So I’ve used the Easter break to examine the performance of the GB electricity system over the past financial year using the data available from Grid Watch*. When I model electricity systems I try to ensure that the modelled system is secure, i.e. the lights won’t go out more often than the security standard dictates. This really comes down to examining three classes of parameters, annual energy, firm capacity and flexibility, so I thought it would be useful to examine the GB grid in the same light. I look at the first two here, the third requires more effort so will form part of a later blog.
This is the parameter that most commentators report, but on its own can give an over simplistic view of how well we’re doing, and how easy it is to completely decarbonise, so I won’t dwell on this. It does look good, more than half the energy came from low carbon sources, although of course there is some debate around whether all of these are really free from carbon emissions. Notable is just how little coal generation is now providing just 3.7% of the mix, down from 6.8% last year, although given its dramatic decline the past few years the last few percent seem to be difficult to get rid of – more about that later. Gas is now responsible for more than 80% of the remaining emissions and reducing this, rather than coal, is the natural target for near term policies. A crude calculation of carbon emissions (using emission intensities of 0.9 and 0.4 t/MWh for coal and gas) gives an emissions intensity for the grid of 0.2 t/MWh, which is actually very low for a system that does not have access to plentiful hydro resources.
So the question here is which plant types were available when we needed them, at times of peak demand? I’ve looked at this from two different perspectives. Firstly which plant were actually generating at the Triads – the three highest peaks of the winter that are separated by at least 10 days. The answer, as you might expect, gives quite a different answer to the annual energy charts. For the sake of transparency I need to point out that the First Triad of the year (around 17:00 on November 22nd) is missing from the otherwise comprehensive GridWatch dataset so the peak from the previous day (the highest demand of November in the data) was taken instead. The chart shows that coal, far from being on the way out, is the second most important technology class during peak demand. Predictably solar did not contribute anything, and although wind did contribute it was at quite a low level, in fact for two of the triads the wind load factor was less than 10%.
The interconnectors are worth discussing as they each behave differently. BritNed was the most helpful from a GB system point of view providing at least 730MW and its maximum capability of 1060MW for one of the triads. The French interconnector also ran at full capacity for two of the triads, but for one it was at zero and in the process of reversing which it did 10 minutes later. The interconnectors connecting GB to the Single Electricity Market (SEM) of Ireland were both exporting, almost to maximum extent over all three Triads, so these were effectively an extra GW of demand for the GB system to find.
A second way to assess the benefit provided by each technology in meeting peak demand is to subtract from demand the power provided at each point in time and see how much the remaining peak demand is reduced. This shows how much capacity was provided towards meeting peak demand, or in other words how much extra firm capacity would be needed if that technology was unavailable. The barchart illustrates the capacity credit of each technology calculated by this methodology. For example if nuclear did not exist then another 6.0 GW of firm capacity would be needed to meet the residual peak. Using this measure coal drops to third place, behind nuclear. Wind helped reduce demand by 1.3GW, so makes a small but not insignificant contribution.
The interconnectors are interesting, again, the BritNed interconnector is shown to be most helpful, but the others increase the need for firm capacity, i.e. the GB system would be better off (from a capacity perspective) without the French and Irish/NI interconnectors.
The final series of half-hourly scatter charts (also showing a 100 point moving average in black) looks at interconnector behaviour across the year. It can be seen that on the whole the Dutch and French Interconnectors mostly import, although at high demand, reversal is more likely. In fact exports to France can cause the net peak demand to increase. The two interconnectors to the SEM on Ireland show a strong negative correlation with demand, i.e. power generally goes the ‘wrong way’ to be helpful to GB. However this is probably because the SEM is suffering (benefiting from) similar weather and similar supply-demand imbalances which it is less able to deal with than the GB system.
It is to be hoped that this is taken into account when they are assigned capacity credit in modelling needed to determine capacity market volumes.
The GB system continues to make progress towards decarbonisation with more renewables, a further decline in coal and strong use of continental imports. However at peak times the firm capacity of gas, nuclear, coal and biomass (in that order) are key to keeping the lights on. Of the interconnectors only BritNed is clearly helping meet peaks, The French link delivers large energy inflows but has a tendency to disappear or reverse at times of highest demand. The Irish interconnectors are of little benefit to GB, they mostly import when their energy is not required and are likely to be exporting during high demand periods.