The value of electricity storage to the UK power system
I’ve got around to looking at the value of storage and have prepared a simple model to assess this for the GB system, copy attached (in Excel 2007 containing a couple of macos).
I’ve taken some sample days of historic price profiles from the www.bmreports.com website – two for each month of the past year to represent business and non business days. The first thing to note is that for the UK system there are two quite different profiles for the system buy and sell prices, with the buy price often being a lot higher than the sell, especially at peak price times. This is important for storage, as the energy is purchased at the buy price and then sold at the sell price. This compounds the effect of the overall cycle efficiency.
The other data used is the size of the storage system (not important as it’s all scalable and assumed not to influence price), cycle efficiency (currently set to 70%) and storage capacity (currently set to 5 hours at full load). All of these are variables so can be changed.
The basic assumption is that the operator can determine from price forecasts when it is likely to be profitable to charge the storage system and when to discharge for the day ahead. The model then looks at each sample day and determines the optimum charge/discharge regime to gain the maximum income after allowing for the cycle efficiency. In general and as we might expect, the charging tends to be done during the early morning low price period and then discharged during the peak periods later in the day. The income is the difference in the income from selling at the peak and buying during the trough periods. The cycle efficiency is allowed for by assuming that the amount purchased is the discharge amount divided by the efficiency. The model produces a range of incomes for each day by varying the charge/discharge period and selects the one that gives the maximum income for the day
Running the calculation through for all 24 days and then multiplying up by the relevant number of month and day types in the year gives the net income for the year. For a cycle efficiency of 70% and a maximum storage capacity of 5 hours at full load, the annual income is £13/kW pa. Increasing the cycle efficiency to 80% increases the income to £19/kW pa, so as we might expect, the income is sensitive to cycle efficiency. Increasing the hours of storage has little effect – in fact increasing it to 24 hours has a negligble effect on the
income even at a higher efficiency. This is because there are limited periods when it is profitable to arbitrage the price and this is often less than the maximum storage capacity of 5 hours and rarely exceeds it.
We should note that the assessment is an approximation as it is based on a few sample days and assumes that they are reasonably representative. In addition, as the prices are historic, ii assumes perfect forecasting and optimisation, which in the real world is unlikely to happen. There may be some value in carrying forward some stored energy for more than a day eg from a weekend to a weekday. However, this again requires good forecasting and this time more than a day ahead. Overall, in practice I would expect the actual income from storage to be lower than estimated by the model.
So, there is some income to be had by storage, but not a lot – probably barely enough to cover the O&M. So, there is no economic case to be made for building new storage systems with the income based solely on arbitraging the UK market. Other income would be needed to make it a worthwhile investment. If we look at other markets where the buy and sell prices are much closer, then we could see an increase in the income. However, it would be very dependent on the profile of the price across the days. I have assessed storage in another country where there is a single buy/sell price, but the profile was less peaky – the result was pretty much the same as we have for the UK, with barely enough income to cover the costs of O&M.
There are cases for isolated communities (eg islands) where storage may be a good option. In these cases, if the capital investment of storage replaces some other capex required eg put in/replace some diesels or a transmission connection, then storage could become the most economic solution.
I hope that this helps with the debate on storage.
LAS Consulting Limited
Tele/fax +44 (0) 1225835419
Mobile +44 (0) 7909680897
E mail firstname.lastname@example.org