annual energy cost review€¦ · figure 11: country energy efficient frontiers for 2009/10...

© Frontier Economics Pty Ltd., Melbourne

Annual energy cost review FINAL REPORT

PREPARED FOR THE INDEPENDENT PRICING AND REGULATORY TRIBUNAL

February 2009

i Frontier Economics | February 2009 |

Contents Frontier Economics - Energy cost review - Final.doc

Annual energy cost review

1 Introduction ......................................................................................... 1

1.1 Background...................................................................................................1

1.2 IPART’s 2007 Final determination ...........................................................1

1.3 IPART’s 2008 review ..................................................................................2

1.4 This report ....................................................................................................2

2 Requirements of the annual reviews....................................................5

3 Context of this review...........................................................................6

3.1 Regulated load ..............................................................................................6

3.2 ETEF.............................................................................................................7

3.3 industry changes...........................................................................................9

3.4 Drought.........................................................................................................9

3.5 Greenhouse policies ................................................................................. 12

3.6 Implications for modelling approach..................................................... 12

4 Modelling ........................................................................................... 14

4.1 Modelling approach.................................................................................. 14

4.2 Modelling assumptions ............................................................................ 16

4.3 Modelling results ....................................................................................... 18

5 Summary and conclusions .................................................................26

ii Frontier Economics | February 2009 |

Tables & figures

Annual energy cost review

Figure 1: Regulated load as a share of NSW load........................................................7

Figure 2: Schedule for the roll off of the ETEF..........................................................8

Figure 3: Capacity reductions under NEMMCO drought modelling.................... 10

Figure 4: Energy reductions under NEMMCO drought modelling ...................... 11

Figure 5: Three-staged electricity modelling process ............................................... 15

Figure 6: NSW average price forecasts ($2006/07).................................................. 18

Figure 7: d-cypha base swap contract prices for 2009/10 (nominal) .................... 19

Figure 8: d-cypha peak swap contract prices for 2009/10 (nominal).................... 20

Figure 9: d-cypha base swap contract prices for Q1 (nominal).............................. 20

Figure 10: d-cypha peak swap contract prices for Q1 (nominal) ........................... 21

Figure 11: Country Energy efficient frontiers for 2009/10 ($2006/07) ............... 22

Figure 12: EnergyAustralia efficient frontiers for 2009/10 ($2006/07) ............... 22

Figure 13: Integral Energy efficient frontiers for 2009/10 ($2006/07) ................ 23

Figure 14: Summary of electricity purchase cost allowance cost for 2009/10 ($2006/07)............................................................................................................. 24

Table 1: NSW energy for 2009/10 (GWh)................................................................ 16

Table 2: Electricity purchase cost allowance cost for 2009/10 under ETEF at 100% scenario....................................................................................................... 24

Table 3: Electricity purchase cost allowance cost for 2009/10 under ETEF rolls off scenario............................................................................................................ 24

1 Frontier Economics | February 2009 |

Introduction

1 Introduction

1.1 BACKGROUND

Frontier Economics has been retained by the Independent Pricing and Regulatory Tribunal (IPART) to advise on the market-based electricity purchase cost allowance for standard retailers in NSW. This advice is provided as part of IPART’s annual review mechanism for the market-based electricity purchase cost allowance.

Frontier Economics has previously advised IPART in regard to both the original determination and the first annual review of the market-based electricity purchase cost allowance. Frontier Economics’ advice in relation to the original determination is set out in Frontier Economics’ March 2007 Final Report1 and Frontier Economics May 2007 report on the impact of the drought,2 both available from IPART’s web site. Frontier Economics’ advice in relation to the first annual review is set out in Frontier Economics’ March 2008 Final Report3 and Frontier Economics’ May 2008 Response to Submissions,4 both available from IPART’s web site.

This report sets Frontier Economics’ advice to IPART in regard to the second annual review of the market-based electricity purchase cost allowance.

1.2 IPART’S 2007 FINAL DETERMINATION

IPART’s 2007 Final Determination on regulated electricity tariffs for small customers, released in June 2007, recognised that the roll-off of the Electricity Tariff Equalisation Fund (ETEF) imposes additional risks on standard retailers supplying regulated customers.5 Standard retailers will have to increasingly rely on purchasing electricity at volatile, unregulated prices while being required to sell an unknown quantity of electricity to regulated customers at fixed regulated prices. In this environment, IPART is required to estimate the likely electricity purchase costs ahead of when the retailers have to buy this power and, on the basis of this purchase cost estimate, determine a retail price. This retail price will remain fixed for at least the following year. The risk to retailers is that their actual electricity

1 Frontier Economics, Energy costs, Public Report, March 2007.

2 Frontier Economics, Analysis of recent changes in NEM wholesale electricity prices, May 2007.

3 Frontier Economics, Annual energy cost review, Final Report, March 2008.

4 Frontier Economics, Annual energy cost review, Response to submissions from stakeholders, May 2008.

5 IPART, Promoting retail competition and investment in the NSW electricity industry, Regulated electricity retail tariffs and charges for small customers 2007 to 2010, Electricity - Final Report and Final Determination, June 2007.

http://www.ipart.nsw.gov.au/files/Electricity%20Retail%20Review%20-%20Final%20report%20and%20determination%20-%20Web%20version%20-%20June%202007.PDF


Introduction

purchase costs may exceed the level at which they are obliged to sell this same electricity to regulated customers.

In recognition of this changed environment, IPART has developed a series of related measures to manage the additional risks to which retailers are exposed as the ETEF rolls off. These measures include:

allowing for a higher retail margin to compensate the standard retailers for the additional risks they face as the ETEF rolls off;

basing the energy cost allowance for each retailer on the most conservative point on the efficient frontier (which reflects the high level of hedging that a risk averse retailer would prefer);

making an allowance for the cost to retailers of holding sufficient working capital to withstand cash flow variations;

providing for reviews of the market-based electricity purchase cost allowance to be conducted annually during the determination period. The annual reviews only consider the movement of the market-based electricity purchase cost allowance, based on the load data provided by the standard retailers for the purposes of the original determination. No other cost movements – for example, green costs or compensation for volatility – are considered in the annual reviews; and

providing a cost pass-through mechanism, which allows retailers to pass through to customers material increases or decreases in costs associated with regulatory or taxation change events that could not be anticipated at the time of the determination.

1.3 IPART’S 2008 REVIEW

In 2008, IPART undertook the first annual review of the market-based electricity purchase cost allowance.6 IPART’s 2008 Final Report concluded that there should be no change to the electricity cost purchase allowance for 2008/09 or 2009/10 from that set out in the 2007 Final Determination.

1.4 THIS REPORT

This report describes the modelling and analysis undertaken by Frontier Economics in modelling the market-based electricity purchase cost allowance for 2009/10 for the purposes of this second annual review.

6 IPART, Market-based electricity purchase cost allowance – 2008 review; Regulated electricity retail tariffs and

charges for small customers 2007 to 2010, Final Report, May 2008.

http://www.ipart.nsw.gov.au/files/Final%20Decision%20-%202008%20Review%20of%20Market%20Based%20electricity%20purchase%20cost%20allowance%20-%2020%20May%202008%20-%20website.PDF


Introduction

This report is structured as follows:

Section 2 sets out the basis for the annual reviews of the market-based electricity purchase cost allowance as set out in IPART’s 2007 Final Determination;

Section 3 sets out the context within which this annual review is being undertaken, and discusses the implications of this for the modelling approach;

Section 4 presents and briefly discusses the results of the updated modelling undertaken for the purposes of this second annual review of the electricity purchase cost allowance; and

Section 5 presents the conclusions of this second annual review of the electricity purchase cost allowance.


Requirements of the annual reviews

2 Requirements of the annual reviews

As discussed, as part of its 2007 Final Determination, IPART determined that it was appropriate to conduct annual reviews of the market-based electricity purchase cost allowance (although not the greenhouse and renewable energy cost allowance). IPART’s 2007 Final Determination set out the following decision rule:

“If the results of a review of the market-based electricity purchase cost allowance for a future year in the determination period show a change (positive or negative) of 10 per cent or more compared to the Tribunal’s most recent allowance amount for the market-based electricity purchase cost, then the Tribunal will notify the Standard Retailers and publish the new electricity purchase cost allowance amount.”7

While accommodating the uncertainties of the market through annual reviews, IPART has limited the scope of the annual reviews in the following way:

“In undertaking the reviews, the Tribunal will adopt the same approach it used in making the 2007 Determination. That is, the Tribunal will:

adopt a conservative approach to estimating the market-based electricity purchase cost, and focus on changes to the spot and contract prices for electricity

use the same the load profiles as it used in the 2007 Determination – it will not update the load profiles.

The reviews will not reconsider the long-run marginal cost of electricity generation, the volatility allowance, green energy costs, NEM fees, energy losses, retail operating costs, customer acquisition costs or the retail margin.”8

IPART has also committed to relying on expert advice in undertaking the annual reviews:

“The Tribunal concluded that, for the purpose of setting the market-based electricity purchase cost allowance, the use of expert advice is superior to relying on the publicly available information. Therefore, for each annual review of this allowance, the Tribunal will engage an expert to advise it on the appropriate future market based electricity purchase cost allowance”9

7 IPART, Promoting retail competition and investment in the NSW electricity industry, Regulated electricity retail

tariffs and charges for small customers 2007 to 2010, Electricity - Final Report and Final Determination, June 2007, p87.

8 ibid, p87.

9 ibid, p86.


Context of this review

3 Context of this review

This second annual review of the market-based electricity purchase cost allowance is conducted at a time of significant uncertainty in the wholesale electricity market. This uncertainty has an impact on participants’ behaviour in the wholesale electricity market and, as will become clear in this report, on outcomes in the wholesale electricity market.

This section provides an overview of some of the more significant factors that are currently affecting the wholesale electricity market, and sets out the implications of this for Frontier Economics’ modelling of the market-based electricity purchase cost allowance, and the way this uncertainty is accounted for in Frontier Economics’ modelling.

3.1 REGULATED LOAD

IPART cannot observe the actual energy purchase cost of the retailers, so IPART can never really know whether their Retail Price Determination errs on the side of generosity to the retailers, or otherwise. However, IPART can observe the effect of their Retail Price Determination. If the regulated price is too low, customers who have the choice of switching freely between market based and regulated tariffs will tend to choose to return to regulated tariffs, all other things being equal.

A useful aspect of ETEF is that it measures the quantity of regulated load each settlement period. This allows IPART to observe the size of the regulated market as a share of the total NSW market. Using the ETEF settlement data, Figure 1 shows the size of the regulated load in NSW as a share of total NSW load from the commencement of ETEF to the end of December 2008. Note, this data has not been corrected for temperature variations or other anomalous influences, so the results in individual years should be treated with caution. In this respect it is more appropriate to draw lessons from the data trends.

Figure 1 suggests at least two things:

the regulated retail price set by IPART over the years has not prevented churn away from regulated tariffs; and

in the most recent year of data (2008), the churn from regulated to market based arrangements has slowed and may have reversed (noting that the data has not been normalized for influences such as climate).

This last point suggests that there may be some evidence that the regulated price may be low relative the underlying market based costs.



Figure 1: Regulated load as a share of NSW load

3.2 ETEF

As discussed in previous reports by Frontier Economics, the ETEF is relevant to the market-based electricity purchase cost allowance. In particular, as the ETEF progressively rolls off, the standard retailers will, almost inevitably, hedge the risks associated with volatile spot prices by entering into financial hedging contracts. The roll off of the ETEF is therefore expected to have an effect on spot prices, and these spot price changes are expected to have an effect on the price and quantity of contracts. The impact of the ETEF on spot and contract prices is taken into account in Frontier Economics’ modelling.10

The ETEF was originally scheduled to begin rolling off from the fourth quarter of 2008. However, the commencement of the roll off of the ETEF has been delayed by Government until the second quarter of 2009.11 The original schedule

10 As discussed in Frontier Economics’ response to submissions for the first annual review, the impact

of the ETEF is taken into account in Frontier Economics’ modelling of spot prices, but the impact of the ETEF is not taken into account in Frontier Economics’ modelling of the efficient portfolio for the standard retailers.

11 Since the modelling work for this report was undertaken, the Government has announced that the commencement of the roll off of the ETEF will be delayed again, until 27 September 2009, at which point the original roll off schedule will resume.

0%

5%

10%

15%

20%

25%

30%

35%

40%

2001 2002 2003 2004 2005 2006 2007 YTD Dec 2008

Regulated load as % NSW Load (CY)

Regulated load as % NSW Load (FY)



for the roll off of the ETEF and the first revised schedule for the roll off of the ETEF are shown in Figure 2.

Figure 2: Schedule for the roll off of the ETEF

While the first revised schedule for the roll off of the ETEF does not change the extent to which standard retailers are covered by the ETEF in 2009/10, the fact that the schedule has been changed once creates uncertainties about future changes to the roll-off schedule. Since the ETEF effects the proportion of load hedged this can effect the relative levels of spot and contract prices. Frontier account for this in two ways.

First, as set out in Section 4, Frontier Economics have modelled the market-based electricity purchase cost allowance under two scenarios for the roll-off of ETEF: one scenario in which the ETEF rolls off according to the first revised schedule and a second scenario in which the ETEF does not roll off at all until after 2009/10.

Second, the uncertainty created by the revision of the roll off schedule for ETEF is taken into account, along with the other uncertainties discussed in this section, in Frontier’s modelling of the spot price. Frontier Economics’ methodology for accounting for this uncertainty is discussed in Section 3.6.

As Frontier Economics’ modelling results indicate, the schedule for the roll-off of the ETEF does not have much impact on forecast spot prices, suggesting that uncertainty about the roll-off schedule for ETEF is also unlikely to have a substantial impact on market outcomes. This suggests that uncertainty about the future of ETEF is unlikely to be a material influence on the determination of retailers’ energy purchase costs.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2

2007 2008 2009 2010

Original ETEF schedule

Revised ETEF schedule



3.3 INDUSTRY CHANGES

Beginning in late 2007 NSW Government embarked on a policy to sell their retailers and lease the generators.. As part of this process, consideration was given to providing the Government-owned generators and retailers with vesting contracts to ensure that a proportion of load was contracted prior to the sale or lease of the assets (as identified in the publicly available draft legislation). As a result of this process, and some uncertainty about whether these arrangements will feature in the NSW Government’s amended policy to sell the retailers and to lease out the generation trading rights, generators and retailers are uncertain about whether they will be provided with additional financial hedging contracts in the future. As a result, generators and retailers may consider that concluding financial hedging contracts may unduly increase their future risk position in the event that any vesting contracts imposed by the Government does not appropriately acknowledge their preferred risk position. The businesses will seek to manage this risk by moderating their current contract position both in terms of quantity and duration (i.e. contract for shorter periods).

3.4 DROUGHT

As discussed in previous reports by Frontier Economics, drought conditions have been seen to have an impact on contract prices. During the first half of 2007, at the time that Frontier Economics was modelling the market-based electricity purchase cost allowance for the original determination, contract prices increased significantly as a result of drought conditions effecting hydro and thermal generators. In particular, the imposition of restrictions on the use of water by thermal power stations in the south-east of Queensland had a substantial impact on the market.

As part of its advice to IPART for the original determination, Frontier Economics modelled the impact of the drought on spot prices over the period to 2009/10. Frontier Economics’ modelling suggested that there was a very high likelihood that contract prices would return to normal by 2009/10. At the time, Frontier Economics suggested that there were two key reasons for this:

there was strong evidence at the time that rainfall patterns would return to normal over the following 12 months, which would assist in replenishing the storages of hydro generators;

modelling showed that if the delivery of recycled water to the Tarong power station were to occur as planned, and Kogan Creek were commissioned on time, this would relieve the shortage of capacity and raise competitive pressures for generators across the NEM.

By the time of the first annual review, the water restrictions on Tarong power station had been lifted, and Kogan Creek was commissioned. As a result, contract prices had substantially moderated from the peaks of 2007. Nevertheless, the effects of the drought were lingering, and Frontier Economics’ modelling for the first annual review took account of the continued impact of the drought on the water storages of hydro generators, particularly Snowy.



Recent observation indicates that Snowy’s water storages during 2008 were very similar to its water storages during 2007. Based on this, Frontier Economics’ modelling for this second annual review adopts the same assumptions about the drought’s impact on Snowy’s availability as were adopted for the first annual review. That is, Snowy is restricted to producing at 80 per cent of its long-term average production, and there is significantly more pumping in off-peak times.

However, there remains some uncertainty about the likely impact of water restrictions on generators over the period to 30 June 2010. Since 2007, NEMMCO has published regular reports on the likely impact of drought conditions on generation in the NEM. NEMMCO’s drought reports set out likely impacts of drought conditions under two scenarios:12

Low rainfall – based on rainfall between 1 July 2006 and 30 June 2007;

Short term average rainfall – based on the average rainfall recorded over the past 10 years.

The capacity reductions forecast by NEMMCO under each of these scenarios are set out in Figure 3. The energy reductions forecast by NEMMCO under each of these scenarios are set out in Figure 4. Two things stand out from these Figures. First, at least under the low rainfall scenario, the drought is still forecast by NEMMCO to have a substantial impact on generator availability. Second, there is significant variation between the impact of the drought under the low rainfall scenario and the short term average rainfall scenario.

Figure 3: Capacity reductions under NEMMCO drought modelling

12 NEMMCO, Drought Scenario Investigation, December 2008 Update, December 2008.

0

200

400

600

800

1000

1200

1400

1600

1800

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

2009 2010 2009 2010 2009 2010 2009 2010

NSW QLD TAS VIC

MW

.

Low rainfall

Average



Figure 4: Energy reductions under NEMMCO drought modelling

Modelling undertaken by Frontier Economics indicates that the generator availability outcomes under the average rainfall scenario do not have a significant impact on spot prices. For this reason, and because of the uncertainty associated with rainfall outcomes over the forecast period, Frontier Economics has not incorporated the average rainfall scenarios in its modelling.

However, generator availability outcomes under the low rainfall scenario undoubtedly would have a significant impact on spot prices. Frontier Economics does not incorporate the restrictions under the NEMMCO low rainfall scenario in its modelling of the spot price because of the very low likelihood that these conditions will occur. Furthermore, if low rainfall conditions do come to pass, even NEMMCO notes that jurisdictions may consider allocation of additional water to the affected power stations if customer electricity supplies are threatened.13

While the substantial reductions in generator availability forecast by NEMMCO in its average and low rainfall scenarios are not directly incorporated in Frontier Economics modelling, the uncertainty associated with the impact of drought conditions is taken into account by Frontier Economics. Given the impact of the drought during 2007, the continuing dry conditions in parts of south-east Australia, and NEMMCO’s regular drought reports which have continued to indicate the possibility of significant impacts on generator availability due to

13 NEMMCO, Drought Scenario Investigation, December 2008 Update, December 2008, page 8.

0

500

1000

1500

2000

2500

3000

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

Q1

Q2

Q3

Q4

2009 2010 2009 2010 2009 2010 2009 2010

NSW QLD TAS VIC

GW

h .

Low rainfall

Average



drought, generators and retailers face uncertainty about the ongoing impact of the drought. This uncertainty will itself have an impact on the behaviour of market participants, and the effect of this uncertainty, along with the other uncertainties discussed in this section, is incorporated into Frontier Economics’ modelling of the spot price.

3.5 GREENHOUSE POLICIES

The Commonwealth Government has committed to introducing an emissions trading scheme by 2010 and to increasing the renewable energy target for the electricity industry, both as part of a policy to limit greenhouse gas emissions.

Policies to reduce the greenhouse gas emissions of the electricity industry will have a substantial impact on outcomes in electricity industry, including spot prices and contract prices. Carbon-emitting generators will face substantial additional costs and will, to the extent possible, pass these on through higher spot prices and contract prices. While it seems clear that an emissions trading scheme will be introduced, there is market uncertainty about the timing, nature and scope of an emissions trading scheme. The result is that both generators and retailers are uncertain about the impact of greenhouse policies on spot prices and contract prices. Generators and retailers are currently negotiating contracts to cover the period that straddles the expected introduction of an emissions trading scheme, and will take future greenhouse policies into account when contracting for this period. This appears to be effecting current contract prices as well as future contract prices.

3.6 IMPLICATIONS FOR MODELLING APPROACH

As discussed in the sections above, there are a number of factors that are currently leading to uncertainty among market participants. This uncertainty is expected to have an impact on electricity trading behaviour in the market, and as a result, spot prices and contract prices.

In order to understand Frontier Economics approach to incorporating the impacts of this uncertainty, it is first useful to think about how these uncertainties would be expected to affect contracting behaviour. When considering whether to enter into a contract, participants will consider the opportunity cost of doing so. This opportunity cost will depend, among other things, on the expected future spot price. For instance, if a generator considers that there is some probability of an increase in spot prices in future – for instance, because of the effects of a drought or greenhouse policies – then the opportunity cost to the generator of entering into the contract will be greater than if spot prices were not expected to increase. As a result, the generator will look for a higher contract price to compensate for the opportunity cost of entering into the contract or, to put this another way, will enter into fewer contracts at the lower price. The incentives of a retailer can be thought of in the same way. If a retailer considers that there is some probability of an increase in spot prices in future because of a drought, then the retailer will be prepared to pay more in order to enter into the contract.



The opportunity cost of entering into contracts will not just depend on the expected future spot price, but also on the existing contract position of participants. For instance, the opportunity cost to a generator of entering into a contract will depend on how that contract fits within the generator’s portfolio. If the generator considers that there is some probability that it will find itself over-contracted in the future, then the opportunity cost to the generator of entering into an additional contract will be greater than if that were not the case due to the heightened risk of unfunded contract difference payments. As a result, the generator will look for a higher contract price to compensate for the opportunity cost of entering into the contract or, to this another way, will enter into fewer contracts at the lower price.

This suggests that the nature of the uncertainties that are currently affecting the market are having the effect of increasing the price that generators will look for in order to enter into a contract or, to put this another way, causing generators to enter into fewer contracts at the lower price.

There are a number of ways the increased uncertainty, or risk, can be accounted for in regulated tariffs, and a number of ways that risk is incorporated in IPART’s 2007 Final Determination, including in the retail margin. As part of its work for the original determination, Frontier Economics adopted an approach to estimating the retail margin that reflected the energy purchasing risk associated with Frontier Economics’ modelling of the efficient frontier. For the purposes of these annual reviews, the only cost component that is open to review is the market-based electricity purchase cost allowance; the retail margin and other cost components are not open to review. However, as discussed, the increased uncertainty discussed in this section also impacts on spot prices and contract prices, and should therefore be accounted for in forecasts of spot prices and contract prices.

As discussed in Section 4, Frontier Economics models spot prices using SPARK – a game theoretic model that identifies optimal generator bidding patterns. SPARK works by allowing generators to bid a range of possible different strategies, rather than forcing generators to bid according to a specified strategy. The model then finds combinations of bidding strategies, and the associated spot price outcomes, that are optimal.

In order to account for the impact of uncertainty on generators’ bidding behaviour, the range of bidding strategies that are available to generators in the modelling for this second annual review has been expanded to incorporate bidding strategies that include a lower level of contracting than had prevailed in the past. Importantly, generators are not forced to adopt these new bidding strategies. Rather, these bidding strategies impact on the expected spot price only to the extent that they form part of an optimal bidding solution.


Modelling

4 Modelling

This section sets out the modelling that Frontier Economics has undertaken to estimate the market-based electricity purchase cost allowance for 2009/10. This section provides an overview of Frontier Economics’ modelling approach, sets out the changes to the modelling assumptions that have been made since previous modelling has been undertaken, and provides the results of Frontier Economics’ modelling.

4.1 MODELLING APPROACH

Consistent with the requirements for the annual reviews set out in IPART’s 2007 Final Determination, Frontier Economics has modelled the market-based electricity purchase cost allowance cost for 2009/10. The same modelling approach that was used for the original determination has been adopted.

Typically, Frontier Economics uses a three-staged modelling approach to determine the market-based electricity purchase cost allowance:

Stage 1 – Involves modelling the future characteristics of the NEM power system over the modelling period, focusing on the requirement for new generation capacity by location to meet the NEM reliability criteria. This stage involves the use of Frontier’s long-term investment model, WHIRLYGIG. The structure and operation of WHIRLYGIG is summarized in Frontier Economics’ Draft Methodology Paper, released in October 2006.14

Stage 2 – Involves modelling the price outcomes in the NEM based on the market structure derived in the Stage 1 modelling using Frontier’s market model, SPARK. This stage focuses on identifying optimal (i.e. mutually profit maximizing) generator bidding patterns and using these to determine future prices. The results of this modelling are a set of half hourly spot price forecasts for the modelling period. Contract prices can be derived from these spot forecasts by adjusting the spot prices to account for the contract premium, as discussed in Section 3.6. The structure and operation of SPARK is summarized in Frontier Economics’ Draft Methodology Paper.

Stage 3 – Involves drawing together the spot and contract price data derived from Stage 2 modelling, together with each standard retailer’s regulated load, into Frontier’s portfolio optimisation model, STRIKE. STRIKE determines the efficient mix of energy purchasing instruments (i.e. spot and contracts of various kinds) for each level of risk. The structure and operation of STRIKE is summarized in Frontier Economics’ Draft Methodology Paper.

The STRIKE analysis requires a correlated time-series of load and price data as an input. It is important that the data capture the likely correlation of load

14 Frontier Economics, Draft methodology for energy costs consultancy and retail costs/margin consultancy,

October 2006.

http://www.ipart.nsw.gov.au/files/Frontier%20Economics%20-%20Draft%20methodology%20paper%20-%20FINAL%20-%2027%20October%202006.PDF


Modelling

and spot price, as this relationship will ultimately impact on the efficient energy purchase costs. For example, a given load shape with a relatively high correlation to spot prices will have a higher efficient purchase cost than the same load shape with a lower correlation to spot prices.

As indicated in Section 2, the annual review is to be conducted using price estimates supplied by Frontier Economics that are updated to reflect any material changes to the NEM (not already taken into account in the modelling) since the modelling was concluded for the energy cost allowance incorporated in IPART’s 2007 Final Determination. However, as required, the retailer-specific load data used is the same used in the estimation of the energy cost allowance incorporated in IPART’s 2007 Final Determination. The analysis is forward looking and not an ex-post review of the retailers’ actual energy costs.

This three-staged modelling process and the relationship between each model is summarised in Figure 5.

Figure 5: Three-staged electricity modelling process

For the purposes of Frontier Economics advice to IPART for the original determination, Frontier Economics applied this three-staged modelling approach discussed above. For this second annual review, however, the first stage of the modelling approach – long-term investment modelling – is not required. Since the modelling period for this second annual review is only the 2009/10 financial year, and investment decisions for generation projects that will be available to the market for 2009/10 have already been made, there is no need to model investment decisions for 2009/10.

DemandNetwork

Existing plantNew plant costs

Regulatory policies

New plant buildContract levels

Industry structureStrategic players &

bid options

Price distributionsPlant dispatch

Customer loadContract bookForward curve

Plant build Pool pricesPlant output

Efficient frontier’soptimal portfolio

DemandNetwork

Existing plantNew plant costs

Regulatory policies

New plant buildContract levels

Industry structureStrategic players &

bid options

Price distributionsPlant dispatch

Customer loadContract bookForward curve

Plant build Pool pricesPlant output

Efficient frontier’soptimal portfolio


Modelling

4.2 MODELLING ASSUMPTIONS

Since Frontier Economics’ modelling for the original 2007 determination and the first annual review, further information about structural conditions in the market has become available. This section provides on overview of the modelling assumptions that have been updated for the modelling for this second annual review.

4.2.1 Demand

Demand forecasts for 2009/10 have been updated to reflect the latest information available from NEMMCO, from the 2008 Statement of Opportunities (SOO).15 The differences in NSW energy demand between the 2006 SOO (used for the modelling for the original determination) the 2007 SOO (used for the modelling for the first annual review) and the 2008 SOO (used for the modelling for this second annual review) are shown in Table 1. The most recent information forecasts a material reduction in energy demand for NSW for 2009/10. Maximum demand forecasts for 2009/10 are also lower in the 2008 SOO than they were in the 2007 SOO

2006 SOO 2007 SOO 2008 SOO

NSW energy (2009/10)

78,160 78,000 76,120

Table 1: NSW energy for 2009/10 (GWh)

Energy forecasts and maximum demand forecasts for 2009/10 for other NEM regions have also been updated to reflect the 2008 SOO.

4.2.2 Supply

Supply forecasts for 2009/10 have been updated to reflect the latest information from NEMMCO’s 2008 SOO. In NSW there has been some decrease in capacity available in 2009/10 in the latest SOO. However, this is within the context of a number of new generation projects being commissioned in NSW prior to 2009/10, including Uranquinty and Colongra.

Changes to generation available in 2009/10 for other NEM regions has also been updated to reflect the 2008 SOO.

15 NEMMCO, Statement of Opportunities for the National Electricity Market, October 2008.


Modelling

There have also been a number of changes to assumptions about ownership of generation plant since Frontier Economics’ modelling for the original determination. Structural changes to the market incorporated in Frontier Economics’ modelling include the following:

Torrens Island – has moved from TRUenergy to AGL;

Hallett – has moved from AGL to TRUenergy;

Bairnsdale – has moved from AGL to BBP;

Oakey – has moved from Enertrade to BBP;

Gladstone – has moved from Enertrade to Stanwell;

Mt Stuart – has moved from Enertrade to Origin; and

Uranquinty – has moved to Origin.

These structural changes have an impact on the strategic bidding outcomes of Frontier Economics’ SPARK modelling.

4.2.3 Drought

As discussed, Frontier Economics has modelled spot prices using the same drought modelling scenario used for the purposes of Frontier Economics’ modelling for the first annual review in 2008. That is, it is assumed that:

Snowy produces at 80 per cent of long-term average and there is significantly more pumping in off-peak times;

Dartmouth and Eildon power stations do not run at all.

4.2.4 ETEF

As discussed, Frontier Economics has modelled spot prices using two scenarios for the roll-off of ETEF:

ETEF rolls off according to the first revised schedule;16 and

ETEF does not roll off at all until after 2009/10.

Results for both scenarios are presented.

4.2.5 Load forecasts

Consistent with the requirements for the annual reviews set out in IPART’s 2007 Final Determination, Frontier Economics has modelled the market-based electricity purchase cost allowance cost using the same load profiles for the standard retailers that were used for the purposes of Frontier Economics’ March 2007 Final Report.

16 As discussed, the second revised schedule for the roll off of the ETEF was not announced until

after the modelling work for this report was completed. However, as discussed in Section 5 of this report, assumed variations in the schedule for the roll off of the ETEF have little impact on the forecast market-based electricity purchase cost allowance.


Modelling

4.3 MODELLING RESULTS

4.3.1 Spot and contract price modelling

As discussed, Frontier Economics has modelled spot prices under two scenarios:

the ETEF rolls off according to the revised schedule; and

the ETEF does not roll off at all until after 2009/10.

The spot price forecasts vary to some extent between these scenarios, reflecting differences in contract outcomes under the two scenarios. Figure 6 sets out Frontier Economics spot price forecast for 2009/10 under these two scenarios. Figure 6 also includes spot price forecasts from the original determination and from the drought modelling undertaken in 2007 and, for the purposes of comparison, d-cypha base load swap contract prices as of 6 February 2009.

$0

$5

$10

$15

$20

$25

$30

$35

$40

$45

$50

FE - original FE - originaldrought

FE - ETEF at100%

FE - ETEF rolls off D-cypha prices

NS

W a

vg p

rice

($/M

Wh)

.

Figure 6: NSW average price forecasts ($2006/07)

Some care is needed in comparing spot price forecasts with d-cypha contract prices from a particular trading day. After all, d-cypha prices can move significantly from day to day. For this reason, Figure 7 and Figure 8 show d-cypha prices for base load swaps for 2009/10 and peak load swaps for 2009/10 for each trading day over the period to 12 February 2009. These figures show that contract prices for 2009/10 have remained at higher levels than were observed prior to the impact of the drought beginning to significantly impact on prices during 2007.


Modelling

Interestingly, while d-cypha prices for 2007/08 contracts and 2008/09 contracts returned to more normal levels following the impact of the drought during 2007, d-cypha prices for 2009/10 have not. Figure 9 and Figure 10 show d-cypha contract prices for quarter one base load swaps and for quarter one peak load swaps for each of 2007/08, 2008/09 and 2009/10. The prices for contracts in 2007/08 and 2008/09 increased substantially over 2007 as a result of the impact of the drought, but then returned to more normal levels during 2008. In contrast, the prices for contracts in 2009/10 did not increase as much during 2007, but have remained around these higher levels since then. Frontier Economics’ modelling suggests that an explanation for this is the uncertainty that has been effecting the market over the past year.

$0

$20

$40

$60

$80

$100

$120

$140

$160

$180

3-O

ct-0

5

28-N

ov-0

5

27-J

an-0

6

24-M

ar-0

6

23-M

ay-0

6

18-J

ul-0

6

12-S

ep-0

6

7-N

ov-0

6

5-Ja

n-07

5-M

ar-0

7

3-M

ay-0

7

29-J

un-0

7

24-A

ug-0

7

19-O

ct-0

7

14-D

ec-0

7

14-F

eb-0

8

14-A

pr-0

8

11-J

un-0

8

6-A

ug-0

8

1-O

ct-0

8

26-N

ov-0

8

27-J

an-0

9

Con

trac

t pric

e ($

/MW

h)

.

2009/2010 - Q1

2009/2010 - Q2

2009/2010 - Q3

2009/2010 - Q4

Figure 7: d-cypha base swap contract prices for 2009/10 (nominal)


Modelling

$0

$20

$40

$60

$80

$100

$120

$140

$160

$180

3-O

ct-0

5

28-N

ov-0

5

26-J

an-0

6

23-M

ar-0

6

22-M

ay-0

6

17-J

ul-0

6

11-S

ep-0

6

6-N

ov-0

6

4-Ja

n-07

2-M

ar-0

7

2-M

ay-0

7

28-J

un-0

7

23-A

ug-0

7

18-O

ct-0

7

13-D

ec-0

7

13-F

eb-0

8

11-A

pr-0

8

10-J

un-0

8

5-A

ug-0

8

30-S

ep-0

8

25-N

ov-0

8

23-J

an-0

9

Con

trac

t pric

e ($

/MW

h)

.

2009/2010 - Q1

2009/2010 - Q2

2009/2010 - Q3

2009/2010 - Q4

Figure 8: d-cypha peak swap contract prices for 2009/10 (nominal)

$0

$20

$40

$60

$80

$100

$120

$140

$160

$180

1-A

pr-0

431

-May

-26

-Jul

-04

20-S

ep-0

415

-Nov

-04

13-J

an-0

510

-Mar

-05

6-M

ay-0

51-

Jul-0

526

-Aug

-05

21-O

ct-0

516

-Dec

-05

16-F

eb-0

613

-Apr

-06

12-J

un-0

67-

Aug

-06

2-O

ct-0

627

-Nov

-06

25-J

an-0

723

-Mar

-07

23-M

ay-

19-J

ul-0

713

-Sep

-07

8-N

ov-0

78-

Jan-

085-

Mar

-08

5-M

ay-0

81-

Jul-0

826

-Aug

-08

21-O

ct-0

816

-Dec

-08

Con

trac

t pric

e ($

/MW

h)

.

2007/2008 - Q1

2008/2009 - Q1

2009/2010 - Q1

Figure 9: d-cypha base swap contract prices for Q1 (nominal)


Modelling

$0

$20

$40

$60

$80

$100

$120

$140

$160

$180

1-A

pr-0

431

-May

-26

-Jul

-04

20-S

ep-0

415

-Nov

-04

13-J

an-0

510

-Mar

-05

6-M

ay-0

51-

Jul-0

526

-Aug

-05

21-O

ct-0

516

-Dec

-05

15-F

eb-0

612

-Apr

-06

9-Ju

n-06

4-A

ug-0

629

-Sep

-06

24-N

ov-0

624

-Jan

-07

22-M

ar-0

722

-May

-18

-Jul

-07

12-S

ep-0

77-

Nov

-07

7-Ja

n-08

4-M

ar-0

82-

May

-08

30-J

un-0

825

-Aug

-08

20-O

ct-0

815

-Dec

-08

Con

trac

t pric

e ($

/MW

h)

.

2007/2008 - Q1

2008/2009 - Q1

2009/2010 - Q1

Figure 10: d-cypha peak swap contract prices for Q1 (nominal)

4.3.2 Efficient purchase cost modelling

For the purposes of this second annual review, efficient frontiers have been modelled for three sets of contract prices:

contract prices based on Frontier Economics’ modelling under the scenario in which the ETEF rolls off according to the first revised schedule

contract prices based on Frontier Economics’ modelling under the scenario in which the ETEF does not roll off at all until after 2009/10;

contracts prices from d-cypha as of 6 February 2009.

Note that efficient frontiers based on d-cypha contract prices are presented for the purposes of comparison only. Consistent with the requirements for the annual reviews set out in IPART’s 2007 Final Determination, the market-based electricity purchase cost allowance is not based on publicly available data on contract prices. Rather, it is based on the same modelling approach used by Frontier Economics for the purposes of the original determination.

Figure 11 through Figure 13 provide the efficient frontiers for each of Country Energy, EnergyAustralia and Integral Energy. For the purposes of comparison, the efficient frontiers from the original determination are also included in these figures. It is clear from this figures that both the spot prices forecasts for this second annual review and the d-cypha data result in a higher efficient purchase cost.


Modelling

$0

$10

$20

$30

$40

$50

$60

$70

$80

$0 $1 $2 $3 $4 $5 $6 $7 $8 $9 $10

Risk (stdev $/MWh)

Exp

ecte

d co

st (

$/M

Wh)

.

FE - ETEF at 100%

FE - ETEF rolls off

D-cypha prices

FE - original

Figure 11: Country Energy efficient frontiers for 2009/10 ($2006/07)

$0

$10

$20

$30

$40

$50

$60

$70

$80

$0 $1 $2 $3 $4 $5 $6 $7 $8 $9 $10

Risk (stdev $/MWh)

Exp

ecte

d co

st (

$/M

Wh)

.

FE - ETEF at 100%

FE - ETEF rolls off

D-cypha prices

FE - original

Figure 12: EnergyAustralia efficient frontiers for 2009/10 ($2006/07)


Modelling

$0

$10

$20

$30

$40

$50

$60

$70

$80

$0 $1 $2 $3 $4 $5 $6 $7 $8 $9 $10

Risk (stdev $/MWh)

Exp

ecte

d co

st (

$/M

Wh)

.

FE - ETEF at 100%

FE - ETEF rolls off

D-cypha prices

FE - original

Figure 13: Integral Energy efficient frontiers for 2009/10 ($2006/07)

Taking the most conservative points on these efficient frontiers, as used in IPART’s 2007 Final Determination, the estimated market-based electricity purchase cost allowance for 2009/10 for each retailer is summarised in Figure 14. For both of Frontier Economics’ price forecasts, and for the d-cypha contract prices, the 10 per cent threshold for the market-based electricity purchase cost allowance has been exceeded for 2009/10 for each of the standard retailers. The electricity purchase cost allowances for 2009/10, and the percentage increase on the forecasts for 2009/10 from the original determination, are set out in Table 2 and Table 3. The percentage increases from the original determination are determined in real terms (in 2006/07 dollars). The electricity purchase cost allowances for 2009/10 from this second annual review are also presented in 2009/10 dollars.


Modelling

$0

$10

$20

$30

$40

$50

$60

$70

$80

2009/10 2009/10 2009/10

CE EA IE

Ene

rgy

cost

(co

nser

vativ

e po

int $

/MW

h) .

Orig (showing 10% threshold)

ETEF at 100%

ETEF rolls off

d-cypha strikes

Figure 14: Summary of electricity purchase cost allowance cost for 2009/10 ($2006/07)

Original determination

($2006/07)

FE – ETEF at 100%

($2006/07)

Increase FE – ETEF at 100%

($2009/10)

Country Energy $43.7 $51.3 17% $55.7

EnergyAustralia $49.8 $57.8 16% $62.8

Integral Energy $52.2 $60.6 16% $65.9

Table 2: Electricity purchase cost allowance cost for 2009/10 under ETEF at 100% scenario

Original determination

($2006/07)

FE – ETEF rolls off

($2006/07)

Increase FE – ETEF rolls off

($2009/10)

Country Energy $43.7 $50.7 16% $55.1

EnergyAustralia $49.8 $57.0 14% $61.9

Integral Energy $52.2 $59.8 15% $65.0

Table 3: Electricity purchase cost allowance cost for 2009/10 under ETEF rolls off scenario


Modelling


Summary and conclusions

5 Summary and conclusions

Based on Frontier Economics’ modelling work, and observation of recent d-cypha prices, Frontier Economics considers that the market-based electricity purchase cost allowance for 2009/10 differs by more than 10 per cent from the allowance incorporated in IPART’s 2007 Final Determination.

Given that current policy is for the roll-off of ETEF to progress during 2009/10, Frontier Economics considers that the appropriate forecasts of the market-based electricity purchase cost allowance are those set out in Table 3. Frontier Economics recognises that the schedule for the roll off of the ETEF under this scenario is slightly different from the second revised schedule for the roll off of the ETEF that was recently announced by Government, but considers that, given there is only a slight difference between the forecasts under the two ETEF scenarios that are modelled, the impact of this slight difference will be minor. Based on this, Frontier Economics considers that the market-based electricity purchase cost allowance for 2009/10 differs by between 14 per cent and 16 per cent from the allowance incorporated in IPART’s 2007 Final Determination, depending on the retailer.

The Frontier Economics Network

Frontier Economics Limited in Australia is a member of the Frontier Economics network, which consists of separate companies based in Australia (Melbourne, Sydney & Brisbane) and Europe (London, Cologne and Brussels). The companies are independently owned, and legal commitments entered into by any one company do not impose any obligations on other companies in the network. All views expressed in this document are the views of Frontier Economics Pty Ltd.

Disclaimer

None of Frontier Economics Pty Ltd (including the directors and employees) make any representation or warranty as to the accuracy or completeness of this report. Nor shall they have any liability (whether arising from negligence or otherwise) for any representations (express or implied) or information contained in, or for any omissions from, the report or any written or oral communications transmitted in the course of the project.

annual energy cost review€¦ · figure 11: country energy efficient frontiers for 2009/10...

Documents