SUMMARY and CONCLUSIONS
Daily peak demands on GB non-wind generators for the weathers of the past nine winters, with 2016/17 wind generators, are found to have the following properties:
- The reductions due to wind power of highest peak demands on non-wind generators are clustered around 1GW, the nine values fall between 0.6 and 1.6GW
- The day (day_A) of highest peak demand on non-wind generation is always a day of very low wind power, and is sometimes different from the day (day_B) of highest daily peak consumption
- The reduction due to wind power of highest peak demand on non-wind generators is the sum of two parts, firstly the difference between the peak consumptions on day_B and day_A (if they differ), secondly the wind power on day_A
Thus, for the weather of the past nine winters, the highest peak demands on non-wind generators (around 55GW) are reduced by wind power by only around 2%, a figure that barely responds to increasing wind capacity, because day_A always has very low wind power, as a result of the effect of 15GW of wind power on windy day demands. Increasing wind power capacity from 2016/17 levels is not the way to increase the capacity margin, or to compensate for its reduction due to closure of conventional plant.
DATA SOURCES and PROCESSING
Demand data files were downloaded from a National Grid (NG) website (see the NATIONAL GRID DATA page of this blog), the associated data for “embedded wind” was added to “demand” to get as close as possible to “consumption”, and then “demodulated” to remove (approximately) the demand reduction at weekends and holidays. The demodulation was done by temperature-weighted interpolation using the mean of the daily Central England Temperature (HadCET).
For winters 2011/12 and later, data for metered wind power was downloaded from the gridwatch website ( http://www.gridwatch.templar.co.uk/ ), and scaled-up to levels observed in 2016. For winters before 2011/12 metered wind power was assumed to be a multiple of embedded wind power, and also scaled-up to 2016/17 levels.
Embedded wind data is available in the NG data files for all nine winters, and was scaled-up to 2016/17 levels using ratios of capacities, with a bit of manual adjustment to get consistent levels for all nine winters.
Total wind power (metered plus embedded) was subtracted from consumption to get the demands on non-wind generation.
The following set of figures shows, for each winter, the daily peak demodulated consumption in black, and the corresponding demands on non-wind generation in red. Horizontal lines mark the maxima of the two sets of data. NOTE: There has been no investigation of possible higher demands on non-wind generation at times other than those of highest consumption.
Conclusions: See the start of the post.