IEA technology analysis and roadmap for energy storage Luis Munuera
IEA s programme of work in energy technology Where do we need to go? Where are we today? How do we get there?
Technology roadmaps provide answers Engage cross-section of stakeholders Identify a baseline Establish a vision Identify technical, regulatory, policy, financial, public acceptance barriers Develop implementation action items for stakeholders
A choice of 3 Futures 6DS/Current Policies where the world is heading under current policy with potentially devastating results 4DS/New Policies reflecting pledges by countries to cut emissions and boost energy efficiency 2DS/450 a vision of a sustainable energy system of reduced Greenhouse Gas (GHG) and CO 2 emissions The 6 C Scenario The 4 C Scenario The 2 C Scenario To achieve the 2DS, energy-related CO 2 emissions must be halved until 2050 Source: Energy Technology Perspectives, IEA 2014
Power sector is central for deep decarbonisation 60 50 Sectors 6DS 60 50 Technologies 6DS 40 40 Gt CO2 30 20 10 2DS 30 20 10 2DS 0 2011 2020 2030 2040 2050 Power generation 41% Industry 19% Transport 19% Buildings 13% Other transformation 8% 0 2011 2020 2030 2040 2050 End-use fuel and electricity efficiency 38% CCS 14% End-use fuel switching 9% Renewables 30% Power generation efficiency and fuel switching 2% Nuclear 7% 50/50 chance of two-degree temperature rise by 2100 requires 40% of the burden to be carried by the power sector
Decarbonising the global electricity sector 6DS TWh 50 000 40 000 30 000 20 000 10 000 2DS 0 2011 2020 2030 2040 2050 Other Wind Solar Hydro Nuclear Biomass and waste In the 6DS, fossil technologies continue to dominate electricity generation with a share of 66% in 2050, whereas in the 2DS more than 90% of electricity are generated from low-carbon sources (renewables, nuclear, CCS). Oil Gas Coal
Three pillars for low carbon transformation of power supply Technology spread 1. Geographic Let wind and spread solar play their part Design of power plants System friendly VRE 3. Take a system wide-strategic approach to investments! 2. Make better use of what you have Investments Operations Source: The Power of Transformation, IEA 2014
Current grid-connected electricity storage dominated by Pumped Storage Hydropower Installed capacity in MW Lithium-ion 100 Lead acid 70 PSH 140 000 Other 976 Sodiumsulphur 304 Nickel-cadmium 27 Flywheel 25 Redox-flow 10 CAES 440
More flexibility required! PWh 45 40 35 30 25 20 15 10 5 Electricity generation 0 2009 2020 2030 2040 2050 non varre varre Hydro Nuclear Bio / Waste Oil Natural gas Coal PLDV annual sales (millions) 200 180 160 140 120 100 80 60 40 20 Annual light-duty vehicle sales 0 2000 2010 2020 2030 2040 2050 FCEV Electricity Plug-in hybrid diesel Plug-in hybrid gasoline Diesel hybrid Gasoline hybrid CNG/LPG Diesel Gasoline 22% from variable renewables in 2050 Almost 60% of LDV sales are EVs or PHEVs in 2050 Increased end-use electrification: heat, industrial end-uses
Role of storage in the energy system Improving energy system resource use efficiency Helping to integrate higher levels of variable renewable resources and end-use sector electrification Supporting greater production of energy where it is consumed Increasing energy access Improving electricity grid stability, flexibility, reliability and resilience. OECD/IEA 2012 OECD/IEA 2013
Heterogeneous mix of technologies
With highly varying technical performance
Future role of storage will defined according to its applications
Costs of electricity storage technologies
Cost of electricity storage depends on its application (1) 2 000 1 600 Inter-seasonal arbitrage (120 h; 500 MW; 5 #/year) 1 000 800 Daily price arbitrage (8 h; 300 MW; 0.75 #/day) 1 000 800 T&D investment deferral (5 h; 100 MW; 1 #/day) 1 200 600 600 800 400 400 400 200 200 0 0 0 PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW 1 000 Load following (1 h; 50 MW; 3 #/day) 1 000 q y g ( ; ; Frequency regulation (10 min; 10 MW; 30 #/day) 1 000 g ( ; ; / y) Off-grid (3 h; 5 kw; 1 #/day) 800 800 800 600 600 600 400 400 400 200 200 200 0 0 0 PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW Price setter 2012 2030 LCOE 2012 2030
Cost of electricity storage depends on its application (2) 2 000 Inter-seasonal arbitrage (120 h; 500 MW; 5 #/year) 1 600 1 200 LCOE 2012 2030 800 400 0 Price setter 2012 2030 PHS H2 CAES NaS LA VRB Li-ion FW
Cost of electricity storage depends on its application (3) 1 000 Daily price arbitrage (8 h; 300 MW; 0.75 #/day) 800 600 LCOE 2012 2030 400 200 0 Price setter 2012 2030 PHS H2 CAES NaS LA VRB Li-ion FW
Cost of electricity storage depends on its application (4) 1 000 g Off-grid ( ; (3 h; 5 ; kw; / 1 #/day) 800 LCOE 2012 2030 600 400 200 Price setter 2012 2030 0 PHS H2 CAES NaS LA VRB Li-ion FW
Analysis of storage potential in 2DS Regions/Countries: China, EU, India, US Analysed at 1h time resolution (arbitrage, load following, integration of solar PV and wind) Detailed dispatch model: 3 cases: 2DS: estimation of storage requirements Breakthrough: as 2DS plus agressive cost reductions for storage comparable to CCGT for arbitrage (e.g. LCOE of USD 90/MWh compared in EU) EV/DR: 2DS plus DR; 2DS projection of number of EVs (e.g. for the US: 45 million EVs in 2050 resulting in a total storage capacity of 1 TWh)
Potential for storage in key regions Attractiveness of storage is highly context-specific, but increases greatly across all scenarios Depending on the system, flexibility from demand response could provide greater returns Long-term model results for storage in 2DS 68 GW 15 GW 22 GW 53 GW
Key changes in wholesale price duration curve in low carbon future Avoidance of extreme prices during scarcity periods with more DR, storage Reduced curtailment, other technologies setting price during periods where low SRMC was price setter
GW 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 GW Does flexibility have a value today? planning ahead for storage 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 1 3 5 7 9 11 13 15 17 19 21 23 Total generation 0 2 4 6 8 10 12 14 16 18 20 22 Conventional Wind PV GW 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0 2 4 6 8 10 12 14 16 18 20 22 Conventional Wind PV Medium-term: PV reduces value of pumped storage Long-term: Mutual increase of market value But long lead times for many storage projects, learning by doing Source: IEA, renewable energy division
Storage is one of a suite of options for providing flexibility
Roadmap Key Findings Storage can support energy system decarbonisation, but strategic planning required Some technologies already competitive in some applications. For biggest markets, storage still too expensive Additional R&D still needed to reduce costs, but must be targeted: key knowledge gaps Optimal role for storage varies widely across regions Power markets are ill-equipped to compensate storage for suite of services they can provide Thermal energy storage systems could make better use of wasted heat
Key actions over the next 10 years Economics of retrofitting existing of storage facilities Develop markets and regulatory environments that enable accelerated deployment i.e benefits-stacking Support targeted demonstration projects and R&D Establish a comprehensive set of international standards Establish international and national data co-operation Analysis required: assessments to quantify the value of storage in specific regions and energy markets
Next phase of work: Multi-carrier energy system integration Smart grid roadmap 2015 Smart cities ETP 2016
Thank you! Any questions? luis.munuera@iea.org
Extra slides
Long-term ETP-TIMES Model
Sharp Declines in Costs Needed 1 200 Levelised cost of electricity (USD / MWh) 1 000 800 600 400 200 0 PSH Hydrogen CAES Sodium Sulphide Lead Acid Vanadium Redox Current cost range 2DS cost target Breakthrough cost target Lithium-ion
Deep decarbonisation in other sectors impacting structure of electricity demand and DR potential PLDV annual sales (millions) 200 180 160 140 120 100 80 60 Electrification of transport FCEV Electricity Plug-in hybrid diesel Plug-in hybrid gasoline Diesel hybrid Gasoline hybrid CNG/LPG Change in industrial structure DR in industry: Captive Generation reducing electricity imports in low carbon scenarios 40 Diesel 20 Gasoline 0 2000 2010 2020 2030 2040 2050 Almost 60% of LDV sales are EVs or PHEVs in 2050 Decisions taken today could pose operational challenges in the future how to advise? Swiss knife: how do investments in energy technologies affect future operation, and how does future operation affect investment outlook today? Need to look in detail at time we use 4 typical days for each of the 28 regions, at a one hour resolution
The swiss knife: Evaluating long-term decarbonisation of the electricity sector Enhanced ETP-TIMES model (long-term; horizon up to 2050): 3h-load segments for a typical day (8 per day, four typical days per year) Large-scale storage: electricity, heat, hydrogen Considering other flexibility options for the electricity system: Flexible generation technologies Inclusion of demand response, e.g. V2G Investment decisions in generation technologies and first estimate on flexibility needs Linear TIMES dispatch model for operational analysis (short-term; one year): 1h-timeslice resolution Analysing operation of electricity system within a year for specific region with investment decisions for generation technologies from long-term model Additional operational constraints (ramp-up/-down, min load, min up/down times) Improved analysis on flexibility needs and role of competing flexibility options
complementing a long-term planning model Fuel costs Flexible generation Technical and economic characteristics Demand side management DSM Electricity and heat demands Potentials Load curves DSM DSM Fuel demand Electricity storage Pumped storage Energy storage CAES District heat storage Generation mix New capacities Emissions Electricity prices Average generation costs Process heat storage Flexible uses in conversion sector
Optimal dispatch analysis of ETP longterm decarbonisation scenarios Operational constraints for a range of plants: Ramping Minimum up/down times Turn-down ratios Part-load efficiency approximation Storage modelled for: Load following Arbitrage/load-levelling Seasonal storage DR at the individual technology level Heat pumps/thermal storage Flexible district heating DR in industry (captive generation) Grid-to-vehicle
Complementing a long-term model wtih a linear dispatch model Enhanced ETP-TIMES model (long-term; horizon up to 2050): 3h-load segments for a typical day (8 per day, four typical days per year) Large-scale storage: electricity, heat, hydrogen Considering other flexibility options for the electricity system: Flexible generation technologies Inclusion of demand response, e.g. V2G Investment decisions in generation technologies and first estimate on storage needs Linear TIMES dispatch model for operational analysis (short-term; one year): 1h-timeslice resolution Analysing operation of electricity system within a year for specific region with investment decisions for generation technologies from long-term model Additional operational constraints (ramp-up/-down, min load, min up/down times) Improved analysis on storage needs and role of competing flexibility options
Long-term ETP-TIMES model for the power sector Fuel costs Flexible generation Technical and economic characteristics Demand side management DSM Electricity and heat demands Potentials Load curves DSM DSM Fuel demand Electricity storage Pumped storage Energy storage CAES District heat storage Generation mix New capacities Emissions Electricity prices Average generation costs Process heat storage Flexible uses in conversion sector
Challenge of integrating increasing shares of variable renewables Share of variable renewables in electricity generation (%) 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 2011 2030 2050 2011 2030 2050 2011 2030 2050 2011 2030 2050 China India European Union United States
Significantly different structure of wholesale markets -> large impact of constraining flexibilty More DSI and lower cost storage Ramping/part-load efficiency constraints -> high impact on SMP curves DSI most attractive option for reducing total system costs
Example output: Operational constraints in 2050 + Pumped storage + DR (EVs, DHW) Gas turbines Pumped storage output Pumped storage input Demand response Gas (CCS + no-ccs) Coal (CCS + no-ccs) Nuclear
Increasing role of electricity in the 2DS Electricity demand growth differs between industrialised and industrialising countries, but an increasing share of electricity in the overall energy mix follows similar trends.
We have the flexible resources Four sources of flexibility Grid infrastructure Dispatchable generation Storage Demand side integration No one flexible resource meets all the needs.
Prices under deep decabonisation
We have the flexible resources Four sources of flexibility Grid infrastructure Dispatchable generation Storage Demand side integration No one flexible resource meets all the needs.
Merit order of flexibility options vary across time and space
Future role of storage will defined according to its applications
Cost of electricity storage depends on its application (1) 2 000 1 600 Inter-seasonal arbitrage (120 h; 500 MW; 5 #/year) 1 000 800 Daily price arbitrage (8 h; 300 MW; 0.75 #/day) 1 000 800 T&D investment deferral (5 h; 100 MW; 1 #/day) 1 200 600 600 800 400 400 400 200 200 0 0 0 PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW 1 000 Load following (1 h; 50 MW; 3 #/day) 1 000 q y g ( ; ; Frequency regulation (10 min; 10 MW; 30 #/day) 1 000 g ( ; ; / y) Off-grid (3 h; 5 kw; 1 #/day) 800 800 800 600 600 600 400 400 400 200 200 200 0 0 0 PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW PHS H2 CAES NaS LA VRB Li-ion FW Price setter 2012 2030 LCOE 2012 2030
Cost of electricity storage depends on its application (2) 2 000 Inter-seasonal arbitrage (120 h; 500 MW; 5 #/year) 1 600 1 200 LCOE 2012 2030 800 400 0 Price setter 2012 2030 PHS H2 CAES NaS LA VRB Li-ion FW
Cost of electricity storage depends on its application (3) 1 000 Daily price arbitrage (8 h; 300 MW; 0.75 #/day) 800 600 LCOE 2012 2030 400 200 0 Price setter 2012 2030 PHS H2 CAES NaS LA VRB Li-ion FW
Cost of electricity storage depends on its application (4) 1 000 g Off-grid ( ; (3 h; 5 ; kw; / 1 #/day) 800 LCOE 2012 2030 600 400 200 Price setter 2012 2030 0 PHS H2 CAES NaS LA VRB Li-ion FW
Long-term ETP-TIMES model for the power sector Fuel costs Flexible generation Technical and economic characteristics Demand side management DSM Electricity and heat demands Potentials Load curves DSM DSM Fuel demand Electricity storage Pumped storage Energy storage CAES District heat storage Generation mix New capacities Emissions Electricity prices Average generation costs Process heat storage Flexible uses in conversion sector
Linear dispatch model: Model formulation (1) Objective function Min ( AnnInvCst p + FomCst p ) VAR_NCAPp + p Annualised investment costs Fixed OM costs Existing capacity p 1 Eff p FuelPric p + VarOMCst Fuel price Variable OM costs Efficiency p VAR_GEN t,p Constraints VAR_CAP = CapResid + VAR_NCAP Total installed capacity p p p p PkCnt Decision variables Electricity generation VAR_GEN t,p VAR_GENt,p p p VAR_CAP ( = ) VAR_CAP Demand VAR_GEN, VAR_NCAP t,p p p PkRsv p New capacity t p Demand Peak reserve margin AvailFac t t,p Indices p t Generation-capacity constraint Demand constraint Peaking constraint generation technologies hours Availability factor
Operational constraints MW Available capacity for a technology Online capacity 1 t MW Generation 2 t MW Minimum operation time 6 Part-load efficiency 8 9 Start-up costs Ramp-up constraint 4 Minimum down time 7 Minimum load 3 Ramp-down constraint 5 t
Linear dispatch model: Model formulation (2) Defining capacity online Minimum generation 1 2 3 VAR_CAPON t,p VAR_CAP VAR_GEN t,p VAR_CAPON t, VAR_GEN t,p VAR_CAPON p AvailFac t,p p t,p MinGen p Ramp-up and down constraints Minimum operation hours 6 VAR_CAPON t,p 4 5 VAR_GEN VAR_GEN t,p t 1,p t VAR_GEN 1,p t 1,p VAR_GEN t,p VAR_CAPON VAR_CAPON τ,p t,p t,p RmpUp RmpDn VAR_CAPON + VAR_CAPON τ t MinOpTime p p p Minimum downtime hours constrain start-up of online capacity VAR_CAPON VAR_CAPON = VAR_CAPONSTR 7 t,p t 1,p VAR_CAPONSTR t,p VAR_CAP t,p t,p VAR_CAPON τ,p τ t MinDnTime p Approximating part-load efficiency by adding fuel cost penalty term in objective function 1 Start-up costs to bring capacity online 8 8 EffOnl p FuelCst p VAR_CAPON NCAP_STRUPCST VAR_CAPON p t,p t,p
Modelling of storage Electricity Storage in term of capacity ( MW ) Pumped storage Pump Turbine Electricity stored Storage reservoir Storage in term of energy ( MWh ) DNI category Concentrated solar power plants with thermal storage DNI Heat Electricity Solar field Thermal storage Power generation For existing pumped storage, lack of data on storage volume in terms of energy CSP plant broken down in solar field, heat storage and turbine block
EV charging and PV integration Solar PV Solar PV and uncontrolled EV charging Electricity demand Solar PV generation EV charging Net load Solar PV and controlled EV charging Adapted from Denholm et al. (2013)
Example: Dispatch for a summer week in the US (EV/DR case)
Impact of EVs in the US (1) DR from EVs displacing storage during peak demand periods
Impact of EVs in the US (2)
Findings, limitations of analysis and future work Conclusions: Storage is an important option, but not the sole solution for flexibility needs, as it competes with other options (DR, flexible generation, grids) DR from EVs in this analysis appears attractive, but hinges on smarter grid infrastructure Key question for DR is hidden/transaction costs and uncertainty over achievable DR take-up Limitations: Flexibility from interconnections through enlarged balancing areas difficult to consider, without explicitly consider spatial system structure Uncertainty over profile of future load curves, particularly in emerging economies, coupled with difficulty of modelling individual end-uses Ideas for future work: Including investment decisions for analysis of a greenfield system Expanding DR options Including flexibility options outside of electricity sector (heat, fuels)
Thank you! www.iea.org/etp2014