Load sharing for solar PV and synchronous generators

Hello

If there are two synchronous generators installed, one is operating in isochronous mode and the other in droop, then I understand that as grid load changes, the generator in isochronous mode will regulate fuel to its prime mover to maintain system frequency. Speed of generators in droop will mode will not cahnge if the generator in isochronous is doing it job properly. I understand that the generator in droop is typically used to accomodate load forecasts by being able to vary its output power based on grid demand, such that the generator in isochronous is not inadvertantly tripped if load fluctuates too much.

If we connect a renewable source (such as solar PV or battery facility) to the grid and we want to export power from these facilities based on power availability, is it best to keep the same arrangement with the synchronous generators? My understanding is that the renewable facilities will act as virtual synchronous generators. So, if we increase the power from the renewable facilities, then the isochronous generator will decrease power output by speeding up/slowing down based on the change in grid frequency. Operators can then equalise the generators (if required) by adjusting the power output of the generator in droop mode. Is this typically how these control systems operate ?
 
polup,

I'm only familiar with home, roof-mounted solar PV cells and controllers--not industrial-sized PV installations, so I don't really know how the output of a solar PV installation would be controlled by operators. I suppose if there's some large rectifier bank that would be capable of varying the amps/volts out of the array that is a definite possibility. But, that's just a SWAG (Scientific Wild-Arsed Guess). My reference would be large AC drives which have a variable frequency output, as well as variable voltage/current to control speed and/or load; they have a converter to make DC that then gets chopped up into variable voltage and -frequency AC. But how that works with a paralleled synchronous AC grid and a solar PV array (the output of which is kind left to produce as much power as possible when the sun is shining) is not something I've worked on or even studied.

You mention "the grid" and I would have to think if the configuration uses an Isoch unit it is not a large ("infinite") grid, but a smaller, power island isolated from multiple generators...?

Personally, I would keep the typical arrangement when exporting power from the PV array, so that if clouds suddenly obscure the array the Isoch unit can "kick in" and help to maintain the frequency as the output current from the solar PV array decreases.

And, yes--you are correct about the Droop machine's function. It simply sits there and puts out the power it is adjusted to produce as long the Isoch unit keeps the grid frequency stable. If the load on the Isoch unit approaches full rated output (prime mover rating) the load on the Droop unit should be increased, which will decrease the load on the Isoch unit making it more capable of responding to a large change in load or generation. Conversely, if the load on the Isoch unit gets close to zero, the load on the Droop unit will have to be reduced to get the Isoch unit to take on more load so it doesn't get tripped on reverse power.

Finally, if the Isoch unit can't keep frequency stable for some reason the Droop unit will increase its power output as frequency decreases in an effort to help support the Isoch unit (and any other generation source) to power the load on "the grid."

What I don't know is what happens to the frequency of the solar PV array when "the grid" frequency decreases. It can't continue at 50- or 60 Hz if the system frequency isn't at 50- or 60 Hz. But, is there a way to "manually" increase or decrease the frequency of the solar PV array output? That I don't know. I just don't know how solar PV arrays respond or are supposed to respond to grid frequency disturbances.

Do you know if the output of the solar PV array can be varied by a human operator (through some kind of control system)?
 
For the wind farms and solar farms in operation today, there is no good answer for “how.” There are many, many “ok” answers. 10-15 years ago, regulation of renewables and especially inverter based resources was... not great. But penetration of IBRs was almost nothing and they contributed very little to the BPS. That’s not the case anymore. It’s very region dependent, but a 1000 yard summary is they want them to operate like thermal plants—be able to ride through grid frequency and voltage excursions and contribute to maintaining grid stability.

Consistency in application is still a few years out, and consistent modeling/dispatching/regulation even more years out. Today, IBR plants operate with headroom, so they never produce more than 90-95% of what they could be so that they can increase real power if frequency dips, or shed load of frequency goes up. New plants in North America are required to operate as described below, but older plants and plants in other parts of the world have a multitude of operational modes.

“FERC Order No. 842 requires all newly interconnecting generating resources within its jurisdiction to install, maintain, and operate a functioning governor or equivalent controls as a precondition of interconnection, effective May 15, 2018. FERC Order No. 842 requires new generation units to have functioning primary frequency response capability. The FERC Order also requires resources to respond to frequency excursion events when plant POM frequency falls at least outside a ± 0.036 Hz deadband, and to adjust output in accordance to a maximum of 5% droop.

This response must be timely and sustained rather than injected for a short period and then withdrawn. However, reserving generation headroom to provide frequency response for underfrequency events is not mandated by FERC Order No. 842. These resources should respond to overfrequency excursion events outside the deadband by reducing active power output in accordance with the 5% droop specification.
The NERC Reliability Guideline: BPS-Connected Inverter-Based Resource Performance outlines recommended dynamic response characteristics. The closed-loop dynamic response of the active power-frequency control system of the overall inverter-based resources, as measured at the POM (or possibly the POI), should have the capability to meet or exceed the performance specified in Table 2.1.49 TOs may consider using or adapting these specifications based on a technical basis (i.e., system studies). The requirements defined should not conflict with any inverter-based resource protection systems.”

From:
https://www.nerc.com/comm/PC_Reliab...rter-Based_Resource_Performance_Guideline.pdf.
 
polup,

I'm only familiar with home, roof-mounted solar PV cells and controllers--not industrial-sized PV installations, so I don't really know how the output of a solar PV installation would be controlled by operators. I suppose if there's some large rectifier bank that would be capable of varying the amps/volts out of the array that is a definite possibility. But, that's just a SWAG (Scientific Wild-Arsed Guess). My reference would be large AC drives which have a variable frequency output, as well as variable voltage/current to control speed and/or load; they have a converter to make DC that then gets chopped up into variable voltage and -frequency AC. But how that works with a paralleled synchronous AC grid and a solar PV array (the output of which is kind left to produce as much power as possible when the sun is shining) is not something I've worked on or even studied.

You mention "the grid" and I would have to think if the configuration uses an Isoch unit it is not a large ("infinite") grid, but a smaller, power island isolated from multiple generators...?

Personally, I would keep the typical arrangement when exporting power from the PV array, so that if clouds suddenly obscure the array the Isoch unit can "kick in" and help to maintain the frequency as the output current from the solar PV array decreases.

And, yes--you are correct about the Droop machine's function. It simply sits there and puts out the power it is adjusted to produce as long the Isoch unit keeps the grid frequency stable. If the load on the Isoch unit approaches full rated output (prime mover rating) the load on the Droop unit should be increased, which will decrease the load on the Isoch unit making it more capable of responding to a large change in load or generation. Conversely, if the load on the Isoch unit gets close to zero, the load on the Droop unit will have to be reduced to get the Isoch unit to take on more load so it doesn't get tripped on reverse power.

Finally, if the Isoch unit can't keep frequency stable for some reason the Droop unit will increase its power output as frequency decreases in an effort to help support the Isoch unit (and any other generation source) to power the load on "the grid."

What I don't know is what happens to the frequency of the solar PV array when "the grid" frequency decreases. It can't continue at 50- or 60 Hz if the system frequency isn't at 50- or 60 Hz. But, is there a way to "manually" increase or decrease the frequency of the solar PV array output? That I don't know. I just don't know how solar PV arrays respond or are supposed to respond to grid frequency disturbances.

Do you know if the output of the solar PV array can be varied by a human operator (through some kind of control system)?
Thanks for the response. I just think of the PV inverters as an virtual synchronous generators. Like yourself, I am not intimate with their inner workings, but I presume DC input to the inverters is being converted to an AC output at a power angle which differs to the grid. Varying the power angle would vary the amount of power exported and the isochronous generator adjusts accordingly. You are right, this is not an infinite grid with a few generators present. There is no connection to the grid.

If the grid frequency decreases (i.e. load drop), the PV inverter would decrease its power output via a control loop. Operators can always override this, but in a normal scenario, this would be controlled via the control system. I'm assuming that if this did not happen, the grid voltage would increase and the inverter would eventually trip on over voltage.
 
polup,

"If the grid frequency decreases (i.e. load drop), the PV inverter would decrease its power output via a control loop. ..."

Frequency decrease is usually the result of either a sudden increase in load, or a loss of generation (a unit is or units are shut down or tripped). They're both kind of the same thing: The load(s) on the grid exceed the amount of generation available.

Under this scenario the desired condition would be for the Isoch unit to pick up the load, but if it weren't capable of doing so then the Droop units would increase their output (if possible) to try to provide the power required by the load(s). If the Isoch unit returns the frequency to nominal, the Droop units will reduce their output to return to their pre-disturbance load(s).

I presume there is no control system sending load commands to any of the units--just pure speed control without any load control.

If you meant something other than what I understood, please correct me.
 
CSA - yes you are correct, sorry I had the frequency / load relationship the wrong way round. I think that's all I need for now, thanks for your help!
 
Top