Chiller Tripping MCCB

[Adam W]

As title, we have a chiller which is tripping out an MCCB. The chiller itself doesn't have it's own OCPD, which I intend to fit at some point anyway, but that wouldn't necessarily stop anything tripping. As far as I can make out the only other things on the circuit are a few fans, a pump, and a couple of sockets but they are fed from a DB and individually protected; I put a clamp meter around the cables from the MCCB, which were only carrying about 10A per cable, plus nothing was tripping without the chiller running, which is why we're thinking that is where the problem lies. I did the same with the cables from the isolator to the chiller with it running and it was using around 180A per phase.

The strange part is the chiller will run for a couple of weeks, then trip. I've IR'd the cables from about halfway (there's an old isolator in the middle) which seem fine, although given it's such an intermittent problem it's unlikely anything would show up at the exact time of testing.

I've had a look around the internet, which has suggested maybe a loose connection causing overheating at the MCCB. I reset it today and took a reading and it only seemed to be drawing about 10A; my next step will be to keep an eye on it over the next couple of weeks, taking current readings and checking the temperature of the contacts to see if there are any hot spots. A few other things have been suggested, such as low pressure, but they would either cause it just to not work, or trip the breaker straight away.

Does anyone have any other ideas?

 

[Marvo]

So the chiller normally draws 180A per phase when it's running? What size is the MCCB? Is the MCCB set up correctly to accommodate start up current?

A poor termination at the MCCB itself can cause heat to be transmitted into the breaker which causes intermittent tripping via the internal thermal mechanism but a poor termination should be easily visually identifiable with evident heat damage.

There's numerous other possible causes.

Low gas charge could cause rapid cycling of the chiller via the LP controls, rapid cycling could cause the MCCB to trip due to frequent start currents. Rapid cycling is also like to shag the compressor eventually due to overheating.

There could also be a poor connection elsewhere in the circuit on one of the phases causing current overdraw on the other two phases.

Running current is often affected by head pressure so blocked condensers or slow running/intermittent running of condenser fans can cause high run currents.

These are just a few possibilities, there's many more. Unless you're familiar with refrigeration systems I'd suggest you do the usual circuit tests on the supply all the way to the chiller. Visually inspect all contacts and terminations from the supply source to the chiller itself and check end to end resistances as well as insulation tests. If everything test fine electrically I'd suggest you get assistance from an HVACR mech to check the refrigeration side of things.

 

[Adam W]

Thanks for relying Marvo. The MCCB is set at 400A, so there should be plenty of headroom for everything being fed from it. The delay is turned right up, but it isn't tripping on startup anyway, it runs for a couple of weeks first. The only thing that isn't turned right up looks to affect the short circuit rating which I'm wary of changing - to my mind if there's a short circuit it needs to disconnect.

I took some current readings at the breaker this morning and it was pulling between about 10 and 17A, across different phases, so clearly not running at full capacity. As such the temperature readings I took weren't much use - about 20°C, compared with 34°C on the MCCB above.

I wouldn't say I'm familiar with refrigeration systems at all - I'd basically just done what you suggested and checked the obvious things an electrician would check; I suspect the same - that it's a refrigeration issue rather than an electrical one. Presumably I got asked to look at it because we haven't got a fridge guy in-house.

 

[Marvo]

The 10-17Amps you were seeing was probably when the compressor wasn't running and it was just the load of the internal fans and lights etc.

Have you checked the on-load voltages at the compressor? One of the tell tale signs of a poor termination or even a failing contactor would be the difference in voltage between on and off load.

If it's a 400A MCCB on a 180A compressor then it might be a bit close but it's also possible that the MCCB would tolerate the start current depending on it's curve and delay settings. I'd expect the start current to be somewhere in the region of 5 times the running current albeit only momentarily. Has the unit been in operation without tripping for some time or is it a relatively new installation? If it's relatively new you can't rule out a design issue, if it's been running for some time without problems you're almost certainly just looking for a fault.

With larger chillers such as this there may be an unloader system that prevents the compressor starting under head pressure, failure of the unloader would cause high start currents but you'd probably need to get assistance to test if it's operating correctly.

With the MCCB being rated at >double the run current it rules out general high head pressure faults I mentioned previously. It would more likely be a compressor mechanical issue such as bearing failure in that case.

Is the compressor run via a drive or is it direct on line?

With intermittent problems like this I'd usually put a data logger on the unit and get some figures over a week or two for voltage and current. 9 times out of 10 just from the voltage/current graphs you can figure out what the problem is or at least they point you in the right direction.

 

[Adam W]

Yes, I'm almost certain the 10-17A was what you say. It was fairly cool outdoors this morning so it probably wouldn't have needed to cool anything, never mind work at full capacity.
The installation is pretty old - it's wired in old colours and judging by the state of it I'd say it's at least 20 years old.
From what I can make out there has been some work done on it fairly recently - I think the compressor was changed or fixed, and the contactors tested. Apparently before that everything was working fine. One thing I noticed was the phase colours were the wrong way round, but that seems to be par for the course across the site and just the hand I've been dealt.

It's direct on line, although I'm sure we have spare drives available - do you reckon it'd be better run through a drive?

I will look into the possibility of getting a data logger on it tomorrow, thanks for the suggestion.

 

[Marvo]

Whatever you do don't reverse the rotation, if the wiring colours are not correct just leave them. Standard reciprocating compressors will often run in either direction but certain types of compressor will be badly damaged if you run them in reverse.

If the unit is old then leave it DOL. Some older compressors won't have a high enough insulation grade to handle being run on a drive so retrofitting one could well be more trouble than it's worth.

Let us know what you find with a data logger. In the mean time I'd suggest you go on a fact finding mission regarding full load running current and max permissible starts per hour etc.

 

[topquark]

Out on a wee bit of a limb here, but Is it possible that you're getting any condensation anywhere it shouldn't be causing a dead short? You'd only need a couple of degrees temp differential to get condensation.