Comment by londons_explore | Hacker News Reader

Comment by londons_explore | Hacker News Reader

londons_explore

Nearly all of the purposes for these flywheels could equally be met with existing infrastructure with software only changes.

Solar inverters can produce/absorb reactive power easily - but while that can frequently be configured from the control panel of individual units, rarely is that ability exposed to grid operators.

Inverters at DC undersea links can do the same. As can some kinds of wind turbine.

All three can also absorb harmonic power and emit it as fundamental frequency power.

Battery energy storage systems can also do all of that, plus also provide 'simulated spinning reserve', and also provide brief huge increases in energy output to deal with various collapse conditions.

Wind can also do this (with caveats).

However, despite lots of 'green' energy sources having the necessary abilities, it frequently is lacking software support, particularly in the 'control plane'. Part of that is because there is no legal mandate to have these abilities, and there is (in most markets) no payment for providing this service. Therefore operators don't bother to enable these features.

In the case where a system splits and one area is 6.3 GW in deficit how is an inverter going to conjure up that power?

The only thing that can help in that case is batteries.

98% of the power plants I’ve been to run the voltage regulator in voltage control mode with 5% droop. Solar and wind farms can do that too, and I am sure it is required by the system operators to allow them to connect.

If grid operators want more or less reactive power they call us and tell us to change our voltage setpoint on our plants. Can’t they do the same with solar and wind? The grid operator does not need real time control of the setpoint since the control loop responds in real time.

This isn't to deal with multi-minute power deficits in split grids. That is dealt with with fast acting generation or load shedding.

Solar and wind systems today typically do not have a voltage setpoint or control loop. That's because there is no big gas/steam valve to adjust the amount of power generated. The operator wants to inject all the power from the sun/wind into the grid, whatever the voltage.

An X% droop would be X% of the total generated energy wasted in the name of stability. Thats a lot of money, and they won't waste it unless required to.

Instead, power deficits in the 10's of millisecond range can be dealt with by inverters (which typically have storage capacitors for ~20ms worth of output). This covers large fault currents, and sudden inflows as bits of grid are reconnected.

Unfortunately, most of today's inverters, which have the capability to do that, will not, due to having code which will only operate the 'happy case'.

> Solar inverters can produce/absorb reactive power easily

> Solar and wind systems today typically do not have a voltage setpoint or control loop. That's because there is no big gas/steam valve to adjust the amount of power generated. The operator wants to inject all the power from the sun/wind into the grid, whatever the voltage.

reactive power being easily controlled is the same as a voltage control loop with droop when connected to a larger system.

Control of the traditional prime mover by way of devices like steam throttle valve controls real power (MW) and not voltage or reactive power (MVAR).

I am curious how an inverter operating away from unity power factor fails to deliver roughly the identical amount of energy as at unity. Can you provide a link that explains it? Traditional generators pump out the same MW barring losses due to increased current no matter the power factor. MW in mechanical * efficiency = MW out electrical. And of course there is a limit to how out of phase the voltage and current waveforms can be, usually +/- 0.9 pf.

A 60Hz cycle is 16.7 ms, A 13 kV breaker takes 80ms to open, to detect a fault (except for differential element) let alone clear it takes a lot longer than 20ms.

> Nearly all of the purposes for these flywheels could equally be met with existing infrastructure with software only changes.

While that's true, sometimes it's easier to have a single-purpose infrastructure, so that you don't have conflicting requirements.

For example, if a wind farm needs to feather its turbines due to excessive wind (which could damage the turbines), that could conflict with a simultaneous objective for frequency regulation.

Reply on news.ycombinator.com