ESTIMATED ENERGY COSTS

Transcription

1 REPORT TO QUEENSLAND COMPETITION AUTHORITY 9 MAY 217 ESTIMATED ENERGY COSTS RETAIL TARIFFS FOR USE BY THE QUEENSLAND COMPETITION AUTHORITY IN ITS FINAL DETERMINATION ON RETAIL ELECTRICITY TARIFFS

2 ACIL ALLEN CONSULTING PTY LTD ABN WAKEFIELD STREET ADELAIDE SA 5 AUSTRALIA T LEVEL FIFTEEN 127 CREEK STREET BRISBANE QLD 4 AUSTRALIA T F LEVEL TWO 33 AINSLIE PLACE CANBERRA ACT 26 AUSTRALIA T F LEVEL NINE 6 COLLINS STREET MELBOURNE VIC 3 AUSTRALIA T F LEVEL ONE 5 PITT STREET SYDNEY NSW 2 AUSTRALIA T F LEVEL TWELVE, BGC CENTRE 28 THE ESPLANADE PERTH WA 6 AUSTRALIA T F ACILALLEN.COM.AU SUGGESTED CITATION FOR THIS REPORT ESTIMATED ENERGY COSTS FOR RETAIL TARIFFS: FINAL DETERMINATION, ACIL ALLEN, MAY 217 RELIANCE AND DISCLAIMER THE PROFESSIONAL ANALYSIS AND ADVICE IN THIS REPORT HAS BEEN PREPARED BY ACIL ALLEN CONSULTING FOR THE EXCLUSIVE USE OF THE PARTY OR PARTIES TO WHOM IT IS ADDRESSED (THE ADDRESSEE) AND FOR THE PURPOSES SPECIFIED IN IT. THIS REPORT IS SUPPLIED IN GOOD FAITH AND REFLECTS THE KNOWLEDGE, EXPERTISE AND EXPERIENCE OF THE CONSULTANTS INVOLVED. THE REPORT MUST NOT BE PUBLISHED, QUOTED OR DISSEMINATED TO ANY OTHER PARTY WITHOUT ACIL ALLEN CONSULTING S PRIOR WRITTEN CONSENT. ACIL ALLEN CONSULTING ACCEPTS NO RESPONSIBILITY WHATSOEVER FOR ANY LOSS OCCASIONED BY ANY PERSON ACTING OR REFRAINING FROM ACTION AS A RESULT OF RELIANCE ON THE REPORT, OTHER THAN THE ADDRESSEE. IN CONDUCTING THE ANALYSIS IN THIS REPORT ACIL ALLEN CONSULTING HAS ENDEAVOURED TO USE WHAT IT CONSIDERS IS THE BEST INFORMATION AVAILABLE AT THE DATE OF PUBLICATION, INCLUDING INFORMATION SUPPLIED BY THE ADDRESSEE. UNLESS STATED OTHERWISE, ACIL ALLEN CONSULTING DOES NOT WARRANT THE ACCURACY OF ANY FORECAST OR PROJECTION IN THE REPORT. ALTHOUGH ACIL ALLEN CONSULTING EXERCISES REASONABLE CARE WHEN MAKING FORECASTS OR PROJECTIONS, FACTORS IN THE PROCESS, SUCH AS FUTURE MARKET BEHAVIOUR, ARE INHERENTLY UNCERTAIN AND CANNOT BE FORECAST OR PROJECTED RELIABLY. ACIL ALLEN CONSULTING SHALL NOT BE LIABLE IN RESPECT OF ANY CLAIM ARISING OUT OF THE FAILURE OF A CLIENT INVESTMENT TO PERFORM TO THE ADVANTAGE OF THE CLIENT OR TO THE ADVANTAGE OF THE CLIENT TO THE DEGREE SUGGESTED OR ASSUMED IN ANY ADVICE OR FORECAST GIVEN BY ACIL ALLEN CONSULTING. ACIL ALLEN CONSULTING 217

3 C O N T E N T S 1 Introduction 1 2 Overview of approach Introduction Components of the energy cost estimates Methodology 2 3 Responses to submissions to Draft Determination Introduction Overall approach Inclusion of updated data in analysis for the Final Determination Estimation mechanisms Cost structure LGC costs 8 4 Estimation of energy costs Introduction Estimation of the Wholesale Energy Cost Estimation of renewable energy policy costs Estimation of other energy costs Estimation of energy losses Summary of estimated energy costs 32 FIGURES FIGURE 3.1 ENERGEX SYSTEM DEMAND SOLAR PV IMPACT, 1 FEBRUARY FIGURE 3.2 ENERGEX NSLP ESTIMATED SOLAR PV IMPACT 1 FEBRUARY FIGURE 4.1 ACTUAL AVERAGE TIME OF DAY QLD WHOLESALE SPOT PRICE ($/MWH, NOMINAL) AND LOAD PROFILE (MW, RELATIVE) TO FIGURE 4.2 DAILY STTM GAS PRICE ($/GJ, NOMINAL) - BRISBANE 12 FIGURE 4.3 FIGURE 4.4 FIGURE 4.5 FIGURE 4.6 ACTUAL ANNUAL AVERAGE DEMAND WEIGHTED PRICE ($/MWH, NOMINAL) BY TARIFF AND QUEENSLAND TIME WEIGHTED AVERAGE PRICE ($/MWH, NOMINAL) 29-1 TO ANNUALISED QUARTERLY BASE, PEAK AND CAP CONTRACT PRICES ($/MWH) FINAL DETERMINATION AND PREVIOUS FINAL DETERMINATIONS 13 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND BASE FUTURES 15 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND PEAK FUTURES 16

4 C O N T E N T S FIGURE 4.7 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND $3 CAP CONTRACTS 17 FIGURE 4.8 TOP ONE PERCENT HOURLY DEMANDS QUEENSLAND 18 FIGURE 4.9 TOP ONE PERCENT HOURLY DEMANDS ENERGEX NSLP 19 FIGURE 4.1 ANNUAL TWP FOR QUEENSLAND FOR 56 SIMULATIONS FOR COMPARED WITH ACTUAL ANNUAL OUTCOMES IN PAST YEARS 2 FIGURE 4.11 FIGURE 4.12 FIGURE 4.13 FIGURE 4.14 FIGURE 4.15 COMPARISON OF UPPER 1 PERCENT TAIL OF SIMULATED HOURLY PRICE DURATION CURVES FOR QUEENSLAND AND HISTORICAL OUTCOMES 2 ANNUAL AVERAGE CONTRIBUTION TO THE QUEENSLAND TWP BY PRICES ABOVE $3/MWH FOR QUEENSLAND IN FOR 56 SIMULATIONS COMPARED WITH ACTUAL OUTCOMES IN PAST YEARS 21 ANNUAL DWP FOR ENERGEX NSLP AS PERCENTAGE PREMIUM OF ANNUAL TWP FOR QUEENSLAND FOR 56 SIMULATIONS FOR COMPARED WITH ACTUAL OUTCOMES IN PAST YEARS 22 CONTRACT VOLUMES USED IN HEDGE MODELLING OF 56 SIMULATIONS FOR FOR ENERGEX NSLP 23 ANNUAL HEDGED PRICE AND DWP ($/MWH, NOMINAL) FOR ENERGEX NSLP FOR THE 56 SIMULATIONS FIGURE 4.16 LGC PRICES FOR 217 AND 218 ($/LGC, NOMINAL) 26 TABLES TABLE 3.1 REVIEW OF ISSUES RAISED IN SUBMISSIONS IN RESPONSE TO DRAFT DETERMINATION 5 TABLE 4.1 ESTIMATED CONTRACT PRICES ($/MWH) 14 TABLE 4.2 ESTIMATED WEC ($/MWH, NOMINAL) FOR AT THE QUEENSLAND REFERENCE NODE 24 TABLE 4.3 ESTIMATING THE 218 RPP VALUE 27 TABLE 4.4 ESTIMATED COST OF LRET TABLE 4.5 ESTIMATED COST OF SRES TABLE 4.6 TOTAL RENEWABLE ENERGY POLICY COSTS ($/MWH) FINAL DETERMINATION , DRAFT DETERMINATION AND FINAL DETERMINATION TABLE 4.7 NEM MANAGEMENT FEE ($/MWH) TABLE 4.8 AEMO PRUDENTIAL COSTS TABLE 4.9 TABLE 4.1 HEDGE PRUDENTIAL FUNDING COSTS BY CONTRACT TYPE 3 HEDGE PRUDENTIAL FUNDING COSTS FOR ENERGEX NSLP 31 TABLE 4.11 TOTAL PRUDENTIAL COSTS ($/MWH) TABLE 4.12 TABLE 4.13 TOTAL OF OTHER COSTS ($/MWH) FINAL DETERMINATION , DRAFT DETERMINATION AND FINAL DETERMINATION ESTIMATED TRANSMISSION AND DISTRIBUTION LOSS FACTORS FOR ENERGEX AND ERGON ENERGY S EAST ZONE 32

5 C O N T E N T S TABLE 4.14 ESTIMATED TEC FOR FINAL DETERMINATION 33 BOXES

6 Introduction I N T R O D U C T I O N 1 ACIL Allen has been engaged by the Queensland Competition Authority (the QCA) to provide advice on the energy related costs likely to be incurred by a retailer to supply customers on notified retail prices for Retail prices generally consist of three components: network costs energy costs costs associated with retailing to end users. ACIL Allen s engagement relates to the energy costs component only. In accordance with the Ministerial Delegation (the Delegation), and the Consultancy Terms of Reference (TOR) provided by the QCA, the methodology developed by ACIL Allen provides an estimate of energy costs to be incurred by a retailer to supply customers on notified prices (non-market customers) for Although the QCA s determination is to apply only to the Ergon Energy distribution area, the TOR specifically requests that ACIL Allen s analysis cover the same tariff classes as covered in the analyses for the , , and determinations, and therefore includes residential and small business customers in south east Queensland. This report provides estimates of the energy costs for use by the QCA in its Final Determination. These estimates have been revised slightly since the Draft Determination by taking into account feedback from the Draft Determination and updated market data. This report also provides responses to submissions made by various parties following the release of the QCA s paper, Draft determination: Regulated retail electricity prices for (February 217), where those submissions refer to the cost of energy in regulated retail electricity prices. 1

7 Overview of approach O V E R V I E W O F A P P R O A C H Introduction In preparing advice on the estimated energy costs, ACIL Allen is required to have regard to the actual costs of making, producing or supplying the goods or services which in this case are the customer retail services to be supplied to non-market customers for the tariff year 1 July 217 to 3 June 218. In undertaking the task, ACIL Allen has not been asked to provide advice on: the effect that the price determination might have on competition in the Queensland retail market the Queensland Government uniform tariff policy time of use pricing any transitional arrangements that might be considered or required. ACIL Allen understands that these matters will be considered by the QCA when making its Determination. 2.2 Components of the energy cost estimates Energy costs comprise: wholesale energy costs (WEC) for various demand profiles costs of complying with state and federal government policies, including the Renewable Energy Target (RET) National Electricity Market (NEM) fees, ancillary services charges and costs of meeting prudential requirements energy losses incurred during the transmission and distribution of electricity to customers. 2.3 Methodology ACIL Allen s methodology follows the methodology used to provide advice to the QCA for the , , , and Determinations (refer to ACIL Allen s report for the Draft Determination 1 and the Final Determination 2 for more details of the methodology). The ACIL Allen methodology estimates costs from a retailing perspective. This includes wholesale energy market simulations to estimate expected pool costs and volatility and the hedging of the pool price risk by entering into electricity contracts with prices represented by the observable futures 1 T.aspx 2 2

8 market data. Other energy costs are added to the wholesale energy costs and the total is then adjusted for network losses Wholesale energy costs As with the , , and reviews, ACIL Allen continues to use the market hedging approach for estimating the WEC for We have utilised our: stochastic demand model to develop 46 weather influenced simulations of hourly demand traces for each of the tariff profiles using temperature data from to and demand data for to stochastic outage model to develop 11 hourly power station availability simulations energy market models to run 56 simulations of hourly pool prices of the NEM using the stochastic demand traces and power station availabilities as inputs analysis of contract data to estimate contract prices hedge model taking the above analyses as inputs to estimate a distribution of hedged prices for each tariff class. We have then analysed the distribution of outcomes produced by the above approach to provide a risk adjusted estimate of the WEC for each tariff class. We have continued to rely on the Australian Energy Market Operator (AEMO) as a source for the various demand data required for the analysis. The QCA provided ACIL Allen with access to ASX Energy data, and OTC data from TFS Australia for the purpose of estimating contract prices. The peak demand and energy forecasts for the demand profiles are referenced to the current AEMO demand forecasts for Queensland and take into account past trends and relationships between the NSLPs and the Queensland region demand. It is our assessment that the AEMO medium series demand projection for provided in AEMO s 216 National Electricity Forecasting Report (NEFR) continues to be the most reasonable demand forecast for the purposes of this analysis. However, unlike the Draft Determination, ACIL Allen no longer assumes the closure of the 6 MW base load Portland aluminium smelter in August 217 in Victoria. Instead, the demand forecast assumes the continuation of operations at Portland throughout the determination period. No changes to the supply side assumptions have been made since the Draft Determination. Our modelling suggests that the continued operation of Portland in , and hence assuming demand in Victoria is on average 6 MW higher in the Final Determination than in the Draft Determination, results in a substantial increase wholesale prices in Victoria. Further, this change in the demand-supply balance, is projected to increase wholesale spot prices in Queensland by about $8/MWh in , all other things equal, relative to the Draft Determination Renewable energy policy costs Energy costs associated with the Large-scale Renewable Energy Target (LRET) and the Small-scale Renewable Energy Scheme (SRES) have been estimated using the latest price information from AFMA and TFS, and renewable energy percentages published by the Clean Energy Regulator (CER). Retailer compliance with these schemes operates on a calendar year basis and hence estimates are required for both 217 and 218 calendar years, with the costs averaged to estimate the financial year costs. To estimate the costs to retailers of complying with both the LRET and SRES, ACIL Allen uses the following elements: historical Large-scale Generation Certificate (LGC) market prices sourced from AFMA and TFS currently legislated LRET targets (in GWh) for 217 and 218 the Renewable Power Percentage (RPP) for as published by the CER 3 The CER is obligated to publish the official RPP for the 217 compliance year by 31 March 217 in accordance with Section 39 of the Renewable Energy (Electricity) Act 2 3

9 estimate of the RPP for 218 the binding Small-scale Technology Percentage (STP) for under the SRES as published by the CER estimate for the STP for 218 under the SRES the fixed clearing house price for Small-scale Technology Certificates (STCs) Other energy costs Market fees and ancillary service costs are estimated based on data and policy documents published by AEMO. Prudential costs, both AEMO and representing capital used to meet prudential requirements to support hedging take into account: the AEMO assessed maximum credit limit (MCL) the future risk-weighted pool price participant specific risk adjustment factors AEMO published volatility factors futures market prudential obligation factors, including: the price scanning range (PSR) the intra month spread charge the spot isolation rate. ACIL Allen undertook a separate analysis of the Retail Operating Cost and Margin ROC and ROM for the QCA for the Determination, and has been engaged to assist the QCA with an update for the Determination, and that separate analysis takes as an input the estimated prudential costs from this report (so as to avoid any double counting). ACIL Allen is of the opinion that prudential costs are associated with purchasing and hedging electricity and would apply to any NEM customer. Therefore they should be treated as a component of the energy cost rather than a cost associated with retailing Energy losses The estimated wholesale energy costs resulting from the analysis is referenced to the Queensland Regional Reference Node (RRN). These estimates need to be adjusted for transmission and distribution losses associated with transmitting energy from the Regional Reference Node to endusers. Distribution Loss Factors (DLF) for Energex and for the Ergon Energy east zone and average Marginal Loss Factors (MLF) for transmission losses from the node to major supply points in the distribution networks are applied to the wholesale energy cost estimates to incorporate losses. Since the Draft determination, the MLFs and DLFs used in the calculations have been updated based on the final MLFs and DLFs published by AEMO on 31 and 3 March 217 respectively. 4 The CER is obligated to publish the official STP for the 217 compliance year by 31 March 217 in accordance with subparagraph 4A (3)(a) of the Renewable Energy (Electricity) Act 2. This is an annual target and does not directly represent liable entities quarterly surrender obligations under the SRES. 4

10 Responses to submissions to Draft Determination R E S P O N S E S T O S U B M I S S I O N S T O D R A F T D E T E R M I N A T I O N Introduction The QCA forwarded to ACIL Allen a total of nine submissions in response to its Draft Determination. ACIL Allen reviewed the submissions to identify issues that required our consideration. A summary of the review is shown below in Table 3.1. The following sections in this chapter address each of the relevant issues raised in the submissions. TABLE 3.1 REVIEW OF ISSUES RAISED IN SUBMISSIONS IN RESPONSE TO DRAFT DETERMINATION Id Stakeholder Wholesale energy costs Contract prices / hedge model Renewable energy policy costs NEM fees Prudential 1 Australian Sugar Milling Council Nil Nil Nil Nil Nil Nil 2 Canegrowers ISIS Ltd Nil Nil Nil Nil Nil Nil 3 CANEGROWERS Yes Nil Yes Nil Yes Nil costs Energy losses 4 Chamber of Commerce and Industry Queensland (CCIQ) Nil Nil Nil Nil Nil Nil 5 Energy Queensland Yes Yes Yes Nil Nil Nil 6 Queensland Consumers Association Nil Nil Nil Nil Nil Nil 7 Queensland Council of Social Service (QCOSS) Yes Nil Nil Nil Nil Nil 8 Origin Energy Nil Yes Nil Nil Nil Nil 9 Cotton Australia Nil Nil Nil Nil Nil Nil Note: Yes = an issue was raised that required ACIL Allen s consideration SOURCE: ACIL ALLEN ANALYSIS OF QCA SUPPLIED DOCUMENTS 3.2 Overall approach A number of the submissions supported the continuation of ACIL Allen s approach for the purposes of consistency. For example, Energy Queensland note on page 2 of their submission: Energy Queensland supports a market based approach for determining energy costs. 5

11 Additionally, the Queensland Council of Social Service (QCOSS) notes on page six of their submission: In regard to energy (generation) costs, QCOSS is not aware of any new information which would suggest alternative approaches to those used in the price determination might be more appropriate. QCOSS supports the QCA s estimation of energy costs for being based on the application of the same methodology that was used in Inclusion of updated data in analysis for the Final Determination A number of submissions made note of the need to include updated data for the Final Determination. For example, Energy Queensland on page two of their submission note: We note that the Draft Determination has accounted for energy costs up to November216. However, given the very high prices experienced between December 216 and March 217 we anticipate that this will result in a higher energy cost allowance. And Origin Energy on page one of their submission note: It is noted that the QCA s Draft Determination only includes energy cost data relevant to November 216. It is understood the Final Determination will include updated wholesale energy cost data. As the QCA would be aware, the forward price for electricity has risen considerably since November 216. It is anticipated that the energy costs included in the Final Determination will be considerably higher than that set out in the Draft Determination. As with previous final determinations, the Final Determination takes into account market information made available since the Draft Determination. This includes: Updated contract prices Updated LGC prices Updated input assumptions to the market simulations. 3.4 Estimation mechanisms Energy Queensland on page two of their submission suggest that the recent increase in price outcomes observed in the market highlights a deficiency in the methodology : Recent high prices in the spot and contract market has again highlighted the limitations in the methodology used for determining energy costs when only using exchange information as the key input to the calculations. This is especially evident in the Large Renewable Energy Target (LRET) costs. ACIL Allen does not agree with this suggestion. The methodology takes into account the latest information available at the time of the analysis. Extension of the cut-off date for contract price data to April 217 for the Final Determination, two months before the commencement of the determination year, minimises the exclusion of further contract price data changes on a volume weighted basis. Further, the methodology utilises exchange based contract price data which is verified by OTC contract price data. This has been the approach for each determination and has consistently shown a very close alignment between the two sources of contract price data. However, for the Final Determination, ACIL Allen has utilised LGC price data from TFS to supplement the previous LGC price data from AFMA (as discussed in more detail in section 4.3.1). 3.5 Cost structure CANEGROWERS suggest that the Ergon NSLP should be used to calculate the energy costs and prudential costs, on pages one and three of their submission: In calculating the underlying cost structure, QCA has applied Energex network costs and applied these to the Ergon network. It has not taken account of the fact that many of the costs in the Energex area do not apply in the Ergon area. A significant factor is that the Energex load profile is peakier than the Ergon 6

12 load profile. This means that the cost of energy in Energex load profile is higher than that faced by Ergon. CANEGROWERS recommends the QCA calculate prudential costs based on the Ergon NSLP, not the Energex NSLP. ACIL Allen has been engaged to estimate the wholesale energy costs for each profile, including the Ergon NSLP, as provided in this report. CANEGROWERS suggestion is a matter for the QCA to consider, as we have not been engaged to provide advice on the manner in which these estimates are used in the determined tariffs. CANEGROWERS, on page two of their submission, disagree with our suggestion that further rooftop solar PV installation will not have an impact on the peak demand of the Energex NSLP: In the Draft Determination QCA reports Over the past few years, the Energex NSLP has become peakier due to increased solar generation reducing daytime demand but having no effect on the evening peak demand. Figure 8 shows how the NSLP has become peakier over time. This observation is at odds with the information contained in Energex s Distribution Annual Planning Report, 216/17-22/21, Volume The Energex NSLP shows that even on a cloudy day, PV is materially reducing the system peak. This is the opposite to the claim made in the QCA Draft Determination. CANEGROWERS reproduce the graph in Figure 3.1 in their submission, which is included in the Energex Distribution Annual Planning Review, which shows the time of day demand for the Energex system on 1 February 216. ACIL Allen agrees with CANEGROWERS rooftop solar PV has reduced the Energex system wide peak demand. However, our comment refers to the Energex NSLP, which is a subset of the Energex total system demand. Figure 3.2 shows the time of day NSLP demand for the same date 1 February 216. The graph clearly shows that when adding back the estimated generation from rooftop solar PV, the peak demand would have been the same given it occurs at around 7:pm 7:3pm. It is important to remember that the NSLP excludes large (commercial) consumers with interval meters, and hence has a different shape to the total Energex system demand. Regardless, our comment is more about the impact of further installations of rooftop solar PV, beyond the current installed stock (given that the current stock has already had its impact). Figure 3.2 shows that the NSLP peaks in summer well into the evening, and hence further installations will not have an impact on the peak demand but will continue to hollow out demand during daylight hours. FIGURE 3.1 ENERGEX SYSTEM DEMAND SOLAR PV IMPACT, 1 FEBRUARY 216 SOURCE: ENERGEX DAPR 216/17-22/21 7

13 12: AM 12:3 AM 1: AM 1:3 AM 2: AM 2:3 AM 3: AM 3:3 AM 4: AM 4:3 AM 5: AM 5:3 AM 6: AM 6:3 AM 7: AM 7:3 AM 8: AM 8:3 AM 9: AM 9:3 AM 1: AM 1:3 AM 11: AM 11:3 AM 12: PM 12:3 PM 1: PM 1:3 PM 2: PM 2:3 PM 3: PM 3:3 PM 4: PM 4:3 PM 5: PM 5:3 PM 6: PM 6:3 PM 7: PM 7:3 PM 8: PM 8:3 PM 9: PM 9:3 PM 1: PM 1:3 PM 11: PM 11:3 PM (MW) FIGURE 3.2 ENERGEX NSLP ESTIMATED SOLAR PV IMPACT 1 FEBRUARY 216 3, 2,5 2, 1,5 1, 5 Time of day Energex NSLP - With solar (as recorded) Energex NSLP - Without solar (As estimated) Solar generation SOURCE: ACIL ALLEN ANALYSIS OF AEMO DATA 3.6 LGC costs CANEGROWERS on page three of their submission re-introduce the argument that a prudent retailer will have entered into long term contract for LGCs and that these should be taken into account: QCA proposes to increase the charges for LGCs at a punitive 49.9% higher in than in This begs the question for which retailers in the Ergon distribution system have the proposed charges been calculated, a new marginal retailer or an incumbent retailer? The charges identified appear to be set for a marginal retailer that has no long-term offtake contracts in place and has not made an investment in renewable energy capacity. Aware of the Australia wide policy preference for renewable energy generation and a stronger policy push towards renewable energy in Queensland, an efficient prudent retailer with evergreen customer contracts (Ergon Retail) would have managed this risk by being long with respect to renewables. Just as retailers in the Energex network have managed their LGC price exposure, as the long-term incumbent retailer in regional Queensland, to manage its exposure Ergon retail will have longstanding purchases of Large-scale Renewable Energy Target (LRET) and Small-scale Renewable Energy Scheme (SRES) certificates. The QCA approach to setting the LGC component of prices is likely to deliver Ergon a windfall trading gain in the purchase of its certificates. CANEGROWERS recommends the QCA calculates the LGC component of prices based on the behaviour of an efficient long term incumbent retailer. This argument has been raised in previous determinations by retailers. However, retailers raise this issue when spot and futures prices are less than the levelised cost of electricity of new investment thermal or renewable generation. In this instance, given that spot and futures prices have increased to levels greater than the levelised cost of electricity of new generation, the suggestion has come from the consumer s side. ACIL Allen recognises that in practice retailers build a portfolio of LGCs from a number of sources including: Direct investment in renewable generation projects PPAs written with renewable generators Spot and forward purchases transacted through brokers and direct trades with counterparties. 8

14 Of these, the only one which is traded regularly with observable pricing are the spot and forward contracts transacted through brokers. ACIL Allen on page 27 in its report for the Final Determination 5 addressed this issue in detail. ACIL Allen continues to hold the view that the prices within the spot and futures market represent the most reliable indicator of the current market consensus view of the price of LGCs faced by retailers in attempting to satisfy their obligations under the LRET. Certainly, ACIL Allen is not of the view of cherry-picking the approach simply because market prices have increased or decreased to be above or below the levelised cost of electricity of new entrant generation

15 Estimation of energy costs E S T I M A T I O N O F E N E R G Y C O S T S Introduction In this section we apply the methodology described in Chapter 2 and summarise the estimates of each component of the Total Energy Cost (TEC) for each of the tariff classes for Historic energy cost levels Figure 4.1 shows the average time of day pool (spot) price for the Queensland region of the NEM, and the actual average time of day load profiles for Queensland, tariffs 31, 33 and 11, and the Ergon NSLP for the past five years. The graphs are useful in understanding the dynamics of the absolute and relative wholesale electricity price changes in the tariffs. It is worth noting the uplift in spot prices in , and again to date in , across most periods of the day, compared with This is a result of an increase in the underlying demand in Queensland due to the ramping up of production associated the LNG export facilities in Gladstone, as well as an increase in gas prices into gas fired generators (as shown by the ramp up in gas prices on AEMO s short term trading market (STTM) in Figure 4.2). Further, it can be seen that to date in prices are noticeably higher and more volatile during the evening periods this is largely due to the recent strong price outcomes in the protracted summer period driven by strong gas prices over the same period, as well as reduced output from some of the NSW coal fired power stations due to coal supply constraints. In relation to each load profile, we note the following: The annual time of day price profile has been volatile over the past five years with the overall level and shape of the price profile changing from one year to the next. For example, in the time of day profile was very flat compared with In and , prices increased largely because of the carbon tax. Prices have generally peaked in the afternoon and evening, whereas in some years there is also a morning peak. In short, the profile of prices varied from one year to the next noting that these are the annual profiles (seasonal profiles are even more variable over time). The load profile of tariff 31 has been relatively consistent from one year to the next since ramping up from about 9:3 pm, peaking at about midnight and then ramping down to about 3: am. This is inversely correlated with the price profile with load higher at times of lower spot prices. This has resulted, on average, in a relatively low wholesale energy cost for tariff 31, compared with the other tariffs. The load profile of tariff 33 has been relatively consistent from one year to the next for most parts of the day. However, there was some volatility between 5:3 pm and 1:3 pm over the past few years. The load exhibits a morning peak at around 8: am and prices also experience uplift around that time. The load also exhibits an evening peak at around 9:3 pm but this varied from year to year 1

16 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM (MW, relative) (MW, relative) 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM (MW, relative) (MW, relative) 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM 12: AM 1: AM 2: AM 3: AM 4: AM 5: AM 6: AM 7: AM 8: AM 9: AM 1: AM 11: AM 12: PM 1: PM 2: PM 3: PM 4: PM 5: PM 6: PM 7: PM 8: PM 9: PM 1: PM 11: PM ($/MWh, relative) (MW, relative) (note that in and it tends to peak around 8:3 pm). Compared with tariff 31, the load profile of tariff 33 is weighted slightly more towards the daylight hours and the evening peak, and hence it is not surprising that its wholesale energy costs are higher than those of tariff 31. Over the past few years, the Energex NSLP load profile (tariff 11), and to a similar degree, the Ergon NSLP, have experienced a carving out of load during daylight hours with the increased penetration of rooftop solar PV. This results in the load profile becoming peakier over time. The Energex NSLP load profile has a higher weighting towards the peak periods particularly the evening peak and hence it is not surprising that the NSLP has the highest wholesale energy cost out of the profiles. FIGURE 4.1 ACTUAL AVERAGE TIME OF DAY QLD WHOLESALE SPOT PRICE ($/MWH, NOMINAL) AND LOAD PROFILE (MW, RELATIVE) TO Queensland pool prices Queensland load $4. $3. $2. $1. $ Energex NSLP (T11) Ergon NSLP T31 T Note: The term relative MW means the loads for each tariff and year have been scaled so they sum to one. This removes differences in absolute scale between the different tariffs and changes in absolute size over time. This is an appropriate representation of the loads since it is the relative shape of the load profile, not its absolute size, which determines its wholesale energy cost. Values for based on data up to 18 April 217. Insufficient data available for for tariff classes due to lag in release of data by AEMO. SOURCE: ACIL ALLEN ANALYSIS OF AEMO DATA 11

17 $MWh, nominal 1/7/214 1/9/214 1/11/214 1/1/215 1/3/215 1/5/215 1/7/215 1/9/215 1/11/215 1/1/216 1/3/216 1/5/216 1/7/216 1/9/216 1/11/216 1/1/217 1/3/217 ($/GJ, nominal) FIGURE 4.2 DAILY STTM GAS PRICE ($/GJ, NOMINAL) - BRISBANE Date Brisbane (STTM) (AEMO) SOURCE: AEMO DATA Figure 4.3 shows the actual annual demand weighted spot price (DWP) for each of the tariffs compared with the time weighted average spot price in Queensland (TWP) over the past seven years. As expected, the DWPs for tariffs 31 and 33 are below the DWP for the NSLPs in each year, with tariff 31 having the lowest price. Although the rank order in prices by tariff has been consistent in each year, the dollar value differences between the prices has varied from one year to the next. For example, in , the flat half-hourly price profile resulted in the three tariffs having relatively similar wholesale spot prices. However, in , the increased price volatility across the afternoon period resulted in the NSLP spot price diverging away from tariff 31 and 33. Conversely, the increase in off-peak spot prices in has lifted the DWP of tariff 31 and 33 up towards that of the NSLP. The increase in spot price outcomes to date in is quite apparent, with prices on an average time weighted basis of about $95/MWh compared with $6/MWh in (representing an increase of just under 6 per cent). FIGURE 4.3 ACTUAL ANNUAL AVERAGE DEMAND WEIGHTED PRICE ($/MWH, NOMINAL) BY TARIFF AND QUEENSLAND TIME WEIGHTED AVERAGE PRICE ($/MWH, NOMINAL) 29-1 TO $12. $1. $8. $6. $4. $2. $ Qld TWP T31 - DWP T33 - DWP Energex NSLP - DWP Ergon NSLP - DWP Qld - DWP Note: Values reported are spot (or uncontacted) prices. Values for based on data up to 18 April 217. Insufficient data available for for tariff classes due to lag in release of data by AEMO. SOURCE: ACIL ALLEN ANALYSIS OF AEMO DATA 12

18 $/MWh, nominal The volatility of spot prices (timing and incidence) in the Queensland region of the NEM provides the incentive to a retailer to hedge their load, since hedging of the loads reduces a retailer s exposure to the volatility. The suite of contracts (as defined by base/peak, swap/cap and quarter) available to retailers does not really change from one year to the next. However, the movements in contract price is the key contributor to movements in the estimated wholesale energy costs of the different tariffs year on year, as is shown in Figure 4.4. The market modelling undertaken by ACIL Allen, and reported in this chapter, aligns with the market s expectations of price outcomes in Compared with the Final Determination, futures contract prices for , on a trade weighted basis, have increased by: About $16.55/MWh for base contracts About $27.67/MWh for peak contracts And about $3.26/MWh for cap contracts. The market is clearly expecting a strong increase in price outcomes in , particularly across the peak periods due to the closure of Hazelwood power station and continued operation of the Portland smelter. Further, gas prices continue to increase such that the short run marginal costs (SRMCs) of gas fired CCGTs are projected to increase by about $5-$1/MWh between 216 and 218, and the SRMCs of gas fired peaking plant are projected to experience a $1-$2/MWh increase. FIGURE 4.4 ANNUALISED QUARTERLY BASE, PEAK AND CAP CONTRACT PRICES ($/MWH) FINAL DETERMINATION AND PREVIOUS FINAL DETERMINATIONS $12. $1. $8. $6. $4. $2. $ Base Peak Cap SOURCE: ACIL ALLEN 4.2 Estimation of the Wholesale Energy Cost Estimating contract prices Contract prices for Queensland were estimated using the trade-weighted average of ASX Energy daily settlement prices since the contract was listed up until 3 April 217. Table 4.1 shows the estimated quarterly swap and cap contract prices for the Final Determination and compares them with the estimates under the Draft Determination and the Final Determination for

19 TABLE 4.1 ESTIMATED CONTRACT PRICES ($/MWH) Q3 Q4 Q1 Q2 Final Determination Base $63.13 $66.46 $89.73 $62.33 Peak $78.56 $89.23 $ $91.62 Cap $6.87 $13.47 $27.41 $6.75 Draft Determination Base $5.84 $57.62 $77.43 $5.83 Peak $72.17 $84.55 $ $74.26 Cap $5.6 $12.8 $23.4 $5.92 % change from Draft Determination Base 24% 15% 16% 23% Peak 9% 6% 2% 23% Cap 23% 12% 19% 14% Final Determination Base $44.77 $54.11 $71.19 $45.16 Peak $54.14 $68.18 $111.6 $57.25 Cap $4.81 $1.48 $21. $5.1 % change from Final Determination Base 41% 23% 26% 38% Peak 45% 31% 28% 6% Cap 43% 29% 31% 35% SOURCE: ACIL ALLEN ANALYSIS USING ASX ENERGY DATA UP TO 3 APRIL 217 Trade weighted contract prices for are on average 36 per cent higher than for due to an expected reduction in plant capacity and gas-fired generators offering capacity into the market at higher prices: On 3 November 216, Engie formally announced the closure of Hazelwood in Victoria (1,64 MW) from April 217. With no near-term replacement capacity planned, futures prices have significantly increased in all regions as a result. Gas prices continue to increase in the spot market, which improves confidence in the view that over the medium to long term contracted gas prices in power stations are on the rise. In some cases, this has led to planned plant closures, for example, Smithfield power station in New South Wales (17 MW), which is to close after July 217, when its gas supply contract ends. Trade weighted contract prices for the Final Determination are on average around 17 per cent higher than the Draft Determination There are some reasons that may explain this increase: At the time of the Draft Determination in November 216, the futures market had not completely factored in the impact of the Hazelwood closure, which was only formally announced less than a month prior to the cut-off date for the Draft Determination. Further, the futures market had assumed and factored in, to some degree, the closure of the Portland smelter load in the lead up to the Draft Determination. However, since the Draft Determination, there is now certainty that the smelter will continue to operate in , with the Victorian and federal governments announcing a package on 19 January 217 to extend the operation of the smelter, which has certainly increased futures prices, even in Queensland. 14

20 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) Electricity demand reached a level of 9,369 MW on 12 February 217, which aligns reasonably well with the forecast of peak demand published in the 216 NEFR given the high temperature conditions experienced in Queensland in February 217. However, Queensland also experienced a protracted summer in , with a record number of consecutive days above 3 degrees in Brisbane for example, which resulted in strong spot price outcomes. It is likely that the futures market for the determination year has been influenced by these recent events particularly the futures price for Q Futures continued to trade heavily at these higher prices, since the Draft Determination, thereby pushing up the trade-weighted average price for the Final Determination. The following charts show daily settlement prices and trade volumes for ASX Energy quarterly base futures, peak futures and cap contracts up to 3 April 217. It is worth noting that although contract prices have increased since the Draft Determination, they have stabilised or even decreased marginally in the last month of the data collection period. Base futures have traded strongly, with total volumes of 7,15 MW (Q3 217), 6,361 MW (Q4 217), 4,766 MW (Q1 218), and 3,916 MW (Q2 218). Peak futures have also traded strongly with 82 MW (Q3 217), 117 MW (Q4 217), 5 MW (Q1 218) and 55 MW (Q2 218). Cap contract trade volumes have also traded strongly with 1,195 MW (Q3 217), 1,3 MW (Q4 217), 92 MW (Q1 218) and 543 MW (Q2 218). FIGURE 4.5 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND BASE FUTURES ASX Energy QLD base Q3 217 ASX Energy QLD base Q Trade date Trade date Volume Settlement price Volume Settlement price ASX Energy QLD base Q1 218 ASX Energy QLD base Q Trade date Trade date Volume Settlement price Volume Settlement price SOURCE: ASX ENERGY DATA UP TO 3 APRIL

21 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) FIGURE 4.6 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND PEAK FUTURES ASX Energy QLD peak Q3 217 ASX Energy QLD peak Q Trade date Trade date Volume Settlement price Volume Settlement price ASX Energy QLD peak Q1 218 ASX Energy QLD peak Q Trade date Trade date Volume Settlement price Volume Settlement price SOURCE: ASX ENERGY DATA UP TO 3 APRIL

22 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 1/7/214 1/1/214 1/1/215 1/4/215 1/7/215 1/1/215 1/1/216 1/4/216 1/7/216 1/1/216 1/1/217 1/4/217 Settlement price $/MWh Trade volume (MW) Settlement price $/MWh Trade volume (MW) FIGURE 4.7 TIME SERIES OF TRADE VOLUME AND PRICE ASX ENERGY QUEENSLAND $3 CAP CONTRACTS ASX Energy QLD cap Q3 217 ASX Energy QLD cap Q Trade date Trade date Volume Settlement price Volume Settlement price ASX Energy QLD cap Q1 218 ASX Energy QLD cap Q Trade date Trade date Volume Settlement price Volume Settlement price SOURCE: ASX ENERGY DATA UP TO 3 APRIL Estimating wholesale spot prices ACIL Allen s proprietary electricity model, PowerMark was run to estimate the hourly pool prices for for the 56 simulations (46 demand and 11 outage sets). Figure 4.8 shows the range of the upper one percent segment of the demand duration curves for the 46 simulated Queensland demand sets resulting from the methodology, along with the historical demands since The simulated demand sets represent the upper, lower and middle of the range of demand duration curves across all 46 simulated sets. It can be seen that the demand duration curves of the simulated demand sets for have a variation similar to that observed over the past five years - that is, the variation between the simulated demand sets does not just occur at the single peak annual demand but across a reasonable portion of the demands within the given simulation 6. This variation in demand contributes to the variation in modelled pool price outcomes as discussed further in this section. 6 The simulated demand sets for are generally higher than the pre observed demand outcomes due to the step increase in demand due to the in-field compression associated with the LNG export projects in Gladstone. 17

23 MW FIGURE 4.8 TOP ONE PERCENT HOURLY DEMANDS QUEENSLAND 1,5 1, 9,5 9, 8,5 8, 7,5 7,.%.2%.4%.6%.8% 1.% 1.2% Percent of time (%) (to April 217) Simulation - Lower Simulation - Mid Simulation - Upper SOURCE: ACIL ALLEN ANALYSIS AND AEMO DATA Figure 4.9 shows the range of the simulated Energex NSLP demand envelopes recent outcomes and covers an average range of about 6 MW across the top one percent of hours. This variation results in the annual load factor 7 of the simulated demand sets ranging between 27 percent and 34 percent compared with a range of 43 percent to 33 percent for the actual NSLP between 28-9 and There has been an observable fall in the load factor in the actual NSLP in recent years due to an increase in penetration of rooftop solar PV panels the increased penetration no longer reduces the peak demand (since the peak demand now occurs between 6:3pm and 8:3pm) but continues to reduce the average metered demand throughout the middle of the day. All other things being equal, the increased peakiness of the load, which is hedged under the methodology, is likely to result in a larger degree of over hedging across the general day-time peak periods, resulting in a larger degree of over hedging overall on an annual basis, which means estimated hedging costs will increase. 7 The load factor is a measure of the peakiness in the half hourly load profile across a given period of time. The annual load factor is the average of the half hourly loads for the given year divided by the maximum of the half hourly loads for that same given year. 18

24 MW FIGURE 4.9 TOP ONE PERCENT HOURLY DEMANDS ENERGEX NSLP 3,1 2,9 2,7 2,5 2,3 2,1 1,9 1,7 1,5.%.2%.4%.6% Percent of time (%).8% 1.% 1.2% Simulation - Lower Simulation - Mid Simulation - Upper SOURCE: ACIL ALLEN ANALYSIS AND AEMO DATA The modelled annual time weighted pool prices (TWP) for Queensland in from the 56 simulations range from a low of $67.47/MWh to a high of $141.39/MWh. This compares with the lowest recorded Queensland TWP in the last 15 years of $28.12/MWh in 25-6 to the highest of $95.69/MWh to date in Figure 4.1 compares the modelled Queensland TWP for the 56 simulations for with the Queensland TWPs from the past 16 years. Although there have been changes to both the supply and demand side of the market, the graph clearly shows that the simulations cover a wide range in potential prices for when compared with the past 16 years of history. The lower part of the distribution of simulated outcomes sits above a number of the actual outcomes (particularly for the earlier years of the market), but by gas prices are projected to be around $11/GJ, compared with $3 - $4/GJ in recent years, and the operating costs of coal plant have increased since the market s inception, and these, coupled with the assumed substantial demand growth due to the LNG terminals, have the effect of influencing an increase in the lower bound of annual price outcomes. ACIL Allen is satisfied that in an aggregate sense the distribution of the 56 simulations for cover an adequately wide range of possible annual pool price outcomes. 19

25 %.3%.7%.1%.14%.17%.21%.24%.27%.31%.34%.38%.41%.45%.48%.51%.55%.58%.62%.65%.68%.72%.75%.79%.82%.86%.89%.92%.96%.99% 1.3% 1.6% 1.1% 1.13% ($/MWh, nominal) $/MWh, nominal FIGURE 4.1 ANNUAL TWP FOR QUEENSLAND FOR 56 SIMULATIONS FOR COMPARED WITH ACTUAL ANNUAL OUTCOMES IN PAST YEARS $16. $14. $12. $1. $8. $6. $4. $2. $. Simulation Historical range Simulated SOURCE: AEMO HISTORIC POOL PRICE DATA AND ACIL ALLEN RESULTS FROM POWERMARK MODELLING Comparing the upper one percent of hourly prices in the simulations with historical spot prices shows the spread of the hourly prices from the simulations also more than adequately covers the historical spread of spot prices, as shown in Figure It is also notable, that as would be expected, the distribution of simulated price outcomes demonstrates a strong positive skewness. FIGURE 4.11 COMPARISON OF UPPER 1 PERCENT TAIL OF SIMULATED HOURLY PRICE DURATION CURVES FOR QUEENSLAND AND HISTORICAL OUTCOMES $16, $14, $12, $1, $8, $6, $4, $2, $ Percent of time Historical range Simulated - Upper Simulated - Lower Simulated - Mid SOURCE: AEMO HISTORIC POOL PRICE DATA AND ACIL ALLEN RESULTS FROM POWERMARK MODELLING ACIL Allen is satisfied that PowerMark has performed adequately in capturing the extent and level of high price events based on the demand and outage inputs for the 56 simulations. The range in annual average contribution to the TWP, of hourly prices above $3/MWh, for the 56 simulations is consistent with those recorded in history as shown in Figure