More renewables, less energy storage

There is no doubt that to get to 100% renewables we will need storage, either from hydro, or pumped hydro, or batteries (or all three).  But how much we'll need depends on other factors.  This study of the Texas grid suggest that having both wind and solar produces a more stable and less variable output, because the wind and the sun don't stop at the same time.  In addition, the output of wind turbines in one location (west Texas) is also complementary to the output of wind turbines in another (south Texas).  Or at least, so the statistical analysis says.   You can read the full study here, but I'll show you a couple of charts so you can see how well it works.

The first chart shows average capacity factors over the year in blue, capacity factors at the twenty hours of peak demand in summer in red (PACP=peak average capacity factor), and the 20 hours of peak demand in winter in green.  Solar farms on on the left, wind on the right.  Solar obviously helps a lot with peak demand in summer, because it tends to coincide with peak daily temperatures on hot days when a lot of air conditioners are being run.  The high PACP for wind in winter is interesting, suggesting that the peak demand for heating in the coldest hours corresponds with stronger winds. Note also that PACPs are much higher for those periods when demand is highest for both wind and solar, which is extremely interesting.  Moral of the story: wind and solar are seasonally complementary and actually well fitted to help satisfy peak demand in winter and summer.

The chart below shows average capacity factors for west Texas wind (orange), South Texas wind (green) and solar (blue) over 5 years on a daily basis for the 24 hours of the summer and winter solstices.  Note how conveniently wind complements solar, strong overnight when the sun isn't shining, and how in summer south Texas wind complements west Texas wind, conveniently peaking in the afternoon and evening when grid demand is at its highest.  Once again, the moral is clear: wind and solar are complementary, on a daily and on a seasonal basis.  And widely separated wind farms are also complementary, though to a lesser extent.

There is also a table (Table 3: which is too big to show here) which shows the correlation between different pairs of wind/solar farms. As you'd expect from what I've said, the correlation coefficients between wind and solar pairs are negative, while the correlation between different solar farms is strongly positive.  This means that a blend of wind and solar would provide a more stable output with lower variability than either on its own (that conclusion derives from the maths of averages and standard deviations, but I shan't explain it here) .  But the interesting correlations are between wind farms in east vs south Texas.  The correlations are positive, but lower than the correlations between different pairs of solar farms.  Combining two wind farms in these different locations would still reduce variability and increase stability but not as much as combining a wind farm and a solar farm.  Most interesting.

The last chart is fascinating.

This chart shows (on the bottom axis) the percentage of time that supply (in MW, not %) is guaranteed.  So at 20% of the time (2 on the lower axis) 50% of combined capacity is available (30 out of 60 MW; 50% because no solar at night).  For half the time (0.5 the bottom axis) it's still almost 50% of capacity is guaranteed.  However at 87.5% ( the theoretical capacity of a coal power stations after allowing for  maintenance and repair downtime) only 13.5% of total capacity is guaranteed.

The interesting point, though, is this: with aging coal power stations, guaranteed capacity drops, because the machinery has aged.  In Australia, it is the coal power stations which "trip" on hot days.  In other words, if we used a realistic figure for coal, we would move that black line to the left, to something like 60% guaranteed.  And at that point, nearly half the capacity of the combined output of a wind farm and a solar farm is guaranteed.  Given the nighttime zero output from solar, that is remarkable.  In places like the US, Europe and Australia, where coal power stations are old, the guaranteed capacity of wind+ solar is as good as the guaranteed capacity of coal.

Moreover, for 70% of the time, guaranteed capacity of wind+solar is 1/3rd of potential capacity.  In other words, if we tripled the initial capacity, i.e., had significant overcapacity, wind+solar would be able to provide for 100%--guaranteed-- of total demand.  That would be wasteful; batteries or other storage will surely be cheaper.  In fact the 30%, or 7.2 hours, of storage would also bring the guaranteed capacity up to 100% for 78.5% of the time, matching new coal.

To sum up:

1. Wind and solar are complementary, with a negative correlation
2. Wind in widely separated locations is complementary, though less so than wind and solar
3. Solar is particularly useful on the days of highest demand in summer; wind in winter
4. Where coal power stations are old, wind plus solar are as reliable as coal
5. Overcapacity would guarantee 100% supply
6. So would 7 hours of storage.

The grids of the future will surely use all 5 means of guaranteeing supply: some overcapacity, mixed solar plus wind, wind farms separated by 500 or 1000 kms or more, and several hours of storage.