Are Panel Load Calculations done in kVA or kW?

Status
Not open for further replies.
C

cjarvis64

Member
Location
42001
Occupation
Electrical Engineer
Sizing loads for multiple panelboards on a project, and I note that my company's panel schedule excel sheet template lists total load in kW. I have been adding up loads using the load data provided by manufacturers where possible, and where not possible I have been calculating apparent power and using that for my load. It doesn't make sense to me why we would list total load in kW because when you factor in power factor (when you can even find it), it decreases your load. It seems prudent to me to use kVA as it is the larger value. What should I be doing? I need a sanity check I feel like.

Thanks
 
The code (Article 220) speaks of everything in kVA so I agree that that is the correct unit.

Where loads have nameplate amps or VA I use that. Where they only list watts I assume the PF is close enough to 1 that the watts can stand in for the kVA. Mind you I only do resi, but I don't see why it would be different at larger scale. I suppose if you have a really large load that they only gave you kW for and you're concerned about the PF, contact the manufacturer to try to get that info.
 
I agree that the calculation should use units of KVA. Adding KW to KW is likely to give you an invalid (i.e., low) result. Most building loads have a pf at or close to 1.0, and you can treat KW and KVA as equal. For anything with motors, such as fans and HVAC equipment, if you have no actual product data at hand, I would assign a pf of 0.8. That should be sufficiently conservative, and is unlikely to result in an undersized service.
 
I agree with everything said however depending on the loads you’re dealing with you’re already being conservative using NEC rules. I think everyone agrees NEC requires us to make services much bigger than needed, I always figure 50-60% of my calculated load is the kva size of the transformer the poco will likely provide. Imo adding pf just makes it worse. Obviously what type of loads will play a big part so not always the case. EV for example, they could realistically see everything from a dedicated ev service.
 
Also keep in mind in the real world, kVA has phase angles. In most cases NEC loves to add kVA for load calculations without even considering phase angles. And I assume this is because it's for worst case scenarios.
 
Tainted said:
Also keep in mind in the real world, kVA has phase angles. In most cases NEC loves to add kVA for load calculations without even considering phase angles. And I assume this is because it's for worst case scenarios.
Click to expand...
Mmmmm, I don't think so.

First of all, whether you have to consider phase angles depends on how many phases you have, not whether you calculate in VA or W.

Secondly, I'm not aware of any code section that prohibits calculating line currents in three phase systems to arrive at appropriate conductor ampacity for a given kVA in three phase systems. Or put another way, a 100A 208/120 service delivers 35,984VA, not 20,800.

Unless I'm misunderstanding and you are just referring to power factor as discussed above.
 
jaggedben said:
Mmmmm, I don't think so.

First of all, whether you have to consider phase angles depends on how many phases you have, not whether you calculate in VA or W.

Secondly, I'm not aware of any code section that prohibits calculating line currents in three phase systems to arrive at appropriate conductor ampacity for a given kVA in three phase systems. Or put another way, a 100A 208/120 service delivers 35,984VA, not 20,800.

Unless I'm misunderstanding and you are just referring to power factor as discussed above.
Click to expand...
I think you are misunderstanding. I am referring to the angle between the real and imaginary component which is corelated to power factor.

I am saying in the real world, you cannot add total kVA without phase angles (unless power factor is 1)
 
I get you now. You're referring to the real world VA of multiple loads likely being less than the sum of the individual VAs. Agreed, that adds a bit more overestimate.
 
I don't think the NEC adds kVA as a way to be conservative. I think it is simply an industry acceptable approximation.

Nothing prohibits doing the more accurate vector addition of currents. IMHO doing so is rarely if ever justified.

Jon
 
winnie said:
I don't think the NEC adds kVA as a way to be conservative. I think it is simply an industry acceptable approximation.

Nothing prohibits doing the more accurate vector addition of currents. IMHO doing so is rarely if ever justified.

Jon
Click to expand...
I haven't ever seen any examples in annex D that use the vector method. Vector method seems like an unfavorable method
 
Tainted said:
I think you are misunderstanding. I am referring to the angle between the real and imaginary component which is corelated to power factor.

I am saying in the real world, you cannot add total kVA without phase angles (unless power factor is 1)
Click to expand...
Sorry, but I disagree completely. KVA plus KVA gives KVA, and phase angles do not enter that concept. All load calculations should be performed in units of KVA. Once you have the total KVA, you can figure out the facility's total KW. But there is very little use for that result, unless you are sizing a backup generator or unless the utility includes that in their billing rates..

If all you have is KW, then you need to account for phase angles, in order to derive KVA. If you attempted to add KW to KW, in an effort to derive total KW, you would have an incomplete, and thus incorrect, result.
 
charlie b said:
Sorry, but I disagree completely. KVA plus KVA gives KVA, and phase angles do not enter that concept. All load calculations should be performed in units of KVA. Once you have the total KVA, you can figure out the facility's total KW. But there is very little use for that result, unless you are sizing a backup generator or unless the utility includes that in their billing rates..

If all you have is KW, then you need to account for phase angles, in order to derive KVA. If you attempted to add KW to KW, in an effort to derive total KW, you would have an incomplete, and thus incorrect, result.
Click to expand...
Maybe I wasn't clear with what I am trying to point out.

What I am saying is in the "NEC" world, phase angles are not considered, which is OK because it's a standard.

In the "real world" and as physics and math points out is that phase angles are to be considered. Electrical utilities always considers phase angles whereas in the NEC book it does not.
 
charlie b said:
Sorry, but I disagree completely. KVA plus KVA gives KVA, and phase angles do not enter that concept.
Click to expand...

I think you may have swapped kVA and kW in your discussion.

If I have 2 loads, each 1 kW, then when I combine these the total load is 2 kW. I won't know the current drawn, but I know the power consumption.

On the other hand, if I have 2 loads, each 1 kVA, when I combine these loads in parallel the total load is anything between 0 and 2 kVA, and I can't calculate that total without knowing the phase angles.

In this case, standard practice is to assume 2 kVA. But that is an approximation.

Jon
 
winnie said:
I think you may have swapped kVA and kW in your discussion.

If I have 2 loads, each 1 kW, then when I combine these the total load is 2 kW. I won't know the current drawn, but I know the power consumption.

On the other hand, if I have 2 loads, each 1 kVA, when I combine these loads in parallel the total load is anything between 0 and 2 kVA, and I can't calculate that total without knowing the phase angles.

In this case, standard practice is to assume 2 kVA. But that is an approximation.

Jon
Click to expand...
If he indeed swapped them by mistake, then there's no disagreement here
 
Tainted said:
If he indeed swapped them by mistake, then there's no disagreement here
Click to expand...
I did not swap units by mistake. But I believe we are in the middle of the following type of argument:
Person one says, "The sky is blue."
Person two says, "No, no, no. You are wrong. The grass is green."

I was referring to what I believe to be the customary process of load calculations. They should all be done in units of KVA. If you only have KW for some loads and don't have their KVA, you can, load by load, determine (i.e., by research or, if necessary, by intelligent guesswork) the reactive power of that load (i.e., its phase angle), and calculating the KVA value of that load. Then you can proceed by adding all KVA values.

Winnie is right in saying that two loads of 1 KVA each can actually add up to any value from 0 to 2 KVA. You will get a 0 KVA result if one is purely inductive (phase angle +90) and the other is purely capacitive (phase angle -90). But loads we usually see have phase angles not that far from each other. In addition, I can't recall ever having to include a significant capacitive load in any building I designed.

I also agree with Winnie in that adding two 1 KVA loads and getting a total of 2 KVA is no more than an approximation. Indeed, that result will likely be higher than a "true result." If you want true precision, add the KW values for all loads to get total KW, next add the KVAR values for all loads (keeping in mind that capacitive reactance is a negative value) to get total KVAR, and finally use the Pythagorean equation to get total KVA. I am certain that no previous employer of mine would have wanted to pay me to do all that extra work.

That brings me back to adding KVA to KVA to get a value of total KVA.

QED
 
@charlie b I think we agree on the physics, and I absolutely agree that totaling up kVA is 'standard best practice'. When systems get designed you calculate the kVA of each load, and then you add those kVA up to get the system kVA requirement.

What tripped me up is when you said that adding kW doesn't give total kW whereas adding kVA gives total kVA. The _physics of the situation_ is exactly the opposite; the kW of loads add up to give the true total kW; to get total true kVA you need to perform vector addition of some sort. For all practical purposes, unless you are dealing with extremely pathological loads, adding kVA to get total kVA is good enough.

-Jon
 
winnie said:
@charlie b I think we agree on the physics, and I absolutely agree that totaling up kVA is 'standard best practice'. When systems get designed you calculate the kVA of each load, and then you add those kVA up to get the system kVA requirement.

What tripped me up is when you said that adding kW doesn't give total kW whereas adding kVA gives total kVA. The _physics of the situation_ is exactly the opposite; the kW of loads add up to give the true total kW; to get total true kVA you need to perform vector addition of some sort. For all practical purposes, unless you are dealing with extremely pathological loads, adding kVA to get total kVA is good enough.

-Jon
Click to expand...

When Charlie said: "If you attempted to add KW to KW, in an effort to derive total KW, you would have an incomplete, and thus incorrect, result."

Maybe he meant you would have an incomplete answer because there is no angle or kVAR present to get kVA?

My take is:
kW + kW gives you kW
kVAR + kVAR gives you kVAR
kW +kVARj gives you kVA with a phasor angle

If someone doesn't agree with this, I don't know what else to say.
 
Tainted said:
When Charlie said: "If you attempted to add KW to KW, in an effort to derive total KW, you would have an incomplete, and thus incorrect, result."

Maybe he meant you would have an incomplete answer because there is no angle or kVAR present to get kVA?
Click to expand...
Reading the original post, I agree. kW + kW always gives correct total kW, but I agree that the number is incomplete for sizing the system.

-Jon
 
I like to think of it as KW only is relevant at the source and the load. Everything in between needs to be designed for KVA, which will always be equal to or greater than KW

I'm sure if I thought long enough I could come up with a scenario where this would result in me using some component that is slightly oversized. But thinking is hard
 
Status
Not open for further replies.
Top