See the four graphs below - which are based on gridwatch data (data every 5 minutes) for the 12 month period starting on 1st November 2020.
The first graph shows the mean amount of wind generation (mean of all 288 5-min data for the day) each day.
The second graph is the same as the first, but smoothed (by means of a 7-day moving average) to get rid of some of the day-to-day variation and hence make it a little easier to 'read'. Note, however, that 'smoothing out the day-to-day variation' is potentially misleading because it will tend, for example, to 'conceal' short (e.g. a day or two) periods of little (or no!) wind generation. However, it can be seen from the first graph (and the two which follow) that there were no days on which there was no wind generation, very few on which the average was <1 GW and not many on which it was <2 GW (about 7% of average total daily demand).
Even the first graph is hiding detail. It would be interesting to see a distribution of "rate of change of output", or perhaps "change of X GW in less than Y hours" with un-smoothed data - when I've looked at the short term graphs (before BMReports did it's facelift and made things a lot harder to see), there were some massive swings in output which would be significantly reduced once averaging out data to fit days or weeks onto a graph.
The third and fourth graphs show the distribution of the 'average wind-generated GW during a day' during the 12-month period considered - i.e. the percentage of days on which the mean wind generation was a certain figure (third graph) or above a certain figure (fourth graph).
...
... and now the graphs ...
View attachment 248984
The first things I note on that are :
The number of massive power changes over short times - there are several times when output changes between something like 100% of maximum "real" output, and 10%, over only a day or two. Just take a moment to think about that, that's a change from 10% to 100%, or from 100% to 10%, over only a day or two. So "something else" has to ramp down or up to cover 90% of your capacity, or in the order of 10GW, over a short time scale. And that's with the data averaged so in detail there will probably be even more significant swings.
I also note that there are a lot of days with low output. Yes, it's easy to say there were no days with zero output, but from a users' PoV, if the industry told users "you can only use 5% of your normal load today" then that wouldn't be very useful
I can see quite a few days where output, in terms of mean or max, is "quite low" - low enough that in the absence of "something else" to fill the gap would mean lights going out across the UK.
And again, I can pick out some interesting stats from these :
Some 10% of the time, UK windmills produced less than 1GW. And about 15% of the time, they produced over 10GW. That's a massive variation in output which "something else" has to cover. And we need to consider what else is going on. In summer it represent a much greater change (as a % of supply/demand) than in winter. I wonder what a graph of "wind output as a % of total demand" would look like ?
In view of what Simon has recently written (below) the next thing I plan to do when I have a little time is look at the durations of periods of very low Wind generation (probably over periods longer than 12 months - there appears to be about 10 years' data available from gridwatch). I may be proved wrong, but I suspect that there will not have been periods as long as "a week or two" with little/no UK wind generation - although, as I said at the start, we can't put too high a proportion of our eggs in the Wind basket, since there will never be any guarantees.
December 2010 would be a good place to start, if the data goes that far back.
And looking at teh first graph, early Dec 20, and late Jan 21 appear to show periods of "several days" with significantly reduced output - and I'm guessing significantly increased demand due to heating loads.
Found this interesting when listening to some middle of the night radio,
http://isleofeigg.org/eigg-electric/. Capped at 5kW for domestic and 10kW for business usage at any one time.
Interesting, and a few things to note about it which would be useful to bear in mind when discussing the UK wide grid and some of the "more interesting" claims made by some parties.
- Firstly, they have about 4 times the hydro than wind. Hydro is much more predictable, and has significantly less variation over shortish timescales - other than massive downpours which could swell a dryish stream to full flow in a short space of time, any well engineered system out to have variation measured in % of output per day rather than per minute. As such, they are starting with a much more predictable and reliable base capacity.
- They also have battery capacity capable of supplying full load for a day - though it's not really clear how much of the total the batteries can provide.
- And they have 100% diesel backup - which they accept that even with an excess fo renewables provides about 5% of supply over a year.
Scale that up and we'd be needing between 50 and 100GW of additional hydro. Not going to happen - I suspect that even if we could get around the objections to flooding some of our nicest countryside, the capacity just isn't there.
And we'd be needing in the order of 1/4 to 1/2 a terawatt hour of storage. Bear in mind that the UK's largest pumped storage scheme is
Dinorwig, which is rated for 1.8GW and a total of 9.1GWH - it's not easy working out the numbers for some of the other 3, but it looks like they have a total capacity in the order of 12GWHr. So less than one hours worth of UK usage (even if you used the total capacity), and then they need "recharging".
Also of note is that a spinning turbine such as the hydro unit provides significantly better inertia than inverter output wind turbines. That's already an issue on the UK grid - remember the power cut in the south east a little while back ? Again, not insurmountable if you throw enough of our money at it.
And of course, if domestic loads are restricted to just 5kW, where would that leave EV owners with their 7kW chargers, or homeowners with the push for heat pumps which for a larger home will probably need that just for the heating at times. I realise a limited supply isn't novel - in the past I've dealt with colleagues in Italy where sometimes they'd lose office power for a couple of minutes while they went to reset the breaker if running the aircon had tripped it.
