Matching Transformer Primary Conductors to Primary OCPD

[jim dungar]

The NEC says transformers connected into a multi-phase "bank" are treated as a single transformer not as individual ones.

 

[jaggedben]

The NEC says transformers connected into a multi-phase "bank" are treated as a single transformer not as individual ones.

That cements this point.

... it does seem wrong to me that if I connect a single phase 4:1 coil (no center tap) that I need a bigger conductor size than if I connect three identical coils to the same primary voltage via a common neutral point.

 

[david luchini]

Notice I never specified the transformer size, as it's immaterial to this question. We can make it 25 kVA if you like.

Yes, I noticed. And it IS material to the question. You keep coming up with with odd-ball scenarios like "240V : 24V/240V single phase transformers" to try to explain your point. Now it's a 25kVA, 480-120V single phase transformer with a rated secondary current of 208A, even though you started out needing "100A @ 120V."

If you have to keep coming up with crazy scenarios instead of practical applications in order to get your point across, that might just be a hint that your point is wrong.

 

[jaggedben]

Yes, I noticed. And it IS material to the question. You keep coming up with with odd-ball scenarios like "240V : 24V/240V single phase transformers" to try to explain your point. Now it's a 25kVA, 480-120V single phase transformer with a rated secondary current of 208A, even though you started out needing "100A @ 120V."

If you have to keep coming up with crazy scenarios instead of practical applications in order to get your point across, that might just be a hint that your point is wrong.

How about connecting more than two or three sets of secondary conductors to the same transformer to serve separate loads? Is that ever done? (Is it not allowed?)

Actually we know it's been done. (Okay, with sources, not loads. And moot in that particular case because Yg-Yg. So it is ever done with delta-wye?)

 

[david luchini]

We're discussing the minimum ampacity requirements for secondary conductors. That's all.

And the point is the minimum ampacity requirements should be the same for one 10kVA, 480-120V single phase transformer (83A rated secondary current), and three 10kVA 480-120V single phase transformer connect together D-Y (also 83A rated secondary current.)

Make a comparison if you connect both the single phase and three phase transformer in the same manner, primary and secondary protected per T450.3, with the maximum allowed primary OCPD. Calculate the required minimum ampacity of the secondary conductors for each per 240.21(C)(6).

You will find that both transformers, which both have the same rated secondary current, also have the same required minimum secondary conductor ampacity (within some rounding.) And wouldn't you expect them to?

How about connecting more than two or three sets of secondary conductors to the same transformer to serve separate loads? Is that ever done? (Is it not allowed?)

Yes, it is done, and yes, you would follow the rules in the same way.

 

[wwhitney]

And the point is the minimum ampacity requirements should be the same for one 10kVA, 480-120V single phase transformer (83A rated secondary current), and three 10kVA 480-120V single phase transformer connect together D-Y (also 83A rated secondary current.)

OK, this is at least on the face of it a plausible idea and is the most constructive thing you've said so far in response to the issue I've raised. To work out the math for those following along:

One 10 kVA, 480V : 120V single phase transformer has a rated primary current of 20.8A and a rated secondary current of 83.3A. So the primary can be protected at 250% * 20.8A = 52.1A, round up to 60A OCPD (Table 450.3(A), Note 1), if the secondary is protected at 125%* 83.3A = 104.2A, round up to 110A OCPD maximum. With a 60A primary OCPD, the minimum 240.21(C)(6) secondary conductor ampacity is 60A * (480V/120V) / 3 = 80A. So if you needed, say, a 50A 120V circuit and a 60A 120V circuit, you could run two sets of secondary conductors up to 25' to side-by-side panelboards, one protected at 50A and one at 60A (Table 450.3(A), Note 2), but you'd need to use 80A conductors for each set.

A 30 kVA, 480D : 208Y/120V 3 phase transformer (or 3 of the above single phase transformers in a D-Y configuration) has a rated primary current of 36.1A and a rated secondary current of 83.3A. So the primary can be protected at 250% * 36.1A = 90.2A, I'll say we have to round to 90A before rounding up, which puts the maximum primary OCPD at 90A, provided again our secondary OCPD is no more than 125% * 83.3A = 104.2A, round up to 110A OCPD maximum. Then with a 90A primary OCPD, if we say the voltage ratio is 480/208, the minimum 240.21(C)(6) secondary conductor ampacity is 90A * (480V/208V) / 3 = 69.3A. While if we say the voltage ratio is 480/120, the minimum secondary conductor ampacity is 90A * (480V/120V) / 3 = 120A.

[If we had fractional size OCPDs, none of this rounding up to the next standard size, then in the 3 phase example we'd find the maximum primary OCPD to be exactly sqrt(3) times the single phase example, and the minimum secondary conductor ampacity would match the single phase case if we use 480/208, and it would be sqrt(3) times larger if we use 480/120.]

Will respond further as working out this example makes a long enough post.

Cheers, Wayne

 

[wwhitney]

Yes, it is done, and yes, you would follow the rules in the same way.

Hence my example in post #54 is practical, not oddball. The only oddball example I had I labeled as such (absurd was my term).

And the point is the minimum ampacity requirements should be the same for one 10kVA, 480-120V single phase transformer (83A rated secondary current), and three 10kVA 480-120V single phase transformer connect together D-Y (also 83A rated secondary current.)

Why do you think that we should be allowed to have a larger primary OCPD, without increasing the minimum secondary conductor size?

The minimum tap conductor size is always proportional to primary OCPD size. So when the primary OCPD size increases by a factor of sqrt(3), the minimum secondary conductor size should also increase by a factor of sqrt(3).

My expectation is obviously that the tap rules in a given length category provide equivalent partial protection of the tap conductors in all cases.

Cheers, Wayne

 

[jaggedben]

...

One 10 kVA, 480V : 120V single phase transformer has a rated primary current of 20.8A and...

A 30 kVA, 480D : 208Y/120V 3 phase transformer (or 3 of the above single phase transformers in a D-Y configuration) ...

Correct me if I'm wrong but wouldn't the second transformer be rated 17.3kVA with the same current ratings?

 

[jaggedben]

...

Make a comparison if you connect both the single phase and three phase transformer in the same manner, primary and secondary protected per T450.3, with the maximum allowed primary OCPD. Calculate the required minimum ampacity of the secondary conductors for each per 240.21(C)(6).

You will find that both transformers, which both have the same rated secondary current, also have the same required minimum secondary conductor ampacity (within some rounding.) And wouldn't you expect them to?

Again correct me if I'm wrong, but aren't the current ratings and allowable primary OCPD ratings the same? Let's just say the OCPD rating is 100A.

So in the first example, I take the 'voltage ratio' of (480/120=4) times 1/3rd of 100A and get ampacity of 133A.

And in the second example, you say I take the 'voltage ratio' of (480/208=2.3) times 1/3rd of 100A and get an ampacity of 77A.

Sorry, I'm getting a different minimum ampacity, unless I misunderstood what you're saying is the typical interpretation of 240.21(C)(6).

I don't think we've made any progress in the discussion here.

 

[wwhitney]

Correct me if I'm wrong but wouldn't the second transformer be rated 17.3kVA with the same current ratings?

OK, you're wrong. : - ) In that you can't have the same current ratings on both primary and secondary for the single phase case and for the delta-wye case. [You could in a wye-wye with the same transformer as in the single phase case.]

We established that a 10 kVA single phase 480V:120V transformer has a rated primary current of 10,000VA/480V = 20.8A and a rated secondary current of 10,000/120V = 83.3A.

Now if you put 3 of these together in a delta-wye configuration, what happens? On the secondary side, each line conductor is only connected to one transformer, so the rated secondary current is still 83.3A per conductor. And because of the way the current adds on the neutral conductor with the 120 degree phase shifts, the neutral only sees at most 83.3A for typical phase shifts.

On the primary side, each delta line conductor is now attached to two transformers. At rated current, it will see 20.8A from each of the two, but if the loading is balanced those coil currents will be 60 degrees out of phase (A to B is 120 degrees from B to C, but if you are combining currents in B, you need to reverse one of those, which adds a minus sign or 180 degree shift). So the two currents of 20.8A add to 20.8 * sqrt(3) = 36.1A.

With the above, the 3 phase arrangement is rated 3 * 10 kVA as you would expect and as you can check on either the primary or secondary side. The secondary rated current is unchanged, but the primary rated current goes up by a factor of sqrt(3).

Or if you want to substitute a 3 phase transformer with the same primary rated current of 20.8A, then yes it would be rated only 17.3 kVA, a factor of sqrt(3) smaller, and the secondary current would also be a factor of sqrt(3) smaller.

Cheers, Wayne

 

[jaggedben]

Sorry, nevermind my last two posts, I see my mistake.

Make a comparison if you connect both the single phase and three phase transformer in the same manner, primary and secondary protected per T450.3, with the maximum allowed primary OCPD. Calculate the required minimum ampacity of the secondary conductors for each per 240.21(C)(6).

You will find that both transformers, which both have the same rated secondary current, also have the same required minimum secondary conductor ampacity (within some rounding.) And wouldn't you expect them to?

So if they gave me a single 480/120 coil with a primary current rating of 20A then I could protect it with a 100A OCPD and my minimum secondary conductor ampacity would be 66.7A.

If I have 3 such coils in a delta-wye configuration, then primary current rating would be 34.6 (if they didn't round off) and my max ocpd would be 86.5 (again I'll ignore the shift to standard rating) and after multiplying by 480 and dividing by 208 and 3 I get the same answer I get 66.5 which is only different due to rounding off the nominal voltage to 208.

I see your point now. Thank you.

 

[jaggedben]

Can't believe you beat me to it by three seconds Wayne.

(What tripped me up, by the way, was not realizing that they actually regulate the protection of each coil by the secondary current rating rather than the primary rating.)

 

[wwhitney]

On the primary side, each delta line conductor is now attached to two transformers.

BTW, this is the source of the issue in this discussion, and the reason for the limitation in 240.4(F).

If we supplied each of our 3 (single phase) transformers with its own 2-wire supply, so 6 conductors in total, then (a) it would be extra obvious that the primary to secondary voltage ratio is 4:1, and (b) the limitation in 240.4(F) would not apply (at least physics-wise), we could use the primary OCPD to protect the secondary conductors, as long as we multiply the primary OCPD by 4:1.

Cheers, Wayne

 

[wwhitney]

Can't believe you beat me to it by three seconds Wayne.

Took me a lot longer to write that up than 3 seconds.

(What tripped me up, by the way, was not realizing that they actually regulate the protection of each coil by the secondary current rating rather than the primary rating.)

Not following what you mean by that?

Cheers, Wayne

 

[wwhitney]

The NEC says transformers connected into a multi-phase "bank" are treated as a single transformer not as individual ones.

I see that language in 450.3 "Overcurrent Protection," but it is restricted to "as used in this section." Is there other language elsewhere that would make that true for purposes beyond 450.3?

Thanks,
Wayne

 

[ramsy]

I don't think we've made any progress in the discussion here.

Industry standards are being disputed between a mathematician and an engineer.

The engineer is more comfortable with tap-rule convention enforced by AHJ's per industry practice, not re-writing those rules.

The mathematician is more comfortable testing theory & physics of any rule, before accepting adequacy, not just following convention.

Most contractors & EE's subject to adopted standards are not allowed to make up their own rules, nor comfortable debating math proofs.

 

[david luchini]

OK, this is at least on the face of it a plausible idea and is the most constructive thing you've said so far in response to the issue I've raised. To work out the math for those following along:

One 10 kVA, 480V : 120V single phase transformer has a rated primary current of 20.8A and a rated secondary current of 83.3A. So the primary can be protected at 250% * 20.8A = 52.1A, round up to 60A OCPD (Table 450.3(A), Note 1), if the secondary is protected at 125%* 83.3A = 104.2A, round up to 110A OCPD maximum. With a 60A primary OCPD, the minimum 240.21(C)(6) secondary conductor ampacity is 60A * (480V/120V) / 3 = 80A. So if you needed, say, a 50A 120V circuit and a 60A 120V circuit, you could run two sets of secondary conductors up to 25' to side-by-side panelboards, one protected at 50A and one at 60A (Table 450.3(A), Note 2), but you'd need to use 80A conductors for each set.

The highlighted is wrong. Therefore the math is wrong. Perhaps this is where you are going off the rails.

A 30 kVA, 480D : 208Y/120V 3 phase transformer (or 3 of the above single phase transformers in a D-Y configuration) has a rated primary current of 36.1A and a rated secondary current of 83.3A. So the primary can be protected at 250% * 36.1A = 90.2A, I'll say we have to round to 90A before rounding up, which puts the maximum primary OCPD at 90A, provided again our secondary OCPD is no more than 125% * 83.3A = 104.2A, round up to 110A OCPD maximum. Then with a 90A primary OCPD, if we say the voltage ratio is 480/208, the minimum 240.21(C)(6) secondary conductor ampacity is 90A * (480V/208V) / 3 = 69.3A. While if we say the voltage ratio is 480/120, the minimum secondary conductor ampacity is 90A * (480V/120V) / 3 = 120A.

And this is what is ridiculous. If you install a 30kVA, 480-208/120V transformer to supply a 100A MCB panel on the secondary side, the Code isn't telling you that you need to run a 120A conductor.

 

[wwhitney]

The highlighted is wrong.

Ah, thank you. I was looking at Table 450.3(A) instead of 450.3(B), and in 450.3(B), Note 1 on rounding up only applies some of the time, and not to the primary OCPD with using primary + secondary protection.

Therefore the math is wrong. Perhaps this is where you are going off the rails.

As a difference in rounding, the change is minor and is unrelated to what we are debating. The correction:

"So the primary can be protected at 250% * 20.8A = 52.1A, round down to 50A OCPD (Table 450.3(B), Note 1 does not apply), if the secondary is protected at 125%* 83.3A = 104.2A, round up to 110A OCPD maximum (Note 1 does apply). With a 50A primary OCPD, the minimum 240.21(C)(6) secondary conductor ampacity is 50A * (480V/120V) / 3 = 66.7A. So if you needed, say, a 50A 120V circuit and a 60A 120V circuit, you could run two sets of secondary conductors up to 25' to side-by-side panelboards, one protected at 50A and one at 60A (Table 450.3(A), Note 2), but you'd need to use 67A conductors for each set."

Cheers, Wayne

 

[david luchini]

"So the primary can be protected at 250% * 20.8A = 52.1A, round down to 50A OCPD (Table 450.3(B), Note 1 does not apply), if the secondary is protected at 125%* 83.3A = 104.2A, round up to 110A OCPD maximum (Note 1 does apply). With a 50A primary OCPD, the minimum 240.21(C)(6) secondary conductor ampacity is 50A * (480V/120V) / 3 = 66.7A. So if you needed, say, a 50A 120V circuit and a 60A 120V circuit, you could run two sets of secondary conductors up to 25' to side-by-side panelboards, one protected at 50A and one at 60A (Table 450.3(A), Note 2), but you'd need to use 67A conductors for each set."

Or you could do something practical and run the secondary to a 100A mcb panelboard.

And strangely enough, the (3) single phase transformers connected together in a delta-wye, with the same secondary rated current, will have (approximately) the same minimum required conductor ampacity of 67A. Imagine that.

 

[wwhitney]

And this is what is ridiculous. If you install a 30kVA, 480-208/120V transformer to supply a 100A MCB panel on the secondary side, the Code isn't telling you that you need to run a 120A conductor.

Why shouldn't it if your primary OCPD is at 250%? If you don't want to have to oversize the secondary conductors, don't oversize the primary OCPD, which is the only thing protecting those secondary conductors from short circuit/ground fault. Or stick with the 10' rule, 240.21(C)(2), which allows a 10:1 ratio instead of a 3:1 ratio.

I think this question from post #67 sums up our difference of opinion quite simply: "Why do you think that we should be allowed to have a larger primary OCPD, without increasing the minimum secondary conductor size?"

Cheers,
Wayne