Single Current Transformer on Parallel Loads - Effect of Low Power Factor?

You could use two CT’s and a summation transformer.

http://www.rayleigh.com/current-transformers/summation-current-transformers.html

Yeah, looks good. The link I posted in the question indicates that CTs can just be paralleled with no transformer.

What's going to happen to the PF though when the PF of one load drops?

What's going to happen to the PF though when the PF of one load drops?

If the secondaries of the CT are paralleled, the resultant current will be the sum of the two, according to Mr. Kirchoff. Each pump current will be a vector sum of the current in phase with the voltage and 90 degrees out of phase with the voltage. So the paralleled CTs will output a current that will be the vector sum of the total current in phase with the voltage and 90 degrees out of phase with the voltage.

So, the bottom line is, you will see a partial phase shift if one pump starts sucking wind, and a further phase shift if both do.

Caveat: I am assuming that both pumps are on the same voltage phase. (If you have 2 phases, 180 degrees apart, you could reverse one CT.)

Yup, all 3-phase pumps fed from the same feed.

So, what's wrong with a low-level switch to cut off the pumps? Why bother with power factor and all its complexity?

Power Factor isn't a complexity at all, it's part of the standard offering of our product (http://www.multitrode.com/products/multismart-and-probes/multismart-pump-station-manager-and-rtu/), you just need to tick a box and it's enabled with a default of 0.6, which can be adjusted.

My problem is what will happen when only one of the parallel loads goes low PF....

There might well be a Low Level Float which will disable all pumps as well - but drawing air is not the only time you can get a low PF. These are sewage pumps, so they can get ragged up, which will result in a low PF. (We automatically reverse the pumps to clear the blockage then restart.)

<...These are sewage pumps, so they can get ragged up...automatically reverse the pumps to clear the blockage then restart...> Well, that wasn't revealed at the initial posting.

Trying to pump rags is a Bad Thing.

Rate of drop of level in the wet well is another way of doing it.

Changing the pump for a Hidrostal "pre-rostal" type of pump (usual disclaimer) means the pump won't get damaged by either rags or air.

• If the first pump starts pumping air, then won't the second as well?
• If the current in the first pump goes to 115%, what is there to trip it off and prevent its being converted into a convector heater?
• What has accuracy got to do with anything?

From here, this installation looks like the proverbial sledgehammer to crack the proverbial nut!

Summing the kW & kVAR for each pump will decrease the sensitivity of your Power Factor Low Flow Pumping Detection instrument, since you will be treating the two motors as a lump value, as asserted many times above. However, to the point PWSlack made where it is likely that eventually both pumps will be pumping air, eventually you will detect low flow.

You could also use a low kW setting instead of Power Factor, but then you'd need to know the pump & motor rating, and normal performance, to detect low flow. Think of Power Factor as the same thing, with the same reduced sensitivity due to lumping the values.

If the device is used for motor thermal protection, then summing the currents is a bad idea.

You can parallel CTs to measure the total current to the two pumps. We do it all the time in substation applications where two breakers feed one line (such as ring bus or breaker-and-a-half schemes), in which CTs on each breaker are summed together for metering and protection relaying. However, I don't think it is that easy for your power factor measurement. You have to know what the vector sum of the two currents is for your abnormal condition to determine what PF you should trip at. It is not as easy as saying that 0.85 PF for one pump and 0.45 PF for the other will give a 0.65 PF at which you want to trip (or something like that). And if there are a variety of possible scenarios for each pump's PF, you may not be able to define a specific go - no go condition. You'd be better off if you could run each CT into its own input, and be able to adjust settings for each input to trip the pumps when PF on either one is out of whack.

Seems to me you want to measure kWh to two pumps with one device having one current input per phase but also try and protect two motors at the same time with the same device.

This will give you heartache if one pump is sucking air and the other is dragging rag and combined power still looks normal. The customer will blame you if you give inferior protection to standard solutions which give separate monitoring and protection to each motor.

So far as summing accuracy is concerned, low power factor makes higher CT phase errors and hence kW errors on top of your metering double the current, so change due to only one motor is less definite.