From microRenewables magazine, Volume 4, Issue 11, 2019
Published by CREST and Friedrich Ebert Stiftung
Electric utilities are afraid of rooftop expansion often make the claim that the increasing use of solar, which is supposedly an “intermittent” resource, will make the grid unstable. By unstable, they presumably mean that an “intermittent” source will make it harder to balance the supply of electricity with the demand. This way, utilities justify their very restrictive policies against rooftop connections.
Unfortunately, we have also heard this claim echoed by legislators and their staff, energy officials, academics, media, and others that influence policy. Thus, we know that utilities have been campaigning behind the scenes to spread their claim.
It is therefore important to show that this claim is false.
The intermittency of solar – that its output varies unpredictably due to cloud cover – only applies to individual installations, but not to the grid as a whole.
To understand this argument, let us look at other “intermittent” phenomena.
On the demand side, many appliances are also intermittent. You enter a room and turn on the lights. You get out and turn the lights off. It happens all the time. You boil water with your electric kettle, and in a few minutes, the kettle automatically shuts itself down. You wash clothes with an automatic washing machine and the motor for washing and the other one for spinning turn themselves on then off as necessary. The refrigerator is plugged in all the time, but its motor also operates intermittently, under the control of the ref’s thermostat, turning on when the temperature gets too warm, and off when it gets too cold. Similarly, the aircon also runs intermittently.
But all these intermittency issues smooth themselves out into a slow and predictable variation at the grid level. Statisticians call this the “law of large numbers”. What is unpredictable at the individual level becomes predictable at the aggregate level. An appliance turns on in one part of the grid, but another one may turn off somewhere else. Overall, the ons and offs average themselves out. Thus, electric utilities can reasonably predict the grid demand. They can draw what is called a “24-hour load curve” (load is another word for demand). This curve varies slightly from day-to-day and week-to-week, but it is statistically predictable and the changes are slow enough, giving electric utilities the time to adjust their supply in order to balance it with the demand.
Sellers in the wholesale electricity market also rely on it to prepare their hourly offers.
Another intermittent phenomenon involves ban deposits and withdrawals. Individual depositors certainly behave unpredictably, in an intermittent way. Banks never know in advance when a depositor will withdraw or deposit money. But over a whole day, taking the banking system as a whole, depositor behavior is predictable enough that banks can more or less predict how much will be withdrawn on a particular day, how much will be deposited, and how much cash to keep in each branch.
Let us take solar intermittency in particular. The sun is certainly predictable. Weather forecasters know the exact time when it will rise, when it reaches the zenith and when it will set, as well as its position throughout the day. This is the baseline for predicting solar output. Clouds are predictable, too. Weather forecasters regularly get satellite information on cloud cover over the country, or over each region of the country. When the clouds are thin and irregularly spread out, they will shade some solar panels but not others. Thus, taking the grid as a whole, solar output can be statistically predicted based on the position of the sun and the extent of cloud cover. Any departure from expectation will occur slowly enough, giving electric utilities the time to adjust their supply in order to balance it with the demand.
Thus, at the grid level, it is more appreciate to talk about solar’s predictable variability. Any unpredictable intermittency is only true for individual solar installations.
As long as there are other flexible power plants on the grid, i.e., plants whose output can be ramped up or down under operator control, solar variability should not be a reason to restrict its spread.
Some electric utilities, in an effort to make life difficult for solar, even propose that solar power plants should equip themselves with battery storage, to mitigate their intermittency.
Again, this is a very expensive as well as unnecessary measure. As we argued earlier, over the grid as a whole, things even themselves out statistically. A power source may reduce its output briefly, but at the same time on another part of the grid, another power source might increase its output – or some appliances may also turn off – rendering unnecessary any costly plant-level adjustment.
Thus, for the highest efficiency and lowest cost, any unpredictable variability is best addressed not at the individual plant level but at the grid level.
Comments