Transformer and OCPD Sizing for PV

[pvgreeze]

Hi everyone,

I'm sure this question has been asked many times before, but I'm struggling to find resources & explanations for OCPD sizing and coordinating with transformers for PV arrays. I'm familiar with standard transformer OCPD sizing as per NEC Article 450, but I'm confused because NEC Article 450 requires sizing OCPDs for the transformer rated current. Now, this makes perfect sense if the transformer is feeding a load, but as a PV source, wouldn't it make the most sense to size the OCPDs for the rated output of the inverters?

I ask because numerous designs that I have started working on require unconventional transformer sizes (mostly between 45kVA - 75kVA). Instead of requiring custom transformer sizes, would I be able to specify a 75kVA transformer for (as an example) 50kVA of connected PV, and size the OCPD accordingly (ignoring inrush)?

I'm also quite confused by NEC 690.9(D) (for the 2017 NEC) and the requirement that both sides of the transformer be considered the 'primary' as opposed to the PV side and how that relates to OCPD sizing. Does this just exist to say that both sides of the transformer need OCPD protection, and that both sides have a 250% maximum rating?

I know this a lot, and probably a common topic of discussion here, but I would appreciate any help & feedback or explicit code guidance for coordinating PV output with transformer sizing and transformer OCPD sizing.

Thanks!
-pvgreeze

 

[jaggedben]

I'm really not experienced with transformers but from code and safety point of view, I think the general idea is that you have two sets of requirements and your components have to meet both.

Hi everyone,

I'm sure this question has been asked many times before, but I'm struggling to find resources & explanations for OCPD sizing and coordinating with transformers for PV arrays. I'm familiar with standard transformer OCPD sizing as per NEC Article 450, but I'm confused because NEC Article 450 requires sizing OCPDs for the transformer rated current. Now, this makes perfect sense if the transformer is feeding a load, but as a PV source, wouldn't it make the most sense to size the OCPDs for the rated output of the inverters?

Simply put, I think it's both.

I ask because numerous designs that I have started working on require unconventional transformer sizes (mostly between 45kVA - 75kVA). Instead of requiring custom transformer sizes, would I be able to specify a 75kVA transformer for (as an example) 50kVA of connected PV, and size the OCPD accordingly (ignoring inrush)?

Can you explain what problem isn't solved by just upsizing the transformer to the next commonly available size? I'd think that'd be cheaper than ordering custom transformers.

I'm also quite confused by NEC 690.9(D) (for the 2017 NEC) and the requirement that both sides of the transformer be considered the 'primary' as opposed to the PV side and how that relates to OCPD sizing. Does this just exist to say that both sides of the transformer need OCPD protection, and that both sides have a 250% maximum rating?

Regarding OCPD on both sides, see the exception to 690.9(D) Regarding rating, if I read 450.3 correctly it's a 125% maximum rating, since that's the lesser of the two minimums.

Also...
Don't expect 'primary' and 'secondary' to be understood when there are sources on both sides. But that said, my understanding is that you should generally still regarding the utility as the primary where that is specified by manufacturers. Because the utility is the side that energizes the transformer and also provides greater fault current.

I know this a lot, and probably a common topic of discussion here, but I would appreciate any help & feedback or explicit code guidance for coordinating PV output with transformer sizing and transformer OCPD sizing.

Thanks!
-pvgreeze

 

[pvgreeze]

Can you explain what problem isn't solved by just upsizing the transformer to the next commonly available size? I'd think that'd be cheaper than ordering custom transformers.

Thank you for looking through and your reply...I think my issue comes from needing to upsize the OCPDs and conductors for a larger transformer than the expected inverter output. For example, using 50kVA of inverter output, if we have a 50kVA 480-208 transformer, I'd only need to size the OCPDs and conductors for the current rating of both sides of that transformer (60.14A - 138.79A) versus having to size the OCPDs and conductors for the current rating of the transformer of a 75kVA (90.21A - 208.18A). So, my problem really is sizing the OCPDs based off of the current rating of the transformer versus sizing the OCPDs off the current rating of the inverter output. Does that make any sense?

 

[augie47]

IMO, you would need to choose between the lowest of the two numbers.
If you are required to have a breaker rated say 50 amps for inverter protection and the transformer 450.3
protection is 75 amps, you would need the 50 amp breaker and vice-a-versa.
I do see the possibility of an "inrush" problem if you have to upsize the transformer.

 

[pvgreeze]

IMO, you would need to choose between the lowest of the two numbers.
If you are required to have a breaker rated say 50 amps for inverter protection and the transformer 450.3
protection is 75 amps, you would need the 50 amp breaker and vice-a-versa.
I do see the possibility of an "inrush" problem if you have to upsize the transformer.

Yeah, that is what I would like to do! I just wish there was clearer language in the code allowing this type of configuration to be acceptable. I totally understand needing to size the OCPD for the transformer full load current in a traditional transformer application, but with solar production and inverters, we wouldn't anticipate exceeding the current rating of the inverters, so it would be pointless to size 50kVA output of inverters like they could "potentially" produce 75kVA. I just wish this logic was codified and not left ambiguous in the NEC.

 

[winnie]

Don't think of 'primary' and 'secondary' in terms of direction of power flow. Think instead of 'which side gets energized to set system voltage'.

Transformers by their nature are bi-directional. Power can flow in either direction. But when the transformer is initially energized, it will be from one side or the other. Most commonly in a PV setup, the _utility_ side of the transformer gets energized and sets the system voltage, then the line interactive PV inverter senses the transformed voltage and adjusts its output voltage to match.

Thus the utility side is effectively the primary, where you have to deal with inrush current and voltage adjustment taps, but power flows from the secondary side to the primary side.

NEC rules for transformers set _maximum_ values for required OCPD. You are permitted to use smaller values.

If the 'primary' OCPD is too small, then it will likely trip because of transformer inrush current. So there is a practical lower limit to the size of the primary OCPD.

If the 'secondary' OCPD is smaller than the rating of the transformer, then you are essentially leaving a bit of transformer capacity 'on the table', but this will not cause any electrical problem.

My understanding of the 'treat both sides as the primary' is as follows: in normal applications you are permitted to greatly over-rate the primary breaker to permit the transformer to reliably energize without tripping the breaker. However you must then protect the secondary coils from overload, thus the limit on the secondary breaker rating. On top of this transformers can pass power between phases, so for most transformer types if your primary were protected at 125% of rating (or less) the secondary would still not be protected from overload of an individual phase. In the case of a PV transformer it would be possible for the PV system to supply rated power on the 'secondary' side yet one of the primary phases to be overloaded. Thus as a practical matter for PV transformers the 'utility side' OCPD (the 'primary') is limited to 125% of the transformer rating.

In your case with say a 50 kVA PV system, I believe you could reasonably have a 75kVA transformer, 'utility side' OCPD sized at 125% of the transformer 'primary' rating, then 'PV side' OCPD sized for the 50 kVA system.

NB: This is not my field, just my understanding from casual reading.

-Jon
'PV'

 

[pvgreeze]

In your case with say a 50 kVA PV system, I believe you could reasonably have a 75kVA transformer, 'utility side' OCPD sized at 125% of the transformer 'primary' rating, then 'PV side' OCPD sized for the 50 kVA system.

Thank you so much for your detailed response!! I like the idea of a configuration like this, and it definitely makes sense to me. Hopefully it is an acceptable configuration to any electrical inspector as well, because it would be a pain to have to constantly custom order transformers and equally as annoying to oversize the PV side OCPD and conductor & conduit just because the inverter output does not align with a typical transformer size. Thanks again!!

 

[jaggedben]

Thank you for looking through and your reply...I think my issue comes from needing to upsize the OCPDs and conductors for a larger transformer than the expected inverter output. For example, using 50kVA of inverter output, if we have a 50kVA 480-208 transformer, I'd only need to size the OCPDs and conductors for the current rating of both sides of that transformer (60.14A - 138.79A) versus having to size the OCPDs and conductors for the current rating of the transformer of a 75kVA (90.21A - 208.18A). So, my problem really is sizing the OCPDs based off of the current rating of the transformer versus sizing the OCPDs off the current rating of the inverter output. Does that make any sense?

No, it doesn't make sense, at least not to me. 450.3 states maximum limits, not minimum requirements, for OCPDs. So upsizing the transformer does not require you to upsize OCPDs and conductors.

 

[pvgreeze]

No, it doesn't make sense, at least not to me. 450.3 states maximum limits, not minimum requirements, for OCPDs. So upsizing the transformer does not require you to upsize OCPDs and conductors.

I think a lot of my confusion also came from which side was primary and which side was secondary, and also thinking of the coordination as different than a typical transformer application where it is serving a load, whereas in this case it is supplying power on both sides (hence 690.9(D) being a bit unclear to me).

I think I have an idea on a base idea (stated above) to size the 'secondary side' (inverter side) off the inverter output and the 'primary side' (utility side) off of the FLA of the transformer.

 

[ggunn]

Thank you for looking through and your reply...I think my issue comes from needing to upsize the OCPDs and conductors for a larger transformer than the expected inverter output. For example, using 50kVA of inverter output, if we have a 50kVA 480-208 transformer, I'd only need to size the OCPDs and conductors for the current rating of both sides of that transformer (60.14A - 138.79A) versus having to size the OCPDs and conductors for the current rating of the transformer of a 75kVA (90.21A - 208.18A). So, my problem really is sizing the OCPDs based off of the current rating of the transformer versus sizing the OCPDs off the current rating of the inverter output. Does that make any sense?

I don't think so. Your OCPD defines the maximum current through the transformer, not the other way round. No one is going to fault you for overprotecting a transformer.

Can you explain what problem isn't solved by just upsizing the transformer to the next commonly available size? I'd think that'd be cheaper than ordering custom transformers.

And that's not to mention lead time.

 

[winnie]

There are two potential problems that I could see from using a larger transformer than necessary:
1) The system efficiency might not be ideal. (IMHO this will not be a significant issue unless the transformer is wildly oversized.)
2) Transformer inrush current sets a minimum bound for the _primary_ OCPD, and eventually this will cause the transformer protection to be larger than suitable for the load. (IMHO this will not be a significant issue unless the transformer is wildly oversized.)

As a practical matter upsizing to the next commonly available size shouldn't cause any problems at all.

-Jon

 

[wwhitney]

2) Transformer inrush current sets a minimum bound for the _primary_ OCPD, and eventually this will cause the transformer protection to be larger than suitable for the load. (IMHO this will not be a significant issue unless the transformer is wildly oversized.)

I don't understand this concern? Seems like you can just:

1) Determine the minimum utility side OCPD required for the PV supply (and any loads).
2) Determine the minimum utility side OCPD required for transformer inrush.
3) Pick the larger of 1 and 2, and size the utility side conductors appropriately.
4) Size the immediate secondary conductors in accordance with the usual rules
5) Quickly hit secondary side OCPD sized only for the PV supply (and any loads)
6) Size conductors to the PV system based on 5.

Cheers, Wayne

 

[winnie]

I don't understand this concern? Seems like you can just:

1) Determine the minimum utility side OCPD required for the PV supply (and any loads).
2) Determine the minimum utility side OCPD required for transformer inrush.
3) Pick the larger of 1 and 2, and size the utility side conductors appropriately.
4) Size the immediate secondary conductors in accordance with the usual rules
5) Quickly hit secondary side OCPD sized only for the PV supply (and any loads)
6) Size conductors to the PV system based on 5.

Cheers, Wayne

You are correct. For some reason I was thinking that it might be a problem if the primary OCPD was larger than the PV system really needed.

-Jon

 

[Zee]

Dealing with same issue. Good to read this.

 

[Zee]

kinda sums it up:

StackPath

 

[Zee]

How would one determine if the fuses intended for use on the primary (utility) side can handle the inrush current? Is it just assumed?
I don't know much about xfrmrs.

 

[pv_n00b]

How would one determine if the fuses intended for use on the primary (utility) side can handle the inrush current? Is it just assumed?
I don't know much about xfrmrs.

You can do a power study with something like Etap or SKM and it will tell you and tell you if you are protecting the transformer damage curves. If you use 135% or so of the transformer FLA you will probably be good if you have to wing it. Some people will go to the maximum OCPD allowed in NEC 450 to avoid the inrush but that really leaves the transformer with less protection than it should have.

 

[winnie]

I'd never thought about this before, but are there breakers with trip characteristics specifically intended for transformer protection?

I'm thinking something like a manual motor starter with overload heaters that doesn't trip on a very high transient but does trip on an extended overload.

Thanks
Jon

 

[pv_n00b]

I don't understand this concern? Seems like you can just:

1) Determine the minimum utility side OCPD required for the PV supply (and any loads).
2) Determine the minimum utility side OCPD required for transformer inrush.
3) Pick the larger of 1 and 2, and size the utility side conductors appropriately.
4) Size the immediate secondary conductors in accordance with the usual rules
5) Quickly hit secondary side OCPD sized only for the PV supply (and any loads)
6) Size conductors to the PV system based on 5.

Cheers, Wayne

In step 5 remember that if the primary OCPD is more than 125% of the transformer FLA then the secondary OCPD has to be 125% or less of the transformer OCPD.

 

[pv_n00b]

I'd never thought about this before, but are there breakers with trip characteristics specifically intended for transformer protection?

I'm thinking something like a manual motor starter with overload heaters that doesn't trip on a very high transient but does trip on an extended overload.

Thanks
Jon

There are products that are designed to reduce inrush current. They monitor the sinewave and only close at a point that minimizes inrush. Kind of expensive and usually used where we are energizing multiple large MV transformers in parallel.
CBs with digital trip units can be set with a high degree of accuracy to avoid the inrush point while still providing great protection, still pretty expensive though.
Not all fuses of the same rating have the same TCC. Sometimes we can find that one fuse with a more convoluted TCC than others that dances around the inrush point.
Transformer manufacturers can also produce custom transformers designed for lower inrush than is typical. It's never been clear to me how this might impact other aspects of the transformer design, like efficiency.