Sunday, September 24, 2023


I was recently asked, as one of the major contributors, to comment on the recent Royal Society report on large scale energy storage. This had been a major exercise, impressively managed and directed by the lead author Chris Llewelyn Smith, involving the examination of how a UK system based on weather dependent renewables might measure up against the actual weather variations observed over 37 years of weather data. My contribution was confined to general observations on how power systems work, and in particular how operational and investment choices can or should be managed in a market economy.


This was the subject of a previous post, but readers are recommended to refer not to my earlier comments but to the report itself and to the policy briefing. Links are given at the foot of this post.


The results were interesting and to some extent surprising. The work indicates a very large scale of storage requirement, driven primarily not by seasonal but by inter-annual variations – runs of years when wind may be below average, and a host of other interesting findings and questions. I was asked just to comment on what I saw as the core economic and financial implications regarding large scale storage.


Economics in his context is clearly about securing the right combinations of generation and storage, the principles to guide our decisions, and the mechanics of getting where we want to be. The objective is to find market or other mechanisms for the outcomes we want, ie getting to low or zero carbon at an affordable cost compatible with an acceptable level of reliability and energy security. It follows that this has to be much more than just a theoretical optimisation, but has to cover national policies, institutions, coordination, markets, regulation and infrastructure?


There are several particularly important general lessons from the report that have general economic and policy implications:


1.     First is the potentially huge scale of storage. With both scale and major economies of scale, we have typical infrastructure characteristics, that need to be financed as cheaply as possible. 


2.     Second, interactions between storage and generation choices and multiple other factors: including the demand side. The report illustrates just how complex this is.


3.     Third, conversion capacity, for moving energy in and out of storage, will matter and has perhaps hitherto been largely overlooked. 


4.     Fourth is the whole issue of policy and planning for reliability of supply. Traditionally this was mostly about adequate margins of generation capacity required over peak demands – so-called needle peaks. But the new world demands a quite different understanding of reliability, when we are talking about, for example, wind drought. The issue then is of kWh energy rather than kW capacity – a major distinction.


So the report raises some very serious questions. We can treat each of the above in turn.


First, it is clear that the storage need has all the characteristics that we associate with large scale infrastructure. This includes possible or probable incidence of natural monopoly, certainly substantial investment costs, long lived assets that are highly use specific, and a financial necessity for a cost of capital as low as possible. For private capital that would mean a high level of reassurance over future revenue streams and the future market and regulatory environment.


Second is the issue of the very complex choices, and their coordination, in systems that rely on storage. It’s important to recognise that there are two distinct timescales here. One is operational - operating the system as efficiently and economically as possible with whatever is the current mix of assets. The second is about necessary investment -  creating the best mix of assets for the future. In a perfect market efficient solutions on both timescales might be expected to result from market prices.  But in the new low carbon world that looks increasingly like a pipe dream.


The conventional view of power sector markets was that the price signals  in a competitive market derived from the immediate needs for the efficient operation of mainly generation assets, replicating what might happen in a fully optimised system such as the merit order. It also had to provide an incentive for adequate capacity.  Various extra mechanisms have often been added that attempt to put a valuation on reliable supply; this is sometimes referred to as value of lost load or VOLL.  In principle it was hoped that all this collectively would  incentivise the right mix of assets, generation, networks and storage for efficient and affordable future systems. In practice the most that can be said is that experience has been mixed.


So what is new. Traditional spot markets were developed to deal with gas and coal powered generators, and to replicate a merit order based on SRMC. They were also largely designed by the employees of those generators They do not translate or adapt easily to low carbon technologies with more complex, probabilistic, intermittency and operating constraints. Storage adds new dimensions, by being intrinsically multi-period, requiring in addition that attention is paid to conversion capacities, and the very different nature of the reliability issue.


The simple metrics of short run cost that sit behind conventional market mechanisms do not capture the information or the complexity required. Investment choices, on the four-way balances between generation, transmission, storage, and conversion capacity, pose further questions, implying a need for coordination. Discussion at the Royal Society brought out some additional questions on the subject of conversion capacities, my third point.


My fourth point may well be the most important public policy question for the future – the security and reliability of electricity supply. We all know that governments cannot stand aside from issues of energy security, and electricity security in particular, however much they might wish to. However this is another dimension where the economic and policy calculus has to change radically, with some very different metrics.


Historically supply reliability in the UK has been about generating capacity – kW, and occasional insufficiency of kW to meet needle peaks. But future crises, if they relate to sustained weather related shortages, will be about kWh rather than kW.  Threats of months of energy rationing require an entirely different way of thinking about reliability. Possibly once in a generation events, like the 1970s 3-day week, a covid crisis or curtailed gas supplies, may mean looking at not just energy supply planning but also the overall energy resilience of the economy. 


Answering all these questions means great attention to the institutional and market structures of the sector. We have to decide who should own and operate large scale storage, on public or private ownership, integration with grid operation, guarantees for private capital, who should make the decisions on energy security and reliability, and so on.  


All these issues are closely inter-related, and the report offers an indication of where we might find the answers. These must rest on some combination of the following:


·      Novel market mechanisms and incentives to reward provision of storage capacity and conversion capacity.


·      elements of long-term contractual assurance for infrastructure providers eg a regulated asset base approach, or government guarantees.


·      Centrally driven coordination of investment plans. Quite common internationally (eg France’s EDF and Germany’s Energiewende).



·      Enhanced role for the National Grid 


·      The creation of a ‘central buyer’, to procure capacity, but also to buy power from generators and sell to retail suppliers and large consumers.


·      Close cooperation between energy companies who implicitly assume collective responsibility for reliability  (the US ‘power pool’ model) 


In summary the economics for me is about:


·      balancing the roles of markets, thus retaining a role for competition, and central coordination

·      financing storage as essential infrastructure, and 

·      re-evaluating the policy approach to planning for reliable future systems


Possibly the most important observation of all, though, is that all these things take time and the task is urgent. That means starting to address these issues now.



Large-scale electricity storage report


Large-scale electricity storage policy briefing

Thursday, September 21, 2023


The recent move by the UK Prime Minister to dilute policies aimed at reducing emissions is controversial. He has justified it, in essence, by reference to the cost to consumers, the cost of living crisis, and the cost to the UK economy. But this is short-termism of an extreme kind. Any consideration of our medium term future indicates the possibility of truly massive future financial bills arising from delaying our course of action. 

Sunac’s avowed commitment to a 2050 target is quite worthless. Even to take it at face value is to stretch credibility.  Postponing the end of petrol and diesel car sales will hinder progress to meet what is already an ambitious and challenging target. This is the view of much of an industry already heavily invested in the transition, as well as many bodies such as the Grantham Institute .


Moreover the volte-face on the means, promoting the decarbonisation of transport, must cast doubt on the seriousness of any commitment to the ends, of avoiding climate catastrophe, or of maintaining the pretence at international leadership on the issue. 


Kicking the can down the road can be continued as an action avoidance strategy almost indefinitely, at least until 2050 arrives, emphasising yet again one of the reasons why strategy should not be built around arbitrary emissions numbers for an arbitrary year. As argued repeatedly in this blog the true objective indicated by climate science is not a particular figure in a particular year but total cumulative emissions. This can be presented as a global carbon budget responsibility for which individual nations can negotiate or be allocated their own national responsibilities.


It is cumulative emissions that cause global warming. So let’s just imagine a rational world in which each country is required to keep its cumulative emissions, let us say from 1990, or even from now, within an agreed limit. Failure to meet that target would require removal of that excess CO2, or payment to someone else to perform that task through sequestration of CO2 from the atmosphere. There is a set of plausible but expensive technologies for this purpose, so we do have some idea of a plausible cost per tonne, and this could arguably be brought down to about £ 250 per tonne.


Under such a scenario of internationally agreed targets, not wholly implausible as the impact of climate change becomes increasingly an existential and dangerous threat, we can see what the true financial and economic cost to the UK of the Sunac volte face might be. The government has offered no estimates of the extra emissions resulting from its relaxation of policy, but others have made provisional estimates of hundreds of millions of tonnes additional emissions. 


Current UK CO2 emissions from the transport sector are about 112 million tonnes annually. So, for example, postponing by one year a programme scheduled to deliver a constant 5% (of base year) reduction in emissions generates an extra 100 million tonnes over 20 year. A two year delay results in an extra 200 million tonnes and so on.


On the basis of our sequestration costs even one hundred million tonnes extra would imply a cost of £ 25 billion, a significant number to factor into any cost benefit analysis of the policy change. But the implications of the Sunac delay seem likely to be much larger than that.