Isolating generator from solar system

[romex jockey]

Some of the PV system inverters have a generator input now

Not that i know the specifics .....

guess i need an ejocshun...:lol:~RJ~

 

[ggunn]

Some of the PV system inverters have a generator input now

Not that i know the specifics .....

guess i need an ejocshun...:lol:~RJ~

All the ones I have seen that have that are actually battery inverters, not PV inverters per se. PV feeds the batteries though a charge controller. Systems like this are significantly more expensive per kW than are grid tied PV systems.

One more thing to point out is that grid tied PV systems are much simpler than any involving generators and batteries. Grid tied PV is pretty much set it and forget it, while the other type requires that the owner of the system understands how it all works together and how to get it running again when it stops working, which they all will at some point.

 

[pv_n00b]

SMA makes a large generator + PV system for remote power at places like mines. They limit the PV to 20% of the minimum load to keep from causing a problem. The PV just saves a little fuel and the generators are always running when power is needed. So it is possible to do a dual system, just not very useful for most people.
https://www.sma-america.com/industrial-systems/pv-diesel-hybrid-applications.html

 

[winnie]

As I said previously, I expect this issue to become more and more common, and that if customers ask for it frequently enough an inexpensive solution will be provided.

Consider that the PV system inverter is controlled by some sort of microprocessor, and I expect the same is true for an inverter generator (which are becoming more common), and that the issue is a controls problem, not a physical impossibility. To make this work would require the correct and compatible software on the two systems, which would not be trivial, but which would be an engineering expense that doesn't particularly change system production costs.

-Jon

 

[ggunn]

As I said previously, I expect this issue to become more and more common, and that if customers ask for it frequently enough an inexpensive solution will be provided.

Consider that the PV system inverter is controlled by some sort of microprocessor, and I expect the same is true for an inverter generator (which are becoming more common), and that the issue is a controls problem, not a physical impossibility. To make this work would require the correct and compatible software on the two systems, which would not be trivial, but which would be an engineering expense that doesn't particularly change system production costs.

-Jon

See post #21. Most people who ask us how to make a generator and a PV system work at the same time haven't thought it through as to how much it would actually benefit them. Even if the power is out for a couple of days, how much would your PV system have saved you during that time at ~10 cents per kWh? For residential systems most of the time the answer is "not much".

 

[Chamuit]

If the home is using X amount of power and the PV is producing X or less (not more), wouldn't the generator just be available voltage that isn't consumed?

 

[electrofelon]

If the home is using X amount of power and the PV is producing X or less (not more), wouldn't the generator just be available voltage that isn't consumed?

Yeah that would be fine. It's when production exceeds usage that you might have not a super happy fun time. Actually I would speculate that the inverter will trip out before doing any damage, but I'm not going to try it.

 

[GoldDigger]

There is a fundamental difference in function between an off grid PV inverter (which pretty much requires a battery bank for reliable service) and a grid tied inverter.
The former is a constant voltage source with current varying with the load impedance, while the latter is a current source, with the supplied current following the applied voltage waveform and limited only by the instantaneous (1 cycle average?) maximum PV power available.
Some output circuit designs are optimum for the first mode, different ones are optimum for the second mode, independent of the control system. An output circuit that can do both might be more expensive than either alone.

Sent from my XT1585 using Tapatalk

 

[jaggedben]

...
Some output circuit designs are optimum for the first mode, different ones are optimum for the second mode, independent of the control system. An output circuit that can do both might be more expensive than either alone. ...

Developments in the industry lead me to believe that controls and safety standards are far more at issue than core inverter components. There are at least three manufacturers now that offer multi-mode inverters that switch from grid-tied to independent (primary source) within a fraction of a second, and several others that do so in a slightly longer time. Most of these are battery inverters but one of them (Enphase IQ8, not quite to market yet) can do the same as a solar panel inverter. As far as I know, all of these can also throttle their output using input from a compatible meter. With respect to the subject of this thread, Enphase has promised incorporation of a generator into an all-AC-coupled architecture in the near future. So the landscape is shifting pretty rapidly, and some of the pat answers we've provided in the past, while still a good explication of theory, may not describe the practical limitations so neatly anymore.

 

[ggunn]

Developments in the industry lead me to believe that controls and safety standards are far more at issue than core inverter components. There are at least three manufacturers now that offer multi-mode inverters that switch from grid-tied to independent (primary source) within a fraction of a second, and several others that do so in a slightly longer time. Most of these are battery inverters but one of them (Enphase IQ8, not quite to market yet) can do the same as a solar panel inverter. As far as I know, all of these can also throttle their output using input from a compatible meter. With respect to the subject of this thread, Enphase has promised incorporation of a generator into an all-AC-coupled architecture in the near future. So the landscape is shifting pretty rapidly, and some of the pat answers we've provided in the past, while still a good explication of theory, may not describe the practical limitations so neatly anymore.

But the economics questions remain. Unless the manufacturers can hit the same price point as grid tied systems you have to examine the financials and decide if it's worth the added expense.

 

[JoeNorm]

Interesting post and something we have dealt with numerous times. I don't think this is going to be "more and more" common, necessarily. The OP has a pretty unique situation that the PV input and generator are so physically far from one another without a means to get a com wire between the two. Usually, like other posters have said, we're able to manage this situation pretty easily with either a relay or landing the PV in a way that isolates it from the generator.

In this case, the best option seems to be to scratch in the relay wires back to the gen.

If a customer really wants generator and PV, they should have a battery based inverter somewhere in the system as well. This is where you'd input your generator. We are doing more and more AC-coupled systems and rerouting the generator input to the battery-based inverter

 

[jaggedben]

But the economics questions remain. Unless the manufacturers can hit the same price point as grid tied systems you have to examine the financials and decide if it's worth the added expense.

True, in the general sense. But out here in California PG&E's Public Saftey Power Shutdowns this fall have completely changed the way people look at those value propositions, pretty much overnight. It's mostly about backup power, not net-metering.

 

[ggunn]

True, in the general sense. But out here in California PG&E's Public Saftey Power Shutdowns this fall have completely changed the way people look at those value propositions, pretty much overnight. It's mostly about backup power, not net-metering.

Which, of course, is consistent with what I said as well. PG&E is in a world of hurt and its customers have only bad and worse options.

 

[winnie]

But the economics questions remain. Unless the manufacturers can hit the same price point as grid tied systems you have to examine the financials and decide if it's worth the added expense.

I agree that this is a key point, and you or someone else calculated that if you have a grid tie system, and then want to run it as a 'generator tied' system during a power outage, the amount of money one is likely to save (in terms of reduced generator fuel use over the duration of a normal power outage) is likely to be quite small.

However I think that this is an idea that is _attractive_ to the customer, so _if_ the manufacturer can do this in software/firmware with minimal or no change in production costs, then it will happen even if the only true value is as a selling/advertising point.

It might actually have real value if it makes a real difference in the amount of gas that gets used during an outage, not in terms of the number of gallons saved (which I agree is likely to be small), but in terms of the number of trips to the gas station or the number of pauses to fill a tank.

-Jon

 

[ggunn]

I agree that this is a key point, and you or someone else calculated that if you have a grid tie system, and then want to run it as a 'generator tied' system during a power outage, the amount of money one is likely to save (in terms of reduced generator fuel use over the duration of a normal power outage) is likely to be quite small.

However I think that this is an idea that is _attractive_ to the customer, so _if_ the manufacturer can do this in software/firmware with minimal or no change in production costs, then it will happen even if the only true value is as a selling/advertising point.

It might actually have real value if it makes a real difference in the amount of gas that gets used during an outage, not in terms of the number of gallons saved (which I agree is likely to be small), but in terms of the number of trips to the gas station or the number of pauses to fill a tank.

-Jon

Like my dad always said, if a frog had wings he wouldn't bump hisass every time he jumps.

In the 12 years I have been in solar, far and away the most common complaint I hear from customers about PV systems is that their system shuts down when the grid fails. If it were easy and/or cheap to build PV inverters that could run off line without batteries and deliver power on demand, someone (everyone?) would be doing it.

 

[winnie]

Like my dad always said, if a frog had wings he wouldn't bump hisass every time he jumps.

In the 12 years I have been in solar, far and away the most common complaint I hear from customers about PV systems is that their system shuts down when the grid fails. If it were easy and/or cheap to build PV inverters that could run off line without batteries and deliver power on demand, someone (everyone?) would be doing it.

But the original post starting this thread is not about running inverters with no energy storage. The post was about running inverters in parallel with a generator.

If you have solar and inverters alone trying to feed an ordinary home, then you will have to regularly deal with the loads exceeding the available production. Unless you have some very smart load shedding then the system can't work. I absolutely agree that solving this is a tough nut unless you have energy storage.

On the other hand if you have a generator properly sized for the home, then you have a system that could in theory work, as long as the generator and inverter control systems play nice with each other. In this case you have 3 scenarios: 1) generator and PV together undersized for the load, which shouldn't happen if the generator is properly sized. 2) PV production less than the home load, so that some load remains on the generator. This case should work with an ordinary grid tie inverter. 3) PV production greater than the home load, so that there is too much production on the 'island'. This is the case that will break things.

IMHO the 3rd scenario might be solvable in firmware.

-Jon

 

[jaggedben]

On the other hand if you have a generator properly sized for the home, then you have a system that could in theory work, as long as the generator and inverter control systems play nice with each other. In this case you have 3 scenarios: 1) generator and PV together undersized for the load, which shouldn't happen if the generator is properly sized. 2) PV production less than the home load, so that some load remains on the generator. This case should work with an ordinary grid tie inverter. 3) PV production greater than the home load, so that there is too much production on the 'island'. This is the case that will break things.

IMHO the 3rd scenario might be solvable in firmware.

-Jon

We know it's solvable because it's been done, many years ago, by Trace, SMA and others. As I alluded to above, my impression is that the reason more products aren't available to do it has more to do with the development costs of controls and meeting product standards than with fundamental engineering problems. That's especially true when you don't know if there's a market for the product. I think we'll see things gradually change as that market grows.

 

[ggunn]

We know it's solvable because it's been done, many years ago, by Trace, SMA and others. As I alluded to above, my impression is that the reason more products aren't available to do it has more to do with the development costs of controls and meeting product standards than with fundamental engineering problems. That's especially true when you don't know if there's a market for the product. I think we'll see things gradually change as that market grows.

SMA doesn't really count because only ~1000W or so is available from the PV system when it is running off grid no matter how large it is. I don't know about Trace, but I still submit that if were easy and/or cheap to run a PV system off grid without storage and supply up to its full capacity on demand, it would be widely available. It's not like no one wants it. If you know how to do it, do it; it will make you very wealthy.

 

[ggunn]

But the original post starting this thread is not about running inverters with no energy storage. The post was about running inverters in parallel with a generator.

If you have solar and inverters alone trying to feed an ordinary home, then you will have to regularly deal with the loads exceeding the available production. Unless you have some very smart load shedding then the system can't work. I absolutely agree that solving this is a tough nut unless you have energy storage.

On the other hand if you have a generator properly sized for the home, then you have a system that could in theory work, as long as the generator and inverter control systems play nice with each other. In this case you have 3 scenarios: 1) generator and PV together undersized for the load, which shouldn't happen if the generator is properly sized. 2) PV production less than the home load, so that some load remains on the generator. This case should work with an ordinary grid tie inverter. 3) PV production greater than the home load, so that there is too much production on the 'island'. This is the case that will break things.

IMHO the 3rd scenario might be solvable in firmware.

-Jon

The two issues are really the same issue; they are about the ability (or lack thereof) of a grid tied PV system to supply power on demand in response to a load. The problem with all three scenarios you propose is that the home load is not a static quantity; it could conceivably be zero. A voltage source like a generator has no problem tracking the load but a current source like a grid tied inverter (at the present state of the art) is a different animal.

 

[winnie]

The two issues are really the same issue; they are about the ability (or lack thereof) of a grid tied PV system to supply power on demand in response to a load. The problem with all three scenarios you propose is that the home load is not a static quantity; it could conceivably be zero. A voltage source like a generator has no problem tracking the load but a current source like a grid tied inverter (at the present state of the art) is a different animal.

Absolutely!

As you note, a generator is _designed_ as a voltage source. The generator control system operates to regulate its output voltage, and as the load changes the engine throttle gets opened or closed until the voltage is returned to the target level.

Meanwhile a grid tie inverter is _designed_ to supply as much power as the PV system is producing to the grid. The output of the inverter is much better treated as a current source feeding a constant voltage grid.

The only point of discussion is that this 'current source' design is very likely implemented in the firmware of a microcontroller driving the power electronics, and thus open to manipulation without large additional hardware expenses. The engineering of this to work would not be trivial however, and so there would need to be lots of customer requests for this feature.

-Jon