My article for October sets out a framework which quantifies potential future cost reductions for different generation technologies. It specifies a £ per MWh cost target for the country and then recommends that subsidies are distributed according to each technology’s likelihood that it can achieve the cost target in the future. In particular, the article:
- identifies the need for a consistent energy framework in the UK;
- establishes a cost target for the UK in £ per MWh;
- suggests a framework for how to assess the likelihood of costs being lower for each technology;
- quantifies the value of learning using Hinkley Point as a specific example; and
- summaries the framework within a series of recommendations for an ongoing annual Generation Option Assessment (GOA).
1. Government policy needs a framework to consistently evaluate options and to guide decision making.
While the current UK Government is an advocate of the free market, its energy policy is “hands on” whether or not it likes to admit it. The UK Government spends £7.3 billion on generation subsidies and the OBR expects this to almost double by 2020/21 (Figure 1).
The bulk of subsidy spending is on Renewable Obligations and CfDs to incentivise the build of renewable generation which can meet environmental targets. The unintended consequences of these policies are that not a single new conventional power station has been built since 2012, despite the extremely tight capacity margins we experience today in the UK and the introduction of the Capacity Market. Every pound invested in renewables, reduces aggregate wholesale electricity market by 60p.[1] Subsidies have unrecognised consequences, so Government energy policy needs to be consistent with the market over the medium to long run.
Figure 1: Current and forecast spending on emissions related energy policy
Source: July 2015 Economic and Fiscal Outlook: Fiscal Supplementary Tables.
Amber Rudd’s ‘reset speech’ in November 2015 outlined the UK Government’s energy policy and established sensible ambitions: to support technologies through their incubation phase; to provide subsidies in specifically targeted areas; and to focus on new gas, nuclear and offshore wind if their costs can become competitive.[2] It did not, however, explain how the Government intended to practically achieve these principles. Instead, policy has been inconsistent with cancellation of CCS funding and a stop-start approach towards Hinkley Point. What the UK needs is a consistent framework to objectively assess technology options and to apply this on an annual basis. This article sets out a basic suggestion of how to achieve this.
2. Commit to a specific and measureable cost target.
What does success look like for UK generation in 2030? A gold-plated electricity system would meet security of supply requirements, but at unreasonably high cost. Likewise, picking a specific technology winner today creates additional risks to security of supply and could result in high future costs relative to other options forgone.
Considering the behavioural angle can instead provide a more objective target. People do not like to pay more for electricity than they have done in the past because we expect progress rather than decline. Baseload day-ahead power has been between £40/MWh and £60/MWh since 2010 (Figure 2). Adding in the additional costs of peak supply and risk management result in a Weighted Average Cost of Electricity (WACOE) for the ‘Big 6’ supply companies of approximately £60/MWh.[3] Finally, to estimate the total cost of electricity generation, the costs of subsidies (Figure 1) need to be included. This results in total generation costs since 2010 of approximately £70/MWh.
Figure 2: UK baseload electricity price
Source: Ofgem, Electricity prices: Day-ahead baseload contracts – monthly average (GB).
Setting a well-defined, specific and challenging goal for the cost of future electricity generation has two main benefits. First, people like a clear challenge to strive towards. JFK’s goal of sending a man to the moon and bringing him back safely by the end of the decade, resulted in an obsessive catalyst for the entire US nation. Second, it allows for measurement. ‘Expensive’ or ‘cheap’ are no longer subjective statements in the eye of the beholder. Instead, technologies must be assessed relative to a defined cost of, for example, a £70/MWh[4] target so that the UK is are no worse off than our recent past.
3. Assess the likelihood of lower costs in the future
Learning is required for zero carbon generation technologies to supply at scale and reasonable cost. New nuclear technology with EPR is expensive, slow to build and has no operational cases to base decisions on. Carbon Capture and Storage (CCS) has a handful of pilots globally and none of these are in the UK. Battery technology, which could make wind’s intermittent generation more stable, is untested at scale and expensive. Any learning process consists of a number of steps:
- Demonstrating technical feasibility of a new technology;
- Commercialisation of technology (realising value through market mechanisms rather than subsidies);
- Scaling technology; and
- Delivering large capacity projects with incrementally lower costs.
The early steps need extensive research and development funding to create something new that has never existed before. Commercialisation and scaling require pilot studies to check the technologies work within markets on an increasingly large scale. Last, trial and error learning is needed at large scale to squeeze every last cost efficiency out of the construction and operational process.
Three technologies of the future have been nurtured over the past year using a seed funding and pilot approach. Carbon Capture and Storage schemes were competing for a £1 billion pot provided by the Government before they withdrew it in November 2015. Demand Side Response has been nurtured through multiple mechanisms including specific balancing products and its own auction in the capacity market. Most recently, battery storage was supported through the new Enhanced Frequency Response (EFR) product, through which National Grid tendered for 200MW of battery capacity.
Amber Rudd’s reset speech stated that the UK Government’s “intervention has to be limited to where we can really make a difference – where the technology has the potential to scale up and to compete in a global market without subsidy.” This is a sensible approach. The difficulty lies in identifying where likely potential exists, but also knowing when to stop funding and cut your losses when future progress for a technology is unlikely.
One potential approach is a framework which considers how likely costs are to change for each technology option in the future (Figure 3). If cost is highly uncertain (i.e. high volatility and a high likelihood of cost reduction), then there is a good chance that future cost could be significantly lower than the cost today. Small investments in such a technology today could provide value significantly in excess of the initial cost through cost savings of future deployments.
Figure 3: Cost volatility framework
Source: Author adaptation of Luehrman, Strategy as a Portfolio of Real Options, HBR Sep-Oct 1998.
If uncertainty is low (i.e. low volatility and low likelihood of cost reduction), then today’s cost is likely to be equal to the technology’s future cost. This is fine if costs are already low, but if the costs are high and unlikely to change then the technology is not a viable consideration for the future and should be avoided today.
Figure 4 plots a number of generation technologies onto the cost volatility framework. Interconnectors, a technology with seven links in progress that require little or no Government support is likely to be cost effective. The technology is also well established and costs are unlikely to change in the future. It therefore sits in the “green zone” and should be built today.
Figure 4: Cost volatility framework with estimated positions of generation technologies
Source: Author adaptation of Luehrman, Strategy as a Portfolio of Real Options, HBR Sep-Oct 1998.
Most other technologies are not cost effective today. New nuclear technology (EPR) is expensive relative to historic cost. Hinkley Point C is built based on £92.50/MWh. This is however, the starting point for nuclear. The last UK nuclear power plant was commissioned in 1995. It takes time to develop nuclear expertise such that the next nuclear plant is cheaper. Replacement of an ageing nuclear fleet on a global scale is likely to trigger high R&D spending over the coming decade, which could result in new and more cost effective designs, such as the Small Modular Reactor (SMR). Therefore, nuclear technology is placed in the high volatility section of the framework and is a useful option to hold today with the expectation that future costs are likely to change.
Offshore wind costs have been falling over the past 20 years. Its learning rate, the reduction in cost for a doubling in output, has been an impressive 5 to 19%.[5] The offshore wind market is also in relative infancy with only 8% of the EU’s 142 GW of total wind capacity.[6] There is clearly room for more as the deployment continues.
Biomass plants are reliant on Government support and are not viable at current LCOE. It is not however, the capital cost uncertainty but instead the input price uncertainty of the fuel feedstock for any plant. While cost reductions can be made in this area, the learning curve for reducing the costs of biological material is likely to be low. Further, storage costs today are extremely high, but also have high learning curve and likelihood of significant cost reduction in the coming decade.
4. Quantify today’s spending with its expected benefit in the future
The cost volatility framework can quantify the order of magnitude of spending today to see whether it is likely to be worthwhile investment in the future. It is easiest to understand using an example.
Hinkley Point C will cost the UK £92.50 per MWh for the life of the plant. This seems expensive relative to a hypothetical target cost of £70 per MWh, but this fails to account for the learning effects and the likelihood that future nuclear installations in the UK will be made at lower costs. We can put a rough estimate to the value of this learning by applying options pricing techniques.[7]
In the case of Hinkley, we can set the parameters of the Balck-Scholes pricing formula such that the current price (S) is £92.50/MWh, the strike price is £70/MWh, time for the cost reductions to take place of 10 years and a risk free rate (r) of 2% and a volatility of 15%.[8] These assumptions result in an option value of £2.80 per MWh, which when spread across the output of a further 10 GW of potential capacity is worth £10.3 billion of spending today to have the chance of lowering costs through future learning.
Estimates of the costs to the UK for Hinkley Point have increased from £14 billion in 2015 to £37 billion in 2016.[9] The cost uncertainty is caused by assumptions for the wholesale price against which the CfD top up is calculated. Calculating the cost to the UK against a cost of £70 per MWh rather than the much lower wholesale prices (of £40 to 50 per MWh) would reduce the number below £14 billion. Our £10.3 billion estimate for value of learning seems that the UK is roughly paying a fair price for Hinkley Point, but only just.
5. Recommendations
The UK is making major decisions on its electricity generation future and a consistent and objective framework will help significantly in this process. It should consist of:
- a specific cost target, for example generation costs of £70 per MWh, to provide a challenge to aim for and a benchmark to measure against;
- an assessment of each technology option with regards to the likelihood that competitive costs can be reached and by when;
- a portfolio approach to placing bets on technologies by allocating capital to the most promising options and recognising when to walk away from technologies which are no longer viable; and
- a quantification before major spending to assess whether the potential benefits of learning are likely to be matched by the upfront cost.
The Government should repeat this framework regularly. The UK transmission network has an annual Network Options Assessment (NOA). In a similar way, an annual Generation Options Assessment is a sensible way forward.
[1] Good Energy, Wind and solar reducing consumer bills An investigation into the Merit Order Effect, 2014
[2] Amber Rudd’s speech on a new direction for UK energy policy, 18 November 2015.
[3] Ofgem, Energy companies’ Consolidated Segmental Statements (CSS), 2009 to 2015.
[4] A cost per MWh is used because of industry familiarity. A more accurate cost assessment should adjustment for differences in load factors (3 times as many wind plants are required for the same conventional plant on a MW basis) and for total system costs (e.g. intermittency, lack of inertia).
[5] Rubin et al, A review of learning rates for electricity supply technologies, 2015,
[6] Wind Europe, 2015 European statistics.
[7] Basic financial option calculations using the Black-Scholes formula.
[8] The assumption with greatest uncertainty is volatility because of limited data to create an estimate. The option has negligible value below 10% volatility.
[9] The Guardian, Estimated cost of Hinkley Point C nuclear plant rises to £37bn, July 2016.
Energy Envelope 