Tuesday, March 17, 2015

Wind variability

A common argument against renewables is that the wind doesn't blow and the sun doesn't shine all the time.  But in fact, as the chart below shows, if the wind turbines are sufficiently far enough apart, their output is not correlated.  Which means that the power from, say, two separated wind turbine farms is on average stable.  Note that this applies at short and long time periods.  So for example, at an average of 5 minutes' worth of output, wind turbines 100 kms apart are entirely uncorrelated.  At an average of 12 hours' output, turbines 100 kms apart are significantly correlated, but those 600 kms apart are only slightly correlated.

This means that a series of wind farms across a large geographical range can provide a stable source of power.  In effect they can provide baseload power.

The same would apply to solar, but less than with wind because obviously everywhere in a region is dark pretty much the same time at night, so the solar panels would need to be hundreds of kms apart (east/west) to provide power at different times.  But on a short term basis, there may be clouds over my panels but not over yours 10 kms away.  And there is also the situation that the wind and sunshine are not correlated: in fact, I have seen some analysis that says they are negatively correlated, because when the sun isn't shining (say during a thunderstorm), the wind is blowing.  So the output of wind and solar together is more stable than either by itself.

None of this will obviate the need for storage, but what it does mean is that the utility industry's fears that even 5% renewables input into the grid would lead to grid instability is wrong.  Many regions are now above 40% renewables with perfectly stable grids;Spain, Denmark, the state of South Australia.  Which means in turn that even before battery costs decline enough to make mass storage viable, we could still go to 40% renewables everywhere, allowing CO2 emissions to decline.

[Read the original article here]

Update:  a commenter in the Guardian suggested that to allow for the variability you would have to have too many wind turbine farms for it to be cost-effective.  Not so.  Let us suppose that wind farms produce 25% of nameplate capacity (it varies between 20% and 60%).  Now this is already included in the cost estimates: the cost of the wind farm is spread across the electricity generated over 20 years, which takes the capacity factor into account.  Note that capacity factors for wind turbines are variable day to day and week to week but very stable year to year.   So if we have two wind farms, both with a capacity factor of 25%, the average for both together is still 25%.  But the variability of the combined output is halved, if their output is non-correlated, which it would be if they're more than 500 kms apart.    Add a third, also 500 kms away from the others and variability falls again, but average output doesn't (assuming of course that the capacity factor is the same for all of them).  So in other words, wind power from enough wind farms far enough apart can provide base power.  And since wind is now the cheapest source of power, it seems demented not to start switching, given the risks of global warming.

No comments:

Post a Comment