Newbie simple (yea right) question for 3-phase

Status
Not open for further replies.
Smart $ said:
And this last part is what is in question. Under 2014 Code, what size would you use? Please elaborate on your calculation of the minimum required ampacity.
Click to expand...

You tell me what you think is wrong with doing it that way.
 
Smart $ said:
Nothing... just looking for particulars in how you'd calculate minimum ampacity after combined...
Click to expand...
Well, if I had two 6kW inverters and one 7kW inverter configured as I said, I would consider it a 21kW inverter for the purposes of the combined output. If they were on a 208V three phase system...

21kW/(208V)(sqrt 3) = 58.2A, so 60A OCPD and conductors sized accordingly.
 
That would give you a conservative value for the actual current, but will underestimate the current number that the AHJ may insist you use for 120% rule calculations.
 
Last edited:
GoldDigger said:
That would give you a conservative value for the actual current, but will underestimate the current number that the AHJ may insist you use for 120% rule calculations.
Click to expand...

In the MDP, do you mean? How so? If it had actually been a three phase 21kW inverter instead of three single phase inverters totaling 19kW it seems to me that the current would have been more on every phase.
 
What I am saying is that the line current derived from the three phase power formula will match the actual measured current. But the sum of the magnitudes of the line to line currents will be greater by a factor of 2/sqrt(3) than what you calculate.
And an inspector may well want to see you use that value, in the same way that the breaker ratings are just added.
 
Last edited:
ggunn said:
Well, if I had two 6kW inverters and one 7kW inverter configured as I said, I would consider it a 21kW inverter for the purposes of the combined output. If they were on a 208V three phase system...

21kW/(208V)(sqrt 3) = 58.2A, so 60A OCPD and conductors sized accordingly.
Click to expand...
What about the 125%?

Under 2014 and using your example...
125%*7kW/208V=42A... so 45A inverter breaker at combiner.
125%*7kW*3/(208*sqrt3)=73A... so 80A minimum PV System OCPD and bus/conductor​
Anything upstream of that is dependent on utility side OCPD, bus, conductor.
 
GoldDigger said:
What I am saying is that the line current derived from the three phase power formula will match the actual measured current. But the sum of the magnitudes of the line to line currents will be greater by a factor of 2/sqrt(3) than what you calculate.
And an inspector may well want to see you use that value, in the same way that the breaker ratings are just added.
Click to expand...
If you and others believe the combiner bus and ocpd has to be sized using the sum of 125% times inverter output rating, with no consideration for 3? reduction of combined current levels, I'll have to review the Code and perhaps submit a proposal to change that for 3? systems.
 
Smart $ said:
If you and others believe the combiner bus and ocpd has to be sized using the sum of 125% times inverter output rating, with no consideration for 3? reduction of combined current levels, I'll have to review the Code and perhaps submit a proposal to change that for 3? systems.
Click to expand...
Please let us know the results of your review.
What I am going with is that the pre-2014 language essentially said that you add the numeric breaker ratings, without phase angle considerations. And 2014 essentially tells you to do the same thing but using 1.25 times the rated output instead of the breaker rating.
It is at best unclear, IMHO.
 
GoldDigger said:
Please let us know the results of your review.
What I am going with is that the pre-2014 language essentially said that you add the numeric breaker ratings, without phase angle considerations. And 2014 essentially tells you to do the same thing but using 1.25 times the rated output instead of the breaker rating.
It is at best unclear, IMHO.
Click to expand...
Initial review is your assessment is correct. However, in pre-2014 editions, Code stated we had to use the sum of breaker ratings. There is no way to vector sum breaker ratings. In the 2014 edition, it changed to "the sum of 125 percent of the inverter(s) output circuit current". It does not state whether we must use the arithmetic sum or permitted to use the vector sum... leaving it open to interpretation... and of course, I interpret it as allowing the vector sum.

One way to clear it up is to add an exception to 705.12(D)(2) which in effect says we can use the vector sum for 3? systems. Thoughts?
 
At worst the CMP will either turn it down saying it would be wrong or turn it down saying it is obvious already. :)
Or if they are having a particularly good day, both! :angel:
 
Smart $ said:
If you and others believe the combiner bus and ocpd has to be sized using the sum of 125% times inverter output rating, with no consideration for 3? reduction of combined current levels, I'll have to review the Code and perhaps submit a proposal to change that for 3? systems.
Click to expand...

I'd be inclined to support that... But rather than view it as a 'change for 3-phase systems', it might be better to define the method of calculation for any portion of an inverter output circuit current that has more than one inverter connected. Currently the code lacks any language that directly address combined inverter outputs, single phase or three phase.

See 705.60(A)(2).
 
I'm trying to wrap my head around what some of you guys are saying, so let me ask some questions...

First of all, consider the case of three identical 7kW single phase inverters connected phase to phase around a three phase system in a dedicated AC combiner panel. Considering the points you bring up, would the output of that AC combiner be treated any differently with respect to conductors and OCPD at the interconnect than would be the output of a single three phase 21kW inverter?

Would it make any difference if the single phase inverters were connected phase to neutral rather than phase to phase (full sized neutral, of course)?

Now make one or two of the inverters 6kW. If I were to treat the output of the AC combiner the same as if it were three 7kW inverters wrt OCPD and conductor size, could there be any downside other than oversizing some of the conductors and/or OCPD, given that I would never underprotect conductors from fault current coming from the service?

And if the unbalanced system were to be interconnected through a three pole backfed breaker in the MDP, wouldn't this be the only way to safely do it?
 
Let's table for the moment the question of unequal ratings. I have no disagreement with you on that.
If each inverter is connected as one leg of a wye, then the output current will be exactly equal to its contribution to the line current and to the bus current.
If one inverter is connected line to line, its output current will appear in both of the line conductors, but in each of them the current will be out of phase with the line to neutral voltage.
When you connect one inverter C to A and one inverter A to B, the current on A will be the inverter rated current times 1.73, not times 2, and that sum current will be in phase with the line to neutral voltage on A.
All we have been saying is that, for three balanced inverters, the 3-phase line current calculated from the rated power will not be equal to the sum of the rated output currents (magnitudes) associated with each line conductor.
And that this difference works against you when applying the 120% rule as currently worded.
 
jaggedben said:
I'd be inclined to support that... But rather than view it as a 'change for 3-phase systems', it might be better to define the method of calculation for any portion of an inverter output circuit current that has more than one inverter connected. Currently the code lacks any language that directly address combined inverter outputs, single phase or three phase.

See 705.60(A)(2).
Click to expand...
The methods of calculation are already conventional. It's just that previous requirements make the 2014 use of the word sum ambiguous.

Here's what I currently have submitted:
Exception to (D)(2): Where single phase inverters are connected to polyphase systems, rating determination shall be permitted to be calculated using the vector sum of 125 percent of the inverters' output circuit current.​

Statement of Problem and Substantiation for Public Input

Editions prior to 2014 specified using the sum of overcurrent protection device ratings. The 2014 edition changed that to using the sum of 125 percent of the inverter(s) output circuit current. Overcurrent protection device ratings are magnitude only. So prior to the 2014 edition, the arithmetic sum was the only option. However, inverter output circuit currents have both magnitude and phase, where single phase inverters are connected to polyphase systems. One of the advantages of polyphase systems is they can deliver the same amount of power as single phase using smaller conductors. Seems reasonable that photovoltaic systems should also utilize this advantage... and using the arithmetic sum prohibits this advantage


Many electrical professionals, including those representing the authority having jurisdiction (AHJ), are uncertain if the 2014 edition permits using the vector sum (e.g. use 3? calculation methods where single-phase inverters are connected to 3? systems). Of the professionals that realize using the vector sum is a possibility, many resort to using the arithmetic sum because it is the safer approach to this uncertainty. This also creates the potential for a system designer or installer to use the vector sum in his rating determination and then having his plan or installation rejected by the AHJ.


An example is three 40A 208V 1? inverters connected in a delta configuration to a 208Y/120V 3? 4W system. The minimum overcurrent protection device (OCPD) for each inverter is 40A ? 125% = 50A. Prior to 2014 edition, after outputs were combined, provisions required using the sum of the two OCPD's connected to the same bus/conductor (100A). However, the rated output after combined is only 40A ? ?3 = 69A. The bus/conductor ampacity rating could be determined 125% ? 40A ? ?3 = 87A (the conventional equivalent calculation of the vector sum of 125 percent of the inverters output circuit current).


If the Code-Making Panel believes the concept to be obvious without revision, perhaps consider changing the proposed exception to an informational note.
Click to expand...

Comments?

EDIT: Note the "?" before a "3" is a square root character in the actual substantiation.
 
Last edited:
GoldDigger said:
Let's table for the moment the question of unequal ratings. I have no disagreement with you on that.
If each inverter is connected as one leg of a wye, then the output current will be exactly equal to its contribution to the line current and to the bus current.
If one inverter is connected line to line, its output current will appear in both of the line conductors, but in each of them the current will be out of phase with the line to neutral voltage.
When you connect one inverter C to A and one inverter A to B, the current on A will be the inverter rated current times 1.73, not times 2, and that sum current will be in phase with the line to neutral voltage on A.
All we have been saying is that, for three balanced inverters, the 3-phase line current calculated from the rated power will not be equal to the sum of the rated output currents (magnitudes) associated with each line conductor.
And that this difference works against you when applying the 120% rule as currently worded.
Click to expand...

OK, let's work an example. Say we have three SMA SB7000's connected phase to phase to a 208V three phase system. I would not derive the output current of the AC combiner from the rated output current of the component inverters, I'd look at the output of my AC combiner as that of a lone three phase 21kW inverter. For the current I calculate:

(3)(7000W)/(208V)(sqrt3) = 58.29A.

How would you do it and what difference do you see?
 
ggunn said:
OK, let's work an example. Say we have three SMA SB7000's connected phase to phase to a 208V three phase system. I would not derive the output current of the AC combiner from the rated output current of the component inverters, I'd look at the output of my AC combiner as that of a lone three phase 21kW inverter. For the current I calculate:

(3)(7000W)/(208V)(sqrt3) = 58.29A.

How would you do it and what difference do you see?
Click to expand...
As discussed earlier, the NEC requires you to use the rated output current of the inverters, not their rated power.
Given that, the arithmetic sum of the output currents from two adjacent phases will be 2/1.732 times the magnitude of the vector sum.
Your method gives you the magnitude of the vector sum, which is what the line and bus will actually carry, but it arguably conflicts with the calculation procedure used in the NEC.
Pre 2014 Code, there is no question in the code language that you must add the numeric breaker ratings rather than using the vector sum.
 
GoldDigger said:
As discussed earlier, the NEC requires you to use the rated output current of the inverters, not their rated power.
Given that, the arithmetic sum of the output currents from two adjacent phases will be 2/1.732 times the magnitude of the vector sum.
Your method gives you the magnitude of the vector sum, which is what the line and bus will actually carry, but it arguably conflicts with the calculation procedure used in the NEC.
Pre 2014 Code, there is no question in the code language that you must add the numeric breaker ratings rather than using the vector sum.
Click to expand...

If I am interpreting correctly what you say, in my example the contribution to the 120% rule in the MDP would be 270A (six 45A breakers in the AC combiner) even though the backfed breaker in the MDP would only be 80A. That makes no sense to me whatsoever.
 
ggunn said:
I'm trying to wrap my head around what some of you guys are saying...
Click to expand...

Me too. I was initially prepared to defend the 2014 code as being not very ambiguous, but I've decided that that's wrong and that Golddigger is right.

First of all, consider the case of three identical 7kW single phase inverters connected phase to phase around a three phase system in a dedicated AC combiner panel. Considering the points you bring up, would the output of that AC combiner be treated any differently with respect to conductors and OCPD at the interconnect than would be the output of a single three phase 21kW inverter?
Click to expand...

Let me answer that this way...
A single-phase 7kW inverter connected line-line at 208V will have a rated output of 33.7A. If you have two of them connected to a busbar then "125% of the of the inverter(s) rated output" could be interpreted as 67.4A. If you have three of them connected in a balanced fashion, an AHJ could even go crazy and say you have to add all three together, for 101.1A.

By contrast, a 21kW three-phase inverter will have the vector math figured into its rated output, which will thus be 58.4A.

So, it could make a big difference if the AHJ doesn't understand or won't accept vector math being figured into "125% of the inverter(s) rated output".
 
Status
Not open for further replies.
Top