Another lesson learned

[ggunn]

When designing a supply side interconnected PV system you can build it up to the size in backfed amps as the service, right? Not necessarily; you'd best look at the size of the transformer feeding that service. What I learned last week is that the POCO may provide a transformer that is significantly less in kVA rating than (in this case) the sum of the ratings of the two services it feeds. In the jurisdiction in this case a PV system or systems is/are only allowed to backfeed in system kW rating up to 90% of the kVA rating of the transformer. With PV systems connected to both services that fit under the rules for supply side connected PV, an integrator can exceed that 90% limit and trigger a transformer upgrade, which can be expensive.

This may be old news to some, but I thought I'd share it in hopes that others who encounter this will catch it before their systems are built.

 

[iwire]

Seems odd, the size of the transformer is of absoulutly no concern to us for services so the same should hold true for PV. Its an issue for the power company, its not our fault they undersized it in the first place.

 

[GoldDigger]

One difference is that the consumer load is typically far lower than the service size while a PV system sized to 90% of the service size will really produce that power at least part of the time.
The service wires could well be too small also.
With two residential units on a transformer the demand factor for normal loads will be less than 100% while two PV systems will have a demand factor of 1.0.
As to who should pay for the upgrade, all you have going to protect you from bring billed for the equipment upgrade are custom and tradition. And the major precedents set were for distributed and co-generation.

If POCO had to supply 100% rated equipment in every case our power bills would all be higher.

 

[iwire]

One difference is that the consumer load is typically far lower than the service size while a PV system sized to 90% of the service size will really produce that power at least part of the time.
The service wires could well be too small also.

I am aware of the differences.

Picture this, at some point the customer requested a certain size service, the POCO decided not to really provide that capacity. They have good reasons for this, but its not the customers choice, it was not what the customer requested so to go back to that customer and put the up grade on them would only fly if you are a monopoly, you would never get away with that sort of deal as a run of the mill company.

If POCO had to supply 100% rated equipment in every case our power bills would all be higher.

Yes they would be, but we are not talking every case, just the cases where the customer adds solar. And yes that cost would be passed on to all of us.

 

[Carultch]

One difference is that the consumer load is typically far lower than the service size while a PV system sized to 90% of the service size will really produce that power at least part of the time.
The service wires could well be too small also.
With two residential units on a transformer the demand factor for normal loads will be less than 100% while two PV systems will have a demand factor of 1.0.
As to who should pay for the upgrade, all you have going to protect you from bring billed for the equipment upgrade are custom and tradition. And the major precedents set were for distributed and co-generation.

If POCO had to supply 100% rated equipment in every case our power bills would all be higher.

Instead of saying that the sum total of all interconnection overcurrent devices shall not exceed the rating of "the service", would it make more sense to limit this to the 310.15(B)(16) ampacity of the service conductors, as they could be sized at only 83% of the service rating for a typical residential service.

It is very counterintuitive to me why a 200A service doesn't require 200A worth of conductor per standard ampacity calculations in section 310.15. I know the explanation is load diversity, but IMO that should already be accounted for in the algorithm that sizes service ratings.

 

[ggunn]

The point of my OP is not whether the practice of undersizing transformers is justified or not, but that it exists and it can be a gotcha if you don't check it out before installing the system. The POCO engineers may not see what is happening until too late, and they may not think it is their responsibility to catch it and tell you about it in time for it not to be a headache for you.

Forewarned is forearmed.

 

[jaggedben]

Picture this, at some point the customer requested a certain size service, the POCO decided not to really provide that capacity. They have good reasons for this, but its not the customers choice, it was not what the customer requested so to go back to that customer and put the up grade on them would only fly if you are a monopoly, you would never get away with that sort of deal as a run of the mill company.

The point of my OP is not whether the practice of undersizing transformers is justified or not, but that it exists and it can be a gotcha if you don't check it out before installing the system. The POCO engineers may not see what is happening until too late, and they may not think it is their responsibility to catch it and tell you about it in time for it not to be a headache for you.

Forewarned is forearmed.

I agree with iwire in principle. With respect to ggunn's point... I think it depends partly on the legal/regulatory environment in your state. I get the impression that in California, at least, utilities don't have the option of giving this particular reason to refuse an interconnection agreement. I could be wrong though. I think if the reason is more like 'it will adversely affect voltage or phase balance on our distribution feeder' then you can get rejected.

Instead of saying that the sum total of all interconnection overcurrent devices shall not exceed the rating of "the service", would it make more sense to limit this to the 310.15(B)(16) ampacity of the service conductors, as they could be sized at only 83% of the service rating for a typical residential service...

The rating of the 'service' could be the service conductors, the main breaker, the meter socket, service equipment busbars ahead of the main breaker(s) etc. etc., whichever is rated the lowest. So just making a rule based on the service conductor rating would be inadequate.
As for whether the 83% rule should be allowed for solar output... I think it would take a re-write of the code to settle that question. :slaphead:

 

[ggunn]

I agree with iwire in principle. With respect to ggunn's point... I think it depends partly on the legal/regulatory environment in your state. I get the impression that in California, at least, utilities don't have the option of giving this particular reason to refuse an interconnection agreement. I could be wrong though. I think if the reason is more like 'it will adversely affect voltage or phase balance on our distribution feeder' then you can get rejected.

In the case I cited the rejection was, IMO, entirely reasonable. The service OCPD and conductors at the two buildings were individually adequate for the PV systems; the problem was that the POCO had them both fed by the same transformer, which was rated at a kVA significantly less than the sum of the two service sizes. If the loads in both buildings were to be completely shut down on a cool sunny day when the PV system was running full out, the transformer would be overloaded, but just looking at the services individually one would have never seen the problem.

 

[Carultch]

The rating of the 'service' could be the service conductors, the main breaker, the meter socket, service equipment busbars ahead of the main breaker(s) etc. etc., whichever is rated the lowest. So just making a rule based on the service conductor rating would be inadequate.
As for whether the 83% rule should be allowed for solar output... I think it would take a re-write of the code to settle that question. :slaphead:

My point is that if all other service equipment is rated for 200A, and you have 200A of "service conductor" sized as service conductors with the 83% rule (2/0 Copper suppose), are you really limited to 175A worth of total interconnected solar OCPD? I know the code doesn't presently consider this issue, but is there a real life consequence to interconnecting 200A worth of solar OCPD on such a system?

 

[Smart $]

My point is that if all other service equipment is rated for 200A, and you have 200A of "service conductor" sized as service conductors with the 83% rule (2/0 Copper suppose), are you really limited to 175A worth of total interconnected solar OCPD? I know the code doesn't presently consider this issue, but is there a real life consequence to interconnecting 200A worth of solar OCPD on such a system?

I believe it is a potential issue. Probably not too many residential services with supply side connections that are equal to service rating.

 

[jaggedben]

In the case I cited the rejection was, IMO, entirely reasonable. The service OCPD and conductors at the two buildings were individually adequate for the PV systems; the problem was that the POCO had them both fed by the same transformer, which was rated at a kVA significantly less than the sum of the two service sizes. If the loads in both buildings were to be completely shut down on a cool sunny day when the PV system was running full out, the transformer would be overloaded, but just looking at the services individually one would have never seen the problem.

From an engineering standpoint the reasoning is clear. But the difference between 'your application is rejected unless you pay for a new transformer' and 'your permission to operate won't be issued until we upgrade the transfomer at our cost' is a matter of policy and regulations.

 

[ggunn]

From an engineering standpoint the reasoning is clear. But the difference between 'your application is rejected unless you pay for a new transformer' and 'your permission to operate won't be issued until we upgrade the transfomer at our cost' is a matter of policy and regulations.

You got that right, but the solar integrator is caught between the customer and the POCO on this one. Building the system before you know about the problem pretty much relieves you of any bargaining power.

But as the thread title says, it's another lesson learned. Starting now I won't stop at the service but will check things out all the way to and including the transformer when drawing up a design.

Just to be clear, I am not whining that things should be different; they are what they are. I am just passing on the lesson I learned.

 

[mivey]

It is very counterintuitive to me why a 200A service doesn't require 200A worth of conductor per standard ampacity calculations in section 310.15. I know the explanation is load diversity, but IMO that should already be accounted for in the algorithm that sizes service ratings.

Because 200 NEC amps does not equal 200 utility amps, the difference being mostly safety factor. The NEC is notoriously conservative.

The utility uses an economical conductor and equipment size method that has less safety factor than NEC and tries to get closer to metered load and can actually allow for the conductor to have short periods of overload.

 

[Carultch]

Because 200 NEC amps does not equal 200 utility amps, the difference being mostly safety factor. The NEC is notoriously conservative.

The utility uses an economical conductor and equipment size method that has less safety factor than NEC and tries to get closer to metered load and can actually allow for the conductor to have short periods of overload.

Good answer. However, service conductor sizing is governed by the NEC. Not by the utility.

I would think that everything on the customer side of the point of common coupling would be governed by the NEC and a consistent set of rules within it.

 

[Smart $]

Good answer. However, service conductor sizing is governed by the NEC. Not by the utility.

I would think that everything on the customer side of the point of common coupling would be governed by the NEC and a consistent set of rules within it.

The professional in me... :thumbsup:


The kid in me... :rotflmao:

 

[Carultch]

The professional in me... :thumbsup:


The kid in me... :rotflmao:

I like the fake meaning of the NEC initialism as the "National Exception Code". Same with FLA=Full Load Actual and FLC=Full Load Code.

 

[ggunn]

I, too, have had my share of struggles to comply with ambiguous and inconsistent language in the NEC, but I don't think anyone would seriously maintain that we would be better off without it.

 

[DanS26]

Had personal experience with this subject. Installed a 15kW PV system on a 200 amp residential system. Everything looked fine until I looked up at the pole transformer......10kva transformer. That unit is going to get warm on a bright sunny day.

Luckily the POCO swapped out for a 25kva without charge to the homeowner.

Good example of of what ggunn is talking about.

 

[macmikeman]

Why are we stopping at the supply transformer? The primaries that are run back to the substation also are limited by their cross section to carry the extra load to the transformer and your line side tap was designed after they were installed. And of course the sub station itself has to now accept a bunch of new not anticipated electrons coming from your panels and it was not designed for such either. And to continue, there are the high tension transmission lines and tower, and the generator plant also. Your bill sir is presented below, net thirty days, we appreciate your payment .

$1,000,000,000,000.00 please do not pay in copper pennies.

 

[Carultch]

Why are we stopping at the supply transformer? The primaries that are run back to the substation also are limited by their cross section to carry the extra load to the transformer and your line side tap was designed after they were installed. And of course the sub station itself has to now accept a bunch of new not anticipated electrons coming from your panels and it was not designed for such either. And to continue, there are the high tension transmission lines and tower, and the generator plant also. Your bill sir is presented below, net thirty days, we appreciate your payment .

$1,000,000,000,000.00 please do not pay in copper pennies.

I think pure copper pennies would be worth more as scrap metal than face value.