’ a routine maintenance forecast for its existing asset base, thus allowing the forecasts to reflect cyclical requirements and to be adjusted for changes in scope when an asset is replaced with new equipment that requires less maintenance; and
(b) basing its defect maintenance forecast on an expected ratio of defect maintenance hours to routine maintenance hours as an average across all assets in particular asset classes.
269 More particularly, the defect ratios are based on historic performance of asset types as appears from the following Table which appeared in the parties’ Agreed Statement of Facts:
| Work stream | 2005-05 | 2005-06 | 2006-07 | Opex forecast |
| Lines | 109% | 72% | 93% | 95% |
| Substations | 34% | 114% | 116% | 115% |
| Communications | 231% | 232% | 196% | 200% |
| Secondary Systems | 49% | 28% | 29% | 30% |
| Land and Easements | 732% | 82% | 48% | 46% |
Opex associated with new capex is calculated by multiplying the controllable opex associated with each category by the capex growth for each category (as a percentage) and an economy of scale factor (typically 95%).
270 The agreed Statement of Facts (omitting footnotes) states:
Under TransGrid’s opex model an increase in the asset base increases the operating expenditures required to service the increased asset base. An asset growth factor is therefore applied to the routine maintenance units for each maintenance category in each year. The asset growth factor is the amount of capital expenditure on new growth divided by the existing asset base. Therefore, if the asset base is $1 billion and there are new assets in the relevant year of $10 million, then the asset growth factor for that year will be 1%.
271 Applying its model, TransGrid’s May 2008 regulatory proposal forecast defect maintenance of $152 million ($2007-2008). The AER’s draft decision noted that this represents an increase of approximately 28% compared to TransGrid’s defect maintenance expenditure in the current regulatory control period.
The first PB report
272 To assist it in its consideration of TransGrid’s regulatory proposal, the AER sought expert advice from PB on, amongst other things, the appropriateness of the methodology used by TransGrid to forecast its opex.
273 Observing that TransGrid’s opex model assumes that:
(a) the amount of additional operating expenditures is directly related to the increase in new assets under management; and
(b) the business is operating under a ‘business as usual’ scenario,
the November 2008 PB report (the first PB report) contends that significantly larger forecast growth related capital works programs than those in place up until 2006-2007 will impact on the reasonableness of the opex forecasts the model produces.
TransGrid’s response to a draft of the first PB report
274 Provided with an opportunity to comment on a draft of the first PB report, TransGrid’s September 2008 response commenced by explaining that:
(a) its model used defect ratios based on the historic performance of the asset base;
(b) the asset base consists of a range of old and new assets; and
(c) the defect ratio is based on this range in the age of the assets.
275 Building on that explanation, TransGrid submitted that while it is likely that a proportion of new assets will require little in the way of defect maintenance, other new assets will require significant effort to address early life issues. In addition, there will be an increasing number of assets moving from the random failure part of the ‘bathtub curve’ (illustrated below) into an area of increasing probability of failure, therefore increasing maintenance costs.
276 TransGrid’s response went on to submit that its ‘business as usual scenario’, which PB challenged, was supported by TransGrid’s modelling of the impact of the significant increase in new growth assets when included in a network of already aging assets. This modelling found that the average system age, and the average age for most asset classes, remains reasonably stable through the next regulatory period.
277 TransGrid’s response also asserts that as the average age of the significant asset classes is not decreasing substantially over time, the average defect ratio for the range of assets will not change substantially. It notes that:
(a) for a system such as its, with an average age in the centre of the ‘Random Failure Zone’ of the ‘bathtub’ curve, a significant shift in the average age is required to change the failure rate (defect cost); and
(b) while a change in technology may result in changes to the defect ratio overall, expects this to be evolutionary rather than revolutionary and the opex effects will not result in a step change.
The AER’s October 2008 draft decision
278 The AER’s October 2008 draft decision accepts that:
(a) the average age of most of TransGrid’s asset classes may not decline over time; and
(b) new assets may require some defect maintenance expenditure and notes that:
(c) it is condition rather than age that drives defect maintenance; and
(d) it did not consider the required defect maintenance expenditure to be significant.
279 Noting that TransGrid developed its defect ratios based on the historical performance of its assets, the draft decision observes that, as a result of a 200% increase in capex, TransGrid’s asset base will change considerably during the next regulatory period.
280 Agreeing with an adjustment proposed by PB and removing the defect maintenance costs for new growth assets, the AER draft decision states:
(a) this will result in the efficient costs that a prudent operator in the circumstances of TransGrid would require to achieve the opex objectives, as required by cl 6A.6.6(c) of the Rules; and
(b) TransGrid advised that the adjustment results in a reduction of $15 million ($2007-2008) to its forecast.
The SKM report
281 To assist it in preparing its revised regulatory proposal, TransGrid requested SKM to provide an assessment of the first PB report, in particular, the asset growth escalation component of PB’s review. After considering TransGrid’s opex model, its response to a draft of the first PB report and the report itself, the SKM report concludes that:
(a) TransGrid has been prudent and efficient in its modelling; and
(b) the adjustment to its opex forecast recommended by PB is not warranted.
282 Based on pertinent detail in its report, SKM’s conclusion notes, amongst other things:
TransGrid’s modelling is based on historic defect rates (which include defects on new asset) and the proposed capital investment programme does not materially alter the average age of any asset type. Consequently, the defect rate across the network could be expected to remain at the same level as that experienced historically.
TransGrid’s January 2009 revised regulatory proposal
283 TransGrid’s January 2009 revised regulatory proposal submits that there is no reasonable basis for the AER to reduce its forecast opex for defect maintenance of new growth assets because, amongst other things, its proposed acquisitions of new assets will not result in a significant change to the average age of its asset base and therefore defect rates will not be affected by the acquisitions.
284 A graph provided in support of TransGrid’s submission shows that average age of TransGrid’s assets is reasonably stable over the next regulatory period. This, the revised proposal says, leads to a conclusion that there would be no expectation that defect rates would be impacted by the acquisition of any new assets.
285 Challenging the AER’s assertion that new assets will not incur significant defect maintenance expenditure as un-supported by evidence, TransGrid’s revised regulatory proposal provides a graph of the ratio of defect v routine maintenance expenditure against the commissioning date of its assets. This graph shows that defect costs for newer assets are significantly higher across all TransGrid’s asset categories.
286 Responding to the draft decision’s observation that the 200% increase in capex will change its asset base considerably, the revised regulatory proposal submits that:
(a) based on the evidence TransGrid has provided, the increase in new assets would be expected to lead to an increase in defect costs; and
(b) there is no basis for concluding that there should be no allowance for defect maintenance of these new assets.
287 TransGrid’s revised regulatory proposal on the defect maintenance issue concludes by:
(a) referencing and attaching a copy of the SKM report; and
(b) submitting that the evidence strongly supports its position that new assets will require an amount of defect maintenance at least equivalent to the requirement for mid-life assets and that it had included such an amount in its revised opex forecasts.
The second PB report
288 To assist it in assessing the issues raised in TransGrid’s revised regulatory proposal, the AER again sought the assistance of PB. PB provide a report (the second PB report) which generally confirms its earlier recommendation but does allow a moderate amount to compensate for:
(a) the non-routine but regular maintenance included in TransGrid’s definition of defect maintenance; and
(b) some costs associated with TransGrid organising and managing works under warranty.
289 In the absence of specific data to support the magnitude of the amount, the second PB report recommends $300,000 per annum as a reasonable estimation of the amount that the AER should allow.
290 While the second PB report accepts at a macro level the conclusion in TransGrid’s revised regulatory proposal that defect rates would not be impacted by new assets, the report:
(a) notes, amongst other things, that over the period 2009 to 2012 the average age of the critical asset class of substations is reduced slightly; and
(b) states that it believes that the impact on specific asset classes is significant.
291 In support of this statement, the second PB report refers to a graph of TransGrid’s average maintenance cost per switchbay over a two year period against various asset commissioning dates which, it says, shows that the average maintenance costs for newly commissioned switch-bays are lower than those for switch-bays commissioned during previous regulatory periods.
292 The second PB report concludes its review of TransGrid’s revised regulatory proposal by observing that as the TransGrid opex model uses system averages to forecast opex, it will tend to overstate the defect rectification expenditures required for newly commissioned assets.
The AER’s April 2009 Final Decision
293 It is apparent from the AER’s Final Decision that the AER:
(a) reviewed TransGrid’s revised regulatory proposal, the SKM report and TransGrid’s answers to the AER’s questions along with the second PB report; and
(b) responded to each of TransGrid’s submissions as outlined above.
294 The foundation for the AER’s decision to reduce TransGrid’s forecast opex for new growth assets is a belief, based on PB’s advice, that the significant increase (200%) in TransGrid’s capex (in particular $1.58 billion for new growth assets), compared with TransGrid’s opex model’s base year, skews the model’s asset growth input and impacts on the reasonableness of the opex forecasts. That is, based on PB’s view that TransGrid’s model assumes a ‘business as usual’ scenario, it is inappropriate to include new growth assets as an input to the model - this view being modified to allow the ‘reasonable and moderate’ amount of $300,000.
The Tribunal’s consideration of TransGrid’s defect maintenance
295 It is common ground that:
(a) over the period 2005-2009, the average age of TransGrid’s assets was increasing – its capex program in this period was insufficient to maintain a constant system age;
(b) TransGrid has a growing and maturing asset base;
(c) during the next regulatory period, end of life issues will be more significant than has been the case in the past;
(d) the average system age and the average age for most asset classes will remain reasonably stable throughout the next regulatory control period;
(e) in the absence of a capex replacement program, the average age of the assets will progressively increase; and
(f) if there were no expenditure on new assets, over the next regulatory period an increasing percentage of assets would move from the random failure zone of the ‘bathtub curve’ into the wear out zone and average defect maintenance costs would be expected to rise.
296 In reaching its decision, the AER focused on new growth assets and ignored what may occur to TransGrid’s system overall. The new growth assets will merely slow the rate of the system overall, not render it younger than the base period used to calculate TransGrid’s defect ratios. The AER’s approach that more new assets with lower defect rates will necessarily lead to lower defect rates in the next regulatory period ignores what is common ground, namely, that end of life issues will be more significant than has been the case in the past.
297 TransGrid’s defect rates are defect rates averaged across its system as a whole. While in the next regulatory period there will be more new assets, because the system as a whole continues to age, there will also be more old assets. Based on system age alone, the defect ratios in that period will be even higher than in the base year and TransGrid’s assumption of a constant defect ratio may be conservative.
298 As to the issue of what PB say is a key category of substations showing an age trend different from that contended by TransGrid, although the average age of substations may reduce a little in 2010, having regard to the base year, they do not really change much and after 2010 their age commences to increase again. Thus, compared to the base year and the measurement period of 2005 through to 2007 for the defect ratios, no assumption favourable to the AER may be drawn from the substation average life.
299 If TransGrid were replacing aging assets with new assets and adding new growth assets, looked at it statically the average age would reduce. Average age should, however, be examined dynamically over the next regulatory period. The end of the regulatory period (2014) is years on from the base period (2005). While the addition of new growth assets may have a temporary impact in 2009/2010, the system will continues to age because the new assets will age.
300 As described by counsel for TransGrid, PB’s view of ‘business as usual’ is ‘inapt’. The base period was a period in which new assets were insufficient to replace ageing assets and was a period during which the system was ageing. In the next regulatory control period, while there will be less of an ageing system, there will be a system that will age slightly overall. This means that the average defect ratios, calculated over the base period, may be appropriately applied to the system overall for the next regulatory period. The assumption made by PB and the AER in separating out new growth assets, that they would somehow reduce the average system wide rate of defects, is wrong. Applying ordinary mathematical concepts, it is reasonable for TransGrid to apply an average defect ratio to the system that will emerge in the next regulatory period.
301 The AER was wrong to:
(a) exclude defect maintenance in respect of new growth assets;
(b) proceed on a basis that TransGrid would incur zero defect expenditure in respect of new growth assets; and
(c) assume that the existing pool of ageing assets, that is, assets other than the new growth assets, would have the same level of defects as in the base period.
302 To put it another way, by disallowing the majority of defect maintenance opex for new growth assets, the AER assumed a constant set of defects for the remaining assets.
303 In accepting PB’s advice that as TransGrid’s opex model uses system average ages to forecast opex, it tends to overstate the defect rectification expenditures required for new assets, the AER is saying it is wrong to apply an average rate of defects to new assets, because that overstates the defect level of new assets. But that is not what TransGrid is doing. TransGrid is not applying average defect rates to new assets. TransGrid is applying average defect rates to the whole system and it therefore applies them to the old assets, (which have higher levels of defects), the mid-life assets, (which have medium levels of defects) and the new assets, (which have a level, albeit contested, of defects). By removing one element to which the averaging is applied, the AER destroys the integrity of the averaging system as a whole.
304 If the impact of the new growth assets were to bring down TransGrid’s system age, the AER’s approach may be appropriate. But that will not be the impact of the new growth assets.
305 For the above reasons, TransGrid has established that either the AER exercised its discretion incorrectly, or its decision was unreasonable in all the circumstances.
306 As observed, TransGrid submitted that if the Tribunal were to reach a conclusion that the AER was in error on this first issue agitated by the parties, it need not decide the other issues addressed by the parties. TransGrid did, however, also concede that whether new assets had higher or lower rates of defect maintenance costs than the average was a relevant issue, albeit far less important than system average age.
307 That is because, even if average system age stays constant, the question remains whether recently commissioned assets tend to have the same defect maintenance costs as those commissioned some years ago.
308 The AER argued before the Tribunal that if system average age remains constant, defect rates are likely to fall because newer technology will reduce defect ratios of assets commissioned in the 2009-14 regulatory period compared with defect ratios of new assets commissioned in previous regulatory periods.
309 The AER did deal with this in its Final Decision. However, based on PB’s reports, its concern remained that new growth assets have little or no defect maintenance costs – which, as explained above, is not the relevant consideration in excluding them from the averaging system. Consequently, the AER was diverted into consideration of whether any, and if so what, costs are covered by warranty. In dismissing TransGrid’s material suggesting the opposite of PB’s expectations, the AER accepted PB’s position that in the absence of further investigation (for which there was no time), it was not prepared to alter its view. In short, whether defect maintenance costs for newly-commissioned assets are falling over time is unresolved. The relevance of such a trend, if it exists, is its effect on overall average defect maintenance costs. Over the relatively short span between the base period and the current regulatory period, the effect seems likely to be small.
The Tribunal’s conclusion on TransGrid’s defect maintenance costs
310 In light of the above reasons, the Tribunal will set aside the AER’s decision in relation to reducing TransGrid’s forecast opex and remit the matter back to the AER to make the decision again. Unless it has material to quantify any likely decrease in average defect maintenance costs due to growth assets, the AER should make the decision on the basis that TransGrid’s forecast opex is calculated using its opex model with asset growth factors.
GENERAL OBSERVATIONS
311 In the course of submissions, a number of legal issues were raised that have not been necessary to consider in detail or to finally determine.
312 Nevertheless, as the Tribunal has had the benefit of extensive legal analysis upon some issues, it may be useful to state, in summary form, the Tribunal’s position:
(a) As to the construction of cl 6.5.2(c)(2)(i) and (ii) of the Transitional Rules (and the equivalent Rules), the Tribunal considers that once a period has been agreed, there seems to be no mechanism by which that period of time can be altered.
Where a period is specified by the AER, then the AER could alter this decision and specify another period of time.
However, in the case of a specified period, only if the NSP waives the requirement for notification within a reasonable time prior to the commencement of the period, can the specified period be a past period. Otherwise, the period of time specified would need to be a future period of time.
(b) The Tribunal, once an error is found to exist within the meaning of s 71C of the NEL, has the same powers of the AER. In essence, the Tribunal could agree with the proposals of the Applicants (original or otherwise), once the specified period is set aside. This is what the Tribunal has done in this review, agreeing to the revised proposals of the Applicants. Alternatively, the Tribunal could specify a period of time, constrained in the same way as would be the AER.
In making its own decision, the Tribunal is bound to consider the submissions of the parties, and may (not must) allow new information or material to be submitted if the new information or material is of the type described in s 71R(3)(a) and (b).
(c) If an error is found to exist in a constituent decision of the AER, then it may be necessary to consider how it impacts upon the ‘reviewable regulatory decision’ (as defined in the NEL) to the extent that the constituent decision itself is not a reviewable regulatory decision. Therefore, as an example, it is possible that an error of fact may be made in a constituent decision which is not material to the reviewable regulatory decision (as defined), or an error is made in a constituent decision which is unreasonable, but does not make the AER’s reviewable regulatory decision (as defined) itself unreasonable. In the case of an incorrect exercise of the AER’s discretion, an error occurring in a constituent decision may or may not impact upon the reviewable regulatory decision. If it does not impact upon the reviewable regulatory decision in such a situation, it may be that a ground of review is established, but the Tribunal in exercising its powers under s 71P(2) may still affirm the reviewable regulatory decision.
(d) The grounds under s 71C do not directly embrace ‘procedural’ errors, such as failure to comply with s 16 of the NEL, any common law requirement of procedural fairness, or failure to consider submissions or give reasons as specifically required by the Rules. Section 71C(1)(d) specifically focuses on the unreasonableness of the AER’s decision itself. Section 71C(1)(c) focuses on the exercise of the AER’s discretion being ‘incorrect’. Section 71C(1)(a) and (b) are directed to errors of fact in the findings of fact. However, if for instance, the AER does not take into account a submission of a provider, then that may result in the exercise of the AER’s discretion being incorrect, as not taking a relevant matter for consideration, either the submission itself or a matter raised in the submission that was not taken into account by the AER. Depending on the circumstances, a ground of review may be established (such as failing to take into account the submission), but the Tribunal may still affirm the AER’s decision (such as where the matter raised in the submission was independently and properly considered by the AER in any event).
(e) The Tribunal has not relied upon any contested non ‘review related matter’ (see s 71R(1) and (6)) in order to arrive at its determination in this case. However, the Tribunal has considered the contested non ‘review related matter’ for the purposes of considering whether it would impact upon its determination. Even if the non ‘review related matter’ was to be considered by the Tribunal for the purpose of its determination, the Tribunal is satisfied that none of the non ‘review related matter’ would impact upon its conclusions and determination.
Difficult questions may arise as to the scope of ‘review related matter’ where the AER specifically uses or refers to only part of a report or material in making a reviewable regulatory decision. Does that mean that other parts of those reports or material not specifically used or referred to are ‘relied upon’ by the AER in making the ‘reviewable regulatory decision’ within the meaning of s 71R(6)(e)?
No one answer can be given to this question. If the reference made to part of a report or material is dependent, for instance, on another part, either to give meaning or explain the actual reference made by the AER, then presumably the AER has ‘relied upon’ the rest of the report or material to the extent necessary to giving meaning to or explain the actual reference made by the AER.
However, this does not mean, that in an appropriate case, the Tribunal may not need to consider new evidence, if an issue arises as to the scope of the ‘review related matter’. For instance, a factual issue may arise as to whether a report was relied upon by the AER in making the reviewable regulatory decision. Section 71R(1) of the NEL would not prevent the Tribunal from performing its statutory function and employing its own general procedural powers to make a factual finding on this question to determine whether, in undertaking the actual review of the reviewable regulatory decision, the report was able to be considered by the Tribunal.
(f) Looking at the purpose and scope of s 71O(2) and s 71R(1) and (6) of the NEL, the matters to which the Tribunal can have recourse include the subject matters raised, the issues raised and the materials relied upon in support of the position or proposal put forward by the Applicants’ to the AER prior to the final determinations, as being relevant to those determinations. It is only if a matter, whether by way of argument or evidentiary material, cannot be identified as broadly arising out of a matter fairly raised before the determinations under review were made, that it will not be permitted to be raised in the review: see in Application by Epic Energy South Australia Pty Ltd (2003) ATPR 41-932 at [24] and also Application by East Australian Pipeline Ltd (2005) ATPR 42-047 at [9].
(g) Many submissions were received upon the nature and characterisation of the correspondence between the Applicants and the AER on the proposed and revised averaging periods. There was no doubt that agreement by the AER to the original proposed averaging period was withheld, and upon this basis the Tribunal has reached its conclusion. The Tribunal does not need to consider the nature and characterisation of the correspondence, for having come to the conclusion that an error of review under s 71C has been made by the AER in its decision to withhold agreement to the original proposals, the Tribunal is empowered then to perform all the functions and exercise all the powers of the AER under the NEL or the Rules (s 71P(3)). On any view, the Tribunal may now agree to one of the proposed averaging periods put forward by the Applicants. As to which proposal, the Tribunal needs to evaluate the position as at the time of its determination based upon the review related matter, or any new information or material it allows to be submitted pursuant to s 71R(3).
(h) For the purposes of this determination only, the Tribunal has proceeded on the basis that there is no material difference between any versions of the Transitional Rules or the Rules to the extent that any alterations have been made since any of the proposals of the Applicants were submitted to the AER. This was the position accepted by the parties.
(i) The Tribunal is mindful that in considering the reasons of the AER, it is important to recall that the reasons are there to inform and are not to be scrutinised in a over-zealous way seeking to discern error from the way in which the reasons are expressed: see eg Minister for Immigration and Ethnic Affairs v Wu Shan Liang (1996) 185 CLR 259, 272. It is all a matter of degree, as an obvious inadequacy of the reasons may amount in any given case to an error in the process of fact finding or the exercise of a discretion. The Tribunal refers to its comments in the Application by EnergyAustralia (2009) ACompT 7 at [16].
CONCLUSION
313 The Tribunal directs that the parties confer and submit appropriate minutes of determination in accordance with these reasons.
314 The Tribunal again takes this opportunity to express its gratitude to the legal representatives of the parties and those instructing them for their assistance to the Tribunal in this complex review.
| I certify that the preceding three hundred and fourteen (314) numbered paragraphs are a true copy of the Reason for Determination herein of the Honourable Justice Middleton (Deputy President), Mr R Davey and Mr R Shogren. |
Associate:
Dated: 12 November 2009
| Counsel for the Australian Energy Regulator: | Mr P Hanks QC with Mr P Gray, Mr P Wallis and Dr V Prisich |
|
|
|
| Solicitor for the Australian Energy Regulator: | Corrs Chambers Westgarth |
|
|
|
| Counsel for EnergyAustralia: | Mr JT Gleeson SC with Mr P Brereton SC |
|
|
|
| Solicitor for EnergyAustralia: | Gilberg + Tobin |
|
|
|
| Counsel for Southern Sydney Regional Organisation of Councils: | Mr FM Douglas QC with Mr WAD Edwards |
|
|
|
| Solicitor for Southern Sydney Regional Organisation of Councils: | HWL Ebsworth |
|
|
|
| Counsel for Transgrid: | Mr A Meagher SC with Mr C Moore |
|
|
|
| Solicitor for Transgrid: | Gilbert + Tobin |
|
|
|
| Counsel for Intergral Energy: | Mr R Dick SC |
|
|
|
| Solicitor for Intergral Energy: | Gilbert + Tobin |
|
|
|
| Counsel for Transend: | Mr A Payne SC with Mr JA Arnott |
|
|
|
| Solicitor for Transend: | Gilbert + Tobin Lawyers |
|
|
|
| Solicitor for Nyrstar Australia: | O’Donnell Salzano Lawyers |
|
|
|
| Counsel for Country Energy: | Mr JRJ Lockhart SC |
|
|
|
| Solicitor for Country Energy: | Gilbert + Tobin |
|
|
|
| Date of Hearing: | 10, 11, 12, 13, 14, 17, 18, 19, 20, 21 August 2009 |
|
|
|
| Date of Reasons for Determination: | 12 November 2009 |