Generator Problems at a Remote Hospital

[Saturn_Europa]

I recently quite my job that Ive had for 12 years to work with Doctors without Borders. My new job is energy manager at a hospital in Sierra Leone. Its kinda fun and kinda sucks at the same time.

We do have a very modern energy plant. We run on three 220 KVA synced Cat gensets. There is no utility connection. Currently we run on only one generator and our load is around 55 KVA. So its a fraction of one generator.

The problem we have been having is when we switch on 10 mini split HVAC units. All the units have small inverters. This causes -kvar back onto the generator. The generator sees -kvars of 5 - 10%. This caused the generator to shut down on alarm.

Another 50kVA generator was provided to supply just the 10 AC’s, but during commissioning with this 50kVA it shutdown a number of times due to Over Voltage. A Fluke power analyser was used to assess the power quality during the commissioning and we observed a high Lagging Power factor on at least one phase.

We contacted the CAT and they said to set the shut down to 20% -kvar on the 220 KVA gensets. Since then we have been able to run fine.

The problem is we will be bringing a lot more air conditioners online. We are worried this will increase the -kvar and put us into shut down again.

The problem is negative reactive power generated by the mini split HVAC units. I think this is causing distortion power factor. My solution to this would be line reactors. The electrical foreman running the construction thinks we should put in a power factor correction capacitor bank.

Its my first week at the hospital so I dont have a lot of data or experience to go on.

Whats the best solution to this problem?

 

[synchro]

The problem we have been having is when we switch on 10 mini split HVAC units.

Are you turning all of the units on at the same time? If so, does the generator shutdown happen shortly after turn-on or is it delayed significantly? If turning them on all at once causes a quick shutdown perhaps the inrush current is causing a problem despite any pre-charge or similar mitigation in the units.

Did the power analyzer show what you consider acceptable harmonic content (both voltage and current) at the generator output? The rectifiers in the front end of the drives when driven with the relatively high source impedance of generators can lead to harmonic problems.

One problem with using power factor correction capacitors is there's a risk of resonance with the inductance of the generator if care is not taken to address it. I've gathered that VFDs usually have a relatively good power factor at the desired fundamental 50 or 60Hz, but harmonic content may make it a problem. I can see that rectifiers using SCRs could create a lagging power factor if they are scaling down the DC voltage. But my understanding is that SCR rectifiers are not common in lower power drives. I'm sure others on this forum with more in the field experience with HVAC drives could provide more insight.

 

[Hv&Lv]

What is the voltage of the mini splits? And the generator? I’m trying to understand why just one phase is lagging and the inverters.

you see negative kVars and you want to add capacitors?

 

[Saturn_Europa]

What is the voltage of the mini splits? And the generator? I’m trying to understand why just one phase is lagging and the inverters.

you see negative kVars and you want to add capacitors?

The mini splits are 230 v and the gensets are 230/400v 3 phase.

Yes we are currently seeing -10 kvar. The generator is showing 0.97 PF leading . Would adding a capacitor make this problem worse? I think we should be looking at line reactors.

 

[GoldDigger]

A leading power factor corresponds to a capacitive component to the load current. Adding parallel capacitance will make it worse.

Sent from my XT1585 using Tapatalk

 

[Hv&Lv]

Capacitors won’t fix it, it will make it worse as stated above. It also sounds Like the mini splits are loaded on the generator unevenly. If you add more AC units, make sure all three phases are loaded equally. Fix the PF like you are suggesting. The foreman just heard capacitors fix bad PF and he doesn’t understand leading and lagging apparently.

 

[Saturn_Europa]

dd

 

[synchro]

The mini splits are 230 v and the gensets are 230/400v 3 phase.

Yes we are currently seeing -10 kvar. The generator is showing 0.97 PF leading . Would adding a capacitor make this problem worse? I think we should be looking at line reactors.

0.97 PF is normally considered good so I wouldn't think it would be causing your problem, at least by itself.

0.97 PF with -10 kVAR means you're drawing 41.1 kVA and 39.9 kW, with a 14.1 degree phase lead.

One suggestion is to measure the current on the neutral lead to the generator with a true RMS meter. This would indicate if there's too much imbalance on the loads and/or excess harmonic currents.

 

[mbrooke]

My question is why are the mini splits so leading. Do have as single line diagram and a panel schedule? And elevators? A lot of times the mini splits are single phase and aren't balance well in the panel board.

Ultimately you would need shunt reactors if leading PF is the case and can't be fixed.

 

[Saturn_Europa]

0.97 PF is normally considered good so I wouldn't think it would be causing your problem, at least by itself.

0.97 PF with -10 kVAR means you're drawing 41.1 kVA and 39.9 kW, with a 14.1 degree phase lead.

One suggestion is to measure the current on the neutral lead to the generator with a true RMS meter. This would indicate if there's too much imbalance on the loads and/or excess harmonic currents.

All the mini-splits are fed from 1 sub panel. Its a 3 phase panel I double checked that its split evenly. There are 3 AC's wired to phase A, 3 ACs wired to phase B, and 4 ACs wired to phase C. I pulled amps using a Fluke 324 true RMS meter. Phase A was 17 amps, Phase B was 18 amps, Phase C was 24 amps and the neutral was 20 amps. This really did not make sense to me. Can you clarify why the unbalanced load was so high on the neutral?

 

[Hv&Lv]

Without having the angles of the phases, I would guess it’s simply the tip to tail result of the vectors
you did state two were leading, one was lagging, correct?

move one to phase A and see if it reduces.

 

[Saturn_Europa]

Without having the angles of the phases, I would guess it’s simply the tip to tail result of the vectors
you did state two were leading, one was lagging, correct?

move one to phase A and see if it reduces.

Ok! Ill give it a try later this week.

All phases are leading -3 kvar -3 kvar and -4 kvar.

 

[mbrooke]

All the mini-splits are fed from 1 sub panel. Its a 3 phase panel I double checked that its split evenly. There are 3 AC's wired to phase A, 3 ACs wired to phase B, and 4 ACs wired to phase C. I pulled amps using a Fluke 324 true RMS meter. Phase A was 17 amps, Phase B was 18 amps, Phase C was 24 amps and the neutral was 20 amps. This really did not make sense to me. Can you clarify why the unbalanced load was so high on the neutral?

Best guess is none linear load or a crossed neutral in a branch circuit.

 

[synchro]

From the KVAR and currents you measured, I get:

Phase A: 3.9 KVA, theta = -50 deg, PF = 0.64
Phase B: 4.1 KVA, theta = -46 deg, PF = 0.69
Phase C: 5.5 KVA, theta = -46 deg, PF = 0.69

To keep it simple, assume that each AC unit draws 6 amps and the generator phase voltages are 120 degrees apart and equal magnitude.
Then the neutral current due to unbalance of the fundamental (50 Hz?) is:
-sin(30) x 18A - sin(30) x 18A + 24A = -18A + 24A = 6A.
So the harmonic current in the neutral is sqrt [ (20A RMS total)**2 - (6A RMS fundamental)**2 ] = 19.1A RMS (assuming that something else is not causing currents in the neutral).

Someone mentioned this company's products for harmonic reduction and PF correction. Perhaps they could chime in whether they could be helpful in this situation. :
https://transcoil.com/products/

 

[synchro]

If you have the equipment take a look at the waveform of the neutral current to see if there's obvious distortion (like a square wave, ringing at higher than the fundamental freq., etc).