How is power limited from renewable sources

[AtonalControl]

TL;DR Summary

How is the power output of renewable sources limited to achieve a specific output. (Ideally solar farms but any renewable source)

I'm an engineering student and I recently watched a lecture from a wind farm control system engineer. He said that the national grid will call them up and ask for a specific amount of power. He then enters that value into the system and it will automatically adjust to achieve that power output.

My question is how is that achieved. If the substation it's connected to requires a specific voltage, like 33kV, then the voltage can't change. The current then depends on the impedance at the substation which I don't think changes. So if the voltage doesn't change and the current doesn't change then I can't see how the power changes. But it must do because the power output of renewable sources is not constant.

I'm building a small model solar farm and I want to be able to vary the power output. A solar farm array goes into an DC/DC converter to achieve a specific voltage, then to an inverter to convert it to AC. If the national grid were to call up a solar farm and ask for the power to be reduced, how would they achieve that and how does it affect the voltage and current at each stage?

 

[berkeman]

Paging @cjl and @anorlunda

 

[berkeman]

Welcome to PF, @AtonalControl

I don't know if anorlunda will be available soon to reply in this thread, but here is a link to a PF Insights article that he wrote about the subject:

https://www.physicsforums.com/insights/renewable-energy-meets-power-grid-operations/

Be sure to follow the links in that article to some of his other articles about the power grid. Great stuff.

 

[russ_watters]

I think there's two different questions there: how to generate and how to deliver.

For anything with a rotating generator or motor, there is a push or pull (torque) associated with the phase angle, which corresponds to the amperage. So you adjust how much you generate by adjusting the torque applied to the generator.

I believe delivery is a function of voltage, but I'm not sure. The voltage of the grid is not actually constant (I do know that for sure).

 

[cjl]

So with AC power, and particularly grid-scale AC power, it's a little different. You can't look at a single "impedance" at a substation, at least not in the sense of "you feed in a particular voltage and the current is deterministically based only on that voltage". To an extent, that's somewhat true, but you can also control how much of that load any generator sees through the phase angle - a generator will always be slightly leading the grid voltage, and the amount that you lead it by changes the amount of load that that generator actually sees. As Russ said, fundamentally, the way you adjust that lead angle is by controlling torque.

Note that if you do this with one generator, the other generators on the grid see a reduction in load as a result, since you're very slightly pulling the phase ahead with the extra generation. If you have a sustained excess of generation greater than the load on the grid, this eventually manifests as an increase in the frequency of the AC, rather counterintuitively, rather than an increase in voltage. If you're used to thinking in small scale DC, you might expect that adding more generation would increase the voltage, but that's (largely) not true on the grid, rather you can track load events through frequency instead. Similarly, a sudden high load causes a frequency sag, which causes the generators to try to lead by a bit more as they try to maintain RPM which naturally tends to balance the load. This also leads to the nice situation where governed generators will naturally load follow, since an increase in load will cause a slight sag in frequency which the governed generators will then increase output to counteract.

Now, as for how you physically control torque on a wind turbine? That's just done by pitching the blades. Wind turbines have an enormous data table telling them what power, RPM, loading, etc to expect based on blade pitch and wind speed, ranging from the optimum power curve where it makes as much as possible all the way down to the minimum operating speed where it's barely running.

 

[AtonalControl]

Thanks a lot cjl that was incredibly helpful. I think my problem is I've been thinking about solar farms all wrong.

I watched that lecture of a Wind Farm control system and thought I would build a small scale version for my project but instead using solar panels to remove the mechanical component. Turns out I don't think solar farms limit their production in any way at all and are more like a battery, they only produce what there is demand for.

Now I have to rethink my whole project in a very short amount of time. Stress haha.

Thanks berkeman and russ too.

 

[Dullard]

Specific to solar panels, you may want to search:
i-v curve
power point tracking

familiarity with these 2 topics will help you understand how a PV panel is like (and not like) a battery.

 

[russ_watters]

Thanks a lot cjl that was incredibly helpful. I think my problem is I've been thinking about solar farms all wrong.

I watched that lecture of a Wind Farm control system and thought I would build a small scale version for my project but instead using solar panels to remove the mechanical component. Turns out I don't think solar farms limit their production in any way at all and are more like a battery, they only produce what there is demand for.

There's a third question here, on how to decide what the output should be from each plant - you don't just throttle every plant to the same percent output. That's more an economic question than a technical question. We have an insight blog article about that:

https://www.physicsforums.com/insights/renewable-energy-meets-power-grid-operations/

The short version is that high capital cost, low operating cost/marginal cost sources are only economical to run at full output nearly all the time. So almost all of the throttling is with fossil fuels or hydro. However, as the renewables fraction grows while battery/storage lags, there will be more times when over-production will happen if they aren't throttled too.

 

[Rive]

So almost all of the throttling is with fossil fuels or hydro. However, as the renewables fraction grows while battery/storage lags, there will be more times when over-production will happen if they aren't throttled too.

For the first part, you mean: almost all planned throttling is done by fossil and hydro.

Regarding the other part... Digging up relevant graphs is getting harder and harder and they come from further and further away from official sources.

Source

 

[russ_watters]

For the first part, you mean: almost all planned throttling is done by fossil and hydro.

Regarding the other part... Digging up relevant graphs is getting harder and harder and they come from further and further away from official sources.
View attachment 332776
Source

Well, if I'm reading that graph/post correctly, renewable curtailment is at about 1.6% of consumption. I think Germany is about half renewables, so that would make it 3.2% of renewable output. Or a 97% capacity factor. I don't want to quibble on where "almost all" ends but I agree it looks like curtailment is starting to become significant.

 

[sophiecentaur]

Turns out I don't think solar farms limit their production in any way at all and are more like a battery, they only produce what there is demand for.

Exactly. A PV cell is used as a 'voltage source', the emf depending on the light flux arriving from the Sun. PV cells are 'quite happy' to be disconnected from the load. OTOH, a turbine driven generator will speed up (sometimes disastrously) when the load is disconnected and the fuel supply to the turbine needs to be throttled back quickly. A nuclear power station is even more at risk if the grid lines are cut because it takes a long time to back off the nuclear reactions (many minutes / hours) and the reactor can melt down. In some remote nuclear stations, a nearby empty lake is filled up to make use of the excess energy.