AGENDA MPCA Clean Power Plan Stakeholder Technical Meeting Webinar March 16, :00-4:00PM Central

Transcription

1 To Access Online Webinar & Teleconference: AGENDA MPCA Clean Power Plan Stakeholder Technical Meeting Webinar March 16, :00-4:00PM Central PLEASE be sure to mute your phone line as soon as the webinar begins to help keep things running as smoothly as possible! We ll have time for questions after each speaker. Join WebEx meeting Meeting number: Meeting password: webinar Join by phone Call-in toll-free number: (US) Conference Code: :00 PM Welcome and review of the agenda Franz Litz and Stacey Davis, CCAP 1:10 PM The Supreme Court s Stay and what it means for Minnesota Frank Kohlasch, MPCA Ari Peskoe, Harvard Environmental Law Institute 1:45PM Preliminary analyses evaluating the cost-effectiveness for Minnesota and the Midwest of the different trading-ready Clean Power Plan options 1:45 Jennifer Macedonia, Bipartisan Policy Center 3:00 Victor Niemeyer, EPRI 3:55 PM Next Steps 4:00 PM Adjourn aq-rule2-22e

2 Modeling of the Final Clean Power Plan JENNIFER MACEDONIA AND BLAIR BEASLEY MINNESOTA STAKEHOLDER MEETING MARCH 16, 2016

3 TABLE OF CONTENTS Modeling Process High-Level Insights Select Results - Preview: PTC/ITC Extension - Cost comparison of policy options - End-Use Energy Efficiency - Treatment of New Units Appendix 2

4 Modeling Process

5 BASIS OF ANALYSIS This Clean Power Plan analysis relies on economic modeling using the commercial version of the Integrated Planning Model (IPM) run by ICF and is based on assumptions and scenarios defined by the Bipartisan Policy Center in consultation with MSEER and its Modeling Subcommittee IPM is a national dispatch model intended to show broad trends and highlight key drivers through multi-scenario analysis The model determines the least-cost means of meeting electric generation requirements while complying with constraints, such as: air regulations, transmission constraints, and plant-specific operational constraints Caution is important when interpreting localized state-level results IPM is optimized at the regional/national level and may not capture all local or company-specific factors Modeling results should be viewed as a tool to supplement other inputs No single scenario and/or set of assumptions should be interpreted as providing the answer

6 PHASE I AND II MODELING PROCESS Final Rule August 2015 Phase I Modeling Ini8ated September 2015 Phase II Modeling Ini8ated December 2015 Phase I modeling started while updates made to the IPM model The IPM updates allow us to represent more features of the Final Clean Power Plan (CPP), including rate-based subcategory rates Phase II modeling is underway and results are expected this spring Phase I runs provide insights into the impacts of the final rule and include national, regional, and state-level results Phase I results include limited rate-based runs, due to their added complexity, and focus on a scenario analysis of mass-based runs Phase I modeling provides important lessons to inform Phase II Such as: gas price, energy efficiency, nuclear, banking assumptions In addition, ITC/PTC tax extenders passed in late 2015 and recently announced firm plant retirements have been incorporated into Phase II 5

7 PHASE I & II MODELING RUNS No 111(d) Policy Reference EE PTC/ITC Core Policy Runs Rate- based State Goals Rate- based Regional Dual Rate Trading Mass- based State Goals (exis8ng and new units) Mass- based Regional Trading (exis8ng and new units) Mass- Based Na8onal Trading (exis8ng and new units) Patchwork Runs Policy Varia8ons Exis8ng units only Alloca8on Methods Gas Price Sensi8vi8es EE Cost/Supply Nuclear Capacity RE Costs/ Innova8on Note: Boxes with solid outlines indicate Phase I or Phase II runs. Boxes with dashed outlines indicate Phase II runs only. 6

8 POLICY REGIONS MSEER RGGI West Other PJM SPP SERC ERCOT Note: Regional scenarios require assumptions about how states/regions are implementing the final Clean Power Plan. For purposes of modeling regional implementation, all EGUs in a state are grouped together in a single region as shown above for policy purposes. However, EGUs continue to be dispatched according to electricity markets with represented transmission bottlenecks. In Phase II modeling, policy regions are modeled at the interconnect level: Eastern, Western, and Texas. 7

9 PHASE I MODELING LIMITATIONS Phase I runs are homogenous on state implementation choices State runs have all states choosing to go it alone with available compliance options inside state borders Regional runs have all states (except Texas) choosing to trade within a multi-state trading region Rate-based run has all states choosing rate-based trading (within state borders only), and mass-based runs have all states choosing mass-based trading (either just within a state or across states). Limited Phase I Rate-based run Assumes each state is island for CPP compliance with its blended state rate Does not include additional transaction costs for ERC crediting Scenarios where all neighboring states are assumed to do the same thing may not be appropriate for drawing state-level conclusions about the optimal CPP policy pathway for a given state 8

10 SUMMARY OF KEY ASSUMPTIONS Phase I Assump8on Phase II Assump8on Unit- level characteris0cs AEO 2015 & NEEDSv.5.13 AEO 2015 & NEEDSv.5.15 Natural Gas Supply & Costs ICF s 2015 Integrated Gas Module (same input as EPA RIA) AEO 2015 Fuel Supply Curves based on mid- point between AEO Core & High Gas Resource (low gas price) cases Renewable Energy Cost ICF Market Research ICF Market Research Electricity Demand AEO 2015 Demand Forecast AEO 2015 Demand Forecast Cost of Incremental Energy Efficiency Supply of Incremental Energy Efficiency 3- step cost curve ( cents/kwh)* Various levels tested 3- step cost curve ( cents/kwh)* ½ EE supply from EPA Nuclear Re0rements All units can con0nue to run past their 60- year relicensing date (opera0ng costs increase with age) All units re0re at their 60- year relicensing date * cents/kwh represents only 55% of the total resource cost of energy efficiency investments, assumed to be the u0lity por0on of ratepayer- funded EE; the assumed total resource cost is cents/kwh. 9

11 PHASE I & II ASSUMPTIONS: MINNESOTA PHASE I Nuclear Power Plants: There is no age- based re0rements. Plants face increased opera0ng costs as they age. Coal Re0rements: Unit- level data from NEEDSv.13. In MN, assumed firm 2015 re0rements: Black Dog, Silver Lake, and Taconite Harbor Energy Center. In , BAU re0rements ranged from 1.04 GW GW. RPS: Modeled as part of a regional RPS target (MISO) EERS: Approximated as part of the AEO 2015 electricity demand forecast PHASE II Nuclear Power Plants: All plants re0re at 60 years. This includes Mon0cello (2031) and Prairie Island (2034). (addi0onal sensi0vity runs planned) Coal Re0rements: Unit- level data from NEEDSv.15. In MN, assumed firm 2015 re0rements: Aus0n Northeast Sta0on, Black Dog, Silver Lake, and Taconite Harbor Energy Center. 2021: Hoot Lake. In , BAU re0rements are 1.40 GW. RPS: Modeled as part of a regional RPS target (IPM/NEMS zones) EERS: Approximated as part of the AEO 2015 electricity demand forecast 10

12 High-Level Insights

13 HIGH-LEVEL INSIGHTS: UNCERTAINTY The magnitude of impacts from the final CPP, including potential compliance costs, are dependent on market factors and state decisions yet to be made, such as: Gas price Availability of end-use energy efficiency (EE) Retirement/relicensing of existing nuclear units The treatment of new units Policy decisions of other interconnected states (e.g., rate vs. mass, EE policies) 12

14 HIGH-LEVEL INSIGHTS: UNCERTAINTY AND TRADING The use of trading in CPP implementation provides compliance flexibility across a broad range of potential futures and a mechanism to approach least cost Expanding trading regions over larger areas tends to increase the benefits and help mitigate impacts of the unknown For example: unexpected outages/retirements, wide range of potential technology futures, extreme weather such as droughts Under fixed assumptions, states with largest emissions gap see greatest cost benefit from trading Without allowance trading, compliance scenarios tend to require more generation shifts between states in order to reach compliance Trading balances supply/demand across states and increases compliance options to achieve a single regional market price (lower on average compared to no trading) 13

15 HIGH LEVEL INSIGHTS: GAS PRICE AND EE Natural gas price trends are an important assumption, with lower gas prices shifting more generation to natural gas and leading to additional coal retirements Gas price influences how gas-fired generation competes with renewable investments to displace coal in compliance scenarios Phase II modeling will further investigate the combined effects of lower gas price with the PTC/ITC phase-out, as well as with greater future capacity needs driven by potential nuclear retirements Varying assumptions about the availability of incremental end-use efficiency (EE) highlights the importance of EE for cost containment and for smoothing the transition in generation and capacity mix Greater EE reduces coal retirements, total costs, allowance prices, and wholesale electricity price impacts 14

16 HIGH-LEVEL INSIGHTS: TREATMENT OF NEW UNITS Potential risk of/magnitude of leakage is dependent on various assumptions about future uncertainties and factors that may not be fully captured in modeling efforts Inclusion of new units and new source complement (NSC) could lead to more consistent market signals and mitigate market distortions In Phase I modeled compliance scenarios, MSEER region doesn t add much new fossil capacity Thus, potential leakage to new NGCC is not as pronounced in MSEER When states are assumed to use the NSC, there is the potential for leakage to CTs (which are not affected by the Clean Power Plan) The model may exaggerate this finding and not be capturing all of the factors that would dampen an overbuild of CTs in the real world Phase II will further investigate the impact of additional factors, including potential nuclear retirements, on these findings The need to replace retiring nuclear plants in some MSEER states could increase the production costs, the ambition required to comply, and the potential for leakage to new NGCC, while also reducing the incentive for an overbuild of peaking CTs 15

17 PREVIEW of Phase II: Impact of PTC/ITC Extension

18 HIGH-LEVEL INSIGHTS: PTC/ITC (PHASE II) The PTC/ITC extension, modeled in Phase II analysis, accelerates wind and solar deployment (even in the absence of the CPP) Modeling suggests tax extenders could spur an additional 33 GW wind and 25 GW utility-scale solar beyond BAU projections, approaching 180 GW combined wind and solar installed capacity in the next decade Combined effect of tax extenders with CPP policy may lead to greater renewable deployment The PTC/ITC extension decreases coal capacity nationally (by 8 GW) Thus, results in lower national CO 2 emissions compared to a reference case without the PTC/ITC Impacts of the PTC/ITC extension vary by state and region 17

19 WIND CAPACITY The Phase II Reference Case with PTC/ITC approaches 130 GW installed wind capacity by 2022 (adding 55 GW beyond 2015) On top of expected/projected additions, an additional 33 GW of wind is projected to be built by 2022 with the tax extenders U.S. Wind Capacity ( ) GW 33 GW GW Historic Data Model Output Sources: EIA: Electric Power Annual 2012, 2013, and AWEA: U.S. Wind Industry Fourth Quarter 2015 Market Report. Projected data from BPC scenarios using IPM. 18

20 WIND CAPACITY In Minnesota, the Phase II Reference Case with the PTC/ITC extension has about 400 MW more wind capacity in 2020, compared to a Phase II Reference Case w/out PTC/ITC extension By 2030, the gap closes to about 200 MW of increased capacity 6 Minnesota Wind Capacity ( ) 5 4 GW Wind Wind 19

21 SOLAR CAPACITY The Phase II Reference Case with PTC/ITC reaches national 50 GW installed solar capacity by 2023 (adding 37 GW beyond 2015 Q3) On top of expected/projected additions, an additional 25 GW of solar is projected to be built by 2023 with the tax extenders 60 U.S. U0lity - Scale Solar Capacity ( ) GW GW GW Historic Data Model Output Solar Historic Solar Under Construc0on Solar Reference Solar Reference (no PTC/ITC) Sources: EIA: Electric Power Annual 2012, 2013, and SEIA: U.S. Solar Market Insight 2015 Q4. Projected data from BPC scenarios using IPM. 20

22 Phase I Modeling Results * These Phase I results do not include influence of PTC/ITC tax extenders

23 Preliminary Do not cite or quote TOTAL CUMULATIVE ADJUSTED COMPLIANCE COSTS The use of trading in CPP implementation provides compliance flexibility across a broad range of potential futures and a mechanism to approach least cost compliance Expanding trading regions over larger areas tends to increase the benefits *Note, modeled scenarios not shown above that allow incremental EE project lower compliance costs. 22

24 Preliminary Do not cite or quote COMPARING POLICY PATHWAYS IN MSEER STATES States with largest emissions gap see greatest cost benefit from all states trading In states whose reference case CO 2 is closer to CPP goals, some counter-intuitive results Phase II updates (gas price, nuclear) could impact these state results 23

25 Preliminary Do not cite or quote IMPACT OF GENERATION SHIFTS Compliance scenarios without flexibility to trade increase generation shifts Access to out of state allowances allows some marginal coal to operate Percent Share of MSEER Generation Reference State Rate State Mass Regional Mass AR 7.0% 6.9% 6.9% 6.8% IA 4.9% 6.2% 5.1% 4.9% IL 19.0% 18.7% 20.2% 19.0% IN Coal- heavy states lose genera0on as 12.6% 11.6% 10.3% 12.3% KY capacity factors drop for compliance 6.8% 4.9% 6.8% 5.7% LA 6.0% 7.4% 7.6% 6.6% MI 10.7% 10.8% 9.7% 10.1% MN 5.0% 5.9% 5.1% 4.3% MO 7.7% 6.3% 7.3% 7.3% MS NGCCs ramp up to serve demand 5.9% 7.0% 6.7% 8.6% MT 3.1% 3.3% 2.9% 3.2% ND 5.2% 3.0% 4.8% 5.0% SD 0.8% 1.3% 1.5% 1.1% WI 5.1% 6.7% 5.3% 5.1% Percent Share of US Generation Reference State Rate State Mass Regional Mass MSEER 24.8% 23.0% 24.3% 24.6% 24

26 Preliminary Do not cite or quote STATE/REGIONAL/NATIONAL ALLOWANCE COSTS Both compliance costs & generation shifts impact demand for allowances, and thus, their market price Trading balances supply/demand across states and increases compliance options to achieve a single regional market price (lower on average compared to no trading) 25

27 Impacts of End-Use EE on CPP Compliance under a Range of Assumptions

28 SCENARIOS VARY ASSUMED EE AVAILABILITY All Phase I scenarios include the level of EE built into AEO2015 demand forecast In addition, EE sensitivity runs shown below allow the model to choose incremental EE (load reductions with various supply/cost assumptions) if they are cost-effective compared to serving forecasted regional load with supply-side resources. 27

29 Preliminary Do not cite or quote IMPACT OF EE ASSUMPTIONS ON COAL RETIREMENTS Increasing the availability of incremental EE reduces coal retirements driven by the CPP 28

30 Preliminary Do not cite or quote CUMULATIVE COAL RETIREMENTS In general, coal retirements in Minnesota follow the national/regional trend: policies that incentivize additional EE reduce CPP policy-driven coal retirements Scenarios that stray from the overall trend at the state-level are largely driven by generation shifts between states 8 Minnesota Cumula0ve Coal Re0rements ( ) GW Reference Regional Mass Regional Mass (Exis0ng) Reference with 1/2 EPA EE Regional Mass (1/2 EPA EE) Regional Mass (Exis0ng, 1/2 EPA EE) Regional Mass (High Cost 1/2 EPA EE) Regional Mass (EPA EE) 29

31 Preliminary Do not cite or quote CHALLENGE OF MODELING EE AND INTERPRETING RESULTS Increasing the availability of incremental EE drives compliance costs down Interpreting results on the impact of EE depends on what you assume about EE in the reference case Because incremental EE that would be cost-effective is constrained in the no EE reference case, compliance costs as compared to business as usual can be negative When EE is permitted in the reference case and there is no incremental uptake under CPP, relative compliance costs appear higher (i.e., ½ EPA and High Cost EE) The Regional Mass (EPA EE) cost below assumes more EE is available under CPP than in the reference case U.S. Average Annual Compliance Cost Compared to Reference with no EE ( ) 15,000 10,000 Million $ 5, ,000-10,000-15,000 Regional Mass Regional Mass (High Cost 1/2 EPA EE) Regional Mass (1/2 EPA EE) Regional Mass (EPA EE) 30

32 Preliminary Do not cite or quote IMPACT OF EE SUPPLY ON WHOLESALE ELECTRICITY PRICES Increasing the availability of incremental EE reduces the price of CO 2 allowances. This results in lower wholesale electricity prices. 31

33 Preliminary Do not cite or quote Leakage

34 DRIVERS OF LEAKAGE IN MODELING ANALYSIS This Phase I analysis shows limited leakage in MSEER when new units are excluded from the policy in terms of new NGCC builds and decreased existing NGCC capacity factors Leakage is sensitive to modeling assumptions Leakage is expected to be more pronounced when compliance costs are higher and less pronounced when compliance costs are lower Different assumptions would reduce compliance costs. This includes: inclusion of PTC/ITC, lower gas prices, more end-use EE Nuclear assumptions influence modeled outcomes on leakage: scenarios that assume no nuclear units are relicensed at 60 years require more replacement capacity to be built and are more likely to project leakage

35 Preliminary Do not cite or quote IMPACT OF EXCLUDING NEW UNITS ON NEW NGCC BUILDS Excluding new units from compliance leads to the addition of new NGCC units. This effect is less pronounced in MSEER and varies by state. U.S. NGCC New Builds ( ) MSEER NGCC New Builds ( ) GW 30 GW Reference Regional Mass Regional Mass (exis0ng) Reference Regional Mass Regional Mass (exis0ng) 34

36 Preliminary Do not cite or quote WHEN NEW UNITS ARE EXCLUDED, UTILIZATION OF EXISTING NGCC TENDS TO FALL *The scenarios above do not include addi0onal EE beyond AEO2015 demand forecast. Modeling addi0onal incremental EE reduces the need for new NGCC and thus, mi0gates leakage. 35

37 Preliminary Do not cite or quote Appendix

38 RANGE OF GAS PRICE FORECASTS 2015 gas prices, futures, and forecasts continue to shift downward from EPA assumed gas trends in final CPP RIA Updated Phase II assumption: $4-5 gas, derived from the midpoint between AEO2015 Base Case and High Gas Resource Case Higher gas demand from CPP = gas price $5-6+ Phase EPA Reference I reference $5-6 $5-6 Gas gas CPP $4-5 gas sensi0vity Midpoint/Phase II $4-5 gas 2030 Henry Hub Gas Prices Reference 5.85 $4-5 Gas 5.09 Mid Point 4.62 NOTE: The $4-5 Gas case is iden0cal to the Regional Mass case, except gas prices are lower 37

39 EE COSTS Preliminary Do not cite or quote All scenarios are based on AEO2015 demand forecast. In policy scenarios that allow incremental EE* (beyond AEO2015), end-use EE is available to serve electricity demand using an assumed three-step supply curve with cost increasing as the supply available at each step is exhausted. In 2020, costs are: 2.3, 2.6, and 3.2 cents/kwh. Costs in each block increase by.3 cents/kwh starting in An assumed participant portion (45% of the total resource cost of EE) is added separately to the compliance cost EE Cost Units = Cents/KWh Units = $/MWh Step 1 Step 2 Step 3 Step 1 Step 2 Step 3 U0lity Por0on Par0cipant Por0on Total Resource Cost * Except for the High Cost EE scenario, where costs are increased by 50% at each step in the three- step cost curve 38

40 WIND COSTS Preliminary Do not cite or quote U.S. Onshore Wind Overnight Capital Costs & FOM (2012$/kW) Vintage Phase I Phase II ,766 1, ,731 1, ,698 1, ,616 1, ,470 1, ,337 1,337 FOM Average Step 1 Average Levelized Cost of Electricity (2012$/MWh) No PTC PTC- Model Year 2016 PTC- Model Year 2020 Minnesota MSEER U.S

41 SOLAR COSTS Preliminary Do not cite or quote U.S. U8lity Scale Solar PV Overnight Capital Costs & FOM (2012$/kW) Vintage Phase I Phase II ,990 1, ,900 1, ,848 1, ,746 1, ,675 1, ,530 1,377 FOM

42 Preliminary Do not cite or quote SCENARIO DESCRIPTIONS Scenario Rate- Based State Trading Mass- Based State Trading (exis0ng and new units) Mass- Based Regional Trading (exis0ng and new units) Mass- Based Regional Trading Sensi0vity Runs: Descrip8on Each state must comply with its blended state- specific rate- based goal included in EPA s final Clean Power Plan. Trading is permited among sources within the state. Banking of credits is allowed. No incremental EE. Each state must comply with the state mass- based target for exis0ng sources plus the new source complement. Trading is permited among sources within the state and banking is allowed. No incremental EE. Each state is assigned the same target as the mass- based state trading scenarios. Trading is permited among all sources in a given region and banking is allowed. No incremental EE. Exis0ng units only Iden0cal to the mass- based regional trading scenario, except the mass- based target includes exis0ng units only. Exis0ng units only with ½ EPA EE Iden0cal to the mass- based regional trading scenario, except the mass- based target includes exis0ng units only and incremental EE is available at half of EPA s EE supply assump0on. 41

43 Preliminary Do not cite or quote SCENARIO DESCRIPTIONS Scenario Na0onal Mass Descrip8on $4-5 Gas Iden0cal to the mass- based regional trading scenario, except a 25% cost reduc0on is applied to each step of the gas cost curve within ICF Interna0onal s Integrated Gas Module. EPA EE Iden0cal to the mass- based regional trading scenario, except incremental EE is available at EPA s EE supply assump0on. ½ EPA EE Iden0cal to the mass- based regional trading scenario, except incremental EE is available at half of EPA s EE supply assump0on. High Cost EE Iden0cal to the mass- based regional trading scenario, except incremental EE is available at half of EPA s EE supply assump0on and EE costs are increased by 50% at each step in the three- step EE cost curve. Each state is assigned the same target as the mass- based state trading scenarios. Trading is permited among all sources na0onwide and banking is allowed. No incremental EE. 42

44 RUN YEAR MAPPING Calendar Year Model Year Calendar Year Model Year

45 COMPARISON OF PHASE I & II MODELING OF FINAL CPP PHASE I Launched in September 2015 Based on the same IPM modeling structure used for post- proposal modeling Models the final CPP, including final emission goals Incorporates many, but not all, updated modeling assump0ons Includes mass- based runs and sensi0vi0es, as well as a limited rate- based run Scenarios are homogenous all states making similar choices PHASE II Launched in December 2015 Based on updated IPM modeling structure designed to represent addi0onal features of the final CPP Models the final CPP, including final emission goals Incorporates updated modeling assump0ons (e.g., PTC/ITC phase- out) Includes addi0onal sensi0vity runs and rate- based runs with subcategory rates. Includes patchwork scenarios, alloca0on approaches, and retail price impacts 44

46 Preliminary Do not cite or quote COMPONENTS OF TOTAL ADJUSTED COST (TAC) TAC= TSC + EE Participant Costs + Import/Export + Net Allowance/Credit Cost Total System Cost (TSC): Includes all costs associated with generation, such as new capacity, fuel, and other operating & maintenance costs, as well as compliance costs such as the utility portion of end-use energy efficiency. For a state, this includes instate generation only. EE Participant costs: We assume 55% of the total resource cost of an end-use energy efficiency measure is born by the utility and 45% of the cost is paid by the consumer/participant. While the utility portion is included in TSC, and thus impacts wholesale electricity costs, the participant portion is a separate line item. Generation shift adjustment: Some scenarios result in generation shifts between states/regions so that the cost of in-state generation may go down, while the cost of importing power goes up (or vice versa). To better account for total costs to deliver energy, this adjustment estimates the cost associated with changes in net electricity imports/exports. Because IPM uses regional (rather than state-level) electricity demand, state-level imports are estimated compared to the reference case. Net allowance/credit cost: The value of the net position in emission credits or allowances (i.e., to what degree is state a net buyer or seller of credits/allowances in a regional trading program). For state implementation, credits don t cross borders; thus this cost is zero. For regional scenarios, this nets to zero at the regional level. 45

47 Preliminary Do not cite or quote

48 State-Level Modeling of Clean Power Plan Compliance Pathways with EPRI s US-REGEN Model Vic Niemeyer Senior Technical Executive Electric Power Research Institute MPCA Clean Power Plan Stakeholder Technical Meeting March 16, Electric Power Research Institute, Inc. All rights reserved.

49 GW US-REGEN 48-State Version: EPRI s In-House Electric Sector Model for CPP Modeling Capacity Expansion Economic Model, Long Horizon to P a ci fi c Mo unt ain -S N Y N E State-Level Resolution for Policy and Regulation Analysis Innovative Algorithm to Capture Wind, Solar, & Load Correlations in a Long-Horizon Model Electric Power Research Institute, Inc. All rights reserved.

50 Electric Model: Key Features Endogenously builds/retrofits/retires capacity in each model time period according to the economics Coal (+ retrofit to gas, biomass, CCS, co-firing, heatrate improvements), Gas NGCCs, Gas Combustion Turbines, Nuclear, Hydro, Geothermal, Wind (Onshore, Offshore), Solar (CSP, PV, Rooftop PV), Diesel/Oil, Coal/Gas with CCS, new biomass Endogenously builds inter-state transmission if needed and economic We select representative hours to capture load-wind-solar correlations across the year i.e. US-REGEN knows when load is high and there s no wind! Based on a dataset of every unit in the country Last updated November Electric Power Research Institute, Inc. All rights reserved.

51 Renewable Resource Data Wind resource data from AWS Truepower Based on 2010 meteorology Solar resource data from AWS Truepower Separate resource for central station PV/CSP versus rooftop solar Based on 2010 meteorology Geothermal resource data based on NREL (2009) estimates for the Western states New potential additions of ~40GW by 2050 (8GW in CA) Assume capacity factor improves from 50% to 80% due to technical progress Electric Power Research Institute, Inc. All rights reserved.

52 Location of Wind Resource by State Electric Power Research Institute, Inc. All rights reserved.

53 Location of Wind Resources by State State-Level Wind Resource Base 45% 40% State-level Wind Resources by 2010 Capacity Factor 35% 30% Less than 400 MW > 40% CF 25% 20% 15% 10% 5% 0% 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8, Electric Power Research Institute, Inc. All rights reserved.

54 Location of Central PV Resource by State 7 * Assumes the use of up to 1% of each state s available land Electric Power Research Institute, Inc. All rights reserved.

55 US-REGEN vs IPM (used by EPA for CPP design, RIAs) US-REGEN and IPM are both based on the same modeling paradigm Full information, inter-temporal optimization Compared to IPM, US-REGEN Uses 48 state-based regions vs IPM s 60+ regions across state lines Aggregates units more, but uses ~ 6 times as many representative hours to capture renewable intermittency better Uses model years 2015, 2018, 2021, 2024, 2027, 2030, 2035, 2040, 2045, 2050; IPM uses 2016, 2018, 2020, 2025, 2030, 2040, 2050 All models of this type have the same computational limitations; modelers must make tradeoffs as to what elements are important to represent the policy at hand Electric Power Research Institute, Inc. All rights reserved.

56 US-REGEN Models Four Main Compliance Pathways Rate Subcategory Rates Steam units target of 1305 lb/mwh, NGCC units target of 771 lb/mwh (2030) State Rate Steam and NGCC units target equal to the state rate CPP Path Cap Existing and New Units Existing and New Steam and NGCC units emit less than the state mass target + the new source complement target Mass Cap Existing Units Only Existing and Steam and NGCC units emit less than the state mass target Electric Power Research Institute, Inc. All rights reserved.

57 Specific Features for Modeling the Clean Power Plan Detailed representation of ERC sources by type Zero, Fossil, Gas-Shift Inclusion of output-based set-asides for Existing Mass path Endogenous energy efficiency US-REGEN can endogenously build energy efficiency (that counts towards CPP compliance) Currently using EPA CPP proposal costs, could revisit Detailed renewable representation US-REGEN was built from scratch to give a very detailed representation of wind and solar, and their intermittency Other options for coal Co-firing, conversion to biomass or gas, CCS retrofits Electric Power Research Institute, Inc. All rights reserved.

58 Compliance Pathway Determines Trading Partners Rate Subcategory Rates Can trade ERCs with any other Subcategory Rate state State Rate Can trade ERCs with another State Rate state in the same compliance plan CPP Mass Cap Existing and New Units Cap Existing Units Only Can trade allowances with any other Mass-Based State Electric Power Research Institute, Inc. All rights reserved.

59 Caveats for Following Model Results All analyses preliminary CPP highly complex, still testing our modeling Models are highly aggregated simulations but not reality No constraints on gas delivery Not forecasting Choices for states intended to show consequences of alternative pathways in a heterogeneous world, not speaking to what pathways states may choose Many uncertainties not explored here Cost of EE and RE Possible future additional CO2 policy/regulation Ability to deploy added transmission Electric Power Research Institute, Inc. All rights reserved.

60 Uses and Limitations of Economic Models Models like US-REGEN are necessarily numerical abstractions of the complex economic and energy systems they represent. As such, they may contain: Approximation errors Incomplete system dynamics Quality of data Essentially, all models are wrong, but some are useful. -- George Edward Pelham Box When viewing model results, it is important to keep in mind: Analyses are not intended to be viewed as a prediction of a particular outcome or cluster of outcomes. Insights come by running a variety of cases, comparing the results, and asking what if questions. Actual deployment of a model outcome is dependent on many additional factors, such as policy, permitting and siting Electric Power Research Institute, Inc. All rights reserved.

61 TWh Reference Scenario Provides Point of Reference but is Not a Forecast 14 5,000 6,000 4,500 5,000 4,000 3,500 4,000 3,000 2,500 3,000 2,000 2,000 1,500 1,000 1, Ref EEA Generation Reference Generation (US48) (US48) Electric Power Research Institute, Inc. All rights reserved. EE + Price Response New Solar Ex Solar New Wind Ex Wind Hydro Gas Turbine CCS Gas New NGCC Ex NGCC CCS Coal New Coal Ex Coal Other Geothermal New Nuclear Ex Nuclear Scenario Load

62 Island Results Each state must comply relying solely on resources within its own boundary; power flows limited to levels in reference case Electric Power Research Institute, Inc. All rights reserved.

63 Price ($/MMBtu) Natural Gas Price Uncertainty Represented with EIA s Annual Energy Outlook 2015 High and Low Paths 9 8 Average Power Producer's Gas Price (US) 7 6 High Price Path (based on AEO2015 Ref) Low Price Path (based on AEO2015 HEUR) Source: U.S. Energy Low Information High Administration s Annual Energy Outlook for Electric Power Research Institute, Inc. All rights reserved.

64 Price ($/MMBtu in 2010$) Natural Gas Price Uncertainty Represented with EIA s Annual Energy Outlook 2015 High and Low Paths 9 8 Average Power Producer's Gas Price (US) + NYMEX Henry Hub 7 6 High Price Path (based on AEO2015 Ref) Low Price Path (based on AEO2015 HEUR) 2 NYMEX Henry Hub Source: U.S. LowEnergy High Information NYMEX Administration s Henry Hub Annual Energy Outlook for Electric Power Research Institute, Inc. All rights reserved.

65 Emission Rate Credit (ERC)/Allowance Prices for 2030 with Full Island Compliance (Low gas price path) 18 $0 $0 $0 $69.3 $0 $16.4 $17.3 $23.2 $14.2 Rate State Rate Subcategory Mass Full Mass Existing State rate/mass path based on minimum costs of island compliance (based on present value of compliance cost through 2050) $8 $13.7 $16.3 $6.8 $6 $16.4 $17.4 $0 $17.5 $0 $0 $28.6 $11.2 $17.2 $15 $ Electric Power Research Institute, Inc. All rights reserved. $16.2 $19.2 $2.2 $16.1 $16.2 $15.7 $2 $0 $0 $28.8 $15.7 $0 $0 $3.6 $0 VT $0 $0 $16.5 Note: for Rate states (green), prices are for ERCs in $/MWh, $0 $0 $0 $0 For Mass states (brown) prices are Min w Low GasP Price for Allowances in $/metric ton

66 ERC/Allowance Prices for 2030 with Full Island Compliance (High gas price path) 19 $0 $0 $0 $0 $0 $25.6 $13.7 $24.6 $13.9 Rate State Rate Subcategory Mass Full Mass Existing $11.7 $0 $32.3 $0 $0 $29.5 $12 $0 $15.4 $0 $0 $62.5 $13.4 $13.6 $19 $ Electric Power Research Institute, Inc. All rights reserved. $24.3 $18.8 $16.6 $27.7 $27.4 $18.2 $17.4 $14 $0 $23.5 $13.8 $0 $0 $0 $16.6 VT $0 $0 $21.7 Note: for Rate states (green), prices are for ERCs in $/MWh, $0 $0 $0 $0 For Mass states (brown) prices are Min w High GasP Price for Allowances in $/metric ton

67 Observations Simple economics of rate vs mass: rate compliance achieved with investment in renewables (wind) and energy efficiency, gas redispatch mass compliance achieved with more gas generation Zero prices imply states are in compliance in 2030 (though possible need some effort to comply in other time periods) Low prices driven by ease of compliance, in turn driven by Low price of natural gas Low incremental cost of wind (in high-wind states) Energy efficiency credits from existing EE programs Announced/expected post 2012 coal retirements Many states at/near compliance for both Rate and Mass paths Electric Power Research Institute, Inc. All rights reserved.

68 National Uniform-Pathway Results All states choose the same compliance pathway and trade ERCs and Allowances per Rate and Mass Model Rules (also trade power) Electric Power Research Institute, Inc. All rights reserved.

69 TWh Reference Scenario 22 5,000 6,000 4,500 5,000 4,000 3,500 4,000 3,000 2,500 3,000 2,000 2,000 1,500 1,000 1, Ref EEA Generation Reference Generation (US48) (US48) Electric Power Research Institute, Inc. All rights reserved. EE + Price Response New Solar Ex Solar New Wind Ex Wind Hydro Gas Turbine CCS Gas New NGCC Ex NGCC CCS Coal New Coal Ex Coal Other Geothermal New Nuclear Ex Nuclear Scenario Load

70 TWh Generation Mix with Uniform Compliance Under Subcat. Rate Path (with ERC trading) 23 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1, RUn Generation (US48) Electric Power Research Institute, Inc. All rights reserved. EE + Price Response New Solar Ex Solar New Wind Ex Wind Hydro Gas Turbine CCS Gas New NGCC Ex NGCC CCS Coal New Coal Ex Coal Other Geothermal New Nuclear Ex Nuclear Scenario Load

71 TWh Generation Mix with Uniform Compliance Under Exist. Mass Path (with Allowance trading) 24 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1, MXn Generation (US48) Electric Power Research Institute, Inc. All rights reserved. EE + Price Response New Solar Ex Solar New Wind Ex Wind Hydro Gas Turbine CCS Gas New NGCC Ex NGCC CCS Coal New Coal Ex Coal Other Geothermal New Nuclear Ex Nuclear Scenario Load

72 2030 Net ERC Exports if All States Choose Sub Category Rate Path and Trade ERCs (ERC price = $10.6/MWh) Rate State Rate Subcategory Mass Full Mass Existing -8.5 ERC exports in TWh VT Low gas price path RUn Exports Electric Power Research Institute, Inc. All rights reserved.

73 2030 Net Emission Allowance Exports if All States Choose Existing Mass Path (EA price = $11.9/metric ton) Rate State Rate Subcategory Mass Full Mass Existing Allowance exports millions of metric tons VT Low gas price path MXn Exports Electric Power Research Institute, Inc. All rights reserved.

74 CO 2 Emissions Electric Power Research Institute, Inc. All rights reserved.

75 CO2 (million short tons) All Island w Uniform Choice of Paths (Sub. Rate vs. Exist. Mass) with no CPP Trading, no Incremental Power Flows 2500 US48 Electric Sector CO2 Emissions Note that with all-island compliance CO 2 higher in 2030 for Exist Mass path Ref RUi MXi Electric Power Research Institute, Inc. All rights reserved.

76 CO2 (million short tons) Nationally Uniform Choice of Paths (Sub. Rate vs. Exist. Mass) with ERC/Allowance Trading (per Model Rule) 2500 US48 Electric Sector CO2 Emissions With ERC/Allowance trading CO 2 emissions are higher, but now Mass path has lower CO 2 than Rate path Ref RUn MXn Electric Power Research Institute, Inc. All rights reserved.

77 Trading Results Sensitive to National Mix of Pathways Electric Power Research Institute, Inc. All rights reserved.

78 2030 Mix1 ERC/Allowance Pricing with Low Gas Prices $0 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 Rate State Rate Subcategory Mass Full Mass Existing $11.8 $11.8 Allowance prices in $/metric ton $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $11.8 $4.4 $11.8 $4.4 $11.8 $11.8 $5.8 VT $5.8 $5.8 $5.8 $5.8 $11.8 $5.8 $11.8 $5.8 $11.8 $5.8 $11.8 $11.8 $4.4 ERC prices in $/MWh $11.8 $11.8 Mix1 Price Electric Power Research Institute, Inc. All rights reserved.

79 2030 Mix2 ERC/Allowance Pricing with Low Gas Prices $0 $11.2 $11.2 $11.2 $11.2 $11.2 $13.9 $11.2 $11.2 Rate State Rate Subcategory Mass Full Mass Existing $11.2 $11.2 Allowance prices in $/metric ton $11.2 $11.2 $5.5 VT $5.5 $5.5 $11.2 $13.9 $5.5 $11.2 $5.5 $13.9 $11.2 $11.2 $5.5 $11.2 $11.2 $11.2 $11.2 $5.5 $11.2 $5.5 $13.9 $11.2 $11.2 $11.2 $11.2 $11.2 $13.9 $11.2 $13.9 $11.2 $13.9 $11.2 $11.2 ERC prices in $/MWh $11.2 $11.2 Mix2 Price Electric Power Research Institute, Inc. All rights reserved.

80 2030 Mix2 ERC/Allowance Pricing with Low Gas Prices $0 $11 $11 $11 $11 $11 $11 $7.2 $11 Rate State Rate Subcategory Mass Full Mass Existing $11 $11 $11 $11 $11 $7.2 $11 $11 $11 $7.2 $11 $7.2 Allowance prices in $/metric ton $7.2 $11 $11 $11 $11 $11 $7.2 $11 $7.2 $11 $5.1 VT $5.1 $5.1 $5.1 $5.1 $11 $5.1 $11 $5.1 $11 $5.1 $11 $11 $7.2 ERC prices in $/MWh $11 $11 Mix5 Price Electric Power Research Institute, Inc. All rights reserved.

81 Observations Mix scenarios are illustrative samples of many possibilities Assume national markets for ERCs and Allowances ERC price if only new-nuclear states choose Rate is low, but that price may invite other state to go rate Mix2 and Mix5 probably more realistic Many states nominally committed to mass path through existing state polices, e.g., California and RGGI states, would be in compliance with the CPP by choosing rate pathway Reasonable variation in future natural gas prices has greater impact on costs than the Clean Power Plan Electric Power Research Institute, Inc. All rights reserved.

82 Strategic Insights Key decisions for states are Rate vs. Mass, but also reliance on participation in the market Some states appear to have lower costs with Rate, some for Mass, no single universal lowest-cost choice Some states may be net beneficiaries of the CPP Trading creates value on both sides of the transaction The future matters Natural gas prices Renewable and EE costs Market scope and depth Supply/demand for ERCs and Allowances depends on individual state choices for Rate vs. Mass Electric Power Research Institute, Inc. All rights reserved.

83 Together Shaping the Future of Electricity Electric Power Research Institute, Inc. All rights reserved.