The Role of Regulation & Policy in facilitating the optimal lowest cost integration of Variable Renewable Energy

Variable Renewable Energy (VRE) from wind and solar is frequently cited as the solution to UK Electricity decarbonisation. Technology learning curves suggest that, depending on location, wind and solar, which accounted for ~8% of global electricity generation in 2019¹, are currently², or projected to become, the “cheapest” source of new generation in most countries by 2030³.Solar and onshore wind power are now the “cheapest” new sources of electricity generation in countries encompassing at least two-thirds of the world’s population, according to Bloomberg NEF³.In its 2020 World Energy Outlook, the IEA declares that solar power schemes now offer the cheapest electricity in history with the technology cheaper than coal and gas in most major countries^{4 5}. Recent Contract-for-Difference (CfD) auctions in the UK suggest that offshore wind is becoming “competitive” with fossil fuel substitutes, while onshore wind is “already” at that point⁶. The LCOE of Nuclear power trails both wind and solar by an appreciable margin².

However, such simple Levelized Cost (LCOE) analyses and technology learning curve projections do not necessarily convey the full costs of VRE integration or indeed the enormity of the system and deployment challenges. Persistent critics of Renewables, and offshore wind in particular, have also raised concerns that recent falling offshore wind CfD auction prices may reflect a winner’s curse, embedding an unrealistic assessment of operating and decommissioning costs, over-optimistic expectations of load capacity through asset lives, perverse incentives imparted by a poorly structured CfD regime, and speculative investment fuelled by poor lending disciplines^{7 8}. Such critiques have often tended to start from a prejudice against renewables, but concerns raised around the drivers behind and sustainability of low prices bear scrutiny.

Denmark is most commonly cited as a model of VRE success, generating ~50% of its electricity from on and offshore wind, and targeting 100% from renewable sources by 2030⁹. The Danish energy system is also substantially decentralised, with a network of small Combined Heat & Power (CHP) plants, nearly two-thirds of which are fuelled by carbon neutral feedstocks, connected to District Heating networks, generating both grid electricity capable of responding relatively quickly to peaks in demand or troughs in wind energy generation, while utilising waste heat from generation to warm over 60% of Danish households; meaning that overall, ~70% of electricity needs are met from renewable sources, while recourse to traditional fossil fuel building heating solutions is reduced substantially.

Nevertheless, Denmark has the most expensive household electricity in the EU¹⁰, driven substantially by VAT and the PSO green levy¹¹, is a net power importer (~15%) mainly from Germany, still utilises substantial quantities of fossil fuels (~30%) in meeting its power needs, and relies on a complex system of balancing interconnectors with Germany and Scandinavia, supplemented by surplus VRE pump-and-store Hydro-capacity based in Norway and Sweden¹². Pricing to customers should, however, fall as the PSO levy is set to be phased out by 2022. The aspiration to move to 100% Renewables also implies that CHP will only ultimately be available from biomass-fired plants. The extent to which the Danish model, with its easy access to European electricity supply balancing, is economically and geographically unique or sufficiently scalable to be applied to the UK is as yet unclear. Denmark is small with high inter-connectivity, while the UK is large with limited inter-connectivity. Given Denmark’s aspiration to be 100% renewable and their reliance on balancing inter-connectivity, they can only really achieve this in a net sense by exporting surplus renewables and importing predominantly non-renewables at times of deficit. If all Denmark’s neighbours adopt a similar approach, then that model might struggle. The UK is a long way from Denmark’s capacity to utilise CHP meaningfully in system balancing. However, its system stability and security could perhaps benefit at the margin from greater international connectivity. California, an economy more similar in size to the UK, and more reliant on solar than wind, has by contrast been much less successful with VRE integration, encountering substantial power outage problems. These failures appear to materially emanate from poor system integration planning and execution and limited inter-connectivity with other systems^{13 14}.

Clearly, different regions will offer different VRE opportunities with associated technical challenges, larger economies might face more complex VRE system integration challenges but also offer greater capacity for flexibility in addressing them, while divergence in regional backup power and storage solutions seems probable. The means by which integration of VRE is paid for by consumers is clearly an important issue, while the extent to which the integration of VRE can be incremental rather than viewed in the context of whole system energy supply and demand is also highlighted.

In the UK, the VRE integration challenge reduces substantially to the flexibility and “smartness” of the grid, the extent to which a whole energy system approach is taken, the ability to manage demand flexibly, optimal geographic dispersion of VRE capacity, and the system integration costs associated with the provision of adequate spinning reserve, backup capacity, and energy storage¹⁵.

The most well established, efficient and reliable energy storage technology is pump-and-store hydro. Batteries offer short-term storage and capacity but are expensive and not currently a long-term storage solution. Their strength is in their ability to provide high powered output, responsiveness, and frequent cycling with low losses. Compressed air¹⁶ is a potential long-term energy storage technology which the UK might exploit with its access to empty North Sea wells, but it is yet to be demonstrated at that scale. Limitations on recourse to pump-and-store hydro and batteries at current technology levels suggests Open (OCGT)¹⁷ or Combined Cycle Gas Turbines (CCGT) or Gas Engines, utilising green hydrogen, as a more immediately viable source of backup capacity for VRE in the UK. OCGTs and CCGTs are relatively cheap to build, can operate on methane or hydrogen, and can be brought online relatively quickly. The engineering innovation challenges to deliver such a system are broadly incremental i.e. promoting devices to support system frequency stability, reducing electricity storage costs, reducing hydrogen electrolyser capital costs while increasing their efficiency, and perfecting hydrogen ready turbines.

However, the cost of integrating VRE rises with its increasing dominance in the electricity system, as incentivisation of backup capacity becomes more expensive, grid flexibility, system supply and frequency balancing challenges become greater, making the Levelized Costs (LCOE) of VRE generation a questionable metric for determining its optimal penetration. The externalities around fossil fuel uses may justify carbon taxes, but VRE also carries its own system externalities. It is important to think about how these externalities might be charged to arrive at a transparent, lowest cost and ultimately decarbonised system. Optimal penetration will be a function of the marginal cost of increased supply, but these in turn are almost certainly contingent on the flexibility of the system. UKERC’s studies on the costs and impacts of intermittency support this view^{18 19}. Various energy technologies too exhibit divergent cost structures between capital and the marginal costs of operation with implications for their optimal load. A number of studies²⁰ converge on similar conclusions around the integration of VRE: the level of system flexibility can reduce integration costs substantially; provision of back up capacity dominates integration costs; diversification of the renewables portfolio tends to facilitate a higher overall share; and rising integration costs are associated with higher periods of VRE generation in excess of demand, although as the LCOE of VRE falls, so do the costs of untapped excess generation.

The Committee on Climate Change²⁰ acknowledges that although all energy technologies impose such system costs, these are likely to be higher for variable renewables. Nevertheless, its modelling suggests that by 2050, the system integration costs of VRE at 57% “optimal” penetration could be less than 10% of overall system costs and that increasing load factors from offshore wind may imply a higher optimal penetration. It also suggests that VRE might earn additional rents for supplying ancillary services that thermal plants currently provide, like fast frequency response, reserve, reactive power and inertia control. More predictable less correlated renewables like wave and tidal power could also play complementary roles. It is important to note in this context that very long-lived renewable energy assets like Tidal Barrages are hard to evaluate in an LCOE framework and so may be unnecessarily excluded as a part of the decarbonisation solution.

Additionally, there is the critical question of how end-use energy challenges in the wider economy are resolved i.e. in building heating, recourse to a hydrogen repurposed gas network, or combined heat and power generation (CHP) with district heating networks, or one more reliant on heat pumps; in generation, the extent to which nuclear is retained for base load power, CHP is deployed as a flexible supply solution, or VRE with green hydrogen powered OCGT/CCGTs comes to dominate; in vehicles, shipping, short-haul aviation and public transport, the extent to which batteries prevail over hydrogen solutions; in heavy industry, the extent to which blue or green hydrogen²¹  becomes the principal solution. Depending on the answers to these specific questions, the UK’s physical demand for electricity could be well over twice its current levels by 2050^{20 22}. The Committee on Climate Change see low-carbon electricity generation quadrupling by 2050, with renewables providing over half the electricity, and nuclear and decarbonised gas (blue or green hydrogen or methane + Carbon Capture) making up the balance²⁰.

So, a number of obvious questions arise: What proportion of VRE can feasibly dominate electricity generation from a system perspective? How much VRE can be physically accommodated and deployed in absolute terms across and around the British Isles? Over what time horizon can this transformation take place? Should the transformation begin with methane powered turbines, transitioning through blue to green hydrogen backup generation, or seek to make an immediate jump to green hydrogen, or place greater emphasis on other clean energy technologies like bio-thermal or decarbonised gas? What storage technologies should be prioritised? What alternative to thermal plants can be sourced to provide system inertia, voltage and frequency management services? How can associated incremental engineering innovation be promoted? From a marginal cost perspective, what is the optimal penetration of VRE vs. other low or no carbon solutions? Is Marginal Cost an appropriate approach to a system dominated by low or even no marginal cost forms of generation? How should markets and regulations be reformed to provide better incentives for flexibility so that integration costs can be minimised? How can the overall costs of system transformation be minimised? To what extent is a whole system approach required?

Modellers at UCL, Edinburgh & Strathclyde Universities, have done substantial work to investigate the extent to which a dispersed VRE system would suffer from intermittency and to better understand the implications of highly uncontrollable power generation in a decarbonised generation network, for instance: how much excess electricity could be generated by VRE for conversion to green hydrogen for backup power, the extent to which backup generation would be sufficient in times of peak demand or VRE supply troughs, and the means by which provision of sufficient spinning reserve¹⁵ to maintain system frequency could be facilitated. The modelling also explores cost-optimisation of investments in generation and energy storage to meet demands in the context of weather variability over many years. UKERC²⁵ note that models like UK TIMES²³ are relatively low resolution when it comes to matching electricity supply and demand in future net-zero contexts, leaving some scope for supply unreliability; while models like highRES are perhaps better suited to answering questions around system security and reliability²⁶.

Geospatial analysis by the UCL highRES²⁴ team estimates UK offshore wind potential from all technologies to be up to 8450 TWh/year; compared to current UK consumption of around 300 TWh/year and UK TIMES 2050 projected annual demand of 503TWh. The highRES modelling works to an assumption of very high VRE penetration²⁷. It sees potential for the emerging technology of floating offshore wind to complement conventional bottom-mounted offshore wind via reduced backup capacity requirements accrued from greater spatial diversification; noting that policy might also need to incentivise developers to build in system-optimal locations as opposed to just concentrating on total energy output²⁷.

UK TIMES²³ model scenarios have suggested that excess power would be converted to green hydrogen for backup and other uses. The modelling in general suggests with a high level of confidence that 60% of UK electricity could be accommodated from VRE, with backup from a mix of hydrogen and methane OCGTs²⁸ or CCGTs; recourse to methane would only ultimately be viable at very low load factors or by recourse to CCS. VRE penetration above 85% is considered more challenging. It is important to note that Frequency Stability¹⁵ is a key problem at high levels of VRE penetration even when renewables supply exceeds demand. Some of the modelling by Strathclyde suggests that the upper VRE limit might need to be lowered to avoid large changes in frequency. The whole system frequency stability assessment is therefore critical in assessing what higher levels of VRE can be accommodated.

Quite a number of studies²⁰ suggest that a 60-80% penetration of renewables could be integrated into the UK energy system; however, these don’t address the associated integration costs. The greater the penetration of renewables in the system, the more critical the model assumptions associated with them become, especially around system flexibility. Projections of growth in energy demand, trajectories of technology learning and costs around competing energy solutions consistent with decarbonisation pathways, are a hostage to fortune. These uncertainties can throw models off track very quickly, requiring a flexible approach to system evolution. Current CCC sponsored studies²⁰ suggest, with the assumption of improved system flexibility, an integration cost of £10-25/MWh for annual penetrations of 50-65% renewables vs. a current LCOE of wind and solar at £50/MWh.

Renewables, in the round and with considerable intra-year variability, accounted for 37% of UK electricity generated in 2019²⁹, with ~24% coming from wind and solar, ~11.5% from thermal renewables, the balance from Hydro²⁹. Theoretically at least, VRE offers substantial potential for the UK, and there is a clear policy steer in the direction of increasing that proportion significantly.

Such a system will need to incentivise backup capacity, storage, spinning reserve for system stability, while facilitating mechanisms that permit Hydrogen Electrolyser and Electricity Storage providers to bid for excess renewables generation. Traditionally, base load power has been sourced from inflexible high capital/low fuel cost plants like nuclear. The divergence in LCOEs between nuclear and VRE’s, even allowing for VRE system integration costs, also begs the question as to whether the role of nuclear in such a system is made redundant and/or whether Biomass+CCS plants might perform the dual roles of providing baseload and spinning reserve. It is also possible that there might be a role for closed loop geothermal generation. Finally, there is the question as to how the current architecture of regulation and supply can accommodate the move from a system optimised by wholesale electricity markets arbitrating lowest marginal cost producers at any one time supplemented by capacity payments, to one where zero marginal costs of supply pervade, but inflexible supply or generation requires intermittent access to much greater storage, and flexible generation capacity?³⁰

OFGEM²² and National Grid³¹ have both emphasised the importance of net-zero and the importance of system supply flexibility, but there are still important questions about how that translates into specific market design, regulations and policies, including those that can support the development and deployment of associated energy technology innovation. It is also important to determine the limitations on technology neutrality such that all energy flexible generation options are ultimately compatible with net-zero pathways. In turn, one also needs to build sufficient room for uncertainties around the evolution of demand and competing energy technology learning in order to facilitate solutions which minimise costs. The policies, market design and regulatory regime that can promote that dynamic change in a lowest cost manner, remain somewhat elusive. Can the current system of regulation accommodate these requirements through modification or are more fundamental reforms required?

A number of possible questions might be explored around market design, policy and regulation:

1. An electricity system substantially reliant on Variable Renewable Energy (VRE) will encounter system integration costs requiring recourse to electricity supply balancing, flexible demand, interconnection, spinning reserve¹⁵, and storage solutions. What form might these take? Can the current system of regulation and market design accommodate these requirements through modification or are more fundamental reforms required?
2. Is Capacity Markets technology neutrality balanced by targeted subsidy an effective approach to delivering lowest cost decarbonisation solutions in a period of enforced dynamic system change?
3. What are the physical limitations on VRE in satisfying UK electricity power generation needs? To what extent does a Whole Energy System approach need to be taken by Regulators & Policy Makers?
4. How can the system integration costs of VRE be understood, estimated, minimised and transparently charged? Should VRE suppliers be forced to bid for backup capacity to cover their own intermittency?
5. Should inflexible generators like nuclear pay for the complementary measures and technologies required to accommodate them?
6. In a system dominated by VRE, how might Policy Makers & Regulators incentivise the provision of adequate spinning reserve¹⁵, synthetic inertia from non-traditional generation sources, back up, and energy storage capacity, while promoting innovation around these solutions?
7. How might the economics of VRE generation be optimised such that Electrolysers and Electricity Storage owners can bid for excess production?
8. How can Government & Regulators reconcile technological neutrality with uncertainty and the need for speed in decarbonising the energy system?
9. In determining the optimal contribution of VRE or long-lived Renewable Energy assets to the energy system, are Levelized Costs a useful arbiter? What alternative approaches might be adopted?
10. Are VRE financing and incentive regimes leading to speculative investment and deployment with possible adverse cost consequences for electricity consumers and governments in the future? How might this be avoided while promoting electricity system decarbonisation?

© Gerard Fox, Regulatory Policy Institute, November 2020

Footnotes & References:

1. Ember Global Electricity Review 2020: https://ember-climate.org/project/global-power-2020/
2. Lazard Levelized Cost of Energy & Energy Storage, 2019: https://www.lazard.com/perspective/lcoe2019/
3. Bloomberg NEF, (2019): https://about.bnef.com/new-energy-outlook/#toc-download
4. Carbon Brief (2020): Solar is now “cheapest electricity in history”, conforms IEA: https://www.carbonbrief.org/solar-is-now-cheapest-electricity-in-history-confirms-iea
5. IEA (2020), World Energy Outlook 2020, IEA, Paris: https://www.iea.org/reports/world-energy-outlook-2020
6. Carbon Brief, 2019: Record low price for offshore wind cheaper than existing gas plants by 2023 https://www.carbonbrief.org/analysis-record-low-uk-offshore-wind-cheaper-than-existing-gas-plants-by-2023
7. The Costs of Offshore Wind Power: Blindness and Insight (GWPF, 2020): https://briefingsforbritain.co.uk/the-costs-offshore-wind-power-blindness-and-insight/
8. Who’s the Patsy? Offshore wind’s high stakes poker game (GWPF, 2019): https://www.thegwpf.org/content/uploads/2019/07/WindPokerWeb-1.pdf
9. Renewable Energy in Denmark: https://en.wikipedia.org/wiki/Renewable_energy_in_Denmark
10. Eurostat (2020): Electricity price statistics: https://ec.europa.eu/eurostat/statistics-explained/pdfscache/45239.pdf
11. BALTPOOL (2020): What is PSO? https://www.baltpool.eu/en/what-is-pso/
12. Danish Utility Regulator National Report 2019: https://r.search.yahoo.com/_ylt=AwrJ7JfcqWRfKc8A7IhXNyoA;_ylu=Y29sbwNiZjEEcG9zAzQEdnRpZAMEc2VjA3Ny/RV=2/RE=1600461405/RO=10/RU=https%3a%2f%2fforsyningstilsynet.dk%2fmedia%2f6616%2fnational-report-2019-for-denmark.pdf/RK=2/RS=THfnFSXv9vHbwyYkY9FqVNN_p7Y-
13. Bad Policy, Not Renewables to blame for California blackouts (2020): https://cleantechnica.com/2020/08/21/bad-policy-not-renewables-to-blame-for-california-blackouts/
14. Renewables didn’t cause California Blackouts, Experts Correct the Record (2020): https://www.seia.org/blog/renewables-didnt-cause-california-blackouts-experts-correct-record
15. One needs to distinguish between online but unloaded generating capacity (spinning reserve) that can be called within 10 mins, frequency responsive spinning reserve (10 seconds), offline but callable-at-short-notice-supply (10 mins) and offline or curtailed supply that can be called on in short order ( 1 hour).
16. Compressed Air Energy Storage (2019): https://www.sciencedirect.com/topics/engineering/compressed-air-energy-storage
17. Open Cycle Gas Turbines: https://www.sciencedirect.com/topics/engineering/open-cycle-gas-turbine
18. UKERC (2017): https://ukerc.ac.uk/project/the-intermittency-report/
19. Nature Energy (2020): https://www.nature.com/articles/s41560-020-00695-4?utm_campaign=socialtraffic&utm_medium=social&utm_source=NEtwitter
20. CCC (2019), Net Zero Technical Annex: https://www.theccc.org.uk/wp-content/uploads/2019/04/Technical-Annex-Integrating-variable-renewables-into-the-UK-electricity-system.pdf
21. The Clean Hydrogen Future has Already Begun (IEA, 2019): https://www.iea.org/commentaries/the-clean-hydrogen-future-has-already-begun
22. Ofgem decarbonisation action plan, 2020: https://www.ofgem.gov.uk/system/files/docs/2020/02/ofg1190_decarbonisation_action_plan_revised.pdf
23. UCL UK TIMES is an energy system model that identifies decarbonisation pathways across the whole UK economy to meet our net zero emissions target by 2050, but with low temporal and no spatial resolution: https://www.ucl.ac.uk/energy-models/models/uk-times
24. UCL highRES model is an electricity system planning and operations model for the UK, with high spatial and temporal resolution. https://www.ucl.ac.uk/energy-models/models/highres
25. UKERC (2020): https://ukerc.ac.uk/about/
26. UKERC (2018), The Security of UK Energy Futures: https://ukerc.ac.uk/publications/uk-energy-security/
27. The role of floating offshore wind in a renewable focused electricity system for Great Britain in 2050 (2018): https://reader.elsevier.com/reader/sd/pii/S2211467X18300920?token=9E4F8CFFA4309C9AA53E88DC44C9CC53C4305558DFC9F26A7CBDC01C87D041D65D7BE273A703AD55C62D1A6E9B1E9E7A
28. Open Cycle Gas Turbines: https://www.sciencedirect.com/topics/engineering/open-cycle-gas-turbine
29. DUKES 2020, pp87: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/912023/DUKES_2020_MASTER.pdf
30. Helm, D., (2018), Burn Out, pp.204-223
31. National Grid Future Energy Scenarios (2020): https://www.nationalgrideso.com/document/173821/download