Calculating Electrical Usage From Three Phase

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

You can can't really. You also need displacement power factor. You can calculate the kVA on each phase, but not the watts. You cannot just add the three kVA values to get a total kVA. You have to add the watts, add the vars, then do the square root of the sum of the squares (basically square the totale watts add that to the square of the totals vars and then take the square root).

These numbers make no sense. You must be leaving out some information about the test. The sum of the watts on each of the phases should equal the total three phase watts. As stated above, the watts sum across the three phases, as do the vars. The kVA does not.

As I said above, you can't calculate the quantity you desire with only that information.

Charles Perry P.E.

I was there when he did the test, and he did a time calculation of the time for the dial to rotate once on each of the individual phases, then he enabled all three phases and repeated the measurement. It was visually quite obvious that the dial would rotate *much* faster when the three legs were connected, and it really did seam to be faster than the sum of the three speeds of the individual phases. He took the number of seconds for one rotation and plugged that into some formula that gave him the kw consumption based on that rotation speed.

The time for one rotation of the meter dial for the legs were:

Leg A: 104 seconds Leg B: 50 seconds Leg C: 102 seconds

With all three phases connected: 8 seconds

He explained this difference by saying that you could not capture three phase equipment by measuring individual legs, and that the difference was attributable to HVAC or anything that was true three phase. What would the above anomaly suggest?

Separate question: can someone explain why it is important to keep the three legs in a three phase circuit balanced?

The anomaly suggests that whoever did the test doesn't kmow anything!

If he really did disconnect two legs at a time and measure the meter rotation with only one leg connected, he was measuring the single phase line to neutral loads only, and completely disconnecting all the line-to-line loads!

How big of a service is this? How did he "disconnect" the phases that weren't measured? Is this service metered with transformers?

The facts as you have stated them don't add up. There appear to be some pieces missing.

For what it's worth, kWhr meters are marked with a meter constant, label as Kh. This is the meter constant in watt-hours per revolution of the disc. This value must be multiplied by the transformer factors if transformers are used on the meter.

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

Ok, so he actually disconnected the meter and connected only one leg at a time?

Actually, an electronic revenue meter does measure each phase independently and then do the correct math to add them. Of course it is measuring all three phases and the same point in time. The fact that each phase was measured independently, with 100s of seconds between, means that the amout of load could have changed significantly between each test. If your HVAC was not running when doing the single phase tests, and then kicked on for the three phase test, that would explain the difference.

Keeping the currents balanced on all three phases helps to keep the voltages balanced. Most three phase equipment will be more efficient with balanced voltages.

Charles Perry P.E.

Yes there is a transformer, and the conversion factor on the meter is 40.

The person doing the test was a PG&E technician, and if there is a fault it must be with his explanation because they seem to have a standardized procedure and equipment for these meter calibrations.

The meter has next to it a set of ten toggle switches. When engaged, the meter is connected to the load. When disengaged, the meter is bypassed. The toggle switches were laid out in the following sequence:

Red Black Black Red Black Black Red Black Black Red

The tech said that each group of three switches (Red-Black-Black) was connected to one phase.of the load. The last Red one goes to Neutral I believe. The tech disengaged all switches and then engaged the first Red Black Black sequence. He measured. for Leg A. After measuring B and C, he engaged all of the switches and did a reading which gave the 8 second rotation result reported earlier.

In terms of size, this is about 8000 square feet, and we are using about

20,000 kWh per month.

independently

No, the difference is explained by our APC Symmetra UPS systems. These take a single phase 208V input. We have two of them in use, on different panels. This creates a very unbalanced load across our installation because only single phases are used for very significant energy consumption on those two units. On one of our three phase panels, one of the phases draws 0 amps because there is no three phase equipment wired to the panel.

"CHANGE USERNAME TO westes" wrote in message news:yPGdnefMVJa8b snipped-for-privacy@giganews.com...

If you knew what the difference was, why didn't you say so? His testing method does seem to ignore phase to phase loads. Not sure why he would even test in this fashion. You can test a three phase meter using all three phases.

Charles Perry P.E.

I've lost your context at this point. The question I started out asking is how do I explain the result he got where measuring and then adding three single phases gave a different result than measuring the three phase load. I was never trying to understand why our phases are not balanced, therefore I had no reason to provide that information.

If A kw + B kw + C kw should equal the three phase load, how do I explain that the three phase load was 300% of that sum? We had all of the equipment turned on, and I don't believe there was any significant load appearing during the three phase measurement.

Hmmm.

Well, to reply to your good questions, there is really no way to measure power consumption by measuring current alone. Even in a balanced system of known voltage, the power factor is a component of the power calculation and this can not be determined by an ammeter.

As to your second point, the sum of the phase powers in indeed equal to the total power, and your test results, because they do not agree with this statement, suggest that some aspect of the test was in error.

The fun is in figuring out the error! The fact that the discrepancy is nearly equal to a factor of 3 may be a clue.

Exactly what type of service is this? Current rating? Voltage? Wye, Delta? Primary or secondary metered? Can you tell if it uses 2 or 3 current transformers? I would guess that it uses 3 200:5 CTs, but I could be in error.

Thanks in advance.

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

I have no idea. They should add up. Something is wrong. Either it is with the test or with the numbers as recorded, or with the calculation. That is about all I can say at this point.

Charles Perry P.E.

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

Just a thought...

After reading your other post about the 'red black black' switches and all, I think what you have is something like this. Each 'red' switch shorts out a current transformer, and the black-black pair disconnect a line-line voltage transformer. And the odd red switch on the end is for the neutral.

So when just one phase is enabled (one set of red-black-black), the neutral current switch was open?? And with significantly unbalanced loads, the neutral current is probably significant. So during the single phase tests the neutral current wasn't metered and this somehow corrupted the readings???

I'm not too familiar with three-phase metering, but could this be a part of it?

daestrom

The service we have is 208V three phase with three 125V legs. I guess this is a Wye service since there is no 208V leg; all three phases are equal. The service is secondary metered. It uses three transformers and they are

200:5. I know that the service *after* transforming delivers at least 400A to our business.

I guess at this point we need to put our own meters, but does it make sense to have that meter separately do measurements on power utilization of three phases together and each phase individually, so that we can later do comparisons between the sum of the three and the reported three phase measurement?