On 23/08/2019 08:05, The Natural Philosopher wrote:
On 22/08/2019 22:29, tony sayer wrote:
It does seem that this DC to AC inverter software circuitry needs some
further though it's not that stable seemingly?

No. And there is a fundamenal point you are missing

It CANNOT be solved just with software.

A temeporaty rid overlaod must be covered either by a loss of load -
load shedding or load reduction - or by an incease of power.

It could be solved by not putting 100% of everything the wind farm
generated out onto the grid or some local battery storage colocated. The
latter would make some sense since then the large scale DC to AC grid
tie converters could be used for two purposes.

A grid overload can also be handled by allowing the voltage and
frequency to drop. Enough load is resistive that it should balance. It
don't see any reason why an AC to DC converter can't track the phase
right down to the point where conventional spinning kit begins to drop
off. Injecting some power is better than injecting none at all.

In general load reductiuon and slowering freqeuncy go together in
syncrhonrous motors, and lowering volatges will in general lower to load
with resistive lods. Not so with switched mode power supplies.

So what happens with spinnging nennies is that the rotational inertia os
in fact kinetic energu and that is what represents the increase in power
that will be fed into te grid. As the rotors slow, energy is lost, and
that becomesÂ* what keeps the grid up. Until somethinmg else can. Like
openeing the throttle on a coalÂ* or nuclear plant

Because steam boilers also represent energy storage too.

The point is this: althoiugh software that reads mains frequency and if
it slows advances the phase of the inverters to feed more power in, are
possible, there IS NO MORE POWER to be had!

Unless you strap lithium bastteries to the DC output of every wind
turbine. Or supercapaciors. Or use a big ****-off rotary inverter. (DC
from wind turbines drives DC motor attached to AC generators attached to
the mains)

More expense. Less profit.

But enhanced resilience against grid fluctuations. It may be a necessary
thing to add to the larger wind farms so that they can output 110% for a
brief 5 minute period when called upon to do so. Even a minute of extra
reserve might well have been enough to prevent cascade failure.

I have my suspicions something wasn't configured right at Hornsea though
and it tripped because the settings were essentially paranoid. It was a
very long way away from the lightning strike and East Anglia was much
nearer and stayed up as did the toy prototype version of Hornsea.

Come to think of it how does the inverter right out in the North see or
know what the grids running at, or does it just ramp up frequency and
see what currents flowing back to blighty?..

Anyone know?..

Phase is the issue.

Before connection the mains frequency and phase is read and the inverter
will connect at slightly leading phase to it. So it draws power. Then
the power in is monitored and the phase adjusted so that full power is
delivered.

So it ramps up phase, not frequency.

It is a large scale phase locked loop.

The whole grid is phaselocked. If you want to connect a generator to it
you MUST lock phase and advance it a tad until you are feeding it. At
zero phase you are not delivering or taking pwer.

The problem is domestic inverters are set to trip out if the frequency
drops too far as well as having no storage.

They are not serious generators. They are bolt on additions

But I doubt they were a really significant factor in the overall cascade
failure. Most domestic solar PV is rigged to make hot water and pocket
the export FIT for 50% they never export!

--
Regards,
Martin Brown