WECC Capital Cost Recommendations June 4, 2012 Arne Olson, Partner Nick Schlag, Consultant Gabe Kwok, Associate
History In 2009, E3 provided WECC with recommendations for capital costs of new electric generation technologies to use in its 10-year study cycles Prior to this effort, the relative costs of WECC’s study cases could only be compared on a variable-cost basis (using PROMOD results) This effort provided WECC with a framework to quantify relative scenario costs on a basis reflecting their actual prospective costs to ratepayers by combining variable & fixed costs In 2011, WECC asked E3 to review the capital costs to ensure continued accuracy Due to the continued evolution of solar PV technologies, E3 lowered its estimates of photovoltaic capital costs Total Cost = Variable Costs (PROMOD) + Fixed Costs (E3 Capital Cost Tool)
Background In the midst of its 10- and 20-year study plans, WECC has asked E3 to provide guidance on resource cost and performance to use in those studies These capital costs will serve as inputs to the 10- and 20-year studies: Including capital costs in the 10-year study cycles enables comparisons of total costs between scenarios Capital costs will serve as an input to the 20-year study’s LTPT, allowing for the development of robust scenarios through cost minimization
Updates E3 presented its initial recommendations to stakeholders on May 15, 2012 Based on stakeholder feedback and comments, E3 has reviewed its recommendations for wind and solar costs Minor revisions were made to several of the present day solar PV costs to better capture expected cost differentials between system types and sizes E3 has revised some of the data inputs used to forecast of cost declines for solar PV and solar thermal technologies: Forecasts of installed capacity for solar PV and solar thermal have been revised to account for near-term projections of global market dynamics Learning rate for solar thermal has been adjusted to reflect greater potential for technological improvements than originally anticipated
Modeling Framework 2022 Study 2032 Study Resource Portfolio Resource Performance (NREL) PROMOD Total Scenario Costs 2022 Study Capital Costs (E3 Capital Cost Tool) Resource Performance (WREZ) LTPT/NXT Resource Portfolio Total Scenario Cost 2032 Study Capital Costs (E3 Capital Cost Tool)
Scope of Updates E3’s Capital Cost Tool considers a broad range of potential new generation technologies The scope for E3’s update is divided into two phases: Near term (integrated in this year’s study cycle): update costs for wind and solar technologies Long term (integrated into subsequent study cycles): review costs for all technologies This division prioritizes updating those costs that are most likely to have changed given the limited time before the start of this study cycle Technologies in E3’s Capital Cost Tool Technology Subtypes Biomass Biogas Landfill, Other Gas CT Gas CCGT CHP Small, Large Coal Steam, IGCC Nuclear Hydro Small, Large, Upgrade Solar PV Fixed Tilt, Tracking Solar Thermal No Storage, 6hrs Storage Wind Onshore
Modeled in Prior WECC Studies New to This Year’s Study Cycle Technologies Covered E3’s current update encompasses the following technologies—an expanded set compared to the original Capital Cost Tool Solar PV Solar Thermal Wind Modeled in Prior WECC Studies Large Utility (20 MW +) Fixed Tilt Tracking No Storage 6hrs Storage Onshore Small Utility (1-20 MW) Fixed Tilt Tracking Rooftop Commercial Residential New to This Year’s Study Cycle New technology characterizations are needed to represent increasing specificity of photovoltaic resources modeled by WECC, especially in the High DG/DSM Case
Approach Determine the cost to install a power plant today (2012) Given limited time, focus is on wind and solar technologies Prior recommendations for other technologies are carried forward Use learning curves to forecast declines in technology capital costs over the next two decades Determine the appropriate applicability of federal tax incentives for renewable technologies over the 10- and 20-year study cycles Develop and apply updated regional multipliers to capture geographic variations in resource costs around the WECC
Notes on Resource Performance With the limited time available before the commencement of the present study cycles, E3’s present scope of work focuses on updating resource costs WECC staff is developing assumptions on resource performance for use in the current study cycle Over a longer timeframe, E3 will work with WECC to ensure that cost and performance assumptions are consistent with one another and represent our best expectations of future development patterns
Present Day Wind and Solar Costs
Present-Day Costs To derive estimates of present-day wind and solar costs, E3 has reviewed a wide range of recent studies and publications For developing technologies, precise capital costs are a moving target that are difficult to pin down A review of literature provides both… …outdated forecasts of what costs would be today; and …retrospective analysis of actual costs from several years ago E3 has used this information to develop its best estimates of costs to install wind and solar plants in 2012 All costs are expressed in 2010 dollars
Historical Trends in Solar PV Costs Installed solar PV costs continue to decrease: Average U.S. behind-the-meter PV data from 1998-2010 (Left) California Solar Initiative (CSI) data from 2009-2011 (right) CSI is focused on rooftop PV Less data available for utility-scale PV and solar thermal Tracking the Sun IV: An Historical Summary of the Installed Cost of Photovoltaics in the United States from 1998 to 2010 California Solar Statistics
Current Trends in Solar PV Prices Market data and experience have shown substantial movement in PV prices over the past two years, suggesting we are on a relatively steep portion of the “learning curve.” This makes identifying current prices a challenging exercise. Source: Technical Potential for Local Distributed Photovoltaics in California
Historical Trends in Wind Costs Average 2010 installed cost was similar to 2009 2010 Wind Technologies Market Report (June 2011)
Data Sources Author Report Name Publication Date Installation Year Historical or Forward CPUC 33% RPS Calculator Update May 2012 2012 Forward E3/CPUC Technical Potential for Local Distributed Photovoltaics in California Mar. 2012 2009 – 2020 Both B&V/NREL Cost and Performance Data for Power Generation Technologies Feb. 2012 2010 - 2050 NREL Residential, Commercial, and Utility-Scale Photovoltaic (PV) System Prices in the United States: Current Drivers and Cost-Reduction Opportunities 2010 Historical DOE SunShot Vision Study 2010 – 2020 CSI California Solar Statistics Jan. 2012 2009 - 2011 LBNL Tracking the Sun IV: An Historical Summary of the Installed Cost of Photovoltaics in the United States from 1998 to 2010 Sept. 2011 2010 Wind Technologies Report June 2011 Lazard Levelized Cost of Energy Analysis – Version 5.0 Sandia Power Tower Technology Roadmap and Cost Reduction Plan April 2011 2013 EIA Updated Capital Cost Estimates for Electricity Generation Plants (for AEO2011) Nov. 2010 2011 NWPCC Sixth Northwest Conservation and Electric Power Plan Feb. 2010 CEC Comparative Costs of California Central Station Electricity Generation Jan. 2010
Solar PV – Fixed Utility (20 MW+) TEPPC 2011 Current Update Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes RETI 2B $3,230 2010 20 Thin Film LTPP $3,400 EIA $3,963 2011 180 TEPPC 2011 2012 100 Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes B&V/NREL $2,396 2015 100 NREL $3,800 2010 187.5 CPUC $2,380 2012 150 California Recommended $2,550 $117 Capital costs for solar PV technologies shown here are expressed relative to the DC nameplate rating. To convert to an AC capital cost, these costs should be multiplied by 1.18 (assuming DC-AC conversion of 85%).
Solar PV – Tracking Utility (20 MW+) TEPPC 2011 Current Update Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes NWPCC $7,294 2008 25 Crystalline RETI 2B $3,825 2010 20 LTPP $3,995 CEC $4,626 TEPPC 2011 2012 100 Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes B&V/NREL $2,664 2015 100 NREL $4,400 2010 187.5 CPUC $2,800 2012 150 California Recommended $123 Capital costs shown relative to DC nameplate rating
Solar PV – Fixed Utility (1-20 MW) Technology has not been represented in past WECC modeling efforts Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes EIA $5,042 2011 8.4 B&V/NREL $2,877 - $3,538 2010 1 - 10 $2,593 - $3,233 2015 CPUC $2,590 - $2,730 2012 5 - 20 California Lazard $2,750 10 Crystalline Recommended $2,975 $135 1 - 20 Capital costs shown relative to DC nameplate rating
Solar PV – Tracking Utility (1-20 MW) Technology has not been represented in past WECC modeling efforts Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes B&V/NREL $3,142 - $3,844 2010 1 - 10 $2,827 - $3,477 2015 CPUC $3,325 2012 1 - 5 Lazard $3,500 10 Crystalline Recommended $3,225 $138 1 - 20 Capital costs shown relative to DC nameplate rating
Solar PV - Commercial Technology has not been represented in past WECC modeling efforts Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (kW) Notes LBNL $5,800 2010 100 – 500 NREL $4,590 217 CSI $5,622 2011 56 California B&V/NREL $4,870 100 $3,904 2015 Recommended $5,000 $256 2012 Capital costs shown relative to DC nameplate rating
Solar PV - Residential Technology has not been represented in past WECC modeling efforts Author Cost ($/kWDC) Generic LCOE ($/MWh) Installation Vintage Size (kW) Notes LBNL $6,600 2010 5 – 10 NREL $5,710 4.9 CSI $6,472 2011 5.5 California B&V/NREL $6,050 4 $4,413 2015 Recommended $6,000 $301 2012 <10 Capital costs shown relative to DC nameplate rating
Solar Thermal – Without Storage TEPPC 2011 Current Update Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes NWPCC $4,761 2008 100 Trough CEC COG $3,687 2010 250 Trough; California RETI 2B $5,350 - $5,550 LTPP $5,300 EIA $4,714 2011 Trough; Wet-cooled $4,692 Tower; Wet-cooled TEPPC 2011 $5,350 2012 Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes B&V/NREL $4,992 2010 200 Trough $4,799 2015 DOE SunShot $4,500 100 Trough; Wet-cooled Lazard $5,000 - $5,400 2012 250 Recommended $4,900 $187
Solar Thermal – With Storage TEPPC 2011 Current Update Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes RETI 2B $7,650 - $7,850 California LTPP $7,500 TEPPC 2011 Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes B&V/NREL $7,178 2010 200 Trough, 6hrs $6,914 2015 DOE SunShot $7,870 250 Lazard $6,300 - $6,500 2012 Trough, 3hrs $7,158 Tower, 6hrs Sandia $7,427 2013 100 Tower, 9 hrs $5,940 Tower, 6 hrs Recommended $7,100 $199 Generic, 6 hrs
Wind TEPPC 2011 Current Update Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes NWPCC $2,127 2008 100 NW RETI 2B $2,150 - $2,600 CA LTPP $2,350 EIA $2,438 2011 TEPPC 2011 Author Cost ($/kW) Generic LCOE ($/MWh) Installation Vintage Size (MW) Notes CEC COG $2,023 2010 50 CA LBNL $2,148 2009-2010 US B&V/NREL $2,013 Lazard $1,300 - $1,900 2012 100 Recommended $2,000 $63
Recommended Resource Costs Cost Summary (2010 $) Technology Subtype DC Capital Cost ($/kWDC) AC Capital Cost ($/kW) Generic AC Capacity Factor (%) Generic LCOE ($/MWh) Solar PV Fixed (>20 MW) $2,550 $3,000 27% $117 Tracking (>20 MW) $2,800 $3,300 29% $123 Fixed (1-20 MW) $2,975 $3,500 $135 Tracking (1-20MW) $3,225 $3,800 $138 Comm Roof $5,000 $5,900 23% $256 Res Roof $6,000 $7,100 $301 Solar Thermal No Storage n/a $4,900 28% $187 6hr storage 36% $199 Wind Onshore $2,000 37% $63 Capital costs for solar PV are converted from DC to AC by multiplying by 1.18 (assuming DC-AC conversion of 85%).
Comparison of Updated Costs to Prior Recommendations Cost Summary (2010 $) WECC 2011 2012 Update Technology Subtype Generic AC Capacity Factor (%) AC Capital Cost ($/kW) Generic LCOE ($/MWh) AC Capital Cost ($/kW) Generic LCOE ($/MWh) Solar PV Fixed (>20 MW) 27% $4,000 $150 $3,000 $117 Tracking (>20 MW) 29% $4,700 $164 $3,300 $122 Fixed (1-20 MW) n/a $3,500 $135 Tracking (1-20MW) $3,800 $138 Comm Roof 23% $5,900 $256 Res Roof $7,100 $301 Solar Thermal No Storage 28% $5,350 $200 $4,900 $187 6hr storage 36% $7,500 $208 $199 Wind Onshore 37% $2,350 $75 $2,000 $63
Forecasting Future Costs for Wind and Solar
Considerations in Forecasting Technology Cost Technology cost changes As nascent technologies become increasingly mature, they may experience cost declines as a result of learning by doing and increased scale of manufacturing Technology costs are sensitive to other factors as well: Trends in the costs of raw materials Relationship of supply and demand Tax credit expiration ITC for solar technologies is set to expire in 2017 PTC for wind expires in 2013; for other technologies in 2012
Learning Curve Theory Learning curves describe an observed empirical relationship between the cumulative experience in the production of a good and the cost to produce it Increased experience leads to lower costs due to efficiency gains in the production process The functional form for the learning curve is empirically derived and does not have a direct theoretical foundation The learning rate (LR) is used to describe the expected decrease in costs with a doubling of experience The theory of learning curves in economics was formalized by Kenneth Arrow in 1962 in “The Economic Implications of Learning by Doing”. This empirical relationship has since been affirmed in a number of works that span many sectors of the economy.
Learning Curves and Solar PV Declines in solar PV module price have tracked the functional form of the learning curve with a learning rate of approximately 20% since 1976 Source: http://www.q-cells.com/uploads/tx_abdownloads/files/11_GLOBAL_OVERVIEW_ON_GRID-PARITY_Paper.pdf Past performance does not indicate future potential Recent cost reductions have not followed the longer-term trends of historical learning Source: Global Overview on Grid-Parity Event Dynamics (Breyer and Gerlach)
Learning Curves and Solar PV Module costs represent only a fraction of solar PV system costs; total system costs have historically declined at a slightly lower learning rate (~17%) Source: Navigant Consulting
Uncertainty in Future Costs Past trends do not guarantee future declines, and other factors influence technology costs Optimistic Path Pessimistic Path Solar PV continues to reap benefits of a high learning rate Global installed capacity grows rapidly Today’s low prices caused by excess supply followed by a rebound as markets re-equilibrate Cost of raw materials rise Int’l markets saturate and US growth slows as the ITC expires
Uncertainty in Global Installed Capacity Future growth of solar PV can vary widely, as shown by the IEA’s 2010 Energy Technology Perspectives scenarios IEA BLUE, High Renewables: renewables serve 75% of load in 2050 IEA BLUE Map: global CO2 emissions reduced to half of 2005 levels IEA Baseline: business-as-usual; no new policies affect energy sector EPIA: http://www.epia.org/index.php?eID=tx_nawsecuredl&u=0&file=fileadmin/EPIA_docs/publications/epia/EPIA-Global-Market-Outlook-for-Photovoltaics-until-2015.pdf&t=1336194876&hash=5f296196c6d2db12aac083a51bbe4491 EIA: http://www.eia.gov/forecasts/ieo/index.cfm IEA: http://www.iea.org/techno/etp/index.asp
Sensitivity of Learning Curves to Global Installations Forecast The choice of a forecast of future installations has a significant impact on anticipated future cost declines The impact of an additional MW of capacity declines as the cumulative installed capacity increases Forecast declines based on a 10% learning rate
Near-Term Outlook for Solar PV E3 has reviewed additional predictions of trends in global installed capacity for solar PV The European Photovoltaic Industry Association’s Global Market Outlook predicts between 208 and 343 GW of solar PV by 2016 Moderate Scenario: pessimistic market behavior, reduced policy support for PV development Policy Driven Scenario: continuation of support mechanisms (FiTs) and strong political favor for solar PV
Forecasting Solar PV Global Installations Through 2032 Short-term market outlook is generally consistent with IEA’s long-term vision The average trajectory of the EPIA’s forecasts results in approximately 1,000 GW of solar globally by 2030 E3 uses the average of the EPIA-derived long-term forecasts to forecast cost reductions for solar PV
Solar PV Learning Rate Recommendation E3 recommends a learning rate of 10% for solar PV, which is applied to the entire capital cost (not just modules) No guarantee that historical rates (17%) will continue Learning rates for mature technologies (coal & gas) have decreased with technology maturation As balance-of-systems components begin to represent larger shares of system costs, learning rates are likely to decrease Coupled with the EPIA-derived long-term PV forecast, this learning rate yields the following estimates of long-term cost reductions
Comparison of Recommended PV Costs to Other Sources
Forecasting Solar Thermal Global Installed Capacity by 2032 IEA’s BLUE Map Scenario includes 600 GW of solar thermal capacity by 2050 To reach this goal, solar thermal global installed capacity would have to reach approximately 200 GW by 2030 European Solar Thermal Electricity Association’s Solar Thermal Electricity 2025 anticipates a cumulative total between 60 and 100 GW by 2025—substantially less E3 has developed a forecast based on the ESTELA forecast that reflects lower anticipated near-term installations of solar thermal facilities Total global capacity installed by 2032 is forecast to be 51 GW Aspirational Pessimistic
Solar Thermal Learning Recommendation Based on stakeholder feedback, a learning rate of 10% was selected for solar thermal Combined with the forecast of global installed capacity from the prior slide, this learning rate yields the following projection of solar thermal cost reductions:
Wind Learning Recommendation Wind is a much more mature technology than either solar PV or solar thermal, with a global installed capacity of close to 200 GW Estimates of learning rates for wind range from 0% - 15%; E3 has adopted a rate of 5% In combination with IEA’s BLUE Map scenario (2,000 GW of wind by 2050), this assumption results in a 12% reduction in wind capital costs by 2032
Federal Tax Credit Landscape Current federal tax policy provides large incentives to wind and solar developers: Accelerated depreciation (5-yr MACRS for wind and solar) Investment tax credit (30% of capital costs for solar) Production tax credit ($22/MWh for wind)
Expiration of Federal Tax Credits Federal tax credits are scheduled to retire in the near future Investment tax credit reverts from 30% to 10% in 2017 Production tax credit ($22/MWh for wind) expires in 2013 PTC for other technologies expires in 2014 The 5-year MACRS, as part of the general tax code, is assumed to remain in place
Combined Impact of Tax Credit Expiration and Technology Learning Increased resource costs resulting from the expiration of tax credits are largely offset by technological progress over the next two decades
Recommended Resource Costs AC Capital Costs by Installation Year (2010 $/kW) Technology Subtype 2012 2022 2032 Solar PV Fixed (>20 MW) $3,000 $2,322 $2,121 Tracking (>20 MW) $3,300 $2,554 $2,333 Fixed (1-20 MW) $3,500 $2,709 $2,475 Tracking (1-20MW) $3,800 $2,941 $2,687 Comm Roof $5,900 $4,567 $4,171 Res Roof $7,100 $5,496 $5,020 Solar Thermal No storage $4,900 $3,455 $2,992 6hr storage $5,007 $4,336 Wind Onshore $2,000 $1,834 $1,711 Recommendations that have been modified since May 15 are highlighted in orange
Resulting LCOEs Levelized Cost of Energy by Installation Year (2010 $/MWh) Technology Subtype AC Capacity Factor (%) 2012 2022 2032 Solar PV Fixed (>20 MW) 27% $117 $109 Tracking (>20 MW) 29% $123 $122 $114 Fixed (1-20 MW) $135 $134 $124 Tracking (1-20MW) $138 $137 $127 Comm Roof 23% $256 $254 $235 Res Roof $301 $299 $276 Solar Thermal No storage 28% $187 $172 $153 6hr storage 36% $199 $182 $161 Wind Onshore 37% $63 $82 $80
Average vs. Marginal
Average vs. Marginal The cost to install one additional MW of solar in 2022 will not equal to the average cost of the solar resources installed between present day and 2022 A large fraction of the solar resources installed by 2022 will have been installed gradually over the next decade
Recommendations for Installed Cost Vintages To account for the many mitigating factors that will affect resource development over the, E3 recommends using the 2015 installed cost for resources installed in the first decade and the 2027 installed cost for resources installed in the second decade To simplify this analysis, E3 also recommends assuming that the PTC is extended through the same time horizon as the ITC, expiring in 2017 First Decade Second Decade Most new resources (especially solar) will come online relatively soon to claim tax credits The choice of 2015 for “average” resource costing reflects this expectation No resources can claim tax credits Year-by-year development of renewables is highly uncertain, so the midpoint of the range (2022-2032) is used as a basis for installed costs
Regional Multipliers
Regional Multiplier Methodology The original Capital Cost Tool included state-specific estimates of technology costs derived from “regional multipliers” Regional multiplier methodology captures geographical differences in costs of labor and materials As part of this update, E3 has explored several questions related to this subject: With the release of an update to the Civil Works Construction Cost Index System, should the regional cost multipliers be updated? What other factors besides construction cost contribute to geographic difference in resource cost and can easily be incorporated into E3’s capital cost tool?
Regional Multiplier Methodology E3 derives technology-specific regional multipliers based on: The relative proportions of equipment, material, and labor that constitute a plant’s costs The Civil Works Construction Cost Index System’s (CWCCIS) state adjustment factors The CWCCIS is a construction-based cost indexing system developed by the US Army Corps of Engineers State adjustment factors capture approximate geographic cost differences in generic construction projects Since equipment costs represent a larger share of costs in power plant construction than in other construction applications, E3 applies state adjustment factors only to the shares of a plant’s cost associated with materials and labor
Sample Comparison of Regional Multiplier Calculations Example regional multiplier calculations, Gas CCGTs in California and Wyoming A B C D E F CWCCIS State Adjustment Factor Category Percent of Total Costs Percent Variable by Location Total Weight {=D x [B x E + (1-E)]} California 1.21 Equipment 70% 0% 0.700 Materials 10% 50% 0.111 Labor 20% 100% 0.242 Total 1.053 Wyoming 0.90 0.70 0.095 0.180 0.975
CWCCIS Update The Army Corps of Engineers released an update to the CWCCIS in March 2011 E3’s prior work was based on CWCCIS from March 2008 Changes to state adjustment factors are minimal but are easy to incorporate into TEPPC pro-forma
Benchmarking Regional Adjustment Factors E3’s adjustment factors capture the same general regional trends as those used by EIA (created by RW Beck)
Other Factors Affecting Relative Geographic Costs State and local tax codes vary widely by location and that have important implications for plant costs The table below shows different tax policies for wind projects in WECC states and their resulting impact on project LCOEs, which range from $88 to $101/MWh AZ CA CO ID MT NV NM OR UT WA WY State Income Tax (%) 7.0% 8.8% 4.6% 7.6% 6.8% 7.9% 5.0% Sales Tax (%) 2.4% 0.8% 4.3% 6.0% 3.0% 0.7% 3.9% 5.4% Property Tax (%) 0.3% 1.0% 0.4% 1.2% 0.5% 0.6% Gross Receipts Tax (%) 6.5% Tax Credit ($/MWh) $10 $3.50 Excise Tax ($/MWh) $1 Generic Wind Cost ($/MWh) $88 $96 $94 $91 $92 $101 $99 Source of information: Tax policies based on E3’s Wyoming Wind Energy Costing Model Generic wind cost calculated assuming capital cost of $2,000 and a capacity factor of 30%
Incorporating State Income Tax One significant variant in state-by-state tax codes that affects the cost of development is the state income tax Ranges from 0% (NV, WA and WY) to 9% (CA) Can be easily integrated into E3’s updated pro-forma In addition to an update of regional multipliers, E3 proposes to incorporate state-by-state income tax rates into the pro-forma to enhance the geographic differentiation of project costs