Thursday, December 6, 2018


Energy storage is now widely recognised as a necessary component in most options for achieving a low carbon future. For most of our history, energy storage has taken the form of physical stocks of fossil fuel, to be drawn on as and when required. Matching supply and demand in real time has therefore been comparatively simple. Even in the complex world of large power systems, generation can be turned up or down with comparative ease.

That world is changing. Renewable resources (mostly) provide energy according to their own timetable and the dictates of weather and season, most obviously so for the best developed resources of solar and wind power. Nuclear power output can be used to follow load, at a cost, but is still relatively inflexible. Matching these outputs to highly variable consumer demand, for a variety of energy services from heat to transport, as well as appliances and processes of all kinds, is going to be more and more challenging. This implies the fundamental importance of energy storage.

The key technical and economic requirements will in different applications include minimising weight or volume (car batteries), round trip energy efficiency (minimising losses in conversion processes to and from storage), sufficient scale (for large power system applications), low capital cost per kW (unit of power output), and low capital cost per kWh (unit of energy stored). The application determines what is most feasible and economic in each case, and hence also both the choice of existing technologies and the priorities in looking for new approaches.

It’s often widely assumed by commentators that the great advances in battery technology mean that we are well on the way to solving all these problems. However that is far from being the case. Lithium- ion batteries are probably reaching their inherent natural technical limits, and although it may be possible to force down production costs further, they still represent a very high capital cost. As a result cost, as well as any scalability or resource limitations, are likely to inhibit their use other than in premium and high value applications, even though these extend to some high volume uses such as electric vehicles (EVs) and a few specific power system applications.  Currently foreseen battery technology scores well on efficiency and reasonably well on cost per kW, but not on capital cost per kWh.

We need and are going to need a number of very different types of storage, and the requirements differ widely across the spectrum of energy services and other requirements. Here are a few of the key issues and questions.

In practice the really critical distinction is between situations amenable to storage solutions that operate on the basis of a daily cycle or similar, and those that have to meet annual or seasonal cycles.  For the former, high capital costs can be acceptable but conversion efficiencies will matter more. For the latter, the reverse is the case. Low capital costs are essential and conversion efficiency less important.

The Dinorwic pumped storage scheme in North Wales illustrates the issue of scale and cost very well.  Dinorwig, the main storage facility accessed by National Grid in the UK, stores the equivalent of about 9 GWh of energy.  Construction is estimated to have cost c £ 500 million and involved shifting 10 million tonnes of rock.  It provides a significant contribution to managing daily load fluctuations, but is still relatively small in relation to the overall fluctuations in the daily load curve. In practice Dinorwig is sometimes used for the provision of ancillary services such as frequency control, rather than in the more obvious role of storing energy against a daily peak demand. Operating on a daily cycle, the all-in cost of storage is of the order of only a few pence per kWh, allowing the facility to make a valuable contribution to the efficiency of the power system.

The amount of storage necessary to flatten the typical current January daily load curve is of the order of 80 GWh, or about 9 Dinorwigs, in principle still a credible level of investment. However to cope with even the UK’s current annual seasonal variations in electricity consumption, the storage need would rise to about 17000 GWh, or nearly 2000 Dinorwigs. On conservative estimates of the need if UK space heating loads were to be met through electricity (even using heat pumps rather than resistive heating) the seasonal storage requirement could be up to three times higher, or 6000 Dinorwigs.

Recovery of the very high capital costs of storage, through revenue or compensating benefits, on the basis of a once a year store/ draw down cycle, is clearly impossible. The silver bullet of cheap seasonal storage has to come through a technology with extremely low capital costs per unit of storage capacity measured in kWh. The prima facie front runners for high volume seasonal storage are chemical or heat based solutions, including hydrogen/ ammonia, synthetic fuels, and bulk heat or phase change methods.

The nature of heat and chemical storage solutions means that seasonal and indeed all storage choices for the power sector can only be properly evaluated by reference to the totality of the energy system. This is most evident for the two very substantial sectors of heating and transport, whose decarbonisation is an essential part of energy policy, since

·         widespread use of heat networks provides an easier means to tap into low cost seasonal heat storage than attempting to recover the stored energy as electricity to power heat pumps.

·         choices in the transport sector, eg between hydrogen and electric vehicles, have profound consequences for the management of electricity supply.

·         electric vehicles are themselves a potentially substantial source of relatively short term storage, as protection against any intermittency in renewables supply.

But this also emphasises the need for a coordinated approach that treats the energy system as an entity, aims to minimise costs and finds compatible paths forward across the very different elements of power, heat for buildings and transport. Current market structures fall a long way short of that requirement.  

No comments: