Power Factor varies from 75 to 250

[RThorn]

I have been doing a billing analysis for a facility. I am seeing huge variation in Kvar at a facility that gives me unreliable power factor values. I have not been on site so I am not familiar with their loads.

One month I see 16,320 kwh, 8,280 Kvar, and demand of 88.8 kW. That gives a power factor of 93% which is believable.

Two months later I see 33,360 kWh, 19,680 kvar, and a demand of 78.72 kW. That gives me a power factor of 250%.

Other than the sizing of a capacitor bank, what else should we be looking for?

Thanks

 

[retirede]

Somethings wrong with the data you’re being given.

 

[RThorn]

It looks like the problem really started after the utility replaced a meter.

The data i have comes right off of the utility bills. Just double checked every line.

 

[jaggedben]

It looks like the problem really started after the utility replaced a meter.

Possibly the issue is with the new meter.

The data i have comes right off of the utility bills. Just double checked every line.

Are you sure that, for example, 8,280 Kvar isn't 8,280 var which would also be 8.28 kVar?
kVAR should be a fraction of kW so the VAR numbers look off by a factor of 1000.

Maybe the new meter is reporting in the wrong unit, reporting VAR as kVAR.

Also you're calculating wrong. Say the VAR is off by 1000, power factor in the first example would be 88.8÷(88.8+8.28)=0.91.
In the second example 78.72÷(19.68+78.72)=0.80.

Also I wouldn't assume the Kw and kVAR numbers come from the same intervals so may not tell you anything accurate about the typical power factor, though they may give you a general idea.

 

[RThorn]

Possibly the issue is with the new meter.




Are you sure that, for example, 8,280 Kvar isn't 8,280 var which would also be 8.28 kVar?
kVAR should be a fraction of kW so the VAR numbers look off by a factor of 1000.

Maybe the new meter is reporting in the wrong unit, reporting VAR as kVAR.

Also you're calculating wrong. Say the VAR is off by 1000, power factor in the first example would be 88.8÷(88.8+8.28)=0.91.
In the second example 78.72÷(19.68+78.72)=0.80.

Also I wouldn't assume the Kw and kVAR numbers come from the same intervals so may not tell you anything accurate about the typical power factor, though they may give you a general idea.

Could be. The Utility bill says kVAR

 

[jaggedben]

Sorry, my calcs above are wrong, I forgot the relationship is trigonometric. But what we're saying about the kvar and kW figures seeming to be off by 1000 is still correct.

 

[RThorn]

If off by a factor of 1000 on the kVAR then the PF is unbelievably low. Sounds like a meter multiplier issue yet the kWh looks correct

 

[charlie b]

. . . I forgot the relationship is trigonometric.

I agree. Power factor is the cosine of the angle opposite to kvar in the power triangle. You need to first calculate kva using the pythagorean equation. Pf is then kva divided by kw. I don't have time now to do the math.

 

[jaggedben]

If off by a factor of 1000 on the kVAR then the PF is unbelievably low. Sounds like a meter multiplier issue yet the kWh looks correct

If the kvars are actually vars then the PF would be quite high, not low. Maybe unrealistically high. But if they are kvars then PF is unrealistically low.

And again, not knowing if the demand (W) and kvar data are from the same time intervals introduces further unknowns.

Could you install independent metering as a reality check?

Also...
By any chance was a solar or other parallel source installed at this facility? That would affect power factor measurements.

 

[jaggedben]

The more I look at this the more I think it's not even a simple multiplier/units problem and is something deeper that's wrong with the meter data. Back to post #2.

 

[wwhitney]

Sorry, my calcs above are wrong, I forgot the relationship is trigonometric.

Assuming each of the lines of data is one month, and assuming constant power over the month, I get:

Month 1: PF = 16.32 / sqrt(16.32²+8.28²) = 89.2%
Month 2: PF = 33.36 / sqrt(33.36²+19.68²) = 86.1%

Are you sure that, for example, 8,280 Kvar isn't 8,280 var which would also be 8.28 kVar?

I'm not seeing the issue, I assume kVar and kWh are comparable? I.e. 1 VAR = 1 volt * 1 reactive amp, and 1 Wh = 1 volt * 1 real amp.

Cheers, Wayne

 

[jaggedben]

Assuming each of the lines of data is one month, and assuming constant power over the month, I get:

Month 1: PF = 16.32 / sqrt(16.32²+8.28²) = 89.2%
Month 2: PF = 33.36 / sqrt(33.36²+19.68²) = 86.1%


I'm not seeing the issue, I assume kVar and kWh are comparable? I.e. 1 VAR = 1 volt * 1 reactive amp, and 1 Wh = 1 volt * 1 real amp.

Cheers, Wayne

I don't follow. Var is an instantaneous measurement comparable to W not Wh. sqrt(W²+var²)=VA. I don't see how Wh can be figured into the PF calc here.

 

[jaggedben]

Okay, I guess you're assuming that what's given as kvar is actually kvarh. That might make sense.

 

[wwhitney]

I don't follow. Var is an instantaneous measurement comparable to W not Wh. sqrt(W²+var²)=VA. I don't see how Wh can be figured into the PF calc here.

Ah, I assumed it was kVar-hrs and the idea was to calculate the power factor if the load were constant power over the month.

If kVar is demand kVar, and kW is demand real kW, I don't see how you can use those to calculate anything, as there is no guarantee they occur at the same point in time.

I also assume demand kW is an apparent power figure, as the utility would charge you for the total current you cause to flow.

Cheers, Wayne

 

[jaggedben]

...

I also assume demand kW is an apparent power figure, as the utility would charge you for the total current you cause to flow.

I'd make no such assumption.

Fwiw, per this linkedin post I found there are at least eight ways utilities can charge for poor PF. A kvarh charge and a demand charge in VA are both among them. If I were being charged for varh I'd hope any demand charge would be in W so I'm not paying twice for PF. But only the OP can tell us more about the tariff that applies here.

8 Types of Power Factor Penalties: Which One Does Your Utility Use?

1. kVA (Demand) Rate Penalty What it is: You get charged extra if your PF is below 1.

www.linkedin.com

Today's digital meters are certainly capable of transmitting any of the aforementioned datapoints.

 

[herding_cats]

I agree. Power factor is the cosine of the angle opposite to kvar in the power triangle. You need to first calculate kva using the pythagorean equation. Pf is then kva divided by kw. I don't have time now to do the math.

exactly. The PF of 0 is 1.