Should you consider a solar-constrained grid connection?

One of the biggest barriers facing battery energy storage development in Great Britain today is grid connections (or lack thereof). Some DNOs are offering grid connections with the stipulation that standalone BESS assets cannot export at certain times. Specifically, in regions where there is lot of distribution connected PV, and at times of the year when that PV produces a lot of power - ie when solar generation is at its peak. In this piece we assume an export limit is placed on such grid connections during the 40 days of peak generation per year, between 7am and 7pm.

Figure 1 (below) shows the generation profile of those 40 days for every other year, between 2013 and 2021. The graph also highlights the increased capacity of solar in GB - we can see how peak generation rose from under 2 GW in 2013 to over 8 GW in 2019.

In this piece, we look at how this particular type of grid connection could impact BESS revenues. So, should you consider a solar-constrained grid connection for your battery? If you’re struggling to get an unconstrained grid connection, yes. Let’s find out why.

Arbitrage opportunity

Figure 2 (below) shows the average daily wholesale spread, by month, for the period 2017 to 2021, against the frequency of peak solar generation. As you can see, when solar generation is at its highest, wholesale spreads are lower - meaning less arbitrage opportunity for batteries hoping to export at these times.

• 84% of peak solar generation days fell between April and July.
• These were also the four months in which the average daily wholesale spreads were at their lowest.
• As mentioned above, this means that there tends to be less arbitrage opportunity during these months. We explore this further in the model below.

Figure 3 (below) shows a modelled (exclusively) arbitrage strategy. Our battery is a one-hour system, cycling once per day, with an assumed 85% import efficiency. In this model, the asset does not do any wholesale trading on those (40) days of peak solar generation. In this case, our battery - the one with the constraint - makes, at most, 8.95% less than an identical, unconstrained battery.

Frequency response markets

Predominantly merchant trading strategies may become commonplace in the not-too-distant future. However, at the moment, the primary revenue stream for battery energy storage systems in Great Britain is frequency response services. Being unable to export during certain hours of the day would mean being unable to provide low-frequency response services - those in which batteries export power.

Figure 4 (below) shows estimated annual frequency response revenues - for both our constrained battery and its unconstrained twin. (We calculated these revenues using the monthly volume-weighted reference prices, from 2021, for both low-frequency Dynamic Containment and Firm Frequency Response.) In FFR, our constrained battery withdraws from providing the service for EFA blocks 3-5, for each month in which a day of peak solar generation falls (using the 2021 data). In DC, our asset withdraws from the service for EFA blocks 3-5 on the specific days of peak solar generation. Both scenarios assume acceptance into those markets - when it occurs - at full capacity.

• Our constrained battery sees its FFR revenues reduced by 24%, and its DC revenues by 6%.
• This disparity is due to contract length. In FFR, assets have to commit to providing the service for their accepted EFA block(s) for every day of the month. Therefore, our constrained battery has to withdraw for the whole month. DC, on the other hand, is procured day-ahead (again, by EFA block). This means that our battery can still provide the service on those days which don’t contain peak solar generation.

This seems significant now, but things are changing

The retirement of the FFR service is coming - expected towards the end of 2022. This means that, next summer, assets with this particular type of constraint won’t have to worry about dropping out of a frequency response service for a whole month. (The entire Dynamic suite of services is procured on a day-ahead basis.) On top of this, assets with this constraint are only barred from providing low-frequency (exporting) services. The launch of the high-frequency (importing) services provides an opportunity for constrained assets to enter these markets instead, as they are still able import during periods of constraint.

Charging opportunity

As well as opportunities in high-frequency response markets, constrained assets may also opt to charge during those hours of peak solar generation when their export is constrained. In the last five years, there have been nine instances of negative imbalance pricing during those periods in which assets would’ve been constrained (7am-7pm, on the 40 days of peak solar generation). Negative pricing means that assets are paid to charge. The average imbalance price during these periods was £16.59/MWh. Figure 5 (below) shows imbalance prices during those periods.

Should you consider it?

Lucrative arbitrage opportunities tend to occur during winter months, when solar generation is at its lowest. This means that there’s a lower impact on merchant revenues than in summer. In addition, frequency response procurement targets are often higher in summer months as there is less inertia on the system. With frequency response services being procured day-ahead (by EFA block), assets with access to a good solar forecast should theoretically be able to operate freely around those periods of constraint.

So, while this type of export constraint would have some impact on BESS revenues, that impact could be minimised through decent forecasting and clever optimisation. Therefore, for sites that are struggling to obtain an unconstrained grid connection, we believe it is worth considering this type of agreement.