Embed Size (px)
Citation preview
1
Ramping up Renewables: Leveraging State RPS Programs amid Uncertain Federal Support
Executive Summary • America now generates more kilowatt-hours of renewable energy than any other country in the
world. In addition to creating jobs and bettering the environment, zero-fuel cost renewable generation provides a critical hedge against future increases in fossil fuel prices. Having been growing robustly for over a decade, however, renewable energy in America now faces some major headwinds.
• While federal incentives such as the Production and Investment Tax Credits have bolstered the supply of renewable energy, support for renewable energy demand has come chiefly from states in the form of Renewable Portfolio Standard (RPS) mandates for utilities to procure minimum amounts of electricity from renewables sources. Currently in place in 29 states (and Washington DC), RPS mandates have driven creation of 1/3 of current US non-hydro renewable electricity.
• Curtailment of federal incentives would leave RPS mandates as the chief medium-term driver of new investment in renewable energy. This paper finds, however, that annual new renewable generating capacity needed to meet RPS mandates through 2030 amounts to 3.25 gigawatts (GW) – for 61.5 GW of new generating capacity in total; this level of “RPS demand pull” is slightly below its average level for 2008-2011 and is equal to only roughly 1/3 of total new US renewable generating capacity in 2011 (9 GW) - which was supported on the supply side by federal incentives. Hence, going forward RPS programs will continue to make a valuable but limited contribution to deployment of new renewable generating capacity in the US.
• Increasing cost-competitiveness with conventional generating capacity and remaining incentives should enable continued deployment of renewables in line with “RPS demand pull.” Without new or expanded RPS targets, however, RPS-driven demand will level off – suggesting that policy support for investment in renewable energy will be stagnating just as America’s need for the energy diversity and security benefits of renewable energy is increasing (due to investment in new gas-fired generation and rising exposure to volatile fossil fuel prices).
• To diversify America’s energy mix and hedge against uncertainty in the price of fossil fuels (in
particular natural gas), state policymakers can increase long-term support for renewable energy demand. Measures such as strengthening existing RPS targets, procuring more renewable electricity from large-scale projects through reverse auctions1
, and introducing additional incentives such as CLEAN (Clean Local Energy Accessible Now) programs, feed-in tariffs, and performance-based incentive rebates can increase such support for renewable energy. Provided that they protect the energy diversity benefits of distributed generation, policies to broaden interstate trade in Renewable Energy Credits (RECs) may be another option to consider.
• Sustaining robust deployment of renewables will also require access to capital to support the required level of private investment. Particularly given recent stresses in financial markets, existing tax policies are essential to support new investment in the renewable energy sector. In
1 For more detailed discussion of reverse auction mechanisms (RAMs), see Appendix V.
2
the shorter term, those policies include the Production Tax Credit and Investment Tax Credit.2
Going forward, consideration can be given to how to best coordinate transitional, supply-focused federal tax incentives with longer-term, demand-focused RPS programs as part of an integrated renewable energy strategy for America.
2 The Production Tax Credit is set to expire at year-end 2012. For more discussion, see our companion US PREF paper “Clean Energy and Tax Reform.”
3
Introduction – Diversified and hedged energy system Renewable energy in the US has been booming. In 2011 the US generated nearly 5 percent of its electricity from non-hydro renewable sources – up from less than 2 percent in 2001.3 Over the last decade the number of states generating more than 5 percent of their electricity from non-hydro renewable sources has quadrupled – from five states in 2001 to 20 states in 2011. As a result of this growth, in absolute terms the US now leads the world in generation of electricity from non-hydro renewable sources.4 The US has seen massive increases in both wind power (with annual wind electricity generation increasing 17X since 2001) and solar power (with annual new installed solar PV capacity increasing 20X since 2008). Over just the past five years, deployment of renewable generating technologies has attracted over $100 billion of new investment.5 Increased US deployment has helped to accelerate dramatic reductions in the cost of solar PV and other renewable generation technologies.6
Figure 1 Non-hydroelectric renewable share of total net generation by state
Source: Energy Information Administration (EIA), 2012 The benefits of deploying renewable energy sources include jobs for American workers7, a cleaner and healthier environment, and, especially – by reducing exposure to volatile fossil fuel prices - a more diverse and secure US energy supply. Key facts about the US power sector make the energy security dimension of renewable energy particularly valuable. America now generates roughly 65% of its electricity from coal and natural gas.8
3 Non-hydro renewable sources include wind power, solar PV, solar thermal, biomass, and geothermal. US Energy Information Administration (EIA), Electric Power Annual and Electric Power Monthly, March 2012, based on preliminary 2011 data.
Dependence of the US power sector on fossil fuels will continue
4 The US produces about 70% more than Germany, the next largest non-hydro renewable electricity producer. EIA, http://www.eia.gov/energy_in_brief/renewable_electricity.cfm, May 2012. 5 Bloomberg New Energy Finance (BNEF), May 2012. 6 Among other causes, such cost reductions have come as a result of economies of scale in production and learning-by-doing that facilitates technological progress. 7 See DB Climate Change Advisers paper, “Repowering America: Creating Jobs,” October 10, 2011, http://www.dbcca.com/dbcca/EN/investment-research/investment_research_2396.jsp 8 EIA, Form EIA-906, Power Plant Report, March 2012.
4
as a result of America’s latest “dash to gas” – over half of planned new electricity generating capacity through 2015 is gas-fired.9 Because fuel accounts for more than 50% of the cost of electricity from coal and gas-fired generation,10 these investments risk increasing America’s vulnerability to fuel-price shocks just as new sources of demand (from the industrial, transportation, and export sectors) threaten to ratchet the price of natural gas upward. By diminishing the need for new fossil generation, deploying low to zero-fuel cost renewable resources provides a hedge against future increases in the price of fossil fuels.11 Hedging against fossil fuel price shocks is a compelling reason to ensure sustained investment in developing domestic renewable energy sources.12
Figure 2 Planned Generating Capacity Additions from New Generators by Energy Source, 2012-2015
Moreover, the decline in the costs of renewable energy technologies continues to enlarge the scope of feasible near-term investment in US renewable energy.
Sources: EIA, DBCCA analysis, 2012
Federal and state policies have been instrumental in enabling America to realize the benefits of clean energy. On the supply side, incentives such as the Production Tax Credit, Investment Tax Credit, and 1603 Treasury Cash Grant program have helped to attract capital into the sector. On the demand side, 9 This is largely because the capital costs of new gas-fired combined-cycle units are 30-50% lower than those of new coal-fired units. For comparison of the economics of new gas and coal-fired units, see DB Climate Change Advisers paper, “Natural Gas and Renewables: the Coal to Gas and Renewables Switch is on!,” October 10, 2011, http://www.dbcca.com/dbcca/EN/investment-research/investment_research_2395.jsp 10 According to the Lazard Capital Markets’ June 2012 Levelized Cost of Energy Model 6.0, fuel accounts for 51% of the levelized cost of energy from a coal-fired plant and 66% of the levelized cost of energy from a combined-cycle natural gas plant – versus 0% for solar and wind power generating plant. 11 For more detail, see R. Wiser et al., “The Hedge Value of Renewable Energy”, Lawrence Berkeley National Laboratory, 2004. 12 For more detail, see the companion US PREF paper “Clean Energy and Tax Reform”.
5,150
26,462
1,270
7,280 7,709
0
5,000
10,000
15,000
20,000
25,000
30,000
Coal Natural Gas Nuclear Wind Solar (PV and CSP)
MW
5
state-level Renewable Portfolio Standards (RPS) that require utilities to procure minimum percentages of electricity from renewables sources13 – the focus of this paper – have provided critical “demand pull” to promote deployment of new renewable generating capacity even amid sluggish growth in overall power demand.14
Figure 3 Supply and Demand for Renewable Energy and Relevant Federal/State Policies
Source: US PREF analysis Success in reaching RPS targets will hinge in part on greater amounts of capital formation and accessibility to private financing. Unfortunately, just as the national benefits of deploying renewable energy are set to increase, policies that have motivated significant levels of capital formation and private investment in renewable energy are set to plateau or decline. Combined with similar cuts to other programs, expiration of the 1603 Treasury cash grant program (at year-end 2011) and scheduled expiration of the Production Tax Credit for wind power (at year-end 2012) will reduce federal spending on renewables from $44 billion in 2009 to $11 billion in 2014.15
This creates substantial uncertainty in the near and medium-term outlook for renewable energy, which threatens continued US market momentum and ability to meet RPS targets.
Retrenchment on the federal level will leave RPS mandates in 29 states as the chief medium-term policy driver for new investment in renewable energy. To examine what this means for the growth of renewable energy in the US, this paper proceeds in two sections. Section One introduces RPS policies, 13 Most RPS standards specify a minimum percentage of electricity that must be procured from renewable sources; Texas and Iowa, however, have “capacity-based” standards that simply require addition of a minimum amount (MW) of renewable generating capacity. For more detail, see the Database of State Incentives for Renewable Energy (DSIRE). 14 For more description of policies affecting the US power market, see Appendix I. 15 For detail on federal spending, see “Beyond Boom and Bust: Putting Clean Tech on a Path to Subsidy Independence”, Brooking Institution, April 2012.
Clean Energy Manufacturers
Federal Government Programs:• 48c Manufacturing Tax Credit– EXPIRED at end-2010• 1703/05 DOE Loan Guarantee Program – EXPIRED in 2011
- Currently a lack of federal policy support for renewable manufacturers
Federal Initiatives Under Discussion• Clean Energy Deployment Administration (“CEDA”)
Clean Electricity:
(Meets RPS demands)
Clean Energy Developers
Federal Government Programs:• Investment Tax Credit – for solar, valid through 2016• Production Tax Credit –EXPIRES at end-2012 (wind) and end-2013 (other technologies)• 1603 Cash Grant – EXPIRED at end-2011
- Soon to be a lack of federal policy support for renewable project developers (esp. wind)
Federal Initiatives Under Consideration:• National transmission policy
Cash Payment:
(Made possible through project
financing)
Capital Equipment:
(Wind turbines, PV solar cells,
etc.)
End Market: Clean Electricity Users
Federal & State Government Programs:• Renewable Portfolio Standard– present in most (29) states (and DC and PR) – key tensions between RPS enforcement measures & cost caps• Procurement by government entities
Other Drivers:• Overall economic attractiveness – including supply-side policy drivers
Federal Initiatives Under Consideration:• Clean electricity standards (“CES”)
Investment Grade Cash
Flows:
(Provided by PPAs; critical for
project financing)
Transmission
6
describes their compliance and enforcement mechanisms, and projects RPS-driven demand for new renewable generation and capacity by state (or region) out to 2030.16
Section Two considers ways to cost-effectively amplify the “demand pull” from RPS programs – such as increased RPS targets, best-practice procurement mechanisms, and complementary supply-side tax incentives.
SECTION ONE: CURRENT STATE OF RPS INITIATIVES A Renewable Portfolio Standard (RPS) is a state-level policy intended to increase the generation of electricity from renewable sources. Though no two states have exactly the same RPS program – ability to tailor to state-specific conditions is one of the strengths of RPS policies - the core aspect of an RPS program is a mandate that electric utilities within a state must procure a minimum percentage or absolute amount of electricity from renewable sources by a specific date.17
As of May 2012, 29 states (plus Washington DC) have enacted RPS programs that include such time-bound mandates for procurement of renewable electricity (with an additional 7 states having enacted RPS policies that set voluntary goals for renewable electricity procurement). Such programs often include “set-asides” or “carve-outs” specifying minimum percentages of electricity to be generated from particular technologies; for example, the figure below illustrates RPS policies in 17 states to include set-asides for solar PV. Figure 4 Renewable portfolio standards by state, listing the mandated % of electricity to be procured from renewables and the targeted date
Source: US Database of State Incentives for Renewables and Efficiency, Goldman Sachs
16 These projections draw on data from Bloomberg New Energy Finance (BNEF) and Lawrence Berkeley National Laboratory (LBNL). 17 For more detail on the structure of an RPS regime, see Appendix II.
State renewable portfolio standard
State renewable portfolio goal
Solar water heating eligible *† Extra credit for solar or customer -sited renewables
Includes separate tier of non-renewable alternative resources
Minimum solar or customer -sited requirementState renewable portfolio standard
State renewable portfolio goal
Solar water heating eligible *† Extra credit for solar or customer -sited renewables
Includes separate tier of non-renewable alternative resources
Minimum solar or customer -sited requirement
WA: 15% by 2020*
CA: 33% by 2020
NV: 25% by 2025*
AZ: 15% by 2025
NM: 20% by 2020 (IOUs)10% by 2020 (co-ops)
HI: 40% by 2030
TX: 5,880 MW by 2015
UT: 20% by 2025*
CO: 20% by 2020 (IOUs)10% by 2020 (co-ops & large munis)*
MT: 15% by 2015
ND: 10% by 2015
SD: 10% by 2015
IA: 105 MW
MN: 25% by 2025(Xcel: 30% by 2020)
MO: 15% by 2021
WI: Varies by utility; 10% by 2015 goal
MI: 10% + 1,100 MW by 2015*
OH: 25% by 2025†
ME: 30% by 2000New RE: 10% by 2017
NH: 23.8% by 2025
MA: 15% by 2020+ 1% annual increase(Class I Renewables)
RI: 16% by 2020
CT: 23% by 2020
NY: 29% by 2015
NJ: 22.5% by 2021
PA: 18% by 2020†
MD: 20% by 2022
DE: 20% by 2019*
DC: 20% by 2020
VA: 15% by 2025*
NC: 12.5% by 2021 (IOUs)10% by 2018 (co-ops & munis)
VT: (1) RE meets any increase in retail sales by 2012;
(2) 20% RE & CHP by 2017
29 states & DC & PR have an RPS
7 states have goals
KS: 20% by 2020
OR: 25% by 2025 (large utilities)*5% - 10% by 2025 (smaller utilities)
IL: 25% by 2025
OK: 15% by 2015
WV: 25% by 2025*
PR: 20% by 2035
☼
7
Summing across states (and, where necessary, excluding exempted entities), Figure 4 below shows that 50% of the total US electrical load currently has some form of RPS mandate – with the RPS-covered share rising to 56% upon full implementation of all RPS standards.
Figure 5 Percent of total US electrical load with active RPS mandates
Note: Investor-owned utilities (IOUs), publicly-owned utilities (POUs), and energy service providers (ESPs) Source: “The State of the States: Update on the Implementation of U.S. Renewable Portfolio Standards,” Ryan H. Wiser and Galen Barbose, Lawrence Berkeley National Laboratory, 2011 National Summit on RPS, October 26, 2011.
RPS as Strong Drivers of the Growth of US Renewable Generation Of the 200,000 gigawatt-hours (GWh) of US non-hydro renewable electricity in 2011, one-third represents new renewable electricity spurred by implementation of RPS mandates (“cumulative RPS-driven new GWh)”. In other words, this is renewable generation not in existence at the time a given state RPS mandate was passed, but that has since come into existence and is counted toward fulfilling the RPS mandate. Figure 6 below illustrates this portion. The figure below also illustrates “cumulative RPS-supported GWh”; this includes generation that was in existence at the time a given state RPS mandate was passed, but whose continued existence has been supported by the need to fulfill the state’s RPS mandate. Taking this pre-existing generation into account, “cumulative RPS-supported GWh” are equal to roughly two-thirds overall non-hydro renewable electricity in 2011 (i.e. 133,000 GWh).
8
Figure 6 Impact of RPS programs on growth of US renewable power supply, 1998-2011
Note: “Cumulative RE GWh” includes generation from wind, solar, biomass, and geothermal, but excludes conventional hydroelectric power. “Cumulative RPS-driven new GWh” refers to renewable generation not in existence at the time a given state RPS mandate was passed, but that has since come into existence and is counted toward fulfilling the RPS mandate. “Cumulative RPS-supported GWh” this includes generation that was in existence at the time a given state RPS mandate was passed, but whose continued existence has been supported by the need to fulfill the state’s RPS mandate. Source: Lawrence Berkeley National Laboratory, DBCCA analysis, 2012
Given that RPS mandates have supported creation of new renewable electricity generation, they by definition have supported installation of new renewable electric generating capacity. The figure below demonstrates, for 2006-2011, the portion of total US new renewable generating capacity that was needed to meet RPS mandates for that year. From 2008-2011, new RPS-driven capacity averaged 3.5 GW per year while total new renewable capacity additions averaged 9.5 GW per year (in both cases, wind turbines accounted for the majority of new capacity added). Hence, from 2008-2011 RPS “demand pull” accounted for 36% of total additions of new renewable generating capacity in the US; installation of the other 64% (i.e. 6 GW per year) reflects chiefly decreasing capital costs for renewable technologies and the impact of supply-side federal tax incentives such as the Production Tax Credit and Investment Tax Credit.18
18 This calculation simplifies matters slightly, as some portion of the 6 GW per year reflects new capacity added to meet RPS mandates in future years. That said, willingness of developers to build capacity in advance of RPS obligations reflects in part the economic impact of federal tax incentives. As noted above, developers are eligible for such tax incentives irrespective of whether they are selling power/RECs into RPS markets.
0
50,000
100,000
150,000
200,000
250,000
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
GW
h
Cumulative RE GWh Cumulative RPS-supported GWh Cumulative RPS-driven new GWh
9
Figure 7 Impact of RPS programs on growth of US renewable generating capacity, 2006-2011
Note: RE capacity figures includes generation wind, solar, biomass, and geothermal capacity, but excludes conventional hydroelectric capacity. Source: LBNL, DBCCA analysis, 2012 While the totals have been far smaller, RPS carve-outs for solar generation have been similarly strong drivers of the growth in solar generation and installation of new solar generating capacity.
RPS-driven Incremental Demand for Renewable Electricity and Generating Capacity to 2030 RPS mandates should be thought of as providing a floor on deployment of new renewable generating capacity. In the absence of a national carbon price or clean energy standard, state-level RPS programs – mandating that electric utilities procure minimum percentages of electricity from renewable sources by specific dates - will throughout the rest of the decade remain the chief policy driver of demand for renewable generation. Recognizing this raises questions such as (1) in states with RPS mandates, how much must generation of renewable electricity increase in order to satisfy RPS targets through 2030?; and (2) how much additional generating capacity must be built in order to enable this increase in generation?
Mind the Gaps
A state’s “RPS generation gap” for a given year will be equal to the difference between its share of renewable generation for that year and the share of renewable generation needed to fulfill its ultimate RPS target. The figure below illustrates that, as of 2010, the size of such gaps varied from one percentage point in Pennsylvania to twenty percentage points in California. As RPS mandates have
2.9
6.0
8.4
10.1 10.3
9.1
1.0 1.1 0.9
2.7
6.3
3.9
0.0
2.0
4.0
6.0
8.0
10.0
12.0
2006 2007 2008 2009 2010 2011
GW
annual new RE capacity annual new RPS-driven RE capacity
10
ramped up over the past few years, however, utilities in the majority of states have recorded 90-100% compliance with RPS obligations.19
Figure 8 Renewable generation as a % of retail electricity sales (2010)
Note: Exemptions or lower mandates for certain utilities (e.g. publicly-owned municipal utilities and co-ops) and “credit multipliers” for certain classes of generation (i.e. 2X for in-state generation) will in many states reduce the actual RPS target below the figure shown above. As a result, the existing gap between renewable generation and the target level will be smaller than that shown above. Source: EIA, SNL, AWEA, Goldman Sachs Research estimates.
Looking forward, the analysis below draws on projections from Bloomberg New Energy Finance (BNEF) and Lawrence Berkeley National Laboratory (LBNL).20
Table 1 States within regional power markets or REC trading zones
Taking 2012 as a base year, the figure below depicts estimates from BNEF of the growth in renewable generation required for all states to meet their RPS targets through 2030. Since every state RPS program permits utilities to fulfill some portion of RPS obligations via imports of power and Renewable Energy Credits (RECs) from states within the same regional power market, in some cases the figures below aggregate state level estimates up to the level of the Regional Transmission Organization (RTO) or regional tracking system for trading of RECs. For each region contained in the BNEF projections, the table below lists the states with RPS programs within this RTO (excluding any state for which BNEF provides an individual state-level projection).
Region States with binding RPS programs
PJM Delaware, Maryland, Michigan, New Jersey, Pennsylvania, and the District of Columbia (North Carolina excluded).
ISO-NE Massachusetts, Rhode Island, New Hampshire, Connecticut, and Maine.
M-RETS Iowa, Minnesota, Montana, North Dakota, Ohio, and Wisconsin (excludes Illinois).
WREGIS Colorado, New Mexico, Oregon, Utah, and Washington (excludes California and Nevada).
Source: DBCCA analysis 2012.
19 For data on compliance by state in 2009-2010 (and obligation for 2012-2013), see Appendix III. 20 In the aggregate, figures from BNEF and LBNL are comparable to those from the Union of Concerned Scientists.
13%17%
10%5% 5%
9%5% 2%
9% 9% 9% 8%3% 1% 1% 0% 2%
10% 9% 8% 7%1% 1% 3%
12%9% 7% 6% 5%
7%
33%
25% 24% 23%20%
16% 15%13% 12%
10%8%
CA OR MN IL NY NH CT NJ HI NM CO NV UT MD DE DC RI MT WA MA MO AZ NC VA SD VT WI MI ME PA
Renewable generation
Targeted renewable generation
11
Surveying across states suggests that (1) achieving RPS targets through 2030 will require generating roughly 155,000 GWh of new renewable electricity, and (2) the need for new renewable electricity to meet RPS targets is allocated fairly broadly across California, the mid-Atlantic, New England, and the Midwest.
Figure 9 GWh gaps (as of 2012) – incremental renewable electricity generation required for states to meet their RPS targets by 2030
Note: States/regions in green comprise 80% of total. For states with carve-outs just for solar, excludes generation needed to meet solar carve-out targets. PJM is a 13-state Regional Transmission Organization. ISO-NE is the Independent System Operator for New England. MRETS is the Midwest – Renewable Energy Tracking System. WREGIS is the Western Renewable Energy Generation Information System. ERCOT is the Electric Reliability Council of Texas. Source: Bloomberg New Energy Finance, DBCCA analysis 2012.
Solar carve-outs While most required incremental RPS generation can be met by one of several eligible technologies, in many states a portion of RPS mandates are “carved out” to support generation of electricity from particular technologies. The most common such carve-outs are for generation from solar power and/or from distributed generation (with rooftop solar PV being the most common distributed generation technology).21 Drawing on data from Lawrence Berkeley National Laboratory, the figure below shows incremental generation needed to meet solar carve-outs in 17 states through 2030; the cumulative figure - 14,000 GWh – is roughly equal to total estimated US solar-generated electricity in 2012.22
21 States define such carve-outs differently – with some states specifying just solar, others solar electric, others customer-sited distributed generation, and others customer-sited solar PV.
Just
22 Due to additions of new solar electric generating capacity in excess of current-year solar carve-out requirements in years prior to 2012 (e.g. in Colorado and Arizona), the actual addition of new solar generation and generating capacity needed to meet solar carve-out targets through 2030 may be below LBNL projections. Hence, given
35,843
27,622
20,861 19,168
15,335 14,594
5,855 5,686 4,603 3,959 1,771 527
(3,814)
(10,000)
(5,000)
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
GW
h
2012-2030 RPS gap = 155,824 GWh
12
six states (shaded in yellow below) account for over 80% of the new solar generation needed to meet RPS carve-outs through 2030.23
Figure 10 Solar electricity generation needed to meet solar set-aside targets by state, 2012-2030
Note: States in yellow comprise 80% of total. Addition of solar electricity generation in excess of RPS solar carve-out targets in years prior to 2012 may reduce required incremental solar generation through 2030 below the 14,984 GWh figure above. For states with carve-outs just for solar, amounts shown above are not reflected in the total renewable generation figures shown above. AZ, CO, NM, and NV assumed to meet some portion of required solar generation via concentrating solar power (CSP); all other states assumed to meet 100% of required solar generation through solar PV. Source: Lawrence Berkeley National Laboratory, DBCCA analysis 2012
National impact of RPS programs on generation of electricity from non-hydro renewables To put the above state-level figures in a national context, the figure below illustrates how closing the GWh gaps (inclusive of solar carve-outs) will increase total RPS-supported renewable generation from 3.6% of total US generation in 2012 to 7.9% in 2020 and 9.6% in 2030. Hence, over the two decades RPS mandates will enable a substantial but limited increase in generation of electricity from non-hydro renewables.
current RPS mandates, the LBNL forecasts set an upper bound on the incremental new solar generation and solar generating capacity that RPS solar carve-outs will support. 23 Note that states with the most aggressive solar carve-out targets will not necessarily be the largest solar markets on a MW basis. While solar carve-outs have played a key role in driving deployment (particularly of solar PV), several other high-impact incentives – for example net metering policies and the California Solar Initiative “buy down” rebate program – have been enacted separately from RPS solar carve-outs. That said, a key contribution of solar carve-outs is to help solar PV lower supply chain costs and approach cost- competitiveness in states with unusually low retail electricity rates (e.g. Arizona).
5,870
1,9621,704
1,404
794 698478 421 419 374 298 196 123 102 93 24 23
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
NJ IL AZ MD OH PA CO NM DE MA DC NC MO NV NY OR NH
GW
h
2012-2030 solar carve-out gap = 14,984 GWh
13
Figure 11 Closing all RPS gaps increases total RPS-supported demand to 9.3% of total generation by 2030
Note: 1 TWh = 1000 GWh. “Total RPS-supported demand” includes demand to meet solar carve-outs. “All other US generation” includes non-hydro renewable generation not covered by RPS mandates – which in 2012 will be ~2% of total US generation. Source: Bloomberg New Energy Finance, Lawrence Berkeley National Laboratory, EIA, DBCCA analysis 2012
The figure below compares the average annual increase in non-hydro renewable generation required to meet RPS targets through 2030 with the actual annual increase in non-hydro renewable generation in 2011. To meet RPS mandates through 2030, the US must increase generation of electricity from non-hydro renewables by 8,657 GWh per year over the next 18 years; this annual increase is equal to just one-third of the total (i.e. RPS-related and non-RPS related) increase in generation of electricity from non-hydro renewables in 2011 (27, 820 GWh). The comparison shows that meeting RPS mandates through 2030 requires increasing generation of non-hydro renewable electricity each year by just one-third as much as non-hydro generation increased in 2011. Again, the conclusion is that – on a national basis - RPS compliance will make only a moderate contribution to the growth of renewable electricity in the US.
96.4%92.1%
90.7%
3.6%
7.9%
9.3%
3200
3400
3600
3800
4000
4200
4400
2012 2020 2030
TWh
All other US generation Total RPS-supported demand
14
Figure 12 Projected RPS-driven growth of RE and solar generation vs. total new generation in 2011
Note: “Avg annual new renewable generation to meet RPS targets” excludes generation needed to meet solar carve-outs. Source: Bloomberg New Energy Finance, Lawrence Berkeley National Laboratory, EIA, DBCCA analysis 2012
Building the Capacity to Close GWh Gaps Boosting the share of non-hydro renewables to meet RPS targets will require adding new wind, solar, biomass, and geothermal electric generating capacity. Inferring from above that closing America’s collective RPS gap as of 2012 will require 155,824 GWh of extra renewable electricity, how many gigawatts (GW) of new capacity will be required to generate this extra electricity will depend on (1) the mix of technologies used to generate new renewable electricity; (2) the average net capacity factor24
In forecasting required capacity, BNEF assumes a 33% average capacity factor similar to that of a US onshore wind farm (heretofore the dominant renewable generation technology in the US). BNEF also allows for states to meet their RPS requirements via imports of electricity from other states. Below are the BNEF estimates for the required additions of overall RE capacity to meet RPS targets (note that, for states with carve-outs just for solar, the estimates below omit capacity required to meet these solar carve-outs).
of new generation; (3) the impact of set-asides for particular technologies; and (4) at the level of an individual state or region, how much of each state’s RPS gap can be met via imports of electricity from neighboring states.
24 The net capacity factor of a power plant is the ratio of the actual output of a power plant over a period of time and its potential output if it had operated at full nameplate capacity the entire time. To calculate the capacity factor, take the total amount of energy the plant produced during a period of time and divide by the amount of energy the plant would have produced at full capacity.
832
602
8657
27,820
0 5000 10000 15000 20000 25000 30000
avg annual new solar generation to meet RPS targets, 2012-2030
2011 new solar electricity generation
avg annual new renewable generation to meet RPS targets, 2012-2030*
2011 new US non-hydro renewable electricity generation
GWh
15
Figure 13 New renewable capacity needed to meet RPS targets, 2012-2030
Note: States/regions in green comprise 80% of total. For states with carve-outs just for solar, excludes generation needed to meet solar carve-out targets. PJM is a 13-state Regional Transmission Organization. ISO-NE is the Independent System Operator for New England. MRETS is the Midwest – Renewable Energy Tracking System. WREGIS is the Western Renewable Energy Generation Information System. ERCOT is the Electric Reliability Council of Texas. Source: Bloomberg New Energy Finance, DBCCA analysis 2012.
Since carve-outs for distributed or solar generation must be met by a specific type of capacity (i.e. solar PV or, in some cases, CSP), the figure below draws on LBNL projections to illustrate required capacity additions to meet solar carve-outs through 2030.
12.4
9.6
7.2
6.3
5.0
2.0 1.8 1.5 1.40.6
0.2-
2.00
4.00
6.00
8.00
10.00
12.00
14.00
CA PJM IL ISO-NE MRETS WREGIS NV NY AZ HI KS
GW
2012-2030 RPS build = 48 GW
16
Figure 14 New solar capacity needed to meet solar set-aside targets, 2012-2030
Note: States in yellow comprise 80% of total. Addition of solar generating capacity in excess of RPS solar carve-out targets in years prior to 2012 may reduce required incremental solar generation through 2030 below the 13.5 GW figure above. Values converted from AC to DC assuming 77% conversion ratio. AZ, CO, NM, and NV assumed to meet some portion of required solar generation via concentrating solar power (CSP); all other states assumed to meet 100% of required solar generation through solar PV. Source: Lawrence Berkeley National Laboratory, DBCCA analysis 2012
National impact of RPS programs on additions of new renewable generating capacity The figure below (1) projects average annual additions of renewable and solar generating capacity needed to meet RPS targets and solar carve-outs from 2012-2030 and (2) compares these projections with total new installations of renewable and solar generating capacity in 2011. The chief takeaway is that through 2030 RPS mandates will collectively support the addition of only 3.25 GW of new renewable generating capacity per year (with 0.7 GW of this total linked to satisfying solar carve-outs).25 Compared with figure 6 above, this level of RPS-supported demand represents a decline from the level of RPS-supported demand during 2008-2011.26
25 The 3.25 GW figure reflects the sum of 2.55 GW per year of new renewable generating capacity to meet Tier 1 (i.e. non-technology specific) RPS targets (excluding solar carve-outs), and 0.7 GW per year of new solar generating capacity (mostly solar PV) to meet solar carve-outs.
The secondary takeaway is that (including new capacity to fulfill solar carve-outs) average annual RPS-supported capacity additions through 2030 equal only one-third of total new renewable capacity added in 2011. This suggests that, going forward, RPS programs
26 While demand supported by solar carve-outs is projected to continue to grow after 2011, the sum of Tier 1 and solar-specific demand declines after 2011. As discussed below, since in many states solar installations have already begun to exceed carve-out minimum levels, for solar the more relevant number may be the level of RPS-supported demand for renewables overall, rather than for solar in particular.
5.6
1.9
1.31.1
0.80.7
0.4 0.3 0.3 0.3 0.2 0.2 0.1 0.1 0.1 0.0 0.00
1
2
3
4
5
6
NJ IL MD AZ OH PA DE CO MA DC NC NM NY MO NV NH OR
GW
2012-2030 solar carve-out build = 13.5 GW
17
will make a valuable but limited contribution to deployment of new renewable generating capacity in the US.
Figure 15 Projected RPS-driven new additions RE and solar capacity vs. total new additions in 2011
Note: *RE capacity figures exclude capacity needed to meet solar carve-out targets. RE capacity figures assume 33% capacity factor (similar to that of a US onshore wind farm). A portion of new solar capacity to meet carve-outs is projected to come in the form of concentrating solar power, rather than solar PV. Source: Bloomberg New Energy Finance, Lawrence Berkeley National Laboratory, DBCCA analysis 2012.
Note that a plateau in RPS-driven demand – particularly related to solar carve-outs – does not mean that investment in such technologies will necessarily cease to grow. Increasing cost-competitiveness with conventional generating capacity combined with remaining incentives will enable continued deployment of renewables in line with “RPS demand pull.” In the case of solar in particular, continued declines in technology cost plus the continued availability of other incentives in many states (e.g. upfront rebates) suggest that in future years annual new additions of solar capacity will by no means be limited to 0.7 GW per year. That said, maxing out of solar carve-out (and consequent decline in the level of RPS-related incentives) will in many states – particularly those with below-average retail electricity rates – deprive solar PV of enabling incentives before the technology has become fully cost-competitive with conventional electricity sources. This suggests that increasing the size of RPS targets (including related solar carve-outs) can ensure that solar PV continues to progress toward grid-competiveness throughout multiple regions of the US. The plateau in RPS-driven demand for renewables overall, however, does suggest that policy support for investment in renewable energy will be stagnating just as America’s need for the energy diversity and security benefits of renewable energy are increasing as more new gas plants are built and so exposure to uncertain long term fossil prices will continue. That the plateau in RPS-driven demand will coincide
0.61
0.82
1.79
2.1
2.8
9.08
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
avg annual new solar capacity to meet carve-outs, 2021E-2030E
avg annual new solar capacity to meet carve-outs, 2012E-2020E
2011 Total New US Capacity (all solar PV)
avg annual new RE capacity to meet RPS targets, 2021E-2030E*
avg annual new RE capacity to meet RPS targets, 2012E-2020E*
2011 Total New US RE Capacity
GW
18
with scheduled expiration and reduction of key federal incentives makes the potential negative impact on growth of renewables particularly significant.
How does an RPS Drive New Supply to Meet the Demand? The above section establishes that meeting RPS mandates through 2030 will require the addition of at least 3.25 GW of new renewable generating capacity per year (with at least 0.7 GW per year of this capacity coming in the form of solar – chiefly solar PV – generating capacity). Analysis now turns to the question of what mechanisms are available to promote creation of a new supply of renewable generating projects to meet this demand.
Means of RPS Compliance Broadly speaking, there are three ways for a utility or unregulated distributor to comply with an RPS mandate:
1. Contract for the specified quantity of renewable electricity 2. In lieu of contracting for the delivery of actual electrons, purchase renewable energy credits
(RECs) representing the specified quantity of renewable electricity 3. Purchase neither physical electricity nor RECs, and instead pay a $/MWh Alternative Compliance
Payment (ACP) for each MWh needed to meet the RPS target
Given their central role as a means of RPS compliance, it is worth examining further how supply and demand for RECs operates.
Renewable Energy Credits To varying degrees, 27 states allow utilities and distributors to meet RPS obligations through purchases of Renewable Energy Credits (RECs). A REC is a tradable certificate that represents 1 MWh of qualified renewable energy generation; procuring and retiring a REC can demonstrate compliance with one or more RPS requirements.
RECs can be either “bundled” or “unbundled” from the underlying renewable electricity. A REC that is packaged together with the underlying electricity in a power purchase agreement (PPA) between a renewable generator and a utility/distributor is considered to be bundled; a REC that is sold separately from the underlying electricity is considered to be unbundled. Whether a developer is compensated implicitly (via a higher PPA price) or explicitly (via offers into the REC market), revenue received from REC sales will supplement revenue received from federal incentives such as the Production Tax Credit and the Investment Tax Credit. A state’s Alternative Compliance Payment (ACP) sets a de facto ceiling on the price that its utilities/distributors will be willing to pay for unbundled RECs. For non-technology specific RECs, ACP levels set a generalized ceiling of $50-$65/MWh (considerably lower in some states). High variability in REC prices across states reflects a combination of (1) variation in ACPs; (2) variation in demand for RECs as a means of demonstrating RPS compliance; and (3) various geographic or technology-specific eligibility requirements that limit interstate trade in RECs. As discussed below, if executed properly, minimizing barriers to interstate REC trades represents a highly promising route to lowering the cost of RPS compliance.
19
Figure 16 Compliance market (Tier 1) REC prices, Jan 2008 – Dec 2011
Source: Spectron Group, 2012
Enforcement and Cost-Control – How likely is it that states will meet RPS mandates? What distinguishes mandatory state RPS targets from merely voluntary or aspirational goals is that mandatory targets are (1) codified as legal obligations; and (2) backed by legal mechanisms that serve to encourage compliance. The table below lists the penalties that 12 states with significant RPS programs may levy against utilities whose procurement of renewable generation falls short of prescribed RPS levels. In eight of the states, utilities that in a given year fail to reach quantitative RPS targets incur a cost of $40-$50/MWh for every MWh by which they are short (and a significantly higher cost for every MWh by which they fall short of solar-specific targets); New Jersey, Maryland, and others formalize this cost in the regime of an Alternative Compliance Payment (ACP), while California and Arizona and Oregon empower their Public Utility Commissions (PUCs) to impose the fines. As a means of promoting compliance with RPS targets, however, the per-MWh fines provide roughly equivalent incentives across all four states. The magnitude of the fines is also significant for setting an upper bound on the premium utilities will be willing to pay to procure renewable generation (a topic discussed below).
20
Table 2 Compliance and Enforcement Mechanisms in 12 of the States with the Most Significant RPS Mandates
Specified Monetary Penalties (2013 values) Additional Penalties that PUC may Impose
State ACP $/MWH fine Suspend/Revoke License Additional Fines CA $50/MWh, up to $25MM per utility
IL N/A N/A N/A N/A
OH $45/MWH
NJ $50/MWh x x
MN x (not to exceed cost of
needed RECs)
WA $50/MWh
MD $40/MWh x
MA x x
AZ PUC discretion
OR x PUC discretion
NC PUC discretion
MO x x
Note: Ohio, New Jersey, and Maryland also have separate Solar Alternative Compliance Payments (SACP) used specifically for compliance with solar carve-outs in those states. In Massachusetts and New Jersey utilities/distributors can automatically recover the cost of ACP payments in retail rates; Maryland and Oregon allow cost recovery only if the PUC deems ACPs to have been the least-cost compliance option. Ohio does not allow cost-recovery of ACPs. Washington and Arizona only in some cases allows the cost of discretionary penalties to be recovered in retail rates. Missouri and California do not allow automatic recovery of the cost of $/MWh fines. All columns N/A for Illinois because Illinois relies on a state administrative agency to procure renewables on behalf of the state’s utilities. Source: DSIRE, Union of Concerned Scientists, LBNL DBCCA analysis 2012
Whereas per-MWh fines have a high probability of being imposed but a limited impact, the additional penalties outlined on the right-side of the above table are the reverse – potentially very high-impact, but with an uncertain probability of being imposed. Legislation in New Jersey and Massachusetts empowers the PUCs in these states to penalize chronic RPS violators by prohibiting rate-based recovery of eventual compliance costs, barring a utility from accepting new customers or, in some cases, even suspending or revoking a utility’s license to operate. While very forceful if implemented, assessing the “bite” of these measures is difficult since (1) implementation is at the discretion of the PUC; and (2) few cases have yet occurred to test a PUC’s willingness to merit out such stiff penalties.
The ‘additional fines’ listed in the final column of the above table fall somewhere in between the two categories described above: potentially more severe than a $40-$50/MWh fine, but also more likely to be imposed than suspension or revocation of a utility’s operating license. Vagueness as to the magnitude of these fines (except in Minnesota, where they cannot exceed the cost of purchasing needed RECs or procuring the needed amount of renewable energy), however - as well as a lack of cases demonstrating PUCs inclination or disinclination to impose them – again complicates the task of assessing just how strongly they influence utilities’ attitude toward RPS compliance.
Cost Caps Extending the reasoning that a utility’s calculus for RPS compliance will reflect chiefly economic considerations, embedded in RPS legislation is frequently a state’s calculus of how much it is willing to
21
invest to increase the generation of electricity from renewable resources. The two middle columns of the table below enumerate cost caps that, in eight of the twelve states, constrain the total dollar amounts to be spent on RPS compliance. Note that these caps apply to the “incremental” cost of complying with an RPS target – that is, the cost of purchasing required renewable MWhs (or RECs) over and above the cost of purchasing an equivalent amount of non-renewable energy.27
Table 3 Cost-Containment Policies in 12 of the States with the Most Significant RPS Mandates
Such caps can apply either to costs for consumers or to retailers (as noted above, an ACP or prescribed per-MWh fine serves as a de facto cap on how much a retailer will spend on incremental RPS compliance). Seven states in the table below specify the precise level a cost cap should take (whether as a $/MWh amount, a per-account amount, or a percentage of the total electric bill); Arizona, on the other hand, leaves it to the discretion of its PUC to determine whether ratepayer surcharges in support of RPS compliance are “just and reasonable.” In all cases, however, these caps impose some threshold on the amount to be spent on renewable energy.
Consumer Distributor
State Ratepayer Surcharge
de facto Cost Cap
Delay/Escape Clauses
CA N/A N/A N/A
IL 2.015% of 2007 retail
rate
If annual cost cap is met or exceeded
OH $45/MWh Caps compliance costs at 3% premium above cost of non-renewable
electricity
NJ $50/MWh
MN PUC discretion for two-year delay
WA If incremental compliance costs exceed 4% of utility’s annual retail revenue
requirement
MD If cost of purchasing non-solar Tier 1 RECs exceeds 8% of utility’s annual
electricity sale revenues in MD*
MA ACP
AZ PUC discretion PUC discretion
OR If incremental compliance costs exceed 4% of utility’s annual retail revenue
requirement
NC $12.00/account Ratepayer surcharge cap; and PUC discretion
MO 1% of average retail
rate
Note: All columns N/A for California because SB 2 (1X2), signed by California’s governor in April 2011, eliminated the state’s existing cost caps; new cost-containment mechanisms are in development and have not yet been enacted. In Maryland, there is a delay/escape clause for compliance with the state’s solar carve-out if the cost of purchasing solar RECs exceeds 1% of utility’s annual electricity sale revenues in MD. Source: DSIRE, LBNL, Union of Concerned Scientists, DBCCA analysis 2012.
27 While states define “incremental” in different ways, there is some uniformity across states. Washington state’s RPS statute, for example, defines the incremental cost of an eligible renewable resource is determined by calculating “the difference between the levelized delivered cost of the eligible renewable resource, regardless of ownership, compared to the levelized delivered cost of an equivalent amount of reasonably available substitute resources that do not qualify as eligible renewable resources, where the resources being compared have the same contract length or facility life.”
22
Cost-cap proceedings are currently underway (or set to begin shortly) in Ohio, New Mexico, Colorado and Arizona. While most observers expect existing RPS targets and compliance deadlines to emerge from these proceedings largely intact, they illustrate that achievement of RPS targets in contingent on staying within prescribed economic constraints.
Delay/Escape Clauses By limiting the annual or marginal cost of RPS compliance, a cost cap may reduce the amount of renewable generation that a state procures in a given year - potentially delaying achievement of annual RPS targets. Illinois, North Carolina, and Missouri explicitly provide that, if a utility’s actual or projected cost of meeting annual RPS targets exceeds the state cost cap (e.g. $12.00/residential account in North Carolina, or 2.015% of the 2007 residential rate in Illinois), the state PUC may lower the utility’s annual RPS target to the level necessary to stay within the cost cap.
While capping costs for the retailer rather than the consumer, Maryland, Oregon, Washington, and Ohio similarly empower their PUCs to reduce annual RPS obligations if compliance costs exceed prescribed thresholds. In Maryland the relevant threshold is a percentage of the cost of a state’s annual electricity sale revenues in the state (e.g. 8% in 2013). Finally, while not prescribing specific thresholds, Minnesota and Arizona empower their PUCs to temporarily suspend RBS obligations should the cost of compliance exceed a “just and reasonable” level.
The cost caps described above set an upper limit on the incremental cost of new renewable electricity to meet RPS targets. While examining the likelihood that every individual cost cap will prove binding is beyond the scope of this paper,28
SECTION TWO: WAYS TO MAKE RPS PROGRAMS MORE EFFECTIVE
in general the costs of new renewable electricity are sufficiently below the caps embedded in RPS mandates that most states are highly likely to meet their RPS obligations and deadlines out to 2030. Evidence in support of this judgment comes from the success that several states wind-rich states (i.e. Texas, Iowa, Oregon) have had in either meeting their end-state RPS targets or complying with such targets well ahead of mandated deadlines. As declines in the cost of solar technologies broaden another means for cost-effective RPS compliance (already in wide use in California and other western states), the likelihood that states will meeting existing RPS targets without delay or interference only increases.
This paper has so far sought to establish that (1) in the aggregate, RPS mandates through 2030 will support 3.25 GW of new renewable generating capacity per year – a valuable but limited amount; and (2) that – given the incremental costs of RPS compliance and the mechanisms promoting compliance – such mandates are highly likely to be fulfilled.
To help America realize the full long-term economic, environmental and energy security benefits of renewable energy – as well as the more immediate benefits of a vital US renewables industry – there are multiple ways to ramp up and sustain support for renewable energy demand. Four promising measures to consider include: stronger RPS targets, best-practice procurement mechanisms, support for transmission development, and, potentially, broader interstate REC trading.
28 Given the broad latitude that many states accord to their PUCs in determining exactly what these levels are, this suggests that the robustness of a given RPS target depends largely on (1) the cost competitiveness of the state’s renewable resources; and (2) how willing the state’s PUC is to integrate RPS compliance into its traditional “least-cost” mandate. For more detail on potential outcomes for an RPS program, see Appendix IV.
23
Strengthen and broaden RPS targets Since existing RPS mandates appear highly likely to be satisfied (in many states, ahead of schedule), states can examine ways to make their RPS targets more robust. A state might do this by either raising an RPS target for an existing compliance year, coupling a longer compliance timeframe with higher end-state targets, or extending the RPS to cover entities previously omitted. There is precedent for states undertaking such action. Perhaps the best known example is in California, where in 2009 Governor Arnold Schwarzenegger issued an executive order increasing California’s RPS target from 20 percent by 2010 to 33 percent by 2020. In Colorado, the state legislature in 2007 modified the state’s RPS program (which emerged from a 2004 ballot initiative) to include a higher target; the 2007 modifications also added separate renewable energy requirements for electric cooperatives, which previously had been omitted from the state’s RPS mandate.29
In 2010, the Colorado legislature again increased the state’s RPS target, to its current level of 30% by 2020. Moreover, earlier this year the Maryland legislature enacted a bill accelerating compliance deadlines for the state’s solar carve-out (for example, changing the ultimate 2% target date from 2022 to 2020).
Efforts to increase existing RPS targets are currently underway in several states, including Michigan and New Jersey. In Michigan, a 2012 ballot proposition seeks to raise the state’s RPS target (currently 10% by 2015) to 20% by 2025. Supporters of the Michigan ballot proposition point to the neighboring Canadian province of Ontario – which already generates 25% of its electricity from non-hydro renewables – as evidence that the 20% by 2025 goal is achievable. In New Jersey, the state legislature is closely considering a bill to accelerate compliance deadlines for the state’s solar carve-out (similar to the Maryland legislation discussed above).
Given the impressive progress many states have so far made toward meeting existing RPS targets, in many states the time may be ripe to increase or broaden existing targets. Note that broadening targets – via either including entities currently omitted, or increasing targets for entities whose current targets are below the state average – may be a particularly appealing way to boost the demand-pull of RPS programs.
Moving away from states that already have mandatory RPS programs, another option is to make existing voluntary programs mandatory. For example, states such as Virginia and Utah have voluntary renewable energy targets that they are still some ways from meeting.30
Finally, whether by statute, ballot initiative, or administrative action, implementing new, binding RPS targets in any of the 14 states that currently lack either a binding RPS target or voluntary goal would serve to increase demand for renewable energy in these states.
Making such programs mandatory – hence introducing the economic incentives for compliance outlined above – would significantly boost demand for renewable energy in these states.
31
29 In 2010, the Colorado legislature again broadened the reach of the state’s RPS program. For investor-owned utilities, the current target is 30% by 2020; the state’s 30 municipal electric utilities and electric co-ops must meet a target of only 10% by 2020.
30 Conversely, North Dakota and South Dakota have voluntary targets that largely have already been met. 31 Should the political climate prove unfavorable to new or more aggressive RPS targets, state policymakers might also consider promoting deployment of renewable generation via upfront, capacity-based rebates (as in the California Solar Initiative). For example, in 2011 the Texas legislature considered (but did not ultimately pass) a bill to implement a CSI-style rebate program for solar PV in Texas.
24
Enable access to the highest-quality resources: Finance Needed Transmission State public utility commissions (PUCs) can affect the cost of bringing new renewable generation online by virtue of the rules that they adopt for transmission cost-allocation. One strategy for PUCs to do this is to allow cost-recovery regimes that encourage interconnection of the highest-grade renewable resources. In many states the sites with the highest-class renewable resources (particularly for wind and solar) are remote from load centers and unconnected to the bulk transmission grid. Under typical cost-recovery rules, new generation projects must pay to construct their own interconnection (called a “tie-line”). For location-constrained renewable sites, however, the cost of such lines can be very substantial – deterring new investment. Developers are reluctant to build project with no guaranteed transmission; transmission owners are reluctant to build new lines without assurance of demand from generators.
California has pioneered a solution to this chicken-and-egg problem with its “Location Constrained Resource Interconnection Facility” program. Essentially, this allows the state’s three IOUs to build new transmission lines to designated resource-rich areas and recover the costs from ratepayers on terms similar to which they recover the costs of typical “reliability” or “economic” transmission investments (i.e. upgrades that benefit the bulk power system generally rather than a single generator). As renewable projects come online, they are allotted a pro rata share of the transmission costs. By enabling development of projects that otherwise would be uneconomic, California’s LCRIF program offers a creative way to support development of renewable resources at the lowest cost.
Other states might adopt similarly innovative programs to support interconnection of location-constrained renewable resources.
Encourage development of the highest-quality resources: Reverse Auction Mechanisms (RAMs) Renewable energy projects of different sizes require different procurement regimes. To encourage continued cost reductions in large-scale projects (i.e. 20 MW and above), more states might procure renewable energy contracts via reverse auction mechanisms (RAMs), such as the ones in place in California and other states. Assuming innovative mechanisms to finance the needed transmission, there remains the question of how to encourage developers to build projects in the highest-capacity regions. One approach involves a so-called Reverse Auction Mechanism (RAM) – where utilities solicit developers to bid in projects at the lowest $/MWh price. International experience suggests that (1) RAMs are successful at encouraging developers to build in the highest-capacity locations; and (2) RAMs – particularly if implemented without adequate pre-screening criteria – may succeed in elicit low-cost bids, only to see many of these bids fail to be built as the contract prices leave developers without sufficient incentive to build the project to completion.
For more discussion of the pros and cons of RAMs, see the attached case studies on California, Brazil, and Texas in Appendix V.
Broaden interstate REC trading (carefully) Expanding inter-state trade in Renewable Energy Credits (RECs) may be another cost-effective means to meeting higher RPS targets. Each state has the discretion to choose what level of RPS compliance it will allow through out-of-state RECs. Given the potential impacts of increased interstate REC trading on local distributed generation, however, implementing this option may require coordinated regional or even national action. Expanding interstate REC trading is likely to succeed only if pursued in concert with the critical energy diversity and local economic benefits of RPS programs. Hence, if pursued,
25
policymakers might consider coupling greater interstate REC trading with stronger carve-outs for distributed generation technologies (in particular, rooftop solar PV).32
RPS programs and renewable resources such as wind and solar resemble each other in that they are all location-specific: as an RPS mandate imposes responsibilities only on electric utilities within a defined geographic jurisdiction, so too wind and solar generators will produce cost-competitive electricity only if placed in specific locations. Regrettably, these locations do not always coincide. While large states such as California may have diverse and plentiful resources to support pursuit of RPS goals, many smaller states have less ability to meet RPS targets with in-state resources. Moreover, the table below illustrates that the cost of developing site-specific renewable resources varies widely across the US. Whereas utilities in Massachusetts and Delaware have contracted for electricity from first-of-a-kind offshore wind projects at a cost of $185-200+/MWh, bids from onshore wind projects in Kansas’ most recent RPS solicitation reached as low as $30/MWh.
Recognizing that in-state economic development has motivated support for adoption of RPS standards in a variety of states, meeting the goals that have been adopted will for many states require more, not less, interstate trade of RECs. Different states impose different (1) limits on the portion of RPS obligations that may be used through out-of-state RECs; and (2) limits on the portion of RPS obligations that may be met through unbundled RECs.33
The majority of states have adopted relatively flexible REC procurement policies, allowing RECs to qualify for compliance so long as the associated renewable power generated is delivered either within the state’s relevant regional transmission organization (RTO) or independent system operator (ISO) territory. This facilitates intra-regional REC trade – between neighboring states that are part of the same RTO or ISO – with a goal of delivering the lowest cost renewable power to market. It is unusual, however, for a state to allow the import of RECs from another region (outside its RTO or ISO), or for it to allow all RECs to be unbundled (typically requiring a maximum portion of unbundled RECs), so as to ensure that not all the employment and economic benefits of RPS-driven renewable power development are being exported out-of-state. Currently then, the US REC market tends to fall along the lines that define the US grid – as demonstrated in the figure below.
The rules for REC procurement, sale and retirement vary substantially by state, with some states only allowing RECs to cover a portion of the RPS requirement, some disallowing out-of-state RECs, some allowing RECs to be “banked” for future years, some requiring that RECs are bundled or only a portion can be un-bundled, and many other state-by-state variations. The bundling vs. unbundling rules have received considerable attention as these help determine the liquidity of a state or region’s REC market, and also the cost of both renewable power and RECs within that market.
32 To the extent that policymakers are wary that RPS carve-outs will attract Commerce Clause challenges – such as the one that TransCanada filed in 2010 against the Massachusetts carve-out for in-state solar distributed generation (a suit that is the currently on hold owing to a partial settlement) – there are several ways to design a carve-out so that it is more likely to withstand such challenges. These strategies include (1) emphasizing in RPS legislation the state’s legitimate interest in non-protectionist goals such as environmental protection, reliability, and diversity of power sources; (2) phasing in new in-state RPS requirements gradually; and (3) opting for in-region (rather than in-state) location-based eligibility requirements. For a full discussion of RPS programs and the Commerce Clause, see Carolyn Elefant and Edward A. Holt, “The Commerce Clause and Implications for State Renewable Portfolio Standard Programs,” Clean Energy States Alliance, March 2011. 33 For state-by-state rules on REC trading, see Appendix VI.
26
Figure 17 US REC Import Flows Relative to RTO/ISO
Note: ISO-NE is the Independent System Operator for New England; NYISO is the New York Independent System Operator; PJM is a regional transmission organization that coordinates the movement of wholesale electricity in all or part of 13 states and the District of Columbia; SERC is the SERC Reliability Corporation; FRCC is the Florida Reliability Coordinating Council; MISO is the Midwest Independent System Operator; SPP is the Southwest Power Pool; ERCOT is the Electric Reliability Council of Texas; WECC is the Western Electricity Coordinating Council; and CAISO is the California Independent System Operator. Source: “US RPS Markets and Utility Strategies”, Emerging Energy Research, 2010 Despite these state-specific requirements, there has been a general trend among RPS states towards greater flexibility regarding regional trading of RECs. California, for example, in March 2010 decided to allow utilities to use tradable RECs (TRECs) and unbundled renewable power (from out of state) to satisfy the state’s RPS, up to a certain proportion of overall renewable power supply. Among other reasons, the state was concerned it could not meet its ambitious targets34 without allowing greater flexibility, and the policy change also intended to control the costs of renewable power by facilitating greater REC trading within the. While over the next few years California’s RPS program will increase the share of power that must come from in-state resources, its 2010 action served to stimulate broader trading of RECs within the Western Electric Coordinating Council (WECC) states.35
Currently, a lack of public information combined with multiple criteria within RPS states (e.g. technology carve outs) and different criteria among RPS states (e.g. with regard to Alternative Compliance Payments; allowable banking periods, etc.) exacerbates the illiquidity and fragmentation of US REC markets. Even accounting for differences in power prices among states, there is substantial price disparity among REC markets in the US - suggesting that such markets are fragmented and beset by high transaction costs. This finding is supported by an analysis of bid-ask spreads in US REC markets, which are found to be excessively high in order to cover illiquidity risks. This is a particular problem in states
34 Utilities are required to have 33% of retail sales derived from eligible renewable energy resources by 2020, with interim targets of 20% by the end of 2013 and 25% by the end of 2016. 35 Note, however, that as a result of significant available supply, California utilities demand for out-of-state RECs is now almost fully satisfied through 2018.
27
that do not publish regular information on REC market conditions, making market visibility difficulty, and sometimes resulting in consistently high REC prices that are not necessarily reflective of market fundamentals – or how close a state is to meeting its standard without a need for RECs.
Provided that it is implemented in a manner that protects the growth of local distributed generation, for example through carve-outs or other mechanisms, increased interstate trading may help to reduce illiquidity and market “thinness”; if implemented effectively this may help states to achieve RPS goals more cost-effectively. However, to be effective, interstate trading would require timely and detailed public information on renewable generation, RPS requirements, level of banking, etc. in participating states – this would enable market participants to infer REC market conditions and minimize excessive transaction costs. Note also that implementing expanded REC market trading may take coordinated regional or national action, and is likely to take years to come about.
Supporting continued growth of residential distribution generation In addition to the policies described above, there are other ways for states and localities to promote demand for renewable electricity - particularly from distributed generation sources such as solar PV. Recent initiatives in California and Colorado highlight two such measures: net metering and performance-based incentives.
• Net metering: this policy allows owners of distributed generation systems (such as rooftop PV) to sell excess power back to the grid, generally for the full retail price. More than 40 states have some form of net metering, and these policies have contributed substantially to the growth of residential solar in the US. Recently, the California Public Utilities Commission (CPUC) voted to change the measurement formula for the state’s “net metering cap” (which limits the amount of electricity that may be sold back to the grid); analysts predict that the change may double the amount of distributed solar power that is eligible for net metering (i.e. from 5 to 10 percent of California’s aggregate peak load).36
• Performance-based incentives (PBI): Programs that offer homeowners an upfront rebate on the cost of distributed generation technologies (with the dollar size of the rebate linked to the kW size of the system) - such as the California Solar Initiative – have been highly successful at promoting growth of rooftop solar PV in the US. As the cost of rooftop PV systems continues to decline, however, some states are transitioning away from upfront, capacity-based rebates toward performance-based incentives (PBI). As opposed to providing a $/kW rebate when the system is purchased, PBIs offer a $/kWh incentive as an installed system actually generates electricity. For example, in March 2011 Xcel Energy (Colorado’s largest utility) replaced its existing $2,000/kW up-front rebate for small (less than 10 kW) solar projects with a 20-year PBI that begins at $0.15/kWh for customer-owned systems and $0.10/kWh for third-party-owned systems.37 While the program’s reduced incentive levels may put new residential solar systems at a competitive disadvantage with Colorado’s retail electricity rates,38
36 While the PUC voted to change the measurement formula, action of the California legislature will be required to actually change the cap itself.
at a high level Colorado’s
37 Eventually declining to $0.11/kWh for customer-owned small systems and $0.07/kWh for third-party-owned small systems. In total for 2012-2013, the Xcel PBI is available to 19.2 MW of new small solar systems. Colorado Public Utilities Commission, “2012-2013 Solar Incentives in Xcel Energy Territory”, http://www.coseia.org/policy/public-utilities-commission.html. 38 Achievement of Colorado’s solar carve-out in 2012 triggered a reduction in the level of incentives available to residential solar systems. Moreover, from the perspective of maximizing supporting for investment in residential
28
PBI regime suggests a transitional mechanism that can reduce expenses for utilities while continuing to support deployment of residential solar systems.
Don’t forget about wholesale distributed generation: CLEAN and feed-in tariff programs One often neglected market segment is wholesale distributed generation: projects of 1-20 MW in size that – rather than off-setting customer usage (as is the case with residential solar PV) – generate power on the utility-side-of-the-meter and sell at wholesale rates to either a utility or electricity retailer. To the extent that policymakers seek to support growth of this market segment, a promising way to do so is through CLEAN (Clean Local Energy Accessible Now) programs. CLEAN programs (also known as feed-in tariffs) offer standard, fixed price, long-term power purchase agreements; while the offered price in such programs is usually determined up-front, it may then later be adjusted as the market responds. Such programs are particularly promising for promoting the growth of “wholesale distributed generation,” meaning distributed generation of 1-20 MW in size. Following passage of California Senate Bill 32, the CPUC has recently released details of a new CLEAN mechanism in California. The mechanism, known as Renewable Market Adjusting Tariff (Re-MAT) will be available for systems up to 3 MW in size; the Re-MAT programs links payments to owners of renewable energy systems to the weighted average contract price that California’s three investor-owned utilities recorded in their Nov 2011 reverse auction.39
For more detail on CLEAN programs in general and the specifics of California’s new program in particular, see Appendix VII.
In addition, a FERC order in 2011 regarding implementation by the California Public Utilities Commission of a feed-in tariff to support development of combined heat and power generation (134 FERC ¶ 61,044 (2011) (January 20, Order Denying Rehearing) paves the way for even greater use of feed-in tariffs to meet state RPS and other policy objectives. In this order FERC found the concept of a multi-tiered avoided cost rate structure to be consistent with the avoided cost rate requirements set forth in the Public Utilities Regulatory Policy Act (PURPA) and its subsequent regulations.40
Complementary Federal Policies through the PTC and ITC
This ruling affords states greater ability to establish feed-in tariff rates at levels that would support private investment, including in renewable energy generation.
Since more ambitious RPS targets and new procurement mechanisms are unlikely to emerge rapidly, federal tax incentives can complement these longer-term, state-level initiatives with more targeted transitional efforts. The Production Tax Credit and Investment Tax Credit can help to sustain private investment in renewable energy.41
solar, policymakers might consider how to structure PBIs so that – on a net present value basis – the incentive available to customer-owned systems is equivalent to that available to third-party-owned systems.
Going forward, coordinating transitional, supply-focused federal tax incentives with longer-term, demand-focused RPS programs can be explored as part of an integrated renewable energy strategy for America. Understanding more clearly how these two sets of policies interact can help inform what an appropriate synthesis will look like. Entities such as the National Renewable Energy Laboratory and Lawrence Berkeley National Laboratory could examine federal-state policy interaction on renewable energy.
39 California Public Utilities Commission, Decision 12-05-035, May 24, 2012 http://docs.cpuc.ca.gov/WORD_PDF/FINAL_DECISION/167679.pdf 40 Passed by Congress in 1978, PURPA has played a key role in promoting growth of domestic renewable energy in the United States. 41 At present, the PTC is set to expire at the end of this year. For more discussion, see our companion US PREF paper “Clean Energy and Tax Reform”.
29
Figure 18 Volatility of US wind installations in years surrounding PTC renewal, 1999-2012 (MW)
Sources: AWEA, 2012; Bloomberg New Energy Finance, 2012
Conclusion This paper finds that fulfilling RPS mandates through 2030 will support the addition of 3.25 GW of new renewable generating capacity in the US. This represents a slight decrease from the level of RPS-supported new capacity additions in 2008-2011, and amounts to roughly only one-third of total new renewable generating capacity added in the US in 2011 (9 GW). These comparisons inform a central finding of this paper that RPS mandates will make a valuable but limited contribution to demand for new renewable electricity in the US.
Given that deploying renewable generating capacity benefits America’s economy, environment, and energy security, consideration should be given to ways to enhance demand for renewable electricity. Promising measures to consider include strengthening existing RPS targets, procuring more renewable electricity through reverse auctions (for large-scale projects), and introducing more local incentives such as CLEAN (Clean Local Energy Accessible Now) programs, feed-in tariffs, and performance-based incentive rebates. Provided that they protect the energy diversity and local economic benefits of distributed generation, policies to broaden interstate trade in Renewable Energy Credits (RECs) may be another option to consider. In the shorter term, state-led efforts can be complemented with supply-side incentives such as the Production Tax Credit and Investment Tax Credit. Going forward, coordinating transitional, supply-focused federal tax incentives with longer-term, demand-focused RPS programs as part of an integrated renewable energy strategy for America should be explored. The actions outlined above will inject a healthy dose of competition and innovation into the US power sector. Additionally, they will enable the US to continue to reap the benefits of large-scale renewable energy deployment – including lower long-term costs, greater technological diversity, utilities, businesses and consumers, and increased protection from the economic and security risks inherent in fossil fuels.
0
2,000
4,000
6,000
8,000
10,000
12,000
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Annu
al W
ind
Cap
acity
Add
ition
s (M
W)
92% Drop
76% Drop
76% Drop
Production Tax Credit Expiration Years
Wind market likely to drop off in 2013
Forecast / Estimate (with PTC extension)
Forecast / Estimate (no PTC extension)
30
Primary Authors: Mark Fulton and Reid Capalino - Deutsche Bank Climate Change Advisers (DBCCA). Contributing authors: Camilla Sharples and Michael Carboy – DBCCA; Michael Conti - Hudson Clean Energy Partners (Texas RAM case study); Craig Lewis and Ted Ko – CLEAN Coalition (CLEAN Programs case study); and Michael Wheeler - Recurrent Energy (California RAM case study). This paper benefitted from review by the following individuals: Steve Corneli – NRG Energy; Todd Foley and Jeramy Shays – ACORE; Holly Gordon – SunRun; and Bonnie Suchman – Troutman Sanders LLP. ABOUT US PREF US PREF is a coalition of senior level financiers who invest in all sectors of the energy industry, including renewable energy. Members educate the public sector to assure renewable energy finance legislation impacts the market as efficiently and effectively as possible, with the goal of helping to unlock capital flows to renewable energy projects in the United States. US PREF is a program of the American Council On Renewable Energy (ACORE), a Washington, DC - based non-profit organization dedicated to building a secure and prosperous America with clean, renewable energy.
31
Appendix I: Market Structure and Relevant Policies of US Clean Energy Industry Like nearly all other energy markets, the clean energy market in the US operates as a function of supply and demand – with supply being provided by clean energy manufacturers and project developers, and demand coming from users of clean electricity. In this appendix we lay out the fundamental components of both the supply and demand ends of the market, and explain at a high-level the dynamics between the two.
Supply Chain On the supply side there is a clean energy supply chain that can be split into 3 segments: the upstream, midstream and downstream. The upstream consists of manufacturers of materials or components for clean energy products (e.g. solar wafers), while the midstream consists of manufacturers of finished clean energy products (e.g. solar modules). The downstream refers to clean energy project developers – those who actually build and operate clean energy projects in the US – with debt and equity financing sourced from a range of market players (e.g. private equity investors, banks and utilities). These projects can range from large, utility-scale projects (e.g. wind or solar farms), to small, residential scale projects (e.g. rooftop solar), with the former entering into Power Purchase Agreements (PPA’s) – usually with utilities – for fixed terms and prices. This collection of sub-industries can be referred to as the clean energy supply chain, and it is a market segment that is driven by both a combination of government incentives and manufacturing or project economics. Demand Pull: Wholesale vs. Retail On the other side of the market there exists the demand “pull” created by actual users of clean power, and these users can be either wholesale or retail. Wholesale refers to large-scale users such as utilities, who purchase power direct from project developers (be it coal, gas or clean energy power projects) under long-term PPA’s, thereby achieving relatively low (and fixed) prices (the “wholesale market”). Retail, by contrast, refers to smaller-scale users, such as residential (i.e. individual households), commercial or industrial consumers (the “retail market”). These types of buyers typically purchase power from utilities, and the power prices paid by retail users are nearly always higher as they include the costs of transmission and distribution (T&D), as well as a return earned by the utility for providing these services. In addition to these price differentials between wholesale and retail markets, there are also substantial price differences by electricity market (state and/or regional) in the US, depending on various factors including the power mix, age of the distribution system, local taxes, and many others.
Role of Policy in the Supply-Demand Dynamic Underlying US power markets are various policies and regulations that are designed to secure a stable energy supply and fair pricing for consumers. In the case of clean energy, policy support has been particularly crucial as these technologies are relatively new and are still experiencing steep cost declines as the technologies mature and economies of scale are achieved. Figure 2 in the introduction of this document shows the interdependent and complementary nature of clean energy markets in the US – from manufacturers to developers to end consumers – and the main federal and state policies (both current and expired) that underlie these market segments. In this paper we assume policy to continue as currently scheduled: the PTC expiring at the end of 2012 for wind (or 2013 for some other types of renewable power), the ITC for solar expiring at the end of 2016 (although a 10% tax credit will remain in place beyond this date), and the already expired 1603 Treasury Cash Grant, DOE Loan Guarantee Programs, and Manufacturing Tax Credit (MTC)
32
Appendix II: Structure of an RPS regime and frameworks for procurement Enactment and fulfillment of an RPS policy involves five sets of players: state legislatures and governors, electric utilities/electricity suppliers, Public Utility Commissions (PUCs), renewable energy developers, and utility ratepayers.
State Legislatures and Governors Of states that have adopted binding RPS policies, most of these policies have emerged as statutes drafted and signed by state legislatures and signed by state governors.42
The state legislature is where the framework of an RPS program is set and its details decided; given this, efforts to amend existing RPS programs generally must originate from and be passed by a state’s legislature.
Electric Utilities and Electricity Distribution Companies With a very few exceptions (to be discussed below), these are the entities responsible for complying with RPS obligations. As described below, exact methods for complying with RPS obligations vary depending on whether the entity operates as a regulated utility monopoly or as an unregulated distributor of electricity in a market with retail electric competition. Public Utility Commissions (PUCs) As the leading regulatory authorities for state electric power sectors, PUCs are the entities responsible for ensuring that utilities and electricity suppliers meet their RPS obligations. Members of a state’s PUC are typically appointed by and accountable to the state’s governor. The structure of the state electricity market (i.e. regulated versus competitive) will determine how involved a PUC is in overseeing the actual procurement of renewable electricity. Most RPS statutes also empower PUCs to levy penalties on utilities and suppliers who fail to meet their RPS targets; similarly, many statutes empower PUCs to suspend or delay RPS compliance if the costs of such compliance exceed specified thresholds. There is potential tension between a PUC’s role in enforcing RPS-compliance and its traditional responsibility to protect ratepayers by ensuring procurement of least-cost generation; how PUCs manage this tension may in certain states strongly affect how much incremental renewable deployment RPS policies support. Developers of Renewable Energy Projects With the exception of cases where vertically-integrated utilities build their own generation, private developers of renewable energy projects (i.e. independent power producers) are the entities to actually finance and install new wind turbines, solar modules, and other technologies for renewable electricity generation. Interstate trade in electricity and Renewable Energy Credits (RECs)43
enable utilities in most RPS states to meet their RPS obligations partly through contracting with developers outside their own state.
Utility Ratepayers As the ultimate source of funds for operation of the electric power sectors, utility ratepayers reap any benefits or costs from the implementation of RPS programs. As discussed below, most RPS programs limit the annual incremental costs of RPS compliance to a maximum percentage of annual retail electricity sales.
42 Exceptions include Arizona, Colorado, and New York. Colorado's original RPS came out of a ballot proposition (Amendment 37); Arizona’s RPS was implemented by the Arizona Corporation Commission (ACC), Arizona’s public utility commission; and the New York the RPS emerged as an executive order. 43 A REC is a tradable certificate that represents 1 MWh of qualified renewable energy generation; procuring and retiring a REC can demonstrate compliance with one or more RPS requirements.
33
Frameworks for RPS Procurement Surveying the structure of state RPS policies, Ryan Wizer and Galen Barbose of Lawrence Berkeley National Laboratory (LBNL) diagnose three distinct models for how renewable electricity is procured to meet RPS targets.44
1. Retail competition model: In states where there is competition to supply electricity to retail customers, distributors typically are able to choose how best to meet their RPS obligations. For example, a distributor might contract to buy the electricity from a renewable generation project via a Power Purchase Agreement (PPA); alternatively, a distributor might meet its obligations via purchases of Renewable Energy Credits (RECs) that are “unbundled” from the underlying electricity supply (a topic explained in more detail below).
2. Regulated procurement model: In states where electricity is supplied to retail customer via
utility monopolies, utilities typically procure electricity and RECs to meet their RPS requirements under the oversight of regulators. An example of such a regulated, centralized procurement scheme is the Reverse Auction Mechanism (RAM) and feed-in tariff used by California’s three large IOUs.
3. Centralized administrative procurement model: Two states – Illinois and New York – have
transferred authority for RPS procurement to specially designated state agencies. These agencies collect ratepayer funds and distribute incentive payments to support pursuit of RPS goal
44 “Renewable Portfolio Standards in the United States: A Status Report with Data Through 2007,” Ryan H. Wiser and Galen Barbose, Lawrence Berkeley National Laboratory, April 2008.
34
Appendix III: RPS Compliance and Obligations by State, 2009-2013
Sources: MWh and compliance % data from Lawrence Berkeley National Laboratory; 2012 and 2012 compliance data from DSIRE (http://www.dsireusa.org/rpsdata/)
State 2009 MWh Obligation 2010 MWh Obligation 2009 (% Comp) 2010 (% Comp) 2012 % Obligation 2013 % ObligationAZ 824,430 1,015,858 90% 93% 3.50% 4.00%CA 29,537,501 34,116,201 89% 86% 20.00% 20.00%CO 1,646,899 1,520,066 100% 99% 15.00% 15.00%CT no data no data no data no data 16.00% 17.00%DC 591,576 699,887 100% 100% 7.50% 9.00%DE 291,451 410,618 100% 99% 8.50% 10.00%HI 0 957,857 n/a 100% 10.00% 10.00%IA 295,800 295,800 100% 100% 105 MW 105 MWIL 2,283,383 2,747,874 100% 100% 7.00% 8.00%KS 0 0 n/a n/a 10.00% 10.00%MA 3,096,274 5,468,563 82% 74% 14.10% 15.10%MD 2,770,353 3,539,778 100% 100% 9.00% 10.70%ME 3,514,043 3,832,365 100% 100% 35.00% 36.00%MI 0 0 n/a n/a 5.60% 6.80%MN 3,860,255 7,093,647 100% 100% 18% (Xcel); 12% 18% (Xcel); 12% MO 0 0 n/a n/a 2.00% 2.00%MT 346,261 692,167 100% 98% 1.00% 1.00%NC 0 24,867 n/a 100% 3.00% (IOUs) 3.00% (IOUs)NH 608,000 830,347 93% 90% 10.65% 11.70%NJ 5,733,633 6,841,213 99% 100% 9.64% (non-solar) 10.48 (non-solar)NM 852,285 858,705 100% 100% 10.00% 10.00%NV 3,551,815 3,493,644 100% 100% 15.00% 18.00%NY 4,868,849 3,061,948 61% 96% 4.54% (new) 5.60% (new)OH 333,809 602,196 100% 100% 1.50% 2.00%OR 0 0 n/a n/a 5.00% 5.00%PA 829,374 no data 100% no data 10.22% 10.72%RI 316,424 no data 100% no data 6.50% 7.50%TX 6,799,347 9,053,544 100% 100% 3,384 MW 3,384 MWWA 0 0 n/a n/a 3.00% 3.00%WI 2,501,915 3,850,101 100% 100% 5.55% 5.55%
35
Appendix IV: Potential outcomes for RPS targets – meet, delay/reduce/suspend, modify, or miss When a state imposes RPS targets on its electric utilities, at least four outcomes can emerge. The paragraphs below briefly examine each one.
Meet the target With respect to the prospects for clean energy, the most promising outcome is for a state’s electric utilities to meet the target. While Texas is so far the only state to meet a binding RPS target45
, electric utilities in many states are approaching target levels of renewable generation. Owing to impressive additions of wind generating capacity over the past decade, utilities in Oregon appears likely to meet their RPS targets within one or two years. While pursuing less aggressive targets, utilities in Vermont, Wisconsin, Michigan, and Maine also will likely fulfill their RPS targets in the near future. If this occurs, these states would join Iowa, Texas, and North and South Dakota (all of which impose voluntary goals) in the “meet the target” group.
Delay/reduce/suspend/accelerate the target A second outcome is that the PUC or other state officials may elect to delay, reduce, or suspend the RPS target. Most major RPS programs include provisos outlining criteria and mechanisms by which RPS targets may be delayed or temporarily reduced. In all cases program cost is the key criterion for delay or suspension, with many programs also according PUCs additional discretion to determine if a delay in implementation is warranted. While cases of a PUC or legislative body intervening to delay RPS compliance have been very limited, regulators and legislators in several states (Connecticut, Illinois, Maine, Minnesota, and New Hampshire) are currently scrutinizing existing RPS targets. There is the potential for some of these reviews to result in weakened or repealed RPS targets. Note that states may also elect to accelerate existing compliance deadlines, as in Maryland in May 2012 (discussed above). Modify the target A third outcome is that state legislators may intervene to modify the RPS target. If delay involves changes to questions of when and how much, then modification involves changes to questions of how. The most likely route to modifying an existing RPS target is to broaden or narrow the eligible generation technologies.46
For example, many states are currently considering revisions to their definitions of “renewable” resources; for example, Wisconsin, Maryland, and Ohio are currently evaluating whether to qualify certain thermal generation resources (for example, thermal biomass generation) as “renewable.”
Miss the target A final outcome is that a state’s electric utilities may simply miss the target. At this point, depending on the state’s enforcement clauses (and, most likely, the discretion of its PUC), the electric utilities may or may not incur penalties such as per-MWh fines or, in extreme cases, prohibition against acceptance of new customers.
45 Texas is one of two states to have a capacity-based (i.e. MW) target. 46 For example, on the floor of the California State Assembly was recently a bill to expand the definition of “renewable sources” in the state’s RPS law to include hydropower plants over 30 MW. Legislative analysis of the bill suggests that adding output from large-scale hydro facilities would immediately raise the share of “renewable” electricity for California’s three large IOUs to 32.6% (versus 17%, excluding large-scale hydro, in 2010) – by itself almost closing California’s GWh RPS gap. While California legislators overwhelmingly rejected this bill, it illustrates how legislative intervention can make RPS obligations a “moving target.”
36
In general, by 2030 we do expect the US to meet the demand pull from combined state RPS’, barring some potential close misses, or downward revisions of standards. This is largely due to the robustness of nearly all state’s RPS compliance rules (ACPs, REC trading, etc.), the relatively small amount of new annual installations necessary, and remaining (if diminishing) policy support – in particular, the ITC for solar which runs through the end of 2016.
37
Appendix V: How Renewable Energy Reverse Auction Mechanisms (RAMs) Work in Practice - Case Studies from the US and Brazil To satisfy RPS targets, utilities and distributors generally can choose to procure new sources of renewable electricity in several different ways. One promising method of procurement that can work well for relatively standardized products such as large scale renewable projects– is a reverse auction mechanism (RAM).
A RAM “reverses” the structure of a typical auction in that – instead of multiple buyers bidding to buy something from a single seller – multiple sellers (i.e. project developers) offer to sell something (i.e. renewable electricity) to a single buyer (i.e. a utility or distributor); as a result, the competition fostered by RAMs tends to drive offers down rather than up. Reverse auctions, as a means of procuring specific amounts of relatively standardized renewable energy projects, can work in the best interest of all stakeholders by driving volume and installations and further reducing costs to the consumer. In the context of electricity procurement, the key element of a RAM is that the utility fixes the quantity (in MW) and then allows competitive offers by project developers to determine the $/MWh price (perhaps subject to some administrative cap). This differs is several ways from feed-in tariff (FiT) or CLEAN (Clean Local Energy Accessible Now) programs. 47
First, the RAM approach is designed to procure a pre-selected quantity of projects, whereas FiT tariffs may be open-ended and procure as many megawatts of renewable energy as can be built economically for the FiT price. In addition, a RAM approach works best where there are highly comparable projects with similar output and cost structures. For example, a RAM would likely serve poorly for selecting a mix of utility scale and rooftop PV solar, concentrating solar and wind, since it would not recognize the value of the peak coincidence of PV or the enhanced capacity factor of Concentrating Solar Power (CSP) , or the location-based cost advantages of rooftop PV. However, for an RPS program that wants to select a certain amount of a specific resource, or whose primary objective is to minimize the cost of fungible renewable energy credits (RECs), a RAM may be appropriate. With the added policy feature of awarding a Power Purchase Agreement (PPA) or contract for RECs to the winning bidders, the cost of financing renewable energy projects could be reduced, with benefits for both the project and the customers of the utility subject to the RPS requirement.
In this paper, we outline how RAMs work in practice (and how they differ from a conventional Request for Proposal), and provide three case studies surveying the results from recent RAMs in California, Brazil, and Texas. Collectively, these case studies suggest three conclusions about use of RAMs as a means to satisfy RPS-driven demand for renewable electricity.
Conclusion 1: Potential to contract for large volumes of RE at attractive prices: By encouraging transparent, standardized competition among project developers, RAMs can enable utilities to quickly and cost-effectively contract for large volumes of new renewable electricity. In August and December 2011, Brazil’s energy regulator contracted for 2.8 GW of new wind power at an average price of $57-$62/MWh48
47 For discussion of CLEAN programs, see Appendix VII.
; meanwhile, in a November 2011 auction California’s three investor-owned utilities (IOUs) contracted for 312 MW of new renewable electricity (most from solar PV) for an average price below $90/MWh. These impressive volumes and low prices reflect RAMs’ selection of projects based on lowest
48 Which is more than 60% below what Brazil had been paying for new wind power under its feed-in-tariff program.
38
cost per MWh, which logically provides an incentive for renewable developers to build projects in the sunniest and windiest regions and to minimize costs throughout their supply chains. Proposed legislation in the 112th Congress illustrates a growing awareness of how the RAM approach to procurement could contribute to increased renewable energy development across America.49
Conclusion 2: Robust pre-screening criteria essential to ensure that bid = built: Where RAMs have faltered, it is because aggressively low bids have left developers without sufficient economic incentive to actually build the projects for which they have been awarded contracts. Preliminary analysis suggests that this “winner’s curse” may afflict many of the participants in Brazil’s August 2011 auctions – with 40% of projects (in terms of total capacity awarded) appearing to have equity returns below 10%. The key to deterring unrealistically low bids is to “pre-qualify” only those participants that meet robust minimum qualifications. For example, as a result of contracts awarded during previous renewable solicitations going unfulfilled, California’s current RAM requires participants to post security deposits, demonstrate 100% site control, and share initial results from a transmission interconnection study.
50
Conclusion 3: Supply-side incentives leverage RAM potential: A final conclusion to emerge from these case studies is that RAM contracts do not emerge in a vacuum. The offers that renewable developers make into a RAM will reflect the incentives available to promote the supply of renewable energy. For example, $57-$62/MWh wind power contracts in Brazil have been possible only due to the availability of low-cost debt from Brazil’s state-owned development bank; similarly, sub-$90/MWh solar power contracts in California benefit considerably from the federal Investment Tax Credit. By promoting investment in renewable generation projects, supply-side incentives maximize the competition among developers that RAMs seek to encourage. Maintaining such incentives will enhance the potential for RAMs to accelerate declines in the cost of US renewable generation.
By raising the credibility of bids to be evaluated, such pre-screening rules are integral to realizing the potential of RAMs.
49 H.R. 909, sponsored by Representative Devin Nunes (and co-sponsored by 73 other members), calls for establishment of a Federal Reverse Auction Authority (RAA) to initiate and conduct reverse auctions to support the development of domestic, renewable energy sources; RAA would be a private sector non-profit entity overseen by the US Department of Energy (DOE). 50 To be eligible, developers must also demonstrate development experience (i.e. construction or commercial operation of at least one similarly-sized project) and use only commercialized technology.
39
How does a RAM work? Figure 1 shows a simplified flow diagram of power, renewable energy credits (RECs) and money between those with exposure to a reverse auction that is conducted by a central contracting authority. Figure 19 Structure of a Reverse Auction Mechanism (RAM)
Source: Hudson Clean Energy Partners, 2012
Project developers would compete for power purchase contracts with the contracting authority by submitting bids via a technology-based reverse auction platform. In effect, bidders with offers above that of the bid needed to clear the market drop out of the bidding, while all offers at or below that bid level are accepted and given the opportunity to negotiate a power purchase or REC contract with the buying utility. How does a RAM differ from a typical RFP process? When a utility issues a request for proposal (RFP) solicitation for either power or renewable energy credits, it is effectively creating a competitive marketplace where the bidder with the lowest price wins the procurement contract. However, without strict requirements for product and bid uniformity, RFPs typically produce proposals that differ widely in terms of technology, product, pricing structure, and various commercial terms and conditions. While this offers the buyer the advantage of a diverse set of projects to choose from, it makes an “apples –to – apples” comparison more difficult, and may limit transparency for both bidders and the utility’s regulators. By contrast, RAMs – when implemented correctly – are designed for highly standardized projects and products that all meet the qualifying requirements, and thus can be evaluated with a price-based approach and purchased with a more standardized contract. The resulting transparency can create more confidence in the regulatory review and in the integrity of the process, to the benefit of both the utility and bidders.
Power & REC Contracting Authority
Renewable Energy Project Co.
Bank(s) Equity Sponsor(s)
Project Developer
RE
Cs
(PP
A)
kWh
$/kW
h
EPC / PF Loan(s)
Debt Service
Dev Funds
Net Income
Dev
Fu
nds
Net
In
com
e
U.S. & State Treasury
Departments
$/R
EC
RE
Cs
(PP
A)
kWh
$/kW
h
$/R
EC
Power & REC Markets
40
Case Study #1: California’s Renewable Auction Mechanism In 2008 the California Public Utility Commission’s (CPUC) Energy Division led a stakeholder process to discuss the possibility of expanding the Commission’s existing Feed in Tariff (FiT) program. Stakeholders desired a streamlined procurement program leveraging a standard contract similar to a FiT. However, they eschewed administratively determined prices in exchange for market-based pricing through a reverse auction mechanism, referred to in California as the RAM.
Ultimately, the stakeholder process designed a procurement mechanism requiring California’s three large investor-owned utilities (“CA IOUs”), Pacific Gas and Electric, Southern California Edison, San Diego Gas and Electric, to offer simultaneous auctions for pre-determined volumes every six months. Four auction periods occur at pre-determined dates across two years. For each auction, sellers offer the CA IOUs a fixed price for a standard contract for energy and RECs, and utilities sort bundled offers based solely on price. Projects are limited to 20MW and under and require commercial operation within 24 months of CPUC approval.51
Once the finalists are determined, the weighted average price of each solicitation is published once the finalists are determined and the CPUC approval process is fast tracked due to developer and IOU acceptance of the non-negotiable, pre- approved standard contract agreement terms.
Reverse auctions, similar to other price-focused procurement efforts, are known to generate speculative bids. Oftentimes in a “lowest price wins” environment where the supply of bidders significantly exceeds available contract demand, many inexperienced developers use this opportunity to provide unreasonably low bids in order to win a contract, with no intention of following through and building out the facility. The high attrition rates of unviable projects within previous renewable procurement solicitations proves that maintaining a solely “price” focused mechanism erodes the effectiveness of the efforts of all involved. The CA RAM attempted to resolve this with “project viability screens” which include the below minimum qualifications:
• 100% Site Control • Results from a Phase I interconnection study • Development Experience (construction or commercial operation of at least one similarly sized
project) • Commercialized Technology
To satisfy the first two criteria the project developer is implicitly required to have negotiated some version of site control and to have posted significant funds for an interconnection study. Another effective mechanism to filter for unviable developers is the requirement to post a development security deposit that will be refunded only when the project achieves commercial operation.52
These program requirements ultimately leverage the scrutiny of the investment community as they determine level of involvement.
51 Recent modifications to the CA RAM program have extended the original 18-month COD timeline to 24 months. 52 The development security deposit is only for winning bids. It is designed to meet a range of project sizes and technologies. It require a $20/kW development security deposit for projects 5 MW and smaller, and a $60/$90 per kW deposit for intermittent and baseload resources, respectively, for projects greater than 5 MW and up to 20 MW in size. Thus a 20MW solar project, if selected, must post a $1.2 million deposit.
41
The CA RAM program has already received accolades for its innovation and competition focus. Principally, it provides an immediate feedback loop between supply and demand, forcing developers to design competitive projects to meet CA IOU criteria, which is a distinct alternative to the traditional CA IOU procurement solicitations that are bilaterally negotiated on a project specific basis. While the CA RAM program is not a perfect procurement of energy at market prices, it combines several key processes: (i) incentivizing build of utility scale projects through long term contracts; (ii) timing transparency for completing project financing; and (iii) faster pricing feedback signaling to developers which technologies and systems are most efficient for providing renewable energy at the least cost. This RAM therefore provides an additional competitive procurement tool for achieving California’s renewable energy goals.53
.
Early results match expectations CA RAM program advocates expected the program would result in low prices. By providing market transparency and a frequent feedback loop to communicate the average auction clearing price, the program enables price discovery informing participants that in six months the next clearing price will likely be lower. In March 2012, the three CA IOUs implementing the CA RAM program each released a summary of the results of the initial auction held in November 2011, ahead of the next scheduled auction in May 2012.
Figure 20 Technology volumes (i.e. number of bids) in Nov 2011 CA RAM
Note: PG&E is Pacific Gas and Electric; SCE is Southern California Edison; SDGE is San Diego Gas and Electric. Source: California Public Utilities Commission, Recurrent Energy, 2012
The results from the CA RAM I auction demonstrate robust competition in this 20MW and under segment of the renewable market. Across the three CA IOUs the average high clearing price54
53 General solicitations do not always achieve this focused impact on price. While renewable solicitations do typically select based on lowest cost, they often allow utilities to value other project characteristics such as technology diversification and the time and location-based value of production.
was
54 While the individual contract price data is confidential, the CPUC makes public a reference price calculated from the average of the highest price accepted in each IOU auction. While this does not provide discovery of the lower end of successful bids, it does provide a data point to track overall price direction.
110
91
30
3 0 0 4 1 2 0
20
40
60
80
100
120
PG&E SCE SDGE
Solar PV
Wind
Baseload
42
$89.23/MWh for an energy-only product. In each of the IOU territories solar PV dominated the bidding volume, making up over 95% of all RAM I bids. Of the 15 winning projects across the three IOUs, 13 are solar PV projects. Although individual contract pricing is not publicized, we know from the average high clearing price that energy-only solar PV is going for sub $90/MWh prices and is likely headed lower. CA RAM, Grid Parity, and Policy Certainty The CA RAM program combines competitive program elements that drive reductions from the entire cost structure of ground mount PV systems in CA and help accelerate grid parity for solar PV projects that meet the relatively standardized requirements of the program. . Developers know that to win, cost-reducing innovations are required from all aspects of a project, from development to construction to the ultimate financing structures used. Innovation will be paramount in the three remaining RAM auctions55 where potentially 6GW of sub-20MW projects in CA compete for the remaining 688 MW of remaining CA RAM program volume56
During the remaining auctions, solar PV developers will be relying on the regulatory certainty each scheduled auction provides to support aggressive competition while racing toward achieving grid parity. In the same way, they are also relying on the support the federal Investment Tax Credit provides to each project’s cost structure. Based on the authorized program size, the expected auction dates and project approval and build-out timeframes, each CA RAM project contributing toward lower cost design and construction will come online before the ITC sunsets in 2016.
.
The transparency of the CA RAM program, the rules for participating, and the procurement volumes expected over the life of the program, also provide the kind of regulatory certainty required to attract significant private capital to the renewable sector and build an industry of the scale required to squeeze out costs and begin to compete with conventional sources of electric generation. The industry has transformed as module manufacturers have ramped up capacity and brought down the largest cost component of a solar PV facility. Continued declines in module prices have assisted in the path toward grid parity, but of equal importance are the other innovations that have resulted in cut-throat competition amongst a large number of developers in the solar space. Any uncertainty created by the threat of eliminating or accelerating the sunset of the ITC could be a large disruption in the solar PV industry which could reduce or eliminate significant private investment and the competition associated with that private investment. Such a shift could impair the industry’s ability to find new areas for cost reduction and at least temporarily decelerate the trend toward grid parity.
55 The RAM Program was approved as a 1,000MW pilot, with the expectation that if successful, the program could be expanded. 56 In the California Independent System Operator Queue released April 19, 2012 over 2.2GW of 20MW and under solar PV is actively working its way through the CAISO interconnection process.
43
Case Study #2: Brazil Electricity Tenders In 2004, Brazil began using auctions with government established FiT prices to encourage wind power development. By 2009 with experience and prices refinement, Brazil embarked on the use of reverse auctions to obtain a desired amount of installed renewable energy capacity at lowest possible price. This model has proven quite successful in terms of eliciting from the market meaningful volumes of generating capacity at significantly lower prices, though presents implementation risks, discussed later in this paper.
Brazil’s 2010 and 2011 wind tenders provides a clear example of the pros and cons of using tenders to drive wind energy deployment, with these tenders only meeting one of the three key design features – a focus on mature technologies. In terms of penalties, Brazil requires 5% of total project costs to be deposited up front by winning bidders and imposes certain fines in the case of failure to comply with tender rules. Brazil, however, has typically been lax in imposing these penalties. Also several projects appear to be based on excessively optimistic wind resource scenarios for novel and untested equipment deployment, and although the government imposes fines for projects that provide less than 90% of the annual contracted energy, there are insufficient requirements or evaluations of how wind resource assessments are conducted. Pros and Cons The success of wind power bids in Brazil’s latest auctions is increasing interest in auction mechanisms (or “tenders”) as a means to promote investment in low-carbon generation technologies. Proponents explain $57/MWh contract prices for wind as a result of the competition and price discovery that auction mechanisms purportedly create; further, they cite such prices as evidence that auctions can encourage deployment of low-carbon technologies more cost-effectively than a price-based instruments such as a feed-in-tariffs (FiT). Historically, however, in some markets auctions have been “bid” but then not necessarily “built” as bidders find when it comes to execution they cannot deliver at the agreed-upon price. The low prices bid by wind developers in Brazil’s recent regulated auctions have implications for both the cost-effectiveness of capacity auctions as a policy mechanism and the economic competitiveness of wind generation relative to other resources. It therefore becomes critical to determine how the Brazilian bidders managed to achieve winning price of $62-$57/MWh bids. Importantly market analysis has shown the importance of wind capacity factors. Specifically, our analysis further highlights the role of loans from BNDES in reducing financing costs for participants in Brazil’s energy auctions. We find that replacing BNDES loans with commercial debt increases the levelized cost of energy (LCOE) of a typical Brazilian wind project by at least 23%. Indeed the low bid cap that Brazil’s energy regulator has set for its next energy auction tentatively rescheduled to summer 2012 – $65/MWh, $15/MWh below what we calculate as the unsubsidized LCOE of a favorably-sited wind power project in Brazil based on commercial loans – underscores the strong influence of BNDES debt on the contract prices in Brazil’s energy auctions. Improving financing terms for developers of low-carbon generation is a legitimate and valuable focus of government policy. Many other countries have “green development banks” such as KfW, the EIB, potentially the UK Green Bank and Australian CEFC. The role of BNDES loans in driving the success of wind projects in Brazil’s energy auctions, however, yields the following conclusions: (i) unsubsidized “grid parity” for wind generation remains in most markets a work in progress and certainly requires very high capacity factors; and (ii) low-cost debt can effectively stimulate wind developers to bid into auctions at attractive prices – but still does not necessarily ensure that the equity returns on such
44
projects will be sufficient for developers to carry the projects to fruition. Indeed, as a study by BNEF last August showed57
, the equity return on a project bid into the auction are such that there is still considerable risk that some are not built.
Will Bid = Built? Problem of Potentially Inadequate Equity Returns Even given Brazil’s terrific wind resource, the August and December 2011 average contract prices for wind generation of $57-62/MWh are strikingly low. They represent a 30% reduction from the average contract price of $87.6/MWh that wind developers secured in Brazil’s A-3 2009 auction.58 Moreover, these prices are below the $70/MWh level that: (i) analysts estimate as a mid-range scenario for the levelized cost of energy (LCOE) from wind generation in Brazil; and (ii) is thought to have been the “minimum bid price” contained in contracts between Brazilian wind developers and their original equipment manufacturers (OEMs).59
Some developers may be able to renegotiate their supply contracts so that their OEMs accept lower margins; others may succeed in achieving unprecedented capacity factors (e.g. 55-60%) via use of such novel tactics as placing high-efficiency turbines (designed for sites with low wind speeds) at sites with high wind speeds. Uncertainty about such outcomes, however, gives rise to strong prima facie suspicion about the level of equity returns $57-62/MWh contracts will ultimately yield for project developers.
Analysis from Bloomberg New Energy Finance (BNEF) confirms this concern. Firstly we assume a cost of debt of 8.75% is used, which relies on state-sponsored BNDES loans, a key dimension of Brazilian policy we discuss later in this section. Evaluating the 78 wind projects contracted in Brazil’s August 2011 A-3 and capacity auctions BNEF calculates that 32 of these projects – representing 870 MW of new capacity (40% of total capacity tendered) – will deliver an annual return to equity of less than 10%.60 Annual equity returns on many of these projects appear to be below 7.5%. Even taking into account the burden of Brazil’s non-compliance penalties, returns of this magnitude may provide inadequate incentive for developers to actually construct their projects – hence recreating the specter of “bid but not built” projects that has played out in the wake of capacity auctions in the UK and elsewhere. Anecdotal evidence for the unattractive economics of investment in wind generation at Brazil’s recent regulated tariffs can be found in the announcement of Desenvix Energias Renovaveis SA – Brazil’s leading renewable energy developer – that it will forego bidding wind projects into Brazil’s March 2012 auction because “the returns are far too low for us.”61
57 “Brazil’s 2011 Tenders: Low Prices, High Risks,” E Tabbush et al, Bloomberg New Energy Finance, 25 August 2011 58 By contrast, from 2009-2011 global wind turbine prices declined by far less. 59 “Brazil’s 2011 Tenders: Low Prices, High Risks,” E Tabbush et al, Bloomberg New Energy Finance, 25 August 2011 60 Yielding an annual equity return above 10% seems to require a project to have an annual capacity factor of at least 45%; by comparison, for onshore wind in the US, the Energy Information Administration assumes an average annual capacity factor of 34%. 61 “Brazil Desenvix Shifts from Wind to Hydro as Power Prices Fall”, Bloomberg New Energy Finance, 15 Feb 2012, https://www.bnef.com/News/53016. Renova cited a price of $76/MWh as the floor beneath which returns on wind development became uncompetitive.
45
Figure 21 Estimated “Winning Bid” Equity Returns, December 2011 A-3 Auction versus Capacity Factors
Note: Assumes CAPEX costs of nearly $1.9m/MW, fixed OPEX costs of $50,000 per year as well as a $3-$6/MWh hedge structure, 70:30 gearing ratio and a 8.75% cost of debt. Annual inflation fixed at 5% for 20 years. Source: Bloomberg New Energy Finance, 2011. BNDES debt – the 600 bps subsidy embedded in every wind power bid From a policy perspective, BNDES plays a key role fostering renewable energy development. The most powerful policy lever at the bank’s disposal is the ability to provide loans at interest rates that are materially below market-based rates. The second lever is the 60% local content rule that projects must abide by in order to qualify for BNDES financing. As a state-owned lender, BNDES is able to provide low-interest loans (“soft dollar” loans) in order to stimulate growth of target industries such as alternative energy. Since 2000, BNDES has committed roughly $10 billion of loans to support development of Brazil’s wind resource (Figure 4), based on data provided by BNDES, we calculate another USD$4.6 billion to be committed through 2013. BNDES loans to Brazilian wind developers appear to carry interest rates 500-750 basis points (bps) below prevailing commercial rates. Based on current commercial rates for Brazilian wind developers (14% - 15% per annum), BNDES debt reduces borrowing costs for eligible wind projects by roughly 40%. Nearly every wind project that has bid into any of Brazil’s auctions has done so with the benefit of debt from BNDES.62
62 Developers of other low-carbon technologies (e.g. biomass and small hydro) can access BNDES financing on terms similar to those available for wind projects; hence BNDES financing is similarly ubiquitous in the bids from developers of these technologies participating in Brazil’s regulated auctions. BNDES low-cost debt is generally not available, however, to developers of more mature technologies such as natural gas-fired turbines or large hydro facilities.
-5%
0%
5%
10%
15%
20%
30% 35% 40% 45% 50% 55% 60%
A-3 Reserve
High deployment risk
Capacity factor
46
Figure 22 BNDES Wind Power Asset Financing In Brazil
Source: Bloomberg New Energy Finance, 2012 and DBCCA analysis, 2012 BNDES Debt as a Driver of low PPA Prices for Wind Power To assess the impact of BNDES’ below-market lending rates on project economics, we analyzed a hypothetical “typical” wind power project similar to those winning bids in recent auctions. The impact of the BNDES loans appear to us as significant in their ability to reduce LCOE. In our analyses a “base case” LCOE is established taking into account the benefit of BNDES debt.63
Figure 23 BNDES Financing Impact on LCOE
The LCOE represents the present value of a project’s lifecycle costs (construction, financing, operating, fuel, decommissioning) divided by the present value of its lifetime generation; since the resulting $/MWh cost includes the costs of equity and debt, it can be thought of as the PPA price required for a project to deliver a given level of return to its investors. Our base case assumptions are then modified to reflect the impact of debt financing on commercial terms (i.e. approximately a 600 bps premium relative to BNDES rates). We find that use of market-rate debt increases the LCOE of a Brazilian wind project by at least 23%.
With BNDES Debt Without BNDES Debt With Equity Cost Constant Risk Adjusted Equity Cost Debt Cost 8.75% 14.75% 14.75% Equity Cost 12.00% 12.00% 18.00% Capacity Factor 50% 50% 50% LCOE (USD$/MWh) $65 $80 $95
DBCCA analyses (2012) using California Energy Commission Levelized Cost of Energy model, 2010.
63 Note that (1) the LCOE of a wind project is highly sensitive to the quality of its wind resource, as measured by its site’s “capacity factor”63; and (2) variance in capacity factors across different sites can lead to a wide range of LCOE values for wind projects (whether in Brazil or elsewhere). That said, calculating the economics of a “typical” project can still illuminate the major influences on the competitiveness of wind generation.
0
1,000
2,000
3,000
4,000
5,000
6,000
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Amou
nt F
inan
ced,
USD
bill
ions
BNDES New Project Financing (Completed and Committed)
47
Given the capital-intensity of wind projects, the magnitude of this shift is unsurprising. The resulting LCOE of $80/MWh is higher than any of the prices recorded in Brazil’s August or December 2011 auctions (which was $65/MWh price for energy from two combined-cycle natural gas units recorded in August 2011); it is also above the $65/MWh price ceiling that Brazil’s energy regulator has set for Brazil’s forthcoming auction in March 2012. This suggests that – even with excellent capacity factors – market-rate debt would price Brazilian wind out of the very auctions that is has so thoroughly dominated over the past year. Implications of Pervasive BNDES Lending in Brazil’s Regulated Auctions Compared with financing available to wind project developers in the US and Europe, loans that carry an 8.75% annual coupon are hardly “cheap debt”. That said, in the case of Brazil, such loans are very much “below-market” debt. Recognizing this has implications for how one interprets the low contract prices for wind recorded in Brazil’s recent regulated auctions. We concur with BNEF that such low prices may indicate modest equity returns for many developers who have bid into such auctions – and thus presage a wave of project defaults (similar to circumstances following regulated capacity auctions in other countries). Moreover, we believe low contract prices for wind reflect in large measure the ~600 bps savings in annual debt costs enabled by BNDES loans to wind developers.64
Had wind developers been forced to bid into Brazil’s recent regulated auctions using commercial debt finance, it is likely that they would have won contracts for a far smaller volume of energy – and at prices well above $57-62/MWh. It also encourages developers to seek the very highest debt ratios they can.
This latter conclusion relates to lessons one draws from Brazil’s recent auctions for (1) the cost-effectiveness of auction mechanisms as a means to promote deployment of renewable resources; and (2) the cost-competitiveness of wind generation relative to other generation sources. The appropriate lesson appears to be that high capacity wind power can compete effectively with conventional generation sources – provided that it has stable access to debt financing on terms that are favorable (or, at least, not as punitive as a ~15% annual coupon). This suggests, in coordination with outlays from a “green” development bank, large-scale auctions may be well-suited to promote more mature renewable technologies such as wind power (particularly in markets that have developers with established track records). The auctions have proven quite successful in terms of stimulating wind power development at lower prices, though execution risks remain. With slower macro economic growth now forecast in Brazil, the auctions may have become a victim of their own success. On 26 March 2012, BNEF65
reported that with a possible unneeded 2,500 MW’s of wind power capacity scheduled for commissioning by 2015, 2012 auctions may be postponed should six planned thermal power stations be built.
64 It bears repeating that such BNDES also extends such low-cost financing to other alternative technologies such as biomass and small hydro. 65 Stephen Nielsen, “Brazil May Not Auction Power Contracts in 2012 Due to Oversupply,” BNEF, March 26 2012.
48
Case Study #3: U.S. General Service Administration’s Use of Reverse Auctions The U.S. General Service Administration (GSA) is an independent agency of the US Government that was established in 1949 to help manage and support the basic functioning of federal agencies. GSA has used reverse auctions to procure power for federal buildings for over a decade, with the objective of achieving savings for U.S. tax-payers. Power procurement and building energy management for GSA’s Region 7, which is responsible for the States of Texas, New Mexico, Oklahoma, Arkansas, and Louisiana, is handled by an “Energy Team”. Program Manager, Kevin Myles66, with the Public Building Service, is a member of that Team headquartered in Fort Worth, TX. As government employees who are responsible for the energy management of approximately 1,350 Government owned and leased buildings, the “Team” works to reduce energy consumption (in part, to promote national energy security) and to contain utility costs. The Team has used electronic reverse auctions to procure power for roughly 115 government buildings in deregulated parts of the State of Texas67
, with a total combined load (electricity demand) of ~130 million kWh/yr.
When the electricity market in Texas deregulated in 2002, Kevin Myles and his Team conducted a study and review of the procurement literature to determine the most cost-effective way to procure power for the buildings in GSA Region 7. During their analysis, the Team found examples of procurement practices in the auto industry and military where reverse auctions were being used effectively. Reverse auctions, as the Team found, are most effective in markets with a uniform, interchangeable, commodity-like product, where one unit can be easily substituted for another unit. For example, auctions where gallons of paint or barrels of oil, or kWh’s of electricity, where one unit does not differ from another in quality work best under this type of procurement. Reverse auctions also benefit when there are multiple suppliers of these products who can compete on price and if there are low barriers to market entry. In many cases these reverse auctions have saved the procuring entities anywhere between 10-25% of what the cost would have been. At the conclusion of the their review, the Team at Region 7 concluded that it could leverage GSA’s bulk buying power to create its own market for utilities services and reduce the ultimate purchase price using a reverse auction. Region 7 subsequently contracted with a company now called “World Energy Solutions” to conduct the first auction on its behalf. The initial auction was a success, yielding the anticipated savings that the Team at Region 7 expected. However, Kevin and his Team decided to go one step further and set out to develop their own reverse auction software and platform to achieve additional cost savings. The Team applied to the GSA CIO’s “venture capital fund” to request funding to build their auction platform and were ultimately awarded ~$110,000 to build a software platform and purchase a server to conduct reverse auctions. After several months of development, GSA Region 7 began to conduct its own reverse auctions in 2005. The Region typically enters the utility markets and procures power for approximately 115 of its deregulated Texas buildings at periodic intervals. Normally, these “buys” are from 3 to 5 years depending on the length of the contract awarded in the previous round. The current five-year contract
66 Information for this case study was collected during an interview between Kevin Myles, of GSA Region 7, and Michael Conti, of Hudson Clean Energy Partners. 67 Municipal Utilities and Cooperatives are not deregulated and thus do not participate.
49
will expire in 2013 and so the next reverse auction is expected before that date, for the follow on contract. How do GSA’s reverse auctions work? The on-line reverse auction process operates in a similar way that a typical auction takes place on eBay, where there is a single good for sale and multiple bidders, only it works in reverse with the offered price going down instead of up (hence the name “reverse auction”). The party that offers the most favorable price and terms “wins” the Government’s business. To kick off the process, GSA Region 7 presents the market with notice of its procurement requirements including the building load profiles and utility use histories for their buildings. The Government then invites utilities, IPPs, and independent power marketers to participate in the reverse auction. The Government reviews and “qualifies” a selection of potential bidders. To maximize competition, the reverse auction announcement is published on the electronic portal for doing business with the Government (FedBizOpps website68
), as well as via a mail notification to all of the registered retail electricity providers with the Public Utility Commission of Texas.
GSA will then conduct a short-listing round in order to select the companies that are allowed to move forward and participate in the reverse auction – for example, a full credit and financial responsibility analysis, is conducted with an evaluation of technical ability. All of this is done with the goal of power reliability and cost containment in mind. The electricity load for GSA Region 7’s deregulated Texas buildings is broken out into 4 lines and wires service territories, which are effectively smaller markets (or clusters of buildings served) on which the power marketers bid. Marketers are asked to price power on a per kWh basis as both 100% conventional generated power, and alternatively as 50%-100% renewable power. The Team asks the marketers to submit these bids so that they can get a true feel for the price of both power products. In their most recent auction, Region 7 found the price of wind power to be too high in their North Texas service area (4.4% premium to 100% conventional generation) but more reasonably priced in their other 3 lines & wires service areas (2-3% premium to conventional generation). It is anticipated that the bids in their next auction will be lower than in previous years as a result of current low natural gas prices. He also expects wind generation to be competitive at 5 – 6 cents/kWh. Myles indicates that the behavior of the participating bidders during the auction is actually very interesting to observe in real time. Initially, he says, bidders will wait a while until the first bid hits and then the volume increases very quickly as the bidding brings the price down. Prior to and during the GSA auction, it is stated that the sellers should pay close attention to the pricing levels of their bids because the GSA will hold them responsible for whatever price they offer, under the firm bid rule. This helps somewhat to prevent the “Winners Curse” but as a failsafe, GSA also requires bidders to place bonds or “corporate guarantees” and to provide information on their hedging strategy in order to prevent a possible default on power delivery at the winning price. Neither the Government nor the marketer is well served if the price turns out to be one at which the marketer is unable to perform. Myles indicates that the timing of going into these markets is pivotal. The Team at Region 7 typically conducts an auction during the shoulder months of the year when power prices are lowest (spring or fall rather than summer or winter). 68 http://www.fbo.gov.
50
Just how much savings? The cost “premium” of buying 50% wind power for three of four lines and wire areas in the most recent auction (2008) equals an estimated $172,192 per year. However, this must be viewed in the context of a nearly $70 million dollar five-year power purchase contract and the ability of the reverse auction to drive down the price of both conventional and renewable energy. The prior reverse auction contract (2005) only had 5% renewable power and averaged 9.23 cents/kWh. The 2008 contract averaged 10.6 cents/kWh. (This number includes transportation charges, so the Agency is paying a net average of ~8.5 – 10 cents/kWh for the actual power.) This is only a 15% increase over the power price locked in three years earlier – that was awarded in a market where power prices were rapidly increasing. The ultimate buying decision, Myles indicates, is a balance between the purely financial question, ‘what’s the least expensive option?’, and the energy security / environmental question, ‘what option would yield the most renewable power?’ The 2008 5-year contract valued at ~$69-$70 million was considered to be both fair and reasonable and was considered to be in the best interest of the Government. It represents an optimized mix of the twin objectives of savings taxpayer money and advancing clean energy goals. Myles believes that the GSA could do even better if not for a statute in the General Service Administration Act of 1949 that sets a 10 year maximum term for utility contracts (other government agencies, such as the Military, have different authorities and longer contracting term limits). Myles indicates that if this contract term limit were increased to 20 or 30 years, he feels his Team could further reduce the contracted price for power that the Government must pay using tax-payer dollars. Such an extension could also attract a greater number of bids from developers who want to build future clean energy capacity and finance new build against 20 – 30 year PPAs with the Government. Why building energy efficiency management is so important Myles and his Team are also partly responsible for seeing that the government meets Energy Efficiency Guiding Principles by 2015 and achieves zero-net-energy use in buildings by 2030, as mandated by Federal Statutes and Executive Orders 13423 and 1351469
. As Myles and his Team have found, the effectiveness of procuring power using reverse auctions is dramatically enhanced by the Agency’s use of building energy efficiency management practices. This is because Region 7 has good command of its power load and can therefore give reverse auction bidders very accurate load profiles on each of its buildings. In turn, the bidders don’t have to build a load shape risk factor into their ultimate bid in order to account for unexpected load and contingencies. This saves both the buyer and the seller money on the transaction. GSA will also give a letter of authorization to each bidder so that each can request a load profile history from the serving utility for confirmation, and so it can make its own independent assessment of the data.
The Building Owners and Managers Association (BOMA) is an industry trade association whose members have an extensive portfolios of commercial properties around the country. Myles quoted the average amount of energy that is required to operate the buildings under management by members of BOMA as being equal to approximately 78k Btu/gross square foot. He states GSA Region 7’s average energy
69 Federal Sustainable Building and Campus Requirements - http://www1.eere.energy.gov/femp/program/sustainable_requirements.html
51
consumption to be approximately 48k Btu/gross square foot. Relative to industry standards, GSA Region 7 has become quite the model of building operation for energy efficiency to be emulated. Scaling this model across other agencies The cost savings that GSA Region 7 has been able to capture by developing its own proprietary reverse auction platform could possibly be scaled across other branches of government. In fact, The Navy, and some other branches, already use reverse auctions mechanisms to purchase uniform, interchangeable, commodity-like products. Myles indicates that the Energy Team at Region 7 will sometimes approach other agencies to invite their participation in GSA energy buys. Myles indicates that they often receive initial interest, but in many instances, other agencies are either already under existing power contracts, or their building loads are too different for it to be efficient or attractive to be included. The underlying impediment to scaling up the reverse auction power purchasing platform that the Energy Team at GSA Region 7 have developed is what Myles sees as a general need for more people with a “power procurement” skill set at many government agencies. Reverse auctions can help drive renewable energy sales This case study shows how reverse auctions can drive renewable energy sales. The GSA, which is a buyer of bulk power, was empowered with the ability to specify its desire to purchase a certain percentage of its power from renewable energy sources because of its reverse auction tool. This had the dual effect of attracting renewable energy credit sellers to an otherwise inaccessible market and allowed for competitive price discovery that in turn made the renewable energy credits affordable.70
It should be noted that renewable energy credits in Texas trade for extremely low prices because Texas has met and exceeded the State’s Renewable Portfolio Standard, therefore causing the market to be oversupplied. Additionally, the renewable power procured by GSA Region 7 was not from new renewable sources but existing wind capacity. However, as previously noted, if GSA and other Government agencies were able to enter into longer-term contracts, beyond 10 years, these agencies could potentially enter into serious long-term off-take discussions with renewable project developers. Extending contract duration limits on certain Federal Government agencies could have the effect of unlocking a high creditworthy off-take market and potentially facilitate new renewables development.
How reverse auctions can help drive volume and future installations of renewable energy? Reverse auctions can reduce the cost to the consumer of procured, standardized power, and have the potential to serve as a mechanism to help drive additional development of renewable power in the U.S. For this to be effective, reverse auction procurement would need to be coupled with a central counter-party with which renewable energy project developers sign power purchase agreements (PPAs). Providing a commercially attractive, streamlined PPA process would likely have the immediate and positive effect of encouraging faster rates of renewable energy development at lower financing costs. Utilities and large power consumers may find that using technology-based reverse auction tools helps streamline the efficient procurement of power. As renewable energy costs continue to fall, such procurement offers a way to go beyond meeting renewable mandates, towards reducing costs and risks through the increasing use of renewable energy. Federal and state-level reverse auctions for renewable energy could help provide the policy support that renewable energy technologies need to continue along their respective cost curves towards grid parity, enabling sustainable long-term growth of this industry in the U.S.
70 The views expressed by Kevin Myles are his own and not official U.S. Government policy.
52
Appendix VI: Geographic Eligibility and Delivery Requirements for Renewable Energy Credits (RECs)
Sources: NREL 2010, DSIRE 2011, Wiser and Barbose 2008.
53
Appendix VII: CLEAN (Clean Local Energy Accessible Now) Programs Clean Local Energy Accessible Now (CLEAN) Programs, also known as feed-in tariffs (FiTs), offer standard, fixed price, long-term power purchase agreements where the offered price is determined up-front, but is then adjusted for later projects as the market responds.
Experience with longstanding CLEAN Programs, such as Germany’s EEG, suggests that a well-designed policy can drive down costs with competition for supplying energy and the rapid learning curve that comes with significant market scale. By providing developers with Transparency, Longevity, and Certainty (TLC), well-designed CLEAN Programs may be able to help fulfill RPS requirements in a timely, cost-effective manner.
CLEAN Program design has been evolving recently as more countries, states and cities gain experience in implementing their programs. The latest example of a statewide policy design is California’s Senate Bill 32 (SB 32) clean energy law. This law mandates that the state’s three largest invest-owned utilities (IOUs) provide a CLEAN Program for any RPS-eligible generation project up to 3 MW in size. The California Public Utilities Commission (CPUC) approved its interpretation of the law in May.
In this interpretation, the CPUC sought to address the key risk of CLEAN Programs – setting the power purchase price “wrong”. If the price is too low, no projects are developed, but if the price is too high, ratepayers are harmed. A variety of price setting and adjustment mechanisms have been used across the world, such as regular recalculation and automatic annual decreases, but SB 32 choose a market response mechanism that has been gaining popularity: price changes based on pre-defined subscription volume levels.
The mechanism consists of pre-defined capacity “buckets” that limit how many contracts can be executed at a given price. The price offered to developers then changes based on subscription to each bucket. If the bucket is filled, the price then decreases for the next bucket. If the there is minimal response to the offered price, the price is increased. Ultimately, however, the CPUC chose to link payments to the results of auctions by California’s three large investor-owned utilities. Payments will initially be linked to the results of the Nov 2011 auctions.71
Since the California SB 32 program has yet to be launched, its potential to set a successful policy example that helps meet RPS goals is unknown. The Sacramento CLEAN program has resulted in 100 MW of solar projects in a two-year period, all of which is interconnected to the distribution grid for serving local load (i.e., all of the 100 MW avoids all costs associated with transmission grid, which is equal to roughly 3 cents/kWh). Note that 100 MW in Sacramento is equivalent to 2.5 GW across the State of California – suggesting that, if adopted statewide, a CLEAN program could stimulate rapid, widespread deployment of solar in California.
71 Described in more detail in the California RAM section of Appendix V.
54
The material contained in this article is current as of the publication date of the article, is intended for informational purposes only and does not purport to address the financial objectives, situation or specific needs of any individual reader. It has been obtained from sources believed to be reliable, but Deutsche Asset Management cannot guarantee its accuracy or completeness. The use of this article is not a solicitation, or an offer to buy or sell any security or investment product. Moreover, the opinions expressed in this article are not necessarily those of Deutsche Asset Management or its employees.
I-028266-1.0
