2016 ALBERTA ANNUAL ELECTRIC STUDY

Transcription

1 2016 ALBERTA ANNUAL ELECTRIC STUDY THE MARCH OF $0/MWh OFFERS

2 March of the Zero Hours Foreword This document presents a distribution of possible future values for each of demand, supply and the resultant pool price of electricity. These are developed from probabilistically combined scenarios, or collections of assumptions, over a fifteen-year forecast period, expressed in terms of energy and peak demand, for of all consumers in the province of Alberta as well as potential exporters. Each scenario for future demand is convoluted at an hourly level with a matching set of assumptions for electricity supply, including both internal generation and import capacity, to derive probabilistic forecasts of the marginal supply cost of power over time. This is modified by the addition of strategic offer strategies to yield the wholesale electric pool price forecast. The range of possible future outcomes is equal to approximately ±1.3 standard deviation about the mean equivalent to a statistical confidence interval of 80 percent. The P10 and P90 levels represent reasonable upper and lower bounds for these values. Each year, EDCA also presents a special topic. This year s feature chapter, March of the Zero Hours, explores the effects of several market changes on the number of MW of supply being offered in at $0/MWh in any given hour. At the moment, of the domestic fleet nameplate of 16,500 MW, the most ever offered into the power pool is about 13,000 MW, because of seasonal derates and outages. In the hours with the highest coincident outages, the total domestic supply availability can be as low as 8,100 MW. A surprisingly high fraction, as much as 7,000MW of this available supply is regularly offered in at or very near $0/MWh. That amount is comprised of minimum-safe coal generation trying to avoid cycling (2,500 MW), cogens (2,500 MW) whose hosts must have the steam which is generated as a byproduct of electricity generation, wind (1,500 MW), hydro (500 MW) and some biomass, plus whatever imports might be available. In the current supply/demand balance, the number hours when there is barely enough demand to use up even just the $0/MWh offers is growing. With the possible addition of 4,000-8,000 MW more renewables (mostly offered in at $0/MWh), more cogen and distributed generation being built to avoid rapidly inflating transmission and distribution costs and the recent general slowdown in load growth, it is possible that the rapid forward march of $0/MWh supply offers will outpace the sluggish growth in baseload demand, increasing the frequency of hours when the pool price settles at or near $0/MWh. However, as a countervailing effect, the accelerated retirement of coal is subtracting $0/MWh offers from the supply stack. This year s study explores all the mechanisms that influence the race between zero offers and baseload demand. Will $0/MWh supply overtake baseload demand, or will baseload demand pull further away from zero offered supply and increase overall pool price. Will the increase in zero hours be large enough to actually slow down supply additions and lead to reduced reliability and hyper-volatility? This unique longitudinal study will help quantify some of these fundamental questions, with effects profound enough to jeopardize the very existence of the FEOC energy-only Alberta Electricity market. #310, 505 8TH AVENUE SW, CALGARY, ALBERTA, CANADA T2P 1G2 MAIN (403) FAX (403) i

3 An independent consultant, EDC Associates Ltd. (EDCA) prepared this forecast as part of a multi-client study. The information is intended to be used by each participating client for the purpose of business planning for long-term electricity related initiatives in Alberta that may currently be under evaluation. All assumptions, models, processes, historical electric energy data and other public or proprietary data gathered by EDCA, as an ongoing concern, relating to economic, demographic, technological and other factors which affect the utilization of electric energy that have been used to develop the results discussed in this study, are the proprietary property of EDC Associates Ltd., except where noted. Copyright EDC Associates Ltd., 2016 No section of this study may be copied or reproduced in whole or in part without the written consent of EDC Associates Ltd. Contributors The project manager would like to acknowledge the considerable input of several staff for their specific contributions to the research, computer modeling, graphics, writing and final editing of this report: Allen Crowley Alex Markowski Marvin Mah Francisco Chavez Overall Project Manager and Editor, Special Chapter text Electricity price research and forecasting, generation supply review, special chapter research, document assembly, analysis and text Energy demand, oil price and gas price forecast analysis and text Future industrial demand and cogen projects, special chapter research These individuals have spent many hours to deliver this report. We hope you find it both informative and useful in your particular endeavors in the Alberta electric energy industry. COVER ART: Each year, EDCA chooses a theme for the report, based on the Special Chapter. This year, EDCA presents an analysis of the high fraction of energy actually offered in at $0/MWh. This has not been a large concern to date, resulting in only a handful of hours each year where the price actually settled at $0/MWh. However, with the announced Climate Leadership Plan, the potential for a very large increase in $0/MWh offers, especially from intermittent renewable sources, may reach the point that generation developers cannot afford to build enough baseload generation to always meet peak conditions, leading to reliability concerns. Of course, this all depends on how aggressively renewables are added and how soon load growth returns to positive territory. The report cover is a parody of the best-selling novel/film, Zero Dark Thirty, and shows the North American night sky during the great Northeast Electricity Blackout of The Zero reference alludes both to the increase in zero emissions generators and the growth in $0/MWh offers and settlement hours. Dark Thirty reflects the increased potential for reduced reliability and the resultant rolling 30-minute load curtailments. ii

4 Disclaimer The information provided in this report is of a forecast nature and is based on what is believed to be sound and reasonable methodologies and assumptions, however cannot be warranted or guaranteed with respect to accuracy. Therefore, any use of the information by the reader or other recipient shall be at the sole risk and responsibility of such reader or recipient. The information provided in this report and the facts upon which the information is based may change at any time without notice subject to market conditions and the assumptions made thereto. EDC Associates Ltd. is under no obligation to update the information or to provide more complete or accurate information if and when it becomes available. EDC Associates Ltd. expressly disclaims and takes no responsibility and shall not be liable for any financial or economic decisions or market positions taken by any person based in any way on information presented in this report, for any interpretation or misunderstanding of any such information on the part of any person or for any losses, costs or other damages whatsoever and howsoever caused in connection with any use of such information, including all losses, costs or other damages such as consequential or indirect losses, loss of revenue, loss of expected profit or loss of income, whether or not as a result of any negligent act or omission by EDC Associates Ltd. iii

5 Table of Contents The Year in Review... 1 Demand... 2 Transmission Infrastructure... 3 Transmission Loss Factors... 3 Supply... 3 Power Prices... 4 Climate Leadership Plan and SGER Changes... 4 Coal Compensation Negotiations Initiated... 5 Power Purchase Agreements Terminated... 5 Maxim Power Temporarily Suspends HR Milner Production... 6 Study Design and Scope... 6 Risk Analysis Methodology... 7 Short-run Risk Variables... 7 Expanded Long-Run Risk Variables... 7 Chapter Layout... 9 Executive Summary Macroeconomics Electric Load Forecasts When the Wind Blows Supply Resource Development Electricity Price Forecasts Levelized Costs Macroeconomics US Economy Canadian Economic Outlook Alberta Macroeconomic Outlook Crude Oil and Natural Gas Market Outlook World Oil Supply and Demand Outlook Crude Oil Price Forecasts Alberta Crude Oil Production Natural Gas Forecast US Natural Gas Market Outlook Natural Gas Price Forecast Alberta Natural Gas Outlook Crude Oil and Natural Gas Price Distributions Large Industrial Project Profile Macroeconomic Summary Demand Forecast Electric Energy & Demand Forecast Alberta Internal Load versus AIES Demand iv

6 Energy and Demand Growth Residential Energy Sales Commercial Energy Sales Farm & Irrigation Energy Sales Industrial Load Growth Behind-the-Fence Load Forecast Transmission and Distribution Losses Export Sales Electricity Demand Forecast Summary March of the Zero Hours Purpose of Study Layout Why Have $/MWh Offers Missing Money in an Energy Only Market Increases in Zero Offers Decreases in Zero Offers Increase in Baseload Demand Joint Probability of Low Load and Zero Offers Modeling Methodologies Coal Retirement Schedule Cogen Expansion Wind Expansion Imports Model Mechanics Sensitivities Results Frequency of $0/MWh Settled Hours Effect on Offer Behaviour Effect on Pool Price Effect on Price Volatility Reduced Incentive for Generation Additions Reliability Effects Summary and Conclusions Supply Resource Development Current Generation Supply Installed Capacity Forecast Unit Retirements Green House Gas Policy Retirements Fuel Supply & Generation Technology Natural Gas-Fired Generation Coal-Fired Generation Hydro Generation Wind Generation Nuclear Generation Biomass Generation Interconnections v

7 Levelized Unit Cost of Electricity Generation Supply Methodology Project Base Probabilities Project Capacity Project Timing Interaction from Other Stochastic Variables Future Supply Resources Base Forecast Assumptions Total MW Capacity Break Down Stochastic Range of Supply Assumptions Relative Market Share Base Supply/Demand Balance Reserve Margin Supply Resource Forecast Summary Electricity Price Forecasts Historical Alberta Pool Prices Forecast Assumptions Imports and Exports Offer Strategy Supply Cushion Green House Gas Emission Costs Scheduled Supply Outages Electricity Price Forecast Results Forecast Energy Production (Base Assumptions) Forecast Price Levels Forecast Price Distributions Market Heat Rate Forecast Results Forecast Market Heat Rate Levels Forecast Market Heat Rate Distributions Forecast Summary Appendices Appendix A Risk Analysis Methodology Stochastic Modeling Concepts Short-run Risk Variables Expanded Long-run Risk Variables Defining & Interpreting P10 and P90 Values Appendix B - Generating Unit Statistics Appendix C - P10 Data Tables Appendix D - P90 Data Tables vi

8 Table List Table 1 - Electric Energy Forecast Annual Compound Growth Rates Table 2 - US Real GDP Growth Rate Results Table 3 - Canadian Real GDP Growth Rate Results Table 4 - Canadian Forecast Inflation Rate Results Table 5 - Interest Rate Results Table 6 - Canadian Unemployment Rate Results Table 7 - Exchange Rate Results Table 8 - Inventory of Major Alberta Projects Table 9 - Alberta Real GDP Growth Rate Results Table 10 - Alberta Summary Large Industrial Project Profile Table 11 - Alberta Electricity Exports Table 12 - Electric Energy Forecast Annual Compound Growth Rates Table 13 - Annual Energy Forecast by Consuming Sector (P10) Table 14 - Annual Energy Forecast by Consuming Sector (P90) Table 15 - Minimum Stable Generation by Coal Unit Table 16 - Assumed Stages of Price-Responsive Load Table 17 - Results from MSA Pivotal Supply Study in Table 18 - Increase in Pool Price During Pivotal Supply Events, by Number of Pivotal Suppliers Table Generation Changes Table 20 - Current Installed Capacity Base (12/31/2015) Table 21 - Net-to-Grid Capacity & Energy Production by Fuel Type (2015) Table 22 - Coal-Fired Retirement Years Table 23 - Generation Retirement Assumptions Table 24 - Summary of Attributes of Various Generation Technologies Table 25 - Historical Wind Fleet Premium/Discounts Table 26 - Yearly Spot Prices ( ) Table 27 - Pool Price Statistics ( ) Table 28 - Annual Average Heat Rate, Annual Changes in AECO-C and Pool Price ( ) Table 29 - Typical Marginal Fuel Cost Table 30 - Distribution of Coal s Offer Strategies for Table 31 - Monthly Domestic AIES Demand (2015) Table 32 - Recent Alberta Coal Unit Uprates Table 33 - Energy Production by Fuel Source (2016, 2030) Table 34 - Non- Regulated Generation Projects (Completed) Table 35 - Announced Non-Wind Generation Projects and Probabilities Base Case Assumptions Table 36 - Announced Wind Generation Projects and Probabilities Base Case Assumptions Table 37 - Economic Assumptions & Commodity Prices (P10) Table 38 - Population, Household & Employment Statistics for Alberta (P10) Table 39 - Alberta Major Economic Variables (P10) Table 40 - Alberta Electric Energy Sales and Peak Demand Forecast by Month (P10) Table 41 - Alberta Export and Import Forecast by Month (P10) Table On-Peak and Corresponding Off-Peak Pricing (P10) Table On-Peak and Corresponding Off-Peak Pricing (P10) Table 44 - Average Alberta Electricity Power Pool & Natural Gas Price Forecast by Month (P10) Table and 6 16 On-Peak Heat Rates (P10) Table 46 - All Hours System Heat Rate (P10) Table 47 - Economic Assumptions & Commodity Prices (P90) Table 48 - Population, Household & Employment Statistics for Alberta (P90) Table 49 - Alberta Major Economic Variables (P90) Table 50 - Alberta Electric Energy Sales and Peak Demand Forecast by Month (P90) Table 51 - Alberta Export and Import Forecast by Month (P90) Table On-Peak and Corresponding Off-Peak Pricing (P90) Table On-Peak and Corresponding Off-Peak Pricing (P90) Table 54 - Average Alberta Electricity Power Pool & Natural Gas Price Forecast by Month (P90) Table and 6 16 On-Peak Heat Rates (P90) Table 56 - Alberta Implied Marginal Heat Rate (P90) Table 57 - All Stochastic Forecast Seeds vii

9 Figures List Figure 1 - Typical Risk Analysis Process Flow Diagram... 8 Figure 2 - Alberta Electric Energy Sales Forecast Comparison Figure 3 - Alberta Electric Demand Forecast Comparison Figure Minimum and Maximum Availability Figure 5 - Alberta Electricity Pool Price Forecast Summary Figure 6 - Chain of Model Assumptions (Appendix C (P10) and D (P90)) Figure 7 - US Real GDP Historical Growth and Estimated Distribution Figure 8 - Random Strip of US Real GDP Growth Rates Resulting in the P10 Forecast Figure 9 - Canadian and US Historical Real GDP Growth Rates Figure 10 - Examples of US and Canadian Real GDP Modeled Growth Rates Figure 11 - Forecasted Canadian Real GDP Growth and Inflation Figure 12 - Canadian Interest Rate and Real GDP Random Forecasts Example Figure 13 - Historical Unemployment Rate and Real GDP Growth Figure 14 - Unemployment and Real GDP Stochastic Forecast Example Figure 15 - Unemployment Forecast Distribution Figure 16 - WTI Crude Oil Price Base Assumptions ($Real 2002 ) Figure 17 - Histogram of WTI Publicly Available Long-Term Price Forecasts Figure 18 - WTI Historical Price Distribution ( ) Figure 19 - WTI Stochastic Price Forecast Example Figure 20 - Crude Oil Price Scenarios at P10 and P90 (WTI Cushing US$/bbl) (Real and Nominal) Figure 21 - Crude Oil Price Scenarios at Base, P10 and P90 (WTI Cushing US$/bbl) Figure 22 - Base Case and Example of Edmonton Oil Prices and Heavy Differential (per bbl) Figure 23 - Conventional Crude Oil Production Forecast Figure 24 - Synthetic Crude Oil Production Figure 25 - Non-Upgraded Bitumen Production Figure 26 - Alberta Crude Oil and Equivalent Production by Type (P10) Figure 27 - Alberta Crude Oil and Equivalent Production by Type (P90) Figure 28 - US Working Natural Gas in Underground Storage and Natural Gas Prices Figure 29 - Projected US Working Natural Gas in Underground Storage Figure 30 - WTI vs. AECO-C ($/GJ of Equivalent Heat Content) Random Example Figure 31 - Natural Gas Price Results (P10 and P90) Figure 32 - Natural Gas Price Results (Base, P10 and P90) Figure 33 - Alberta Coal Bed Methane Production Figure 34 - Alberta Conventional and CBM Natural Gas Production (P10) Figure 35 - Alberta Conventional and CBM Natural Gas Production (P90) Figure WTI Price Forecast Distribution (2002 US$/bbl) Figure Natural Gas Price Forecast Distribution (2002 $/GJ) Figure 38 - Demand Build-Up Figure 39 - AIES and AIL Energy Definitions (2015) Figure 40 - Example of AIL, AIES and Exports in Real Time Figure 41 - Domestic AIES Demand Seasonality (January to December) Figure 42 - Normal Heating (HDD), Cooling Degree Days (CDD) and Average Daylight Hours Figure 43 - Typical AIES Demand Diurnal Patterns by Month (1.000=Annual Average MW) Figure 44 - Residential Average Customer Usage (kwh/year per Customer) Figure 45 - Residential Energy Forecast Figure 46 - Commercial Energy Forecast Figure 47 - Farm and Irrigation Energy Forecast Figure 48 - AIES Industrial Energy Forecast Figure 49 - Behind-the-Fence Energy Forecast Figure 50 - AB>BC Export Duration Curves of Available Transfer Capacity (ATC) (2015) Figure 51 - AIES Export Forecast Figure 52 - Alberta Electric Energy Sales Forecast Comparison Figure 53 - Alberta Electric Demand Forecast Comparison Figure 54 - Alberta Pool Price Duration Curve ( ) Figure 55 - Alberta Pool Price Duration Curve ( ) Figure 56 Alberta Supply/Demand Balance Figure 57 - Threshold for Provoking Competitive Response viii

10 Figure 58 - Percentage of Hours with Pivotal Suppliers Figure 59 - Fraction of Hours as Pivotal Supplier or Sole Pivotal Supplier ( ) Figure 60 - Duration Curve of Largest Pivotal Supplier vs Next Largest Figure 61 - Pivotal Supplier by Market Participant Figure 62 - Pivotal Hours and Pool Price vs. Number of Concurrent Pivotal Suppliers Figure 63 - Effect of Pivotal Supplier Profit Maximization on Average Annual Pool Price in Figure 64 - Effect of Pivotal Supplier Profit Maximization on Average Annual Pool Price in Figure 65 - Effect of Pivotal Supplier Profit Maximization on Average Monthly Pool Price in Figure 66 - Effect of Pivotal Supplier Profit Maximization on Average Monthly Pool Price in Figure 67 - Effect of Delaying Before Acting on Pivotal Supply Opportunities in Figure 68 - Effect of Delaying Before Acting on Pivotal Supply Opportunities in Figure 69 - Effect of Desired Size of Pivotal Supply Before Acting on Opportunity in Figure 70 - Effect of Desired Size of Pivotal Supply Before Acting on Opportunity in Figure 71 - Effect of % Hours Supplier Acts on Pivotal Supply Opportunity in Figure 72 - Effect of % Hours Supplier Acts on Pivotal Supply Opportunity in Figure 73 - Illustration of Bidding Up in Non-Pivotal Hour Figure 74 - Effect of Profit Maximization on Pool Price in Pivotal and Non-Pivotal Hours in Figure 75 - Effect of % of Hour Suppliers Act on Profit Maximizing Opportunities in Figure 76 - Historical Net-to-Grid Capacity by Fuel Type Figure 77 - Alberta Historical Supply/Demand Balance & Reserve Margin Figure 78 - Northern vs Southern Wind Fleet Discount ( ) Figure 79 - Comparative Raw Levelized Unit Costs (excluding GHG Costs) Figure 80 - Levelized Unit Costs with Differing GHG Offset Prices Figure 81 - Generation Supply Methodology Figure 82 - Annual Net Capacity Additions by Fuel Type (Base Assumptions) Figure 83 - Total Net-to-Grid Generation by Fuel Type (Base Assumptions) Figure 84 - Total Net-to-Grid Generation Capacity (P10, P50 and P90) Figure 85 - Relative Market Share by Fuel/Technology Type (1996, 2015, and 2030) Figure 86 - Relative Market Share by Key Fuel Type (P10, P50 and P90) Figure 87 - Annual Supply/Demand Balance Forecast (Base, P10 and P90) Figure 88 - Simple and Discounted Reserve Margin (Base Assumptions) Figure 89 - Simple and Adjusted Reserve Margin (Base, P10 and P90) Figure 90 - Historical Daily Alberta Electricity Pool Prices ( ) Figure to 2015 Alberta Electricity Prices and Rolling Average Prices Figure 92 - Pool Price Duration Curve ( ) Figure to 2015 Average Heat Rates and Rolling Average Heat Rate Figure 94 - Road Map of Price Forecast Methodology Figure System Import/Export ATC Figure 96 - Historical Imports/Exports from BC, MATL, SK and CMH ( ) Figure 97 - Sample Alberta Merit Order Curve Figure 98 - Distribution of Offer Strategies for Figure 99 - Changes in Coal s Offer Strategies (2015 vs 2014) Figure Minimum & Maximum Availability Figure Daily Coal-Fired Generation (2015) Figure Supply Cushion Frequency and Average Pool Price ( ) Figure Range of GHG Emissions Cost Forecast Stochastics ($/t) Figure /2017 Forecast Maintenance Outages Figure Forecast Generation Output by Plant Type (Base Assumptions) Figure Forecast Capacity Utilization (Base Assumptions) Figure Alberta Electricity Pool Price Forecast Figure Alberta Real 2015$ Electricity Price Forecast Figure Pool Price Distribution for Highest and Lowest Real Pool Price Year Figure Pool Price Forecast Distribution for Widest and Narrowest Dispersion Figure Alberta Marginal System Heat Rate Forecast Figure Lowest and Highest Average Heat Rate Years Figure Narrowest and Widest Heat Rate Dispersion Figure Alberta Electricity Pool Price Forecast Summary Figure Stochastic Forecast Distribution Concept Figure Typical Risk Analysis Process Flow Diagram Figure Final versus Sub-Distributions Figure Schematic of Stochastic Results ix

11 1 Introduction This report is the eighteenth extensive issue on the Alberta electricity market. Since the first issue in 1998, this series of reports has kept readers up to date with ever-changing Alberta electricity market fundamentals, regulatory events and policy changes. It also looks beyond the Alberta jurisdictional boundary, analyzing key geo-political and international economic events that may influence the Alberta electricity market. The Year in Review The Alberta economy, boasting robust GDP growth in 2013 and early 2014, took a severe body-blow in mid-july 2014, as oil prices began a steep decline from over $105/bbl to the current sub-$40/bbl mark. The continuing bleak oil price outlook in 2015 provoked a series of layoff announcements and project deferrals, as inventories continued to grow in spite of steep reductions in rigs. The housing market inventory began to expand and building permits dropped almost 40%. Consumption of electricity weakened in 2015, recording no annual growth (0.0%), partly from the mildest winter weather in decades and from weaker oil and gas and industrial growth. Inmigration in 2015 (38,700/year, preliminary) was down by over 25,000 (40%) from the 2014 level. WTI crude oil decreased in 2015 to average US$48.67/bbl, compared to the US$93.82/bbl in From a high of US$61.36/bbl on June 10, oil fell to a low of US$34.55/bbl on December 21 and continued down to US$26.19/bbl by mid-february, as already record inventories continue to rise in spite of a 70% decline in rigs. Only two years ago, 2013 power prices had averaged $80.19/MWh as supply scarcity pushed prices into triple digit territory for several consecutive months. Since then, power prices have collapsed to a multi-year low in 2015 ($33.34/MWh), as depressed oil prices stalled Alberta s appetite for power at the same time that several large supply additions commissioned. Imperial Oil s Kearl and Nabiye cogens commissioned in January, followed by ENMAX/Capital Power s 873 MW Shepard Energy Center in March. These and other gas-fired units benefited from weak natural gas prices, with healthy levels of production keeping AECO-C natural gas prices under $3/GJ for most of the year. Power Chapter 1 1

12 prices dipped to $20.52/MWh in April, the lowest priced month since the market fully deregulated in The collection of additional newsworthy events are presented in categories corresponding to each section of the Annual Report. Demand In 2015, the Canadian and especially the Albertan economies continued to weaken, with decline in the resource sector offset by growth in the non-resource sectors. US maintained growth with increased private spending and consumption, improved housing starts and prices, and reduced unemployment. China s economy continued its gradual slowing, causing more reaction from the stock market. With ongoing increases in US domestic oil supply, a weaker view of global demand and OPEC s announced intention to maintain their production levels, the oil price floated in the $50-$60 range in the first half of 2015 before descending to mid $30s by year-end. The oil and gas sector continues to be the most important driver of the Alberta s economy in terms of development, business and government investments, employment and operations. After five years of positive provincial GDP growth (4.5% in 2010, 5.7% in 2011, 4.5% in 2012, 3.8% in 2013 and 4.8% in 2014) and energy prices stabilized around the $95/bbl mark, the oil price finally collapsed in mid-2014, spawning rounds of layoff announcements and project delays. The Canadian dollar followed suit, dropping from 0.94 US$/C$ in July to the current 0.75 US$/C$ level. The saga of the TransCanada Keystone pipeline application ended in November 2015 when the Obama administration rejected the project, stating that Keystone will not serve the national interests of the United States. TransCanada then filed legal action against the Obama administration under the North American Free Trade Agreement (NAFTA) to recover $15 billion US in costs and damages on that basis that the denial was not justified. With the demise of Keystone, prospects for TransCanada s Energy East are improving, but that project has its own issues with federal and provincial partners. Concern about the take-away capacity of planned oilsands developments has been at least temporarily abated by a huge increase in the oil-by-rail alternative. Domestic AIES demand peaked at 9,032 MW on January 5 th, hour ending (HE) 18. This was a small increase of 78 MW (0.9%) over the previous record high of 8,954 MW set December AIL demand set the current record of 11,229 MW on January 5 th, 2015 HE 18, up 60 MW (0.5%) from the December 2014 peak demand of 11,169. In 2015, AIES energy sales (i.e. domestic and export energy traded through the Alberta market) reached 62,606 GWh, while AIL energy sales tallied 80,257 GWh. The difference between AIL and AIES growth is largely attributed to the additional behind-the-fence generation at oil sands mines, in-situ bitumen and upgrader projects. Chapter 1 2

13 Several large generator additions, such as Shepard (873 MW combined cycle), Blackspring Ridge (300 MW wind) and Imperial s Nabiye (170MW) and Kearl (85 MW) cogens, increased supply much faster than the increase in AIES load. New cogen also widens the gap between AIL and AIES demand, further prolonging depressed pool prices. Transmission Infrastructure The two largest government-mandated Critical Transmission Infrastructure (CTI) projects, ATCO s Eastern Alberta Transmission Line (EATL) and AltaLink s 500 km, 500 KV HVDC Western counterpart (WATL, now $1.7B) have been energized. After Alberta Utilities Commission (AUC) approval in November 2012 (Decision No ), construction of the $1.8 billion, 485 km 500 kv HVDC EATL line and two converter stations began in December 2012 and was completed in February 2015, with reclamation along the line finalized in November and the line energized December 16, AltaLink completed the 350 km, 500kV HVDC WATL line on December 10, The two lines will reduce line losses and significantly increase system reliability. Southbound flows from Alberta to Montana have drifted up to occasionally top the 250 MW mark, with a total southbound export out of Alberta of 89 GWh in 2015, 30% of the northbound MATL imports of 296 GWh from Montana into Alberta. Transmission Loss Factors In its Order 182, the Commission directed the AESO to change the current Line Loss Rule to implement the Commission s findings in this decision. On February 1, 2016, the AESO filed an implementation plan to develop a revised line loss rule that will operationalize the AUC rulings. The plan promises a process, changed line loss rules and draft 2017 loss factors by July 1, followed by a hearing from July to September in time for a January 1, 2017 implementation. This will be followed by the Module C hearing, likely during November and December 2016, in which the process and values for the retroactive financial remedy to all generators back to 2006 will be determined and approved. The AESO is currently developing a process to operationalize a new AUC-created rule for aggregating the output of generators on the same bus. The AUC ruled that such aggregations would be allowed as long as the units have a single owner, are on the same bus and offer in their allowed seven price/quantity pairs on that aggregated basis. The level of aggregation can be changed annually. With the new ILF loss factor calculation process, the greater the aggregation on a bus, the lower the incremental loss factors will be. This creates several fairness problems between units on the same bus with different owners. Supply ENMAX/Capital Power s 873 MW Shepard combined-cycle plant, the largest ever addition to Alberta s fleet, officially reached commercial operation in March 2015 after having sent intermittent power to the grid since September Imperial Chapter 1 3

14 Oil s Nabiye (195 MW) and Kearl (84 MW) cogens commissioned in January. Although these units use the majority of their power on-site, they have the effect of reducing AIES demand (i.e., the amount of demand traded across the grid, ultimately setting the pool price). Several smaller additions also went into service, such as Northstone s JL Landry, an additional unit at Harmattan, Harvest Bilbo Gas Plant, Mazeppa Gas Plant, Sunshine West Ells and Ralston. Combustion tuning/control system upgrades at Exelon s Northern Prairie Power Project effected a 12 MW increase in capacity. West Fraser Mill s Slave Lake Biomethanation facility began operation and Sundance #3 s uprate raised the unit s capacity from 362 to 368 MW. Towards the end of the year BluEarth s Bull Creek wind farm began producing power, as did the Sunshine Colony wind farm. Although not technically an addition, the AESO increased the combined BC/MATL summer ATC (effective June 2015) and winter ATC (effective November 2015) to 1,110 MW, regardless of Alberta Internal Load, although it may be reduced if the available load under load shed service for imports is insufficient. Power Prices Favorable temperatures, weak natural gas prices, healthy levels of generation and a slow-down in Alberta s economic growth conspired to drive 2015 annual power prices to an all-time low of $33.34/MWh. Prices settled below $35/MWh in 10 of the months, dipping to $20.52/MWh in April the lowest month since deregulation in June, at $97.31/MWh, was the most aggressively priced month of the year, with repeated strings of triple digit pricing in the latter-half of the month as hot summer temperatures increased cooling demand in the face of weak wind and other generation output. As low as prices were in 2015, they have become even softer in early February and March both fell below $20/MWh, with no hour in either month exceeding $50/MWh. Similar to 2015, warmer-than-average temperatures, a flush supply cushion and very low AECO-C natural gas prices (dipping into the $1/GJ range) have all contributed. Climate Leadership Plan and SGER Changes On Sunday November 22, 2015, the provincial NDP government announced their so-called Climate Leadership Plan, calling for a higher tax on carbon, phasing out of all coal by 2030 and growing renewables to account for up to 30% of total electricity production by The government s climate change initiative is a vast departure from the Business as Usual scenario, which would otherwise retain some coal capacity until the end of 2061, with renewables accounting for only 10% of total electricity production by Many of the policy choices for specific design elements of the Climate Leadership Plan scenario are still not precisely specified, even the exact interpretation and implementation of the target quantity of renewables. EDCA identified about a dozen different elements, such as the shape of the accelerated coal retirement schedule, the definition of the target renewables penetration parameter, the Chapter 1 4

15 fraction of retired coal that would be replaced by renewables, the mix of renewables to be incented (percent of wind, solar, hydro, biomass, etc.), the style of Renewable Energy Credits (RECs), the likely load demand growth, the design of the auction of these credits, as well as the method and amount of coal compensation paid. Each element could be set at any number of different levels and the different policy elements could be combined in an infinite number of ways. Until policy specifications are more certain, EDCA will present two forecast paths in this Quarterly report. The first is a Business as Usual scenario that assumes the NDP wishes to fully preserve Alberta s Fair, Efficient and Openly Competitive (FEOC) electricity market and does not pay out-of-market incentives for renewables. SGER is left in place, albeit with the increased stringency of the June 2016 rules, coal retirements occur as per their respective existing federal dates and renewable growth occurs through pure market forces (except for the AESO s first call for power which is assumed will occur). The second forecast path is a Climate Leadership Scenario which quantifies one possible combination of policy outcomes which include a mandated target % renewables, incented by adequate out-ofmarket payments. Alberta s current mechanism for pricing carbon is the Specified Gas Emitters Regulation (SGER). Originally established in 2007, it required facilities that emit more than 100,000 tonnes/year of greenhouse gases to reduce their emission intensity by 12% from a facility-specific baseline intensity, or pay $15/t for any non-compliance. Shortly after the NDP formed the new government, they increased the target reduction percentage to 15% (effective January 2016) and then to 20% (effective January 2017) and raised the cost per tonne to $20/t (effective January 2016) and then to $30/t (effective January 2017). However, as part of the November 2015 Climate Leadership Plan, the government announced a performance based tax on carbon to be introduced in 2018, with payment required on each tonne of emissions in excess of what would have been emitted by a good as best gas intensity level generator. Coal-fired units would be the hardest hit, with a 1.0 t/mwh unit s environmental compliance costs roughly tripling (from $6/MWh to over $18/MWh). To date, only the SGER change has been passed into law. Coal Compensation Negotiations Initiated In March 2016, the provincial government took the first steps towards accelerating coal retirements by engaging Terry Boston, former president/ceo of PJM Interconnection, for a $600,000 fee, to work with coal-fired electricity generators, the AESO and the Government of Alberta to develop options and compensation to phase out emissions from coal-fired generation by Power Purchase Agreements Terminated Battle River #3 and #4 were the most recent PPAs to naturally expire, at the end of 2013 with offer control reverting from ENMAX to ATCO. The Sundance #1 and #2 PPAs are the next to naturally expire at the end of 2017, although their Chapter 1 5

16 retirement dates per the federal schedule is the end of However, in January 2016, ENMAX announced that it was terminating its Battle River #5 contract, stating that the June, 2015 NDP change in SGER parameters constituted a Change in Law as contemplated in Section 4.3 of the PPA, and that such a change rendered their contracts unprofitable, or more unprofitable. The Change in Law provision is written such that it is the single contingency deemed to be a legitimate cause for termination without penalty. TransCanada followed ENMAX s lead and terminated its Sundance A, B and Sheerness PPAs in March 2016, under the same clause. Several weeks later, Capital Power terminated their Sundance C PPA. When the PPAs are terminated, the offer control returns to the Balancing Pool but the unit is not taken out of service. Although the PPAs were written with the clear intent to allow their termination for precisely this type of change in law, the Alberta government will nonetheless investigate if there is any way that the terminations can be disallowed. The balancing Pool has the choice to run the PPAs as though they were the Buyer, or return them to their owners and pay a termination penalty equal to the outstanding book value of the asset. Assuming the Balancing Pool retains offer control of all of the coal-fired PPA units (4,485 MW, including Keephills #1 and #2 and Genesee #1 and #2 but not including 227 MW of Excess Capacity currently controlled by the Owners or 200 MW of Genesee strips) and the market at the end of 2016 will have roughly 15,750 MW of dispatchable capacity (which offer control is compared against), the Balancing Pool would hold control of just over 28% of the market. However, this total could be brought down depending on future generation additions, the subsequent sale of strips and delegating offer control (e.g., ENMAX continues to operate Battle River #5). This will still make the Balancing Pool pivotal in a high fraction of hours. Maxim Power Temporarily Suspends HR Milner Production On March 23, 2016, Maxim Power announced 1 it was temporarily suspending electricity production at HR Milner. The decision was made in response to the record low Alberta power prices, although the company intends to resume operations as market conditions improve, perhaps as early as Q2. HR Milner is one of the oldest and most expensive coal units (not mine-mouth coal). Study Design and Scope This year s Annual Study, as all those in the past, continues to focus on the longterm Alberta electric industry market fundamentals, along with other influencing factors. EDCA deploys a collection of integrated forecasting models to assess future market supply, demand, emissions and price dynamics. Besides developing these forecasts through the use of scenario analysis, using a collection of discrete input assumptions that defined a single deterministic most likely forecast, 1 Chapter 1 6

17 EDCA also incorporates probabilistic Monte Carlo techniques. These techniques allow the reader to also assess and quantify the potential range of future market supply, demand, price and emissions dynamics by describing the various input assumptions and outputs as probabilistic ranges rather than discrete events. The report presents expected P10 and P90 bands (a 10% or 90% probability that the actual value will be lower or higher than the expected mean value). The reader can also derive his own appropriate confidence interval around the mean value of the outcome and better quantify the risk associated with the forecast result. This is not to say that deterministic modeling is outdated or incorrect, but simply that stochastic modeling through the use of Monte Carlo techniques describes the inherent risks more fully. The pool price distribution is intended to represent the range of future possible outcomes resulting from a range in future electricity supply, demand and market rules input assumptions (see Appendix A for a full description and interpretation of the EDCA proprietary probabilistic process, its applications and limitations). Risk Analysis Methodology In order to quantify the potential range of the deviations in the price profile, EDCA uses Monte Carlo techniques to incorporate several key risk elements into its generation dispatch and energy price forecast model. The model convolutes the various stochastic risk elements collectively, to arrive at a composite price profile. Alternatively, each input can be varied in isolation to assess its importance, or sensitivity, to the overall variability. Variables are categorized as either short-run or long-run. Short-run Risk Variables The short-term risk variables reflect the typical range of demand and supply variance resulting from short-run influences such as weather, sunlight hours, work week, intra-month natural gas price volatility, variability of wind energy production and forced outages of generation units and tie-lines. These parameters are varied about historical mean values and typically produce a small dispersion of the total price distribution. Expanded Long-Run Risk Variables The EDCA models are also designed to assess the impact of significant changes to longer term assumptions which can potentially produce a much more dramatic impact in the future electricity price forecast (see Figure 1). Chapter 1 7

18 HELP E D C A S S O C I A T E S L T D. Figure 1 - Typical Risk Analysis Process Flow Diagram Long-term Monte-Carlo Risk Variables Nat Gas Volatility Unit Availability Weather Short-term Monte- Carlo Inputs Wind Energy Production Profile Energy Demand (MWh) Supply Additions (MW, Timing) +/-1% to +/-5% +/-25%, +/-1 Yr Generation Supply Additions & Other Base Case Assumptions Unit Bidding Behaviors Environmental Costs (On/Off) Generation Dispatch & Energy Pricing Model Natural Gas Price ($/GJ) -$2/ to $6/GJ Outputs Mid-C Import Pricing (Heat Rate GJ/MWh) +/-1.25GJ/MWh Unit MWh s Pool Price Distribution EDCA has identified five key assumptions, in rough order of impact, that typically represent the most significant amount of risk in any future electricity price forecast: generation timing and probability, potential environmental costs, natural gas and other fuel price changes, AIES energy and demand growth and Mid-C market prices. EDCA also makes specific assumptions such as strategic bidding behaviors and planned maintenance scheduling on the supply-side, and demand responsiveness on the load side. Typically, each of these inputs is varied on a mutually independent basis, relative to one another and from year to year, but any set of correlation could also be modeled. The quantitative results reported in this study are the output of EDCA s proprietary long-term integrated electricity models. The integrated model set includes several sub-models that assess Alberta s demographic and economic outlook, oil and natural gas production and export potential, electricity demand (by sector, utility and transmission point-of-delivery), generation supply, bulk transmission and tie line availability and pool price, as well as costs for air-borne emissions such as Hg, NOx, SOx, Particulate Matter (PM) and CO 2 equivalent. The EDCA model discretely models the interaction of the Alberta market with its adjacent markets including if tie capacity is increased. The interaction between the supply/demand fundamentals in Mid-C, BC and Saskatchewan markets is also discretely modeled to produce real-time import and export volumes, based on strategic behaviors of market participants and regional market price differentials. Chapter 1 8

19 Feature Chapter Zero Dark Thirty The March of Zero Hours Chapter Layout This year s report contains 8 chapters: Chapter One, this Introduction, presents a collection of newsworthy 2015 happenings in categories corresponding to each section of the Annual Report, any changes in key market rules and regulations, major new facilities, GHG emissions policy pronouncements, plus an overview of report methodology and structure. Chapter Two presents an executive summary of the key findings of the analysis and the quantitative results from each chapter. Chapter Three presents an overview of the underlying macroeconomics and demography in Alberta, including an outlook for the key industry segments and a review of crude oil and natural gas prices. The probabilistic inputs developed in this chapter drive the demand forecast in the next chapter. Chapter Four examines the output of the quantitative electric load growth model with discussion of the key forces driving the results. Domestic load growth across the key consumer groups residential, commercial and industrial is presented and discussed. A commentary on export opportunities and transmission and distribution losses rounds out the total Alberta electricity requirements for domestic generation and import supply. Chapter Five, this year s feature chapter, ZERO DARK 30, explores the effects of several market changes on the number of MW of supply being offered in at $0/MWh in any given hour. A surprisingly high fraction as much as 7,000MW of an hourly available supply between 8-12,000 MW is regularly offered in at or very near $0/MWh. That amount is comprised of minimumsafe coal generation trying to avoid cycling (2,500 MW), cogens (2,500 MW) whose hosts must have the steam which is generated as a by-product of electricity generation, intermittent wind (1,500 MW), run-of-river hydro (500 MW) and some biomass, plus whatever imports might be available. In the current supply/demand balance, the number hours when there is barely enough demand to use up even just the $0/MWh offers is growing. With the possible addition of 4,000-8,000 MW more renewables (mostly offered in at $0/MWh), more cogen and distributed generation being built to avoid rapidly inflating transmission and distribution costs and the recent general slowdown in load growth, it is possible that the rapid forward march of $0/MWh supply offers will outpace the sluggish growth in baseload demand and coal retirements, potentially increasing the frequency of hours when the pool price settles at or near $0/MWh. This year s study explores all the mechanisms that influence the race between zero offers and baseload demand. Will the increase in zero hours be large enough to actually slow down supply additions and lead to reduced reliability, hyper-volatility and potentially the very viability of the FEOC energy-only Alberta Electricity market. Chapter Six discusses electric energy supply fundamentals. The examination of the future electricity supply starts with an overview of existing generation capacity, cost structure and the expected timing of unit retirements. Future supply options Chapter 1 9

20 and availability, generation technologies, supply demand balance and reserve margin are then discussed. The chapter also presents a discussion of the key elements that define the range of future supply additions as well as the key drivers of future generation costs. Chapter Seven combines the effects of supply and demand into a forecast of the wholesale price of electricity and emissions. In this chapter s analysis, all of the various probabilistic parameters of supply and demand are used to define the distribution of future price forecasts. The quantitative results are presented and discussed, noting key assumptions and conclusions. Chapter Eight contains the appendices of the report, including supplementary charts and tables. Subscribers also receive an associated Excel summary file presenting the various macroeconomic input data and results such as natural gas price, electricity peak demand and energy, production by fuel, real and nominal pool prices, heat rates, and generation retirements and additions. All financial forecast data is presented throughout this report in nominal terms or Money of the Day, unless otherwise noted. Chapter 1 10