ISSUE to

Transcription

1 ISSUE 16

2 TABLE OF CONTENTS 1) INTRODUCTION ) THE ROLE OF FORECASTING IN THE NEM ) FUTURE CHALLENGES FACING THE MARKET ) : A YEAR OF TRANSITION? ) Carbon Pricing and a future Emissions Trading Scheme ) Evolution or Revolution: A changing generation mix ) Demand Side Changes the end of the Resources Price Boom? ) LOOKING FORWARD:THE ISSUES OF TOMORROW ) Central Scenario ) Alternative Scenario ) INSIGHTS OF TOMORROW: BASE CASE ) WHOLESALE POOL PRICES ) Response to Queensland s emerging LNG industry ) Understanding the price trace: the impact of reference years and outage patterns ) LARGE-SCALE GENERATION CERTIFICATES ) GENERATION TRENDS ) TRANSMISSION RESULTS ) PROFITABILITY OF NEW PEAKING PLANT ) FUTURE UNCERTAINTY: AN ALTERNATIVE SCENARIO ) WHOLESALE POOL PRICES ) Response to Queensland s extreme pricing forecast ) PROFITABILITY OF NEW ENTRANTS ) Increased QLD Peaking Capacity (1050MW) ) QNI Upgrade and Increased QLD Peaking Capacity (550MW) ) QNI Upgrade and Increased QLD Peaking Capacity (1050MW) ) Increased QLD Peaking and CCGT Capacity (250MW & 500MW respectively) ) QNI Upgrade and Increased QLD CCGT Capacity (500MW) ) GENERATION TRENDS:ALTERNATIVE SCENARIO ) CAUSE AND EFFECT: MAJOR DRIVERS FOR THE FORECAST ) IN REVIEW ) The Supply Side ) The Demand Side ) Pricing Analysis ) Policy Developments ) MULTIPLE REFERENCE YEARS: IGNORANCE IS NO LONGER BLISS ) Demand ) Wind and solar ) KEY CONCLUSIONS AND FUTURE CHALLENGES APPENDIX A) MODELLING METHODOLOGY A.1) 2-4-C MARKET FORECASTING SOFTWARE A.2) TRANSMISSION MODELLING A.3) DEMAND AND ENERGY FORECASTS A.4) MODELLING OF EXISTING PLANT A.5) BIDDING MODELLING A.6) UNIT OUTAGES A.7) MODELLING RENEWABLE ENERGY GENERATORS IN 2-4-C A.8) SOLAR PHOTOVOLTAIC AND SOLAR THERMAL MODELLING A.9) LTIRP MARKET FORECASTING SOFTWARE APPENDIX B) DETAILED MODELLING ASSUMPTIONS B.1) INPUT DATA SOURCES... 83

3 B.2) RENEWABLE ENERGY PLANTING IN RESPONSE TO GOVERNMENT RENEWABLE ENERGY POLICIES B.3) MODELLING THERMAL GENERATORS IN 2-4-C B.4) MODELLING OF RENEWABLE GENERATORS APPENDIX C) MARKET AND FINANCIAL ASSUMPTIONS C.1) MARKET DEVELOPMENT ASSUMPTIONS C.2) ASSUMPTIONS WITH REGARD TO MARKET EXTERNALITIES APPENDIX D) ABOUT ROAM CONSULTING D.1) SERVICES AVAILABLE D.2) PRODUCTS AVAILABLE APPENDIX E) ROAM INSIGHT DATA LIST OF TABLES TABLE 2.1 ROAM INSIGHT SCENARIO COMPARISON... 9 TABLE 3.1 FREQUENCY OF TRANSMISSION CONGESTION AS REPRESENTED BY TRANSMISSION CONSTRAINTS TABLE 3.2 IMPACT OF QNI UPGRADE AND PEAKING CAPACITY ON TRANSMISSION CONGESTION TABLE 4.1 ADDITIONAL CAPACITY REQUIRED ALTERNATIVE SCENARIO TABLE 4.2 PROFITABILITY POTENTIAL FOR NEW ENTRANTS:PEAKING PLANT (1050MW) TABLE 4.3 PROFITABILITY POTENTIAL FOR NEW ENTRANTS: QNI + PEAKING PLANT (550MW) TABLE 4.4 PROFITABILITY POTENTIAL FOR NEW ENTRANTS: QNI + PEAKING PLANT (1050MW) TABLE 4.5 PROFITABILITY POTENTIAL FOR NEW ENTRANTS: OCGT + CCGT PLANT (250MW + 500MW) TABLE 4.6 PROFITABILITY POTENTIAL FOR NEW ENTRANTS: OCGT + CCGT PLANT (250MW + 500MW) TABLE 5.1 FREQUENCY COUNT OF TRADING PERIODS WITH POOL PRICES GREATER THAN OR EQUAL TO $300/MWH BY FINANCIAL YEAR TABLE 5.2 ROOT MEAN SQUARE DIFFERENCE IN DEMAND IN TOP DEMAND PERIODS COMPARED WITH THE AVERAGE TABLE 5.3 AVERAGE WIND GENERATION IN TOP DEMAND PERIODS IN TABLE 5.4 AVERAGE ROOFTOP SOLAR PV GENERATION IN TOP DEMAND PERIODS IN TABLE TH PERCENTILE OF WIND AND ROOFTOP SOLAR GENERATION IN TABLE A.1 GENERATOR OUTAGE MODELLING ASSUMPTIONS TABLE A.1 DESIRABLE CHARACTERISTICS FOR HALF-HOURLY AVAILABLE WIND FARM OUTPUT TIME- SERIES TABLE B.1 ASSUMED LEVELS OF DEMAND SIDE PARTICIPATION LIST OF FIGURES FIGURE 1.1 ROAM S MARKET MODELLING METHODOLOGY... 5 FIGURE 2.1 CARBON PRICE TRAJECTORIES BASE AND ALTERNATIVE SCENARIOS FIGURE 3.1 AVERAGE ANNUAL POOL PRICES FIGURE 3.2 VOLUME WEIGHTED AVERAGE ANNUAL POOL PRICES FIGURE 3.3 IMPACT OF RESPONSE TO LNG ON VOLUME WEIGHTED POOL PRICE:QUEENSLAND FIGURE 3.4 IMPACT OF RESPONSE TO LNG ON VOLUME WEIGHTED POOL PRICE:NEW SOUTH WALES FIGURE 3.5 VOLUME WEIGHTED AVERAGE ANNUAL POOL PRICES: QNI + QLD PEAKING RESPONSE FIGURE 3.6 IMPACT OF OUTAGE ITERATION AND REFERENCE YEAR ON VOLUME WEIGHTED POOL PRICES FIGURE 3.7 LGC PRICE FORECAST FIGURE 3.8 FORECAST POOL REVENUE FOR WIND GENERATORS FIGURE 3.9 MIX OF ENERGY GENERATED (HISTORICAL VERSUS FORECAST) FIGURE 3.10 SHARE OF ENERGY PRODUCTION PER TECHNOLOGY (HISTORICAL VERSUS FORECAST) FIGURE 3.11 REGIONAL SHARE OF ENERGY PRODUCTION (HISTORICAL VERSUS FORECAST) FIGURE 3.12 REGIONAL SHARE OF WIND ENERGY PRODUCTION (HISTORICAL VERSUS FORECAST)... 25

4 FIGURE 3.13 HISTORICAL FREQUENCY OF NEGATIVE OR ZERO PRICE EVENTS PER REGION FIGURE 3.14 REGIONAL SHARE OF WIND ENERGY PRODUCTION (HISTORICAL VERSUS FORECAST) FIGURE 3.15 AVERAGE INTER-REGIONAL TRANSMISSION FLOWS (HISTORICAL VERSUS FORECAST) FIGURE 3.16 FLOW DURATION CURVE OF INTERCONNECTORS IN FIGURE 3.17 FLOW DURATION CURVE COMPARING PRE AND POST QNI UPGRADE FIGURE 3.18 PROFITABILITY OF NEW PEAKING PLANT ACROSS ALL ITERATIONS FIGURE 3.19 PROFITABILITY OF NEW PEAKING PLANT ACROSS ALL ITERATIONS WITH QNI UPGRADE FIGURE 4.1 AVERAGE ANNUAL POOL PRICES ALTERNATIVE SCENARIO FIGURE 4.2 LEVEL OF UNSERVED ENERGY:ALTERNATIVE SCENARIO FIGURE 4.3 VOLUME WEIGHTED POOL PRICE FOR ALTERNATIVE SCENARIO SENSITIVITIES:QUEENSLAND 37 FIGURE 4.4 LEVEL OF USE IN QUEENSLAND FOR EACH ALTERNATIVE SCENARIO SENSITIVITY FIGURE 4.5 MIX OF ENERGY GENERATED (HISTORICAL VERSUS FORECAST) 1050MW OCGT CASE FIGURE 4.6 MIX OF ENERGY GENERATED (HISTORICAL VERSUS FORECAST) OCGT + CCGT CASE FIGURE 4.7 DROUGHT CONDITIONS AFFECT MAJOR HYDRO SYSTEMS IN THE ALTERNATIVE SCENARIOS 43 FIGURE 5.1 COMPARING THE MAXIMUM DEMAND FORECAST (2011 AEMO) VERSUS ACTUAL. 48 FIGURE 5.2 COMPARING THE ANNUAL ENERGY FORECAST (2011 AEMO) VERSUS ACTUAL FIGURE 5.3 EXAMPLE FORWARD CONTRACTS PRICES NEW SOUTH WALES FIGURE 5.4 BREAKDOWN OF RETAIL PRICE (IPART, 2012) FIGURE 5.5 QUEENSLAND TOP 500 DEMANDS FOR 6REFERENCE TRACE FORECASTS FIGURE 5.6 SOUTH AUSTRALIA TOP 500 DEMANDS FOR 6REFERENCE TRACE FORECASTS FIGURE 5.7 CAPACITY FACTOR OF WIND GENERATION IN TOP DEMAND PERIODS FOR EACH REFERENCE YEAR FIGURE 5.8 CAPACITY FACTOR OF ROOFTOP PV GENERATION IN TOP DEMAND PERIODS FOR EACH REFERENCE YEAR FIGURE A C NEM REPRESENTATION FIGURE A.2 OVERVIEW OF ROAM S DEMAND FORECASTING METHODOLOGY FIGURE A.3 COMPONENTS OF FORECASTING DEMAND FOR NSW ON 1NOVEMBER FIGURE A.4 COMPONENTS OF FORECASTING DEMAND FOR NSW ON 1JULY FIGURE A.5 CONSTRUCTING A CONTINUOUS HOURLY TIME-SERIES FROM ACCESS-A NUMERICAL WEATHER PREDICTION (NWP) FORECASTS FIGURE A.6 OBSERVED CAPACITY FACTORS AND WEST PREDICTIONS ON AND FIGURE A.7 COMPARISON OF WEST TIME-OF-DAY TARGETS AND OBSERVATIONS FOR SIX EXISTING WIND FARMS FIGURE A.8 AN EXAMPLE OF A FIVE-DAY PERIOD OF WEST HOURLY WIND FARM GENERATION COMPARED WITH THE OBSERVED GENERATION FOR SNOWTOWN WIND FARM FIGURE A.9 LOCATIONS OF BOM WEATHER STATIONS FIGURE B.1 HISTORICAL AND FORECAST TOTAL NATIVE ANNUAL ENERGY DEMAND MEDIUM AND HIGH SCENARIOS FIGURE B.2 FORECAST TOTAL NATIVE 10% POE PEAK DEMAND MEDIUM AND HIGH SCENARIOS FIGURE B NTNDP GAS PRICE FORECASTS FOR SCENARIO 1, SENSITIVITY FIGURE B.4 RENEWABLE ENERGY PLANTING IN EACH YEAR AND HOW IT MEETS THE LRET FIGURE B.5 WIND ENERGY INSTALLED BY REGION LIST OF BOXES BOX 1 THE PURPOSE OF FORECASTING AND THE ROLE OF ROAM CONSULTING... 6 BOX 2 THE CHALLENGES OF A CHANGING DEMAND... 8 BOX 3 CHALLENGES FACING THE NATIONAL ELECTRICITY MARKET BOX 4 A CAVEAT ON THE ALTERNATIVE RESPONSES TO QLD PRICES BOX 5 THE RESULTS OF THE BASE CASE:A SILVER LINING OR A DARK CLOUD? BOX 6 THE RESULTS OF THE ALTERNATIVE SCENARIO: A REALISTIC ALTERNATIVE TO THE BASE CASE BOX 7 UNDERSTANDING THE CAUSES OF THE CURRENT SYSTEM... 64

5 STRUCTURE OF THIS REPORT Chapter 1 provides a brief introduction to this publication, ROAM Insight, why modelling is a critical part of the energy market, what the key issues are, and a brief introduction to the National Electricity Market. Chapter 2 provides a glimpse into the issues of the near future, in particular of the next six years. The chapter discusses this issue of ROAM Insight in terms of the scenarios and sensitivities investigated to assess the possible challenges in the market in the lead up to 2020, and the reasons for using a scenario based approach to market modelling. Chapter 3 provides a discussion of the modelling outcomes of most general interest. Wholesale pool prices, interconnector flows and generation trends are calculated for the central scenario, as well as the forecast value of Large-scale Generation Certificates. For those with an interest in delving into the details, we recommend a review of the Data Analysis Package which accompanies this report, to allow readers a huge depth of analysis into the underlying data of each of the simulations discussed herein. Chapter 4 provides a discussion on the alternative scenario modelled. The impact of each of the key variables is investigated for each modelled scenario and sensitivity, and potential responses to the market drivers in this scenario are discussed. The chapter provides the reader with an appreciation of the range of possible futures which could eventuate, and therefore the range of outcomes which are plausible over the forecasting period subject to key input assumptions. Chapter 5 provides the reader with an appreciation of how we arrived at this point, and what the major points of interest have been in terms of changes to the supply and demand sides of the market, as well as policy changes which have commenced and are proposed. The reader should obtain a detailed review of the changes and lessons learned for the market during the financial year having read this chapter. Chapter 6 wraps up the report, summarising the key findings of the simulations and the opportunities which exist for the supply side and the demand side over the next six years. We also preview the path to 2020, a point in time where global commitments for emissions reductions are targeted. We consider how the National Electricity Market might evolve in order to contribute to those emissions reductions goals and consider the progress to date in terms of Government policy. The appendices outline the functioning of ROAM s 2-4-C model, and the detailed modelling input assumptions which apply to our ROAM s modelling services. The appendices also include further description of standalone products available to the market. In addition to this report, a Data Package is available which includes the complete set of raw simulation data from which all modelling outcomes shown in this report are derived. This Data Package also enables the reader to obtain the complete half hourly forecast results for all scenarios and sensitivities modelled, allowing alternate analyses and even more in depth analyses than those presented here. All financial values quoted throughout this report are real June 2012 dollars unless otherwise stated.

6 1) INTRODUCTION The National Electricity Market (NEM) is the deregulated, competitive electricity supply market in Queensland, New South Wales, the Australian Capital Territory, Victoria, South Australia and Tasmania. It is managed by the Australian Energy Market Operator (AEMO), regulated by the Australian Energy Regulator (AER), occasionally reformed by the Council of Australian Government s (COAGs) Standing Council on Energy and Resources (SCER) and configured through the rules created by the Australian Energy Market Commission (AEMC). The NEM is a compulsory gross pool market with no capacity payments. All generators must sell and all retailers must buy electricity through the pool. The price is set through the generator bidding process, being the price bid for the last megawatt (MW) of generation dispatched. The price varies every 5 minutes and is settled on a 30 minute basis as the average of the six 5 minute dispatch intervals. Generators and retailers are exposed to significant pool price variations and often manage their risk in a parallel derivatives market. This report focuses on the operation of the pool and forecasts provided relate to the pool (or wholesale) electricity market. In the longer term, contract prices should reflect pool prices. 1.1) THE ROLE OF FORECASTING IN THE NEM Nobody can predict the future with 100% certainty, however an appreciation of what may transpire in the future is essential for making good, informed decisions for investors, policy makers, developers, customers and other interested parties. ROAM Consulting (ROAM) is an energy market modelling business. We forecast the intricacies of the energy market in the finest degree of granularity to ensure that our models produce the most accurate, most reliable forecasts of the future developments of the NEM (and other energy markets), given the assumptions made in developing the input data. At ROAM, we use both a top down and bottom up approach to developing our sophisticated forecasting and analysis models. This includes considering the impacts of all Government policies affecting the energy market (carbon policy, renewable energy policy, wind farm geographic development restrictions, Solar Flagships scheme, GreenPower scheme etc), as well as the current rules and regulations of the NEM (reliability standard, minimum reserve levels, market price cap and market floor price, transmission regulatory decisions etc). Consideration is given to the primary input data of the forecasting models the demand forecasts and the supply side including possible new developments and the best estimate of future generator costs (fuel, capital and other costs). Lastly, but by no means least, ROAM develops bidding strategies for each of the existing and future generators by analysing, from the bottom up, the strategies adopted by generators in the recent past, and reflecting these operational patterns in the bidding strategies going forward, allowing for external influences such as changes to the carbon price and fuel costs. We always use the most credible and most recent input data published from reliable sources so that the outcomes reflect the latest thinking with regards to technical, cost and policy constraints. Our philosophy is to use publicly available, published information as much as possible which has been peer reviewed and industry accepted, but we develop much of the detailed, time sequential data to ensure the model s accuracy and to fill any gaps in the public domain information. For example, we prefer to use the demand forecast targets as published by AEMO, which this year includes a detailed breakdown of the demand and energy targets for various demand categories, and we develop half hourly load traces using our own load forecasting utility. This utility accepts the demand and energy

7 targets and a reference year to grow (or shrink!) the load trace to maintain as much as possible the shape of the underlying load while meeting the demand and energy targets (and our own minimum load criteria). More discussion regarding the load forecasts is provided later in this document. The figure below shows the general process to ROAM s forecasting methodology. Figure 1.1 ROAM s Market Modelling Methodology

8 Box 1 The purpose of forecasting and the role of ROAM Consulting An understanding of what could happen based on sound and robust assumptions and considering a range of potential futures is essential before any decision may be made. This is not unique to the energy industry every major decision that you make should be informed with thorough research and a logical appraisal of all the options. ROAM Consulting is a leading provider of expert services in electricity market modelling to participants in the Australian NEM and electricity markets around the world. At ROAM we apply world s best practice modelling techniques and apply our expertise on the most complex issues in energy market modelling in order to provide the knowledge required by our clients to make well informed decisions. To continue reading ROAM Insight Issue #16 visit roamconsulting.com.au to find out more and order today!

9