1. Optimising energy storage to balance high levels of intermittent renewable generation Paul E. Dodds UCL Energy Institute, University College London Presented at 33 rd USAEE/IAEE North American conference in Pittsburgh 27 th October 2015 Thanks: Ed Sharp, Birgit Fais, Hannah Daly (all UCL)

2. The challenge

7. Storage technologies MECHANICAL / THERMOMECHANICAL/ GRAVITATIONAL Pumped Hydro Heat Pumped Temperature Difference System Liquid Air Energy Storage (LAES) Compressed Air Energy Storage (CAES) CAES Undersea Bags Pumped Hydro with Compressed Air Flywheels Advance Rail Energy Storage ELECTRO-CHEMICAL Rechargeable Batteries (e.g. Lead–acid, Lithium–ion, Sodium–sulfur) Vanadium Redox Flow Batteries Supercapacitors THERMAL Phase Changes Solar Ponds Sensible Thermal Energy Storage: Diurnal and Seasonal CHEMICAL Hydrogen from Water Electrolysis Chemical Reactions (zeolites/water/ inorganic oxides) Power to Gas Large Scale Hydrogen Storage Traditional Energy Storage (natural gas, oil and coal) OTHER Superconducting Magnetic Energy Storage

8. Comparing technologies Specific (output) energy (J/kg) Output energy density (J/m3) Specific power (W/kg) (o/p & i/p) Power density (W/m3) (o/p & i/p) Minimum natural energy & power scales of a single device (J & W) Optimum natural energy & power scales of a single device (J & W) Nominal cost per unit energy & power at optimum scale (£/J & £/W) Marginal cost per unit energy & power at optimum scale (£/J & £/W) Lowest power slew rate at which performance degrades noticeably (W/s) Effective turnaround efficiency etc…

10. Energy system model: UK TIMES (UKTM) Developed from the UK MARKAL energy systems model: Bottom-up Perfect foresight Cost-optimisation – CAPEX, OPEX are model inputs; fuel prices and the price of carbon each year are calculated It helps us to understand: Energy flows through the economy in the future Impact of environmental and other constraints

11. UK high renewable scenario No new nuclear No CCS

12. Energy storage in 2050 Thermal storage includes night storage heaters but excludes hot water tanks. No demand-side response is assumed.

13. Analysing the future role for energy storage Grid-scale electricity models Only the electrical system Fixed electricity demand Only grid-scale storage High temporal resolution 30 mins/1 hour Several studies Energy system models Whole energy system Flexible electricity demand All types of energy storage Low temporal resolution 4 daily periods, 4 seasons No studies

14. What is the impact of low temporal resolution? 1.Calculate excess generation from UKTM electricity portfolio in a high-resolution electricity model 2.Prevent UKTM from using the excess electricity to meet direct electricity demand 3.Check that UKTM is building the required technology capacity for the scenario

15. Energy storage in 2050

16. What is the impact of low temporal resolution? Consequences: Increases electricity consumption by 6% Change mainly affects non-renewable generation: – Renewable reduced by 1% – Baseload up by 10%, flexible up by 5% New investments in CAES and power-to-gas for hydrogen consumption in the wider energy system

17. Impact on hydrogen consumption

18. Conclusions Low-resolution energy system models do not properly represent the difficulties of using intermittent renewable generation. High-resolution electricity system models do not consider the wider types of energy storage in the energy system, which could be cheaper alternatives to grid-scale electrical storage. Several options – more networks, more storage, DSM/DSR We don’t as yet know the most appropriate way to integrate energy storage into the UK energy system in the future.

19. Thank you for listening