PV Inverter backfeeding and damaging generator

[philly]

I was reading somewhere that an PV inverter that could possibly be connected to an emergency system should never be allowed to remain on when the system is running on backup generator power. It was stated that most inverters have an input that will shut the inverter off via an aux contact in the emergency generator when the emergency generator is up and running.

Amoung other reasons, one of the reasons for shutting the inverter down is that if the loads on the system are less than the amount of power that the inverter is supplying, then the inverter will push current back into the generator and damage the generator. Can someone explain to me how this would be the case. I understand that as long as there is sunlight avaliable a PV inverter will output a constant current regardless of system conditions, however I'm not sure I understand how this will backfeed into the genrator. Can someone please help explain?

 

[jaggedben]

I was reading somewhere that an PV inverter that could possibly be connected to an emergency system should never be allowed to remain on when the system is running on backup generator power. It was stated that most inverters have an input that will shut the inverter off via an aux contact in the emergency generator when the emergency generator is up and running.

Short answer is yes, that's true. Also, another option besides an auxillary contact is simply to put the PV system on the same side of an ATS as the utility.

Long answer is that in systems that are designed to work off-grid, it may be possible to have both inverter and generator on at the same time if there are sufficient loads, but that rarely if ever happens on purpose. Usually the generator is there as a back up in case there isn't enough sun for the PV system to keep the batteries charged.

Among other reasons, one of the reasons for shutting the inverter down is that if the loads on the system are less than the amount of power that the inverter is supplying, then the inverter will push current back into the generator and damage the generator. Can someone explain to me how this would be the case.

What will happen is that the current flowing from the inverter will act on the generator as if it were a motor, and cause it to slow down, stop, or theoretically even spin backwards. It is not completely different than if a physical object were slowing down or stopping the generator.

I think in the real world most generators will just stall or otherwise stop producing grid quality power, thus causing the inverter in turn to immediately shut down. Damage may be kind of theoretical: I've yet to hear any stories of a PV inverter actually damaging a generator. But in theory, it's possible, depending on how well the generator can handle being inhibited in this way. It's certainly the case that they will make each other useless, and that any generator or PV inverter manufacturer will tell you that backfeeding a generator with a GT inverter is not covered under the warranty.

I understand that as long as there is sunlight avaliable a PV inverter will output a constant current regardless of system conditions,

"regardless of system conditions" is incorrect or at least a poor choice of words. The inverter will output power at the grid voltage (essentially), and the current will be proportional to the power produced by the array, so not constant. I think you are conflating the properties of PV cells with the properties of inverters. I don't know how useful it is to think of an inverter as a current source instead of a voltage source (I think it's both?), but in any case that's pretty irrelevant to the question.

 

[jaggedben]

I understand that as long as there is sunlight avaliable a PV inverter will output a constant current regardless of system conditions..

I've been refreshing my memory on this...

A grid-tied PV inverter is configured as a constant current source. (The current is constant for a given amount of PV power.)
An off-grid PV inverter is configured as a constant voltage source.

 

[philly]

What will happen is that the current flowing from the inverter will act on the generator as if it were a motor, and cause it to slow down, stop, or theoretically even spin backwards. It is not completely different than if a physical object were slowing down or stopping the generator.

I think in the real world most generators will just stall or otherwise stop producing grid quality power, thus causing the inverter in turn to immediately shut down. Damage may be kind of theoretical: I've yet to hear any stories of a PV inverter actually damaging a generator. But in theory, it's possible, depending on how well the generator can handle being inhibited in this way. It's certainly the case that they will make each other useless, and that any generator or PV inverter manufacturer will tell you that backfeeding a generator with a GT inverter is not covered under the warranty. .

I dont exactly see how the inverter would push current into the generator. If the generaor has a voltage at its terminals would the inverter see the generator as an impedance load similar to how a motor would appear as an impedance load on a normal system?

Or assuming that the generator is viewed as a current source is this a case where a current source in in parallel with a voltae source where the current source (inverter) will develop the voltage source's (generator) voltage across it but will push a constant current into the voltage source?

"regardless of system conditions" is incorrect or at least a poor choice of words. The inverter will output power at the grid voltage (essentially), and the current will be proportional to the power produced by the array, so not constant. I think you are conflating the properties of PV cells with the properties of inverters. I don't know how useful it is to think of an inverter as a current source instead of a voltage source (I think it's both?), but in any case that's pretty irrelevant to the question.

I was of the understanding that inverters were represented as a current source. Since they are not producing voltage but rather just syncing with the grid they are only pushing current out based on what is avaliable from the solar input on the DC side.

 

[jaggedben]

I dont exactly see how the inverter would push current into the generator.

Where else do you think the inverter current will go, if there are not enough loads? The generator's alternator completes the circuit.

If the generaor has a voltage at its terminals would the inverter see the generator as an impedance load similar to how a motor would appear as an impedance load on a normal system?

Maybe I'm wrong - I can't claim to firmly grasp the physics myself - but my understanding is yes.

I was of the understanding that inverters were represented as a current source. Since they are not producing voltage but rather just syncing with the grid they are only pushing current out based on what is avaliable from the solar input on the DC side.

This is basically right, except that, as I clarified, only grid-tied PV inverters are configured as a current source.

 

[RCinFL]

First a few points on GT inverter and generators.

A GT inverter is required by UL1741 spec to have very tight restrictions on grid frequency and voltage limits. Frequency limits are grid (60 Hz) -0.7 Hz to +0.5 Hz. A GT inverter will not activate if frequency of AC is outside this range. A syncronous generator (one that output frequency is based on engine speed) often is not accurate enough to maintain these tight frequency limits. Second, a GT inverter limit is line voltage, typically 205 to 260 vac rms.

There is also a requirement on GT inverters to periodically test the grid to ensure it is still present to avoid "islanding" where GT inverter continues to operate when grid goes open. This is typically done by GT inverter intentionally attempting to wobble its phase lock to grid AC. A valid and present grid is like an immovable object to the GT inverter and any attempt to wobble the phase lock and/or raise or lower voltage will be met with increase and decrease in current flow from GT inverter.

Syncronous generators have a number of different regulator designs that adjust the engine speed and field current to ensure stable output voltage and frequency from the generator. Many newer generators have an electronic control regulator with a stepper motor control of the engine throttle. It is difficult to predict how a given generator's regulator system may reach to an input push of power. Most of the time the output voltage will rise out of the GT inverter high end spec causing it to shut down. It will attempt to reactivate in about 5 minutes.

The closest real world experience I have with back feeding a generator is with a hybrid battery based grid connected bi-directional inverters. These inverters syncronize to AC input and can push or pull power from the AC source. In this situation the inverter will usually release (open its input power relay) when it tries to push power back into generator due to input AC voltage going out of range high in voltage.

But again, this depends on how generator regulator reacts to back push of current. You can be fairly sure that the designer of the generator's regulation system was not considering that a source would be pushing current back into generator. I personally would not allow a current backfeed into a generator for any significant period of time.

 

[PsychicViking]

I was reading somewhere that an PV inverter that could possibly be connected to an emergency system should never be allowed to remain on when the system is running on backup generator power.

I am currently faced with this situation on a current project. We are installing a 200KW PV system on a 3 phase 277/480v system with (2) 1000KW backup diesel generators. We are thinking of installing a shunt trip breaker on the output of the PV sysetm that will operate on a 3 second power outage. We would also have to install a reverse power relay to protect the generators. The reason for the reverse power relay is not because the generator would spin in the reverse direction as a motor load. It would spin the same direction as a motor load, but the thrust would reverse. A generator thrusts in one direction along the shaft when in acts as a source and in the opposite direction when it is a load ie. from the PV system that has a higher output. Reverse power does not damage the generator but the engine. Thrust bearings are usually only designed to absorb thrust from one direction.

 

[Open Neutral]

SMA claims, as I understand it:

A) When on backup inverter ["Sunny Island"] supply, and the generator kicks in [because the batteries are low]; the inverter sync's to the generator and either assists, or if the load is low; it recharges its batteries.

B) Further, their Sunny Boy gridtie units will drop off when Reddy Kilowatt croaks, BUT once their cousins above are going, they'll join back in... so on a sunny day, the Sunny Boy's will charge the batteries/help supply loads.

If I have this wrong, hopefully Tallgirl will correct me.

 

[RCinFL]

SMA claims, as I understand it:

A) When on backup inverter ["Sunny Island"] supply, and the generator kicks in [because the batteries are low]; the inverter sync's to the generator and either assists, or if the load is low; it recharges its batteries.

B) Further, their Sunny Boy gridtie units will drop off when Reddy Kilowatt croaks, BUT once their cousins above are going, they'll join back in... so on a sunny day, the Sunny Boy's will charge the batteries/help supply loads.

If I have this wrong, hopefully Tallgirl will correct me.

SMA has a secondary control system between Sunny Island power inverter and their Sunnyboy GT inverter. The Sunny Island will control the output of the Sunny Boy GT inverter by varying its output frequency. When off grid, if battery is charged and loads cannot consume GT power then the Sunny Island tells the GT inverter to reduce its power output.

When on grid or generator, the SMA system will still backfeed power. When on generator, Sunny Island cannot control Sunnyboy GT inverter because generator is line frequency master.

 

[philly]

If we are indeed saying that a grid-tied inverter is a current source then its output current will take all avaliable paths based on the current divider principal.

Now I have been told that if there is not enough load on the system the inverter will backfeed into the generator and damage it. I would think though that even with enough load on the system the inverter would still feed some current into the generator but maybe not as much based on the following:

As I mentioned a current source will supply its rated current (assuming full DC input for the inverter) no matter what the load is (as long as its not an infinite load). I would think that with lots of load on the system these loads would present less of an impedance to the inverter and therefore the ouput current from the inverter would seek the path of the loads based on the current divider principal and not much current would flow to the generator. Does the generator have a high load impedance as seen by the inverter compared to other loads.

Now if the other loads are shut off this does not provide alternate paths for the current to flow so this means more current will flow to the generator becuase there is no longer other low impecance paths. In this instance the current can now damage the generator. At what value of current or % will the generator become damaged?

Does my reasoning make sense?

 

[jaggedben]

I would think though that even with enough load on the system the inverter would still feed some current into the generator ...

I would think that with lots of load on the system these loads would present less of an impedance to the inverter and therefore the ouput current from the inverter would seek the path of the loads based on the current divider principal and not much current would flow to the generator.

I think everyone agrees that no danger would arise until the PV inverter current exceeded the loads. But obviously from a system design point of view this possibility always has to be accounted for.

I have no idea if there is theoretically always an infinitesimal amount of current that flows to the generator, but I think that practically speaking, if the loads exceed the inverter current, then the amount of inverter current flowing to the generator is zero. It's certainly true that a meter between the loads and the generator wouldn't measure any power flowing in that direction.

At what value of current or % will the generator become damaged?

I would think that that would depend on the generator model.

 

[SOG38]

PV back feeding a generator

PV back feeding a generator

IMO
If the generator and the inverter are both on line and the voltage of both are equal and in phase regardless of system load neither would be a load to the other. Just as if you put two batteries is parallel with each other. If the voltage is the same there is NO current flow.
Either of the units off it will then become a load to the other.

 

[jaggedben]

IMO
If the generator and the inverter are both on line and the voltage of both are equal and in phase regardless of system load neither would be a load to the other. Just as if you put two batteries is parallel with each other. If the voltage is the same there is NO current flow.
Either of the units off it will then become a load to the other.

Sorry, but with a GT inverter this is wrong.

A GT inverter is configured as a current source and will raise (or lower) the voltage to whatever is necessary to push out its current. The voltage of the inverter will not match the generator, it will attempt to be greater. A GT inverter is not like a battery.

 

[RCinFL]

GT inverter acts as an AC current source over the narrow acceptance range of the grid specifications, 60 Hz -0.7Hz +0.5Hz, and 240 vac +/- 17 volts. GT inverter will shut itself down if AC source master is outside these limits. There is also a limit on the rate of change of AC source frequency/phase which many generators will not meet due to sluggish engine rpm feedback control as load varies on the generator.

Since it is possible for the AC voltage to stay within acceptance limits with an open AC grid with just the right amount of load to match the GT inverter injected current, the GT inverter will periodically try to 'wiggle' the phase to ensure the grid is still there. This phase wiggle attempt may not cause the generator AC frequency to go out of spec so GT inverter would not shut down.

It will backfeed power into a generator if AC loads do not consume the PV generated power. The most likely, but not guaranteed, scenario is the generator output voltage will rise due to PV power pushed at it and go outside the GT inverter maximum voltage limit causing the GT inverter to shut down.

After about a 5 minute period the GT inverter will turn itself back on and the process will repeat unless AC loads begin to consume greater then the PV power production (or until generator quits, one way or another).

 

[SOG38]

PV/GTI backfeeding to a generator

PV/GTI backfeeding to a generator

Here is a good explanation of the operation of a GTI
http://solar.smps.us/grid-tie-inverter-schematic.html
It is a voltage source generating a slightly higher voltage so it can feed back onto the grid. To this it would indeed feed back to a gen set operating in parallel with the GTI. The load will take the current necessary up to the GTI max. The current fed back to the Alternator would be determined by the impedance. The amount of current it could handle would be a question for the manufacturer.
I wouldn't think operating them together would be a good idea.