On Nov 18, 4:18 pm, tony sayer <
t...@bancom.co.uk> wrote:
> In article <
0gddp8-06c....@squidward.local.dionic.net>, Tim Watts
> <
tw+use...@dionic.net> scribeth thus
In normal operation, the grid will have a number of power stations on
base load (generally coal and nuclear) and a number of GT combined
cycle plants on part load, operating flexibly to manage the minute to
minute fluctuation in demand. In addition there will be a number of
plant (both big steam and CCGT) in "spinning reserve", meaning they
are running at idle, synched to the grid but not actually supplying
power. After that there are the "hot standby" plant, basically GTs
that are turned on but not synched, and steam plant with steam in the
boilers, ready to go, but not actually connected to the grid. On top
of that there will be open cycle GT peaker plant of various sizes.
The larger peaker plant (of the 100MW-200MW size) can be started from
cold in about 20 minutes or from warm in about 15. Smaller GTs can
start faster, but obviously have a lower output. Ditto diesel
generators. Then there is also wind and solar that supplies as much
as it can (and its fluctuation is balanced by the CCGTs). There is
also the pumped storage that will only be called in when there is a
large demand spike or loss of generating capacity as it doesn't last
very long, but can start in seconds.
Generally, taking the spare capacity in CCGTs and the spinning
reserve, there is enough generating capacity that can be called on if
the largest power station on the grid trips (Sizewell B), at which
point the hot standby becomes the new spinning reserve and other plant
is brought up to become the new hot standby.
In the day in question, Sizewell B tripped, so the above happened.
Unfortunately 5 minutes later another large station tripped (I believe
another nuclear plant). This left a shortfall that would likely be in
the region of 600 MW. When the power supplied to the grid is less
than that drawn out of it, the spinning bits lose energy, so the
frequency drops. Nominally it's at 50 HZ, but anything between 49.5
and 50.5 Hz is tolerable. When the frequency starts to drop the first
thing that happens is that every plant online will be activated in
something called "frequency response mode", which is basically a posh
way of saying "throw the kitchen sink at it". Basically every station
is ramped up as fast as it can be to as high an output as is possible
(including exceeding normal temperature limits and things like that),
in the hope that this will keep things going until new plant can come
online. No doubt whatever pumped storage was available would also be
thrown into the mix. A slight hiccup occurred as the frequency
dropped to about 47 Hz, namely that all the wind and solar generation
triped because it was "under speed". At this point, action is taken
to allow load shedding to begin (though in the event it didn't
happen). This is where the events of this particular occasion were
turned around (new capacity came online and restored balance). Were
that not to have taken place, then load shedding would begin.
Hopefully that could be done in a suitable controlled manner to match
the demand to the available capacity, so that order could be
restored. Basically sacrifice supply to some parts so that the whole
can be saved.
Robin