Up sizing utility transformer for solar install

[tbnorrow]

I am looking at installing a 215 KVA solar array. The existing 150 KVA transformer will be replaced with a spare 300 KVA that I have. No load is being added to this system. I am simply adding the 215 KVA array to offset utility power. Is there any restrictions with this installation? Does changing impedance effect tripping MCB and arc flash?

 

[ron]

I am looking at installing a 215 KVA solar array. The existing 150 KVA transformer will be replaced with a spare 300 KVA that I have. No load is being added to this system. I am simply adding the 215 KVA array to offset utility power. Is there any restrictions with this installation? Does changing impedance effect tripping MCB and arc flash?

If the plan is to offset utility power, why put a 215kVA array? The existing load must be less than 150kVA now, right?

Depending on the voltage, will you get stuck complying with 690.64 and 705.12? That will limit the connection size.

 

[jaggedben]

Is there any restrictions with this installation?

I think you will need to calculate a new OCD size for the transformer, and probably install new conductors, but this will be covered in Article 450 and elsewhere, not article 690.

As far as restrictions regarding what size transformer can be backfed with solar, there are none AFAIK, although consulting with your local utility might be wise.

If the plan is to offset utility power, why put a 215kVA array? The existing load must be less than 150kVA now, right?

215 kW peak power PV will only deliver an average of 20-40 kW over the course of the year. Given the size, this must be a commercial property, and maybe it operates round the clock at a higher average load than that. Still, the numbers suggest an unusual situation. Maybe he is applying for the FIT program in Ontario.

 

[ggunn]

215 kW peak power PV will only deliver an average of 20-40 kW over the course of the year. Given the size, this must be a commercial property, and maybe it operates round the clock at a higher average load than that. Still, the numbers suggest an unusual situation. Maybe he is applying for the FIT program in Ontario.

What do you mean by "an average of 20-40 kW over the course of the year"? Do you mean 20-40kW steady state power averaged over 24 hours/day, 365 days/year (and of course, solar doesn't produce at night), or do you mean 20-40 kW *hours* per some time period?

If the system were here in Austin, at a fixed azimuth of 180 degrees and a fixed tilt of 30.3 degrees, PVWatts predicts that it would produce something in the neighborhood of 300,000 kWh/year, or about 800 kWh/day. If I divide that by 24 hours/day, I get about 33.5 kWh/hour, or 33.5 kW. Is that how you arrived at your number?

 

[tbnorrow]

I am working with our utility provider on this install. I am in the beginning stages. The most accessible transformer is a 150 kva. We are a large commercial site and own all our transformers. The existing load will not change. We are back feeding the grid with the 215 kva of solar. I have 2 other transformers that will not need to be upsized for this application, but they are a little more complicated to access. If the wire size and OCBP are sized accordingly, I should be ok? I will let you know what the utility provider recommends.
Thank you.

 

[broadgage]

I would agree that the transformer needs to be replaced with a larger one, as the O/P proposes.
The existing transformer is sufficient for the load, which is not being increased, but it is not sufficient for the PV backfeed.

Under good conditions, the PV installation should produce its rated output of 215 KVA.
This would overload a 150 KVA transformer.
It is probable that most of the time, some power would be used in the facility, and that this would reduce or eliminate any overloading.
For example with PV production of 215 KVA and a load of 100 KVA, the net load on the transformer would be 115 KVA, which is well within the capacity of the original transformer.

One should however allow for the worst case, which would be full PV production when the plant is shut down and consuming no, or little, power.

If other circumstances allow, it might be a better design to limit the PV capacity to the size of the existing transformer, thereby saving the need for a larger transformer.
If desired, a second PV instalation could be connected to another transformer elswhere in the facility.

 

[jaggedben]

What do you mean by "an average of 20-40 kW over the course of the year"? Do you mean 20-40kW steady state power averaged over 24 hours/day, 365 days/year (and of course, solar doesn't produce at night), or do you mean 20-40 kW *hours* per some time period?

I mean power averaged over the whole year, or kW hours per year divided by hours in the year.

If the system were here in Austin, at a fixed azimuth of 180 degrees and a fixed tilt of 30.3 degrees, PVWatts predicts that it would produce something in the neighborhood of 300,000 kWh/year, or about 800 kWh/day. If I divide that by 24 hours/day, I get about 33.5 kWh/hour, or 33.5 kW. Is that how you arrived at your number?

I multiplied the system size by two assumed capacity factors (15 and 20 percent) that from my studies represent the high and low of the most likely range, and then I rounded off a bit. Quick and dirty, but good enough for a back-of-the-envelope estimate. Glad to see that PVWatts corroborates.

 

[ggunn]

I mean power averaged over the whole year, or kW hours per year divided by hours in the year.

OK, but of course you realize that the system won't be running at or near 30kW very much of the time. Maybe on an overcast day or just on the daylight side of sunrise and sunset. If you could build a system that runs at 30kW 24/365 you'd really have something! ;^)

 

[jaggedben]

OK, but of course you realize that the system won't be running at or near 30kW very much of the time.

Of course I do. But from a net-metering point of view it makes no difference, or not much. If the solar fluctuates hugely but still averages 30kW, and meanwhile the load fluctuates perhaps not hardly at all but also averages 30kW, then energy charges on the bill could be zero. I was trying to explain to those listening why a site with a max load of less than 150 kW might possibly want to install a solar system bigger than that.

 

[ggunn]

Of course I do. But from a net-metering point of view it makes no difference, or not much. If the solar fluctuates hugely but still averages 30kW, and meanwhile the load fluctuates perhaps not hardly at all but also averages 30kW, then energy charges on the bill could be zero. I was trying to explain to those listening why a site with a max load of less than 150 kW might possibly want to install a solar system bigger than that.

Something else to consider is that net metering arrangements differ from utility to utility and the output of a PV system varies a lot month to month depending on the available sun hours. Some agreements let the meter run backwards and forwards all year and settle up at the end of the year (or fiscal year), while others zero it out monthly. If a utility pays you at all for any generation in excess of usage, it is usually a much lesser amount per kWh than retail, so one needs to run the numbers in light of the specific net metering arrangement available.

 

[jaggedben]

Something else to consider is that net metering arrangements differ from utility to utility and the output of a PV system varies a lot month to month depending on the available sun hours. Some agreements let the meter run backwards and forwards all year and settle up at the end of the year (or fiscal year), while others zero it out monthly. If a utility pays you at all for any generation in excess of usage, it is usually a much lesser amount per kWh than retail, so one needs to run the numbers in light of the specific net metering arrangement available.

True. I had in mind an annual "true-up."