Roger Chapman wrote:
On 25/11/2011 11:23, The Natural Philosopher wrote:

snip

This leads to a basic principle that can be expressed thus:

"Costs are proportional to the set of the worst cases. Earnings are
proportional to the average case."

This has a huge impact in terms of renewable generation from
intermittent sources. Basically the engineering has to be able to absorb
the peak generator output (or the source must be throttled back and the
energy discarded) BUT the income is calculated ion the average power
output - in the case of wind that's a 4:1 difference. Or more. That is
on average the size of the generator and link wires is 4 times greater
than on average, it needs to be.

Irrelevant for domestic PV installations. 4KW is well below the load
that even the meanest domestic connection will supply so no additional
infrastructure is required.

It gets worse. The backup likewise has to be as good as the worst case
wind output. One of the things that has become established over the last
year or two is that national wind output can, on occasion, drop to as
near zero as makes no difference.

Solar PV is even worse. It's guaranteed to do this every night.

Solar PV doesn't really need any hot reserve, let alone spinning
reserve. With a multitude of very small generators over an extended area
the actual output could be predicted to very close limits merely on the
basis of cloud cover and date.


it doesn't matter.

You STILL have to ramp up gas to cope at sunset, and that gas needs to
be warmed up on the way, and while it ramps up, its wasting gas, and
when it ramps down, its losing stored heat.

It makes **** all difference if you KNOW what variation you have in
advance - you still have to DEAL with it.


Solar PV may do little to help with the 5.30 pm peak but almost all of
its output (all for most of the year) will fall within the high demand
zone during the day unlike the output of windmills which is much less
predicable and close to random in distribution through 24 hours.


Wrong again, the forecasts are accurate to 70% typically.

It is not the unpredictability of intermittency that is the problem its
the variation, and, in particular, the slew rate.



snip