The Role of Energy Storage with Renewable Electricity Generation (Report Summary)
Paul Denholm Erik Ela Brendan Kirby Michael Milligan
March 2010
NREL/PR-6A2-49396 Outline
• Operation of the Electric Grid
• Electricity Storage in the Existing Grid
• Impacts of Renewables on the Grid and the Role of Enabling Technologies
• Storage and Flexibility Options for Renewable- Driven Grid Applications
• Conclusions
National Renewable Energy Laboratory 2 Innovation for Our Energy Future Operation of the Electric Grid
• Supply must always meet demand
• Large hourly and season variations in electricity demand
• Operating reserve requirements to maintain system stability and reliability
National Renewable Energy Laboratory 3 Innovation for Our Energy Future Large Demand Variability
1.0
60000 0.9 0.8 50000 0.7
40000 0.6 0.5 30000 0.4 Load (MW) 20000 0.3 Summer Maximum Winter Spring Minimum 0.2 10000 0.1 (Fraction Load of Annual Peak)
0 0.0 0 24 48 72 96 120 144 168 Hour
Hourly electricity demand for three weeks in the ERCOT (Texas) Grid in 2005
Requires multiple generator types: baseload, load-following (intermediate load and peaking)
National Renewable Energy Laboratory 4 Innovation for Our Energy Future Reserve Requirements
• Provides stable and reliable operation • Not uniformly defined • Frequency Regulation – Serves the random, rapid variation around the normal load – Highest value • Contingency Reserve (often referred to as spinning reserves) – Quickly replaces a lost generator or transmission line – Infrequently called ~ 1x/week for about 10 minutes
National Renewable Energy Laboratory 5 Innovation for Our Energy Future Frequency Regulation Requirements
System load following and regulation. Regulation (red) is the fast fluctuating component of total load (green) while load following (blue) is the slower trend
National Renewable Energy Laboratory 6 Innovation for Our Energy Future Reserve Requirements Add Costs
3000
2500 Insufficient Peak 2 2000 Reserve Capacity Peak 1 1500 Int Base2 Load (MW) Load 1000 Higher Cost Unit Started Insufficient Flexibility to Base 1 Provide Reserves Early To Provide Reserve 500 Capacity
0 3000 0 4 8 12 16 20 24 Hour 2500
“Ideal” Dispatch 2000 1500 Lower Cost Unit Backed Off To Accommodate (MW) Load 1000 Higher Cost Flexible Generator 500 0 0 4 8 12 16 20 24 • Non-economic dispatch Hour
• Part-load inefficiencies Reserve Constrained • Additional units online Dispatch
National Renewable Energy Laboratory 7 Innovation for Our Energy Future Energy Storage in the Existing Grid
• Historical motivations
• Renewed interest due to emergence of restructured markets
• Applications of energy storage
National Renewable Energy Laboratory 8 Innovation for Our Energy Future Electricity Storage in the Existing Grid
Historical motivations (pre-1980) • Storage provides load following and reserves, while increasing use of low-cost baseload plants • Limited options for peaking and intermediate load generation – Increasing cost of natural gas and oil – Restrictions on use of oil and natural gas (Power Plant and Industrial Fuel Use Act) – Low-efficiency oil and steam gas plants as opposed to today’s efficient gas turbines • Anticipated construction of many nuclear plants providing low cost (but inflexible) baseload energy
National Renewable Energy Laboratory 9 Innovation for Our Energy Future Pumped Hydro
Was an attractive alternative to other load-following generators
14.0 Gas Steam Oil Steam 12.0 Gas CCGT Storage (Coal Charging)
10.0
8.0
6.0 Fuel Cost (2008 Cents/kWh) (2008 Cost Fuel 4.0
2.0
0.0 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Year Variable costs of storage are less than from oil or gas
Pumped hydro costs during this period (before 1980) are comparable to combined-cycle gas generators
National Renewable Energy Laboratory 10 Innovation for Our Energy Future Limited Storage Built after 1980
• Collapse of oil and gas prices
• Repeal of fuel use act
• High-efficiency low cost gas turbines become available
• Storage development limited to ~20GW of pumped hydro storage, 1 CAES plant plus a few batteries and demonstration projects
National Renewable Energy Laboratory 11 Innovation for Our Energy Future Revised Interest in Energy Storage
• Advances in storage technologies
• Increases in fossil fuel prices
• Energy markets
• T&D siting challenges
• Perceived need for storage with renewables
National Renewable Energy Laboratory 12 Innovation for Our Energy Future Value of Storage in Restructured Markets
Historical Values of Energy Storage in Restructured Electricity Markets Market Location Years Annual Value Assumptions Evaluated Evaluated ($/kW)
Energy PJMa 2002-2007 $60-$115 12 hour, 80% efficient device. Range of Arbitrage efficiencies and sizes evaluated[1] NYISOb 2001-2005 $87-$240 10 hour, 83% efficient device. Range of (NYC) efficiencies and sizes evaluated. $29-$84 (rest) USAc 1997-2001 $37-$45 80% efficient device, Covers NE, No Cal, PJM
CAd 2003 $49 10 hour, 90% efficient device. Regulation NYISOb 2001-2005 $163-248
USAe 2003-2006 $236-$429 PJM, NYISO, ERCOT, ISONE
Contingency USAe 2004-2005 $66-$149 PJM, NYISO, ERCOT, ISONE Reserves
a Sioshansi et al. 2009 b Walawalkar et al. 2007 c Figueiredo et al. 2006 d Eyer et al. 2004 e Denholm and Letendre 2007
National Renewable Energy Laboratory 13 Innovation for Our Energy Future Historical Value of Energy Storage in U.S. Markets
4000 Capital Charge Rate 3500 9.8% 3000 12.0% 13.9% 2500
2000
1500 Arbitrage Only Regulation Only 1000
Required Storage Capital Cost ($/kW) 500
Contingency Only 0 0 50 100 150 200 250 300 350 Annual Benefit of Storage ($/kW)
Relationship between total installed cost and annualized cost
Greatest value is frequency regulation – focus of many applications (flywheels, vehicle to grid). Arbitrage alone is generally insufficient to support most storage technologies, which are generally >$1,000/kW
National Renewable Energy Laboratory 14 Innovation for Our Energy Future Applications of Energy Storage
Description Timescale of Operation Application
Load Leveling/ Purchasing low-cost off-peak energy and selling it Response in minutes to hours. Discharge time of hours. Arbitrage during periods of high prices.
Firm Capacity Provide reliable capacity to meet peak system Must be able to discharge continuously for several hours or more. demand.
Operating Reserves
Regulation Fast responding increase or decrease in Unit must be able to respond in seconds to minutes. Discharge time is typically minutes. Service is generation (or load) to respond to random, theoretically “net zero” energy over extended time periods. unpredictable variations in demand.
Contingency Fast response increase in generation (or Unit must begin responding immediately and be fully responsive within 10 minutes. Must be able to Spinning decrease load) to respond to a contingency such hold output for 30 minutes to 2 hours depending on the market. Service is infrequently called.[2] Reserve[1] as a generator failure.
Replacement/ Units brought on-line to replace spinning units. Typical response time requirement of 30-60 minutes depending on market minutes. Discharge time Supplemental may be several hours.
Ramping/Load Follow longer term (hourly) changes in electricity Response time in minutes to hours. Discharge time may be minutes to hours. Following demand.
T&D Replacement Reduce loading on T&D system during peak Response in minutes to hours. Discharge time of hours. and Deferral times.
Black-Start Units brought online to start system after a Response time requirement is several minutes to over an hour. Discharge time requirement may be system-wide failure (blackout). several to many hours.[3]
End-Use Applications TOU Rates Functionally the same as arbitrage, just at the Same as arbitrage. customer site. Demand Charge Functionally the same as firm capacity, just at the Same as firm capacity. Reduction customer site. I
Backup Power/ Functionally the same as contingency reserve, nstantaneous response. Discharge time depends on level of reliability needed by customer. UPS/Power Quality just at the customer site.
National Renewable Energy Laboratory 15 Innovation for Our Energy Future Impacts of Renewables on the Grid
• Storage is often perceived as “necessary” for renewables to achieve a large (>10%? >20%?) penetration.
• Renewables are seen as a source of value for storage
• Can renewables be used without storage?
• How do renewables impact the grid?
National Renewable Energy Laboratory 16 Innovation for Our Energy Future Impacts of Renewables on the Grid
50000 Load Wind Net Load 45000 Ramp Range (Increases in this two-week period from 40000 19.3 GW/day to 26.2 GW/day) 35000
30000
25000 MW 20000
15000
10000
5000
0 1-Apr 8-Apr 15-Apr Uncertainty in wind output Variation in wind output increases net load increases uncertainty in net ramp ate (Increases in this period from load to be met with 4,052 MW/hour to 4,560 MW/hour) conventional generators
Four major impacts of variable generation (VG) on the grid: 1) Increased need for frequency regulation 2) Increase in hourly ramp rate 3) Increase in uncertainty of net load 4) Increase in ramp range
National Renewable Energy Laboratory 17 Innovation for Our Energy Future Costs of Wind Integration
• Simulate system with and without solar and wind – Use unit commitment software includes existing generation mix, transmission system – Use lots of wind and solar simulations to consider spatial diversity – May involve substantial costs • Evaluate costs of: – Additional regulation reserves – Additional load following – Wind uncertainty
National Renewable Energy Laboratory 18 Innovation for Our Energy Future Costs of Wind Integration
Wind Capacity Regulation Load- Unit Total Oper. Penetration Cost Following Cost Commitment Other Cost Impact Date Study (%) ($/MWh) ($/MWh) Cost ($/MWh) ($/MWh) ($/MWh)
2003 Xcel-UWIG 3.5 0 0.41 1.44 Na 1.85 2003 WE Energies 29 1.02 0.15 1.75 Na 2.92 2004 Xcel-MNDOC 15 0.23 na 4.37 Na 4.6 2005 PacifiCorp-2004 11 0 1.48 3.16 Na 4.64 2006 Calif. (multi-year)a 4 0.45 trace trace Na 0.45 2006 Xcel-PSCob 15 0.2 na 3.32 1.45 4.97 2006 MN-MISOc 36 na na na na 4.41 2007 Puget Sound Energy 12 na na na na 6.94 2007 Arizona Pub. Service 15 0.37 2.65 1.06 na 4.08 2007 Avista Utilitiesd 30 1.43 4.4 3 na 8.84 2007 Idaho Power 20 na na na na 7.92 2007 PacifiCorp-2007 18 na 1.1 4 na 5.1 2008 Xcel-PSCoe 20 na na na na 8.56
a Regulation costs represent 3-year average. b The Xcel/PSCO study also examine the cost of gas supply scheduling. Wind increases the uncertainty of gas requirements and may increase costs of gas supply contracts. c Highest over 3-year evaluation period. 30.7% capacity penetration corresponding to 25% energy penetration d Unit commitment includes cost of wind forecast error. e This integration cost reflects a $10/MMBtu natural gas scenario. This cost is much higher than the integration cost calculated for Xcel-PSCo in 2006, in large measure due to the higher natural gas price: had the gas price from the 2006 study been used in the 2008 study, the integration cost would drop from $8.56/MWh to $5.13/MWh.
National Renewable Energy Laboratory 19 Innovation for Our Energy Future Conclusions of Wind Integration Studies (<30% Penetration)
• Challenges are unit commitment, regulation and load following • Integration costs are modest (typically less than $5/MWh) • Increased variability can be accommodates by existing generator flexibility and other “low-cost” flexibility such as increased balancing area cooperation (balancing wind generation and load over larger areas to “share” the increased variability. • Spatial diversity smooth's aggregated wind output reducing short-term fluctuations to hour time scales • Storage would “help” but is not needed, and the integration costs would not “pay” for currently expensive storage technologies.
National Renewable Energy Laboratory 20 Innovation for Our Energy Future Limits to VG Penetration - Curtailment
• At high penetration, economic limits will be due to curtailment
– Limited coincidence of VG supply and normal demand
– Minimum load constraints on thermal generators
– Thermal generators kept online for operating reserves • Results from EWITS and WWSIS, along with other studies, indicate that beyond 30% VG, new sources of flexibility will be needed
National Renewable Energy Laboratory 21 Innovation for Our Energy Future High Penetration Limits
Load Net Load Load Met by Inflexible Generation Wind 70000
60000
50000 Dispatch with low VG
40000 penetration (wind MW 30000 providing 8.5% of load)
20000
10000
0 0 24 48 72 96 Hours Wind Net Load Met by Inflexible Generation
70000 Curtailed Wind Net Load Met by Flexible Generation Load Dispatch with higher VG 60000 penetration (wind 50000
providing 16% of load) 40000
MW 30000 Inflexible systems may be 20000 unable to accommodate 10000 large amounts of wind 0 0 24 48 72 96 Hours
National Renewable Energy Laboratory 22 Innovation for Our Energy Future Current System Flexibility
Limited by Baseload Capacity
250
200
150 Price/Load
100 Relationship in PJM
50 Wholesale Price ($/MWh) 0 0 10000 20000 30000 40000 50000 60000 Load (MW) 35
30 Below Cost Bids 25 20
15
10 Plant operators would rather sell
energy at a loss than incur a Wholesale Price ($/MWh) 5 costly shutdown. Wind may be 0 curtailed under these conditions 18000 20000 22000 24000 26000 Load (MW)
National Renewable Energy Laboratory 23 Innovation for Our Energy Future Increased Flexibility is Required
Flexibility Factor 100% 90% 80% 70% 60% 50% 100% Typical Minimum 90% Load in Current 80% Systems 70% 60% 50% 40%
Generators 30% 20% 10%
% of Energy From Inflexible Energy % of 0% 0% 10% 20% 30% 40% 50% Minimum Load as a Fraction of Peak Load
At current minimum loads, baseload generators provide 60-70% of generation, leaving only 30%-40% for VG.
National Renewable Energy Laboratory 24 Innovation for Our Energy Future Decreased Minimum Load Needed to accommodate greater amounts of VG without significant curtailment
Load Met by Renewable Energy Net Load Met by Flexible Generation 50000 Curtailed Renewable Energy Net Load Met by Inflexible Generation 45000 Renewable Energy Inflexible System - 40000 Minimum Load of 35000 30000 21 GW (65% FF) 25000 MW 20000 15000 10000 Decreased 5000 curtailment 0 0 24 48 72 96 Hours Net Load Met by Inflexible Generation Curtailed Renewable Energy
50000 Net Load Met by Flexible Generation Load Met by Renewable Energy Renewable Energy 45000 More Flexible System 40000 -Minimum Load of 35000 30000 13 GW (80% FF) 25000 MW 20000 15000 10000 Simulations based on 5000 2005 load and weather 0 0 24 48 72 96 Hours
National Renewable Energy Laboratory 25 Innovation for Our Energy Future Curtailment as a Function of Flexibility
50% 21 GW Min Load (65% FF) 45% 12 GW Min Load (80% FF) 40% 6 GW Min Load (90% FF) 35% FF=Flexibility Factor 30%
25%
20%
15% Average VG Curtailment 10%
5%
0% 0% 10% 20% 30% 40% 50% 60% 70% 80% Fraction of Energy From RE
Average curtailment rate as a function of VG penetration for different flexibilities in ERCOT
National Renewable Energy Laboratory 26 Innovation for Our Energy Future Curtailment as a Function of Flexibility (cont.)
50%
45%
40%
35%
30% 21 GW Min Load 25% (65% FF) 20% 12 GW Min Load (80% FF) 15% 6 GW Min Load Marginal VG Curtailment 10% (90% FF)
5% FF=Flexibility Factor
0% 0% 10% 20% 30% 40% 50% 60% 70% 80% Fraction of Energy From RE Marginal curtailment rate as a function of VG penetration for different system flexibilities in ERCOT (marginal curtailment is defined as the percentage curtailed as a function of the incremental penetration)
National Renewable Energy Laboratory 27 Innovation for Our Energy Future VG Curtailment
May Result In Unacceptably High Costs at High Penetration
2.0 21 GW Min Load 1.9 (65% FF)
1.8 12 GW Min Load (80% FF) 1.7 6 GW Min Load (90% FF) 1.6
1.5
1.4
1.3
Relative Cost of VG (Average) 1.2
1.1 2.0 1.0 0% 10% 20% 30% 40% 50% 60% 70% 80% 1.9 Fraction of Energy From Wind and Solar 1.8
1.7
1.6
1.5
1.4 21 GW Min Load (65% FF) Relative cost of VG – average (top chart) 1.3 12 GW Min Load (80% FF) and marginal (bottom chart) – as a Relative Cost of VG (Marginall) 1.2 6 GW Min Load function of VG penetration for different 1.1 (90% FF) system flexibilities in ERCOT 1.0 0% 10% 20% 30% 40% 50% 60% 70% 80% Fraction of Energy From Wind and Solar
National Renewable Energy Laboratory 28 Innovation for Our Energy Future Renewables-Driven Grid Applications
Storage and Flexibility Options
• At high penetration of VG, additional flexibility is required
• What are the options?
National Renewable Energy Laboratory 29 Innovation for Our Energy Future Renewables-Driven Grid Applications
Storage and Flexibility Options (cont.)
While storage provides an “obvious” answer to the problem of supply-demand coincidence, there are a number of options
National Renewable Energy Laboratory 30 Innovation for Our Energy Future Flexibility Supply Curve
The cost of all options has yet to be determined. Currently, energy storage is expensive and incurs the penalty of round-trip losses.
Based on an original by Nickell 2008
National Renewable Energy Laboratory 31 Innovation for Our Energy Future Energy Storage Can Reduce VG Curtailment
50% No Storage 45%
40% 5% Storage Capacity
35% 5% Storage Including Minimum Load Reduction 30%
25%
20%
15% Average RE Curtailment 10%
5%
0% 0% 10% 20% 30% 40% 50% 60% 70% Fraction of Energy From Wind and Solar
FF=80% (12 GW min load) Energy storage can reduce curtailment both by shifting otherwise unusable generation, and also increase system flexibility by providing reserves (reducing the need for partially loaded thermal generators) and replacing “must-run” capacity
National Renewable Energy Laboratory 32 Innovation for Our Energy Future Dedicated Renewable Storage?
• Dedicated renewable storage is generally a non- optimal use • Could have scenarios where one storage device is charging while another is discharging simultaneously in the same system • “Renewable specific” applications are already typically captured in grid operations
RE Specific Application “Whole Grid” Application Transmission Curtailment Transmission Deferral Time Shifting Load Leveling/Arbitrage Forecast Hedging Forecast Error Frequency Support Frequency Regulation Fluctuation Suppression Transient Stability
National Renewable Energy Laboratory 33 Innovation for Our Energy Future Storage Technologies
Common Name Example Applications Discharge Time Required
Power Quality Transient Stability, Frequency Regulation Seconds to Minutes Bridging Power Contingency Reserves, Ramping Minutes to ~1 hour Energy Management Load Leveling, Firm Capacity, T&D Deferral Hours
Not shown – thermal storage in end use and in CSP plants
National Renewable Energy Laboratory 34 Innovation for Our Energy Future Caveats
• Efficiency – Not uniformly defined (should be AC-AC, but sometimes stated in terms of DC-DC, which doesn’t capture conversion) – May not include parasitics – CAES (which uses natural gas) and thermal storage cannot be easily compared to pure electricity storage devices such as pumped hydro • Cost – Many technologies have not been deployed as large scale, so costs are largely unknown – Commodity prices affect estimates from different years – Difficult to compare devices that offer different services (power vs. energy)
National Renewable Energy Laboratory 35 Innovation for Our Energy Future Storage Devices for Power Quality
• Requires response time of seconds or less, discharge time of up to 10 minutes • Flywheels – Currently targeted toward frequency – Regulation applications • Capacitors – Limited deployment to date (low energy capacity) • Superconducting Magnetic Energy Storage (SMES) – Limited deployment to date due to high cost and low energy density
National Renewable Energy Laboratory 36 Innovation for Our Energy Future Storage Devices for Bridging Power
• Rapid response (seconds to minutes) • Discharge time of up to 1 hour • May also provide power quality services • Several battery technologies – lead-acid – nickel-cadmium – nickel-metal hydride – lithium-ion
National Renewable Energy Laboratory 37 Innovation for Our Energy Future Storage Devices for Energy Management
• Discharge time of several hours • Some technologies also provide power quality and bridging power • Would be the most important class of storage devices for decreasing VG curtailment • High energy batteries • Pumped hydro • Compressed Air • Thermal Storage
National Renewable Energy Laboratory 38 Innovation for Our Energy Future High Energy Batteries
• High-temperature batteries – Sodium sulfur – the most widely deployed battery with more than 270 MW installed – Sodium-nickel chloride (early commercialization stage) • Liquid electrolyte “flow” batteries – Feature separate power and energy components – All in the early commercialization stage – Vanadium redox – Zinc-bromine – Other chemistries being pursued
National Renewable Energy Laboratory 39 Innovation for Our Energy Future Pumped Hydro Storage
• Only technology deployed on a gigawatt scale – ~20 GW in the U.S. at 29 sites – ~100 GW worldwide • Installations in the U.S. range from <50 MW to 2,100 MW • Can exceed 75% AC-AC round-trip efficiency • Permitting is difficult and time-consuming, but ~ 30 GW of proposed new capacity in the U.S.
National Renewable Energy Laboratory 40 Innovation for Our Energy Future Compressed Air Energy Storage (CAES)
• Hybrid technology that a gas turbine and burns natural gas • Single point efficiency not easily defined – Better to use electricity energy ratio (0.6-0.8 kWh in per kWh out) and heat rate (4000-4300 BTU/kWh) • Two sites – U.S. (110 MW) and Germany (270 MW) • Alternative designs proposed including adiabatic (no fuel input)
Alabama CAES Plant (110 MW)
National Renewable Energy Laboratory 41 Innovation for Our Energy Future Thermal Storage
• End-use thermal storage in buildings – Hot or cold – Very high round-trip efficiencies (near 100% in some applications) – Commercially available • Thermal storage for CSP – Very high round-trip efficiency (stores energy before Carnot losses) • Both applications are tied to a specific use and difficult to compare to more flexible storage technologies
National Renewable Energy Laboratory 42 Innovation for Our Energy Future Electric Vehicles and Vehicle-to-Grid
• Potential valuable source of distributed storage • More likely to be used for ancillary services than bulk energy shifting due to high battery cost • Active area of research
National Renewable Energy Laboratory 43 Innovation for Our Energy Future Conclusions
• The role of storage is an economic issue – does the value of storage exceed its benefits? • Storage is undervalued in existing markets and it is still difficult to assess the true value and opportunities for energy storage in the current and future grid • Renewables increase the value of storage, but the current grid can accommodate substantially increased amount of renewables with options that appear to be lower cost than new dedicated storage • At penetrations of wind and solar that exceed 30%, increased curtailment will require new sources of grid flexibility • New models and analysis will be required to evaluate the benefits of energy storage in the future electric grid.
National Renewable Energy Laboratory 44 Innovation for Our Energy Future For More Information
Full Report:
Denholm, P.; Ela, E.; Kirby, B.; Milligan, M. (2010). Role of Energy Storage with Renewable Electricity Generation. NREL Report No. TP-6A2-47187.
Available at: http://www.nrel.gov/docs/fy10osti/47187.pdf
National Renewable Energy Laboratory 45 Innovation for Our Energy Future