Flexibility and Productivity of the UK Electricity System

The UK Government is pushing the electricity system towards broadscale adoption of a selected range of expensive and uncertainly net-beneficial flexibility measures, including electricity storage, demand side response, and other so-called “smart” technologies, into which some £20 billion would have to be invested in the shorter term. Examination of the consultants’ report behind this policy reveals that in spite of this large commitment, and the economic rents government must create to incentivise it, Combined Cycle Gas Turbines will continue to provide the backbone of system flexibility and reliability. This is very dubious value for money and will probably erode productivity.Slowly and not always surely, over the last millennia or so humans have become rich for reasons they didn’t understand, but we can now see was an increase in the productivity of the energy sector. During this period the energy sector ceased to be predominantly organic, in other words agricultural, and instead became mineral, in other words fuelled by coal. This new energy system produced large volumes of energy, and also required relatively little of its own output for maintenance, leaving a large margin for the creation of other economic activities. The coal industry, rich though its owners were, never dominated the European economy and society, particularly that of the United Kingdom, in the way that the landed interest had done in the pre-coal era, when land was not only, as the Physiocrats believed, the principal source of wealth, but was in fact the main kind of wealth.We are now in danger of slowly but quite surely becoming poorer because of careless and state mandated reductions in energy sector productivity. For example, on the 24 of July the UK government made a major announcement concerning revisions to the electricity market to increase flexibility in generation and demand (for the press statement see “Plan launched to bring smart energy technology into homes and businesses”, and for the main paper see Upgrading our energy system: smart systems and flexibility plan).Government presented this initiative as “Putting Consumers in Control”, with the implication that this will be of great benefit to domestic consumers. However, careful examination of the “Summary of Actions” in the main document shows that the policy drive is in reality focused on large commercial entities. It is about enabling legislation for storage investors, about Demand Side Response for large non-domestic consumers, and allowing energy aggregation companies to access the Balancing Mechanism. It is not that actions for domestic consumers are absent, far from it, but rather that they are nowhere near as prominent in the Actions as they are in the poetry of the discursive statement that precedes them.This interpretation is confirmed by the details of the consultant’s analysis on which government is relying in this announcement, which is a report for the UK government, An analysis of electricity system flexibility for Great Britain (November 2016) by the Carbon Trust, an organisation that was created with public funds but is now a semi-commercial entity providing consultancy services. The study was written by Carbon Trust staff, but relies on modelling analysis by Goran Strbac, Marko Aunedi and Danny Pudjianto from Imperial College London, whose work in this field is well known, and which has been referred to elsewhere in posts on this blog. Anyone interested in the details, the context and the background of the policy announcement must look beyond the government statements and at the Carbon Trust/Imperial study.For example, government announced in its press release that the flexibility policies had the potential of saving consumers large sums:

[…] new smart technologies like smart meters – and appliances you can control from your mobile phone – along with other improvements to manage the energy system will help the country save up to £40 billion on energy costs over decades to come.

The government’s own main document expands on this to the effect that the potential saving is actually between £17bn and £40bn by 2050, in other words a saving on average of between £500m and something over £1bn a year, which is certainly significant, but not quite as dramatic as the press release makes it sound.Furthermore, reference to the Carbon Trust/Imperial study reveals that this saving is calculated by comparison to a system that proceeds with the current carbon reduction targets, and renewables targets, but does not spend on new system flexibility. Since those carbon reduction policies are costing significantly upwards of £10 billion a year at present in subsidies and system costs, the new flexibility policies will deliver only a modest saving on a large cost that could be almost completely avoided with different policies on carbon reduction and renewables.And of course the flexibility policies are themselves very expensive, implying major capital expenditure, almost entirely in electricity storage and sophisticated Demand Side Response, in the short run in the hope of securing net benefit in the much longer run. The scale of that short run expenditure is nowhere to be found in the government documents, but the Carbon Trust/Imperial study provides very informative estimates, as can be seen in the following figure from the study:Figure 1: Cost differences between the twelve core scenarios and no flexibility scenarios (N1, N2) cumulative to 2050,  Figure 7, in Carbon Trust, An analysis of electricity system flexibility for Great Britain (November 2016), 26.The bars below the axis represent costs, dark blue for capital expenditure on storage and pink for expenditure on Demand Side Response. The bars above the axis represent hoped for savings resulting from those storage and DSR investments. The white diamond in the bars represents the net saving in each case. The scenario assumptions are, needless to say, complicated, assuming differing levels of electricity demand, interconnector construction, and flexibility cost, and I do not intend to discuss them here. The main point to take from the chart is that the capital cost of the flexibility measures ranges from a significant £2 billion to a colossal £20 billion, but that the authors expect this cost to be more than returned by savings. They write:

[…] the costs of deploying storage and DSR (£2-20 bn) are significantly outweighed by the savings from avoided generation CAPEX (£14-19 bn), OPEX (£13-15 bn), distribution CAPEX (£4-13 bn) and transmission CAPEX (£0.04-1.5 bn). These savings represent the avoided investment necessary to meet carbon reduction targets whilst providing security of supply at the most affordable cost, and are explained in detail in the following sections.

A significant driver of these cost savings is that deploying flexibility technologies can result in needing 6-9 GW less low carbon generation to meet carbon targets in 2050, and 3-29 GW less peaking plants to meet peak demand and system stability requirements (p. 27)Two observations must be made: Firstly, these savings are against costs that could be avoided almost completely if policy were not constrained by carbon reduction targets. This is very far from being the cheapest system imaginable, indeed it is almost certainly not even the cheapest clean system available. While it may be true that storage and DSR can trim those costs a little, much greater savings could be made by a thoroughgoing policy redesign.Secondly, since the cost savings are spread out over an extended period of time, during which circumstances and technological progress could dramatically alter the landscape, they must be regarded as uncertain in reality. That compares with the storage and Demand Side Reserve costs, which will be made in the shorter run and are by comparison certain. Indeed, it is quite possible that the consumer might in the short run be forced by policy to spend up to £20 bn on these technologies, but see nothing like the hoped for benefits in the longer run. That would not be particularly surprising. – 2050 is a very long way off. Anything could happen.In any case, the measures proposed offset but do not fully resolve the underlying problem, which is low productivity of the asset base. The Carbon Trust/Imperial paper is candid about this, and remarks that:

The greater variability in electricity supply, due to increased renewable generation, and exacerbated peaks in electrical demand, due to the electrification of heat and transport, will result in reduced utilisation of generation, transmission and distribution assets. This reduction in utilisation increases the capacity required to achieve the same performance, leading to higher system costs. Technology options such as demand side response, energy storage, flexible power generation and electricity interconnectors, can mitigate the negative impacts on asset utilisation. (p. 4)

But it is only a mitigation, a sticking plaster on a gaping wound. The scale of the problem is clearly illustrated in a further chart from the Carbon Trust/Imperial study, a figure that describes the installed capacity and output of Combined Cycle Gas Turbines in one of the twelve scenarios considered, scenario S2 in which DSR and energy storage are both available at low cost.Figure 2: Deployment (GW) and output (TWh/year) of conventional CCGTs to 2050 for S2 Figure 17 in Carbon Trust, An analysis of electricity system flexibility for Great Britain (November 2016), 40.The installed capacity of CCGTs is recorded on the left hand axis, and the energy generated by that plant on the right hand axis. In 2025, the authors project that in this scenario there will be about 30 GW of CCGT generating about 140 TWh of energy, which implies a load factor of about 53%. That is a respectable degree of utilisation though far below the level deliverable from such machines. But this falls rapidly over these decades, and by 2050 the remaining 16 GW of CCGTs, has a utilisation of only 6%. In other words a large fleet of CCGTs is retained to guarantee security of supply, but is hardly used at all. Its capital, operational, and maintenance costs must all be recovered from a very small volume of energy sales, implying high prices, and a very generous capacity contract, a payment to exist.Indeed, it is a key message of the Carbon Trust/Imperial study, but strikingly underplayed, indeed invisible, in government’s headline statements, that even with cheap DSR and cheap battery storage, CCGTs nonetheless remain essential to system flexibility and security.Given this fact, it is obvious that it would be more prudent to provide the flexibility of the future system with a more highly utilized CCGT fleet, without the hazardous and complicating addition of expensive and uncertainly beneficial storage and DSR. Government may reply that they are hoping to stimulate invention and innovation in these two areas. But on the scales contemplated, up to £20bn of capital investment, they seem likely to create only yet another zoo of rent-seeking and productivity eroding government pets.