Rotary phase converter / chiller issue

[collinkonow]

Good afternoon Gents,
Looking to have a troubleshooting discussion. Trying to diagnose an issue of a circuit breaker tripping. None of the original install was done by me. Here's the situation:

Class 320 service.
Main panel feeding a 100A main lug sub-panel. This sub-panel feeds only one load, a rotary phase converter (RPC). I do not have the specs handy, but It is sized properly. This sub-panel is fed by a 2-2-2-4 SER. (Not sized properly).

RPC is feeding three 3 phase chillers. 1- 50A, 1- 40A, 1 30A. (actually running loads are approx. 35A, 25A and 18A respectively). Total connected running load when all three units are running is 79-82 amps per phase. There are no other loads on the RPC

The issue is that the 100A breaker for the RPC is intermittently tripping. There are large amounts of time (weeks) where there are no issues, making it very difficult to troubleshoot. The issue is the 50A unit. If it is off, the issue goes away. If turns on with any other unit, the issue can occur.

Two separate HVAC contractors have said that there are no issues with the unit and are blaming the problem on the undersized SER. I 100% disagree with that and have said that the wire size is not directly correlated to the tripping breaker and that the breaker doesn't care about wire size, only the amperage it receives (hence why properly sized conductors are so important). I'm not an expert in 3 phase chillers, but suspecting an issue with a contactor or something within the unit, but they say everything is fine.

As a side note I spent an hour with the manufacturer of the RPC ensuring it is running properly (and it is) and also had the power company respond to the property to check all connections at transformers and elsewhere. All checked out fine. Also, all connections at all termination points related to all of the above mentioned were checked and tight. The 100A RPC breaker was replaced with a new one.

This system previously ran flawlessly for 3-4 years without any issues. The chiller units were used units when installed on this property.

Any advice on this would be greatly appreciated. I may have missed a few details so if you have any questions let me know. Thanks in advance.

 

[__dan]

Assuming the RPC 100 A breaker is at the line side, lets say kVA through the breaker at 100% is kVA= 100A x 240 V = 24 kVA.

At the load side of the RPC, kVA could be A x V x 1.732. You want to know the equivalent load side Amps at 3 phase at 100% on the single phase line side.

24kVA = 240V x A x 1.732

Load side Amps = 57.7 Amps 3 phase, at 100 Amps line side single phase. The breaker is working correctly if you try to draw 80 Amps 3 phase through the RPC.

Load side kVA = 240 x 80 x 1.732 = 33 kVA unless there is something more for the problem statement.

Possibly the RPC was not sized for all three units at the same time.

 

[Joethemechanic]

What is the voltage as tested at the coolers dropping to during starting?

Have you tested all the capacitors in the RPC panel?

 

[winnie]

I believe @__dan has the answer.

82A on the 3 phase side of the RPC means some 142A on the input side. No surprise that the 100A breaker is tripping.

Jon

 

[Joethemechanic]

Assuming the RPC 100 A breaker is at the line side

For some reason when I read it I pictured it on the load side of the RPC;

I wonder what the RPC's rating is?

 

[winnie]

Also: a 100A breaker is expected to carry more than 100A for a pretty long period of time.

So it is possible that the system was overloaded right from the start but is only now having tripping issues because of factors such as higher outdoor air temperatures, equipment aging and drawing slightly more current, breaker aging from heat (from the overload), and other minor changes that wouldn't make a difference in a properly sized system.

I'd recommend simply amping the supply conductors to the RPC when all 3 chillers are running.

Jon

 

[Little Bill]

Don't you have to include the motor for the RPC along with the loads it supplies?

 

[Joethemechanic]

Don't you have to include the motor for the RPC along with the loads it supplies?

Well it's consuming power, to generate a 3rd phase, so you are getting s good part of what it's using back. But they do get warm, so that is more losses, And I am not sure what power factor they operate at, but I'm sure it's not unity, so that would contribute to the amperage draw on the line side.

To tell you the truth I don't think they are very efficient. I've built a few for guys for machine tools, but the output is never as nicely balanced as real three phase.. But these guys were just happy to be able to run their lathe or bridgeport , I don't think they cared at all about the efficiency of it.

I can't imagine having one run 24/7 to supply some chillers is that good of a design

 

[collinkonow]

What is the voltage as tested at the coolers dropping to during starting?

Have you tested all the capacitors in the RPC panel?

Yes all capacitors were tested at the RPC and were fine

 

[collinkonow]

I believe @__dan has the answer.

82A on the 3 phase side of the RPC means some 142A on the input side. No surprise that the 100A breaker is tripping.

Jon

82A is on line side of 100A feeding the RPC

 

[collinkonow]

Also: a 100A breaker is expected to carry more than 100A for a pretty long period of time.

So it is possible that the system was overloaded right from the start but is only now having tripping issues because of factors such as higher outdoor air temperatures, equipment aging and drawing slightly more current, breaker aging from heat (from the overload), and other minor changes that wouldn't make a difference in a properly sized system.

I'd recommend simply amping the supply conductors to the RPC when all 3 chillers are running.

Jon

I did that. 79-82A on supply conductors to RPC with all 3 units running

 

[retirede]

I did that. 79-82A on supply conductors to RPC with all 3 units running

You stated in post 1 that the 3 phase currents totalled 78A.

It’s virtually impossible that those 2 readings are happening at the same time.

 

[__dan]

Generally RPC's have poor line to line Voltage matching. If you measure (load side) Voltage line to line , A - C, A - B, B - C, there's likely to be a large Voltage imbalance exceeding 1%, which causes a phase current imbalance. you would have to Amp clamp each phase for each load and likely there is a problem. Line to line Voltage imbalance for a three phase motor causes excessive heating in the motor.

Locked rotor starting current on the chillers could likely be an issue. Generally (old code rules), the supply conductors would be 125% but the upstream short circuit protection, the breaker, would be at 250%, to allow carrying motor starting current, applied to the largest motor for motor groups.

At 50 Amps three phase largest chiller, 125% and 250% is 62 Amps conductor size, 125 Amps breaker size.

Translating that through the RPC going from single phase to three phase at the same voltage (240 V ), kVA through the RPC stays the same neglecting efficiency losses, line current goes up by 1.732 at the same load single phase.

So the 62 A, 125 A, x 1.732 goes to 108.25 Amps supply conductor size (possibly the RPC rating also), 216 Amps RPC breaker size.

That's only for the one largest.

In the code rules I don't know if the 250% motor breaker allowance would be allowed upstream of the RPC. But most of the effect of LRC starting currents would be present at the upstream RPC breaker, with some mitigation by the impedance of the RPC itself.

So in sum, the RPC generated voltage imbalance could have an effect which is bad for the motor, including making it harder to start or increasing starting current time. Plus, the RPC breaker is not sized with an allowance to carry that motor starting current.

I don't know what the code would allow for the RPC breaker with motor starting load, with consultation with the manufacturer, though I have installed them myself (not that big).

That is if there is no other issue

An installation like that, operating and maintenance costs, the electric bill, is going to be far larger over time compared to the installed cost. I suspect here is an issue there, especially if the use is continuous year round commercial.

If you or they look at it from the point of view of total costs over the life of the install, possibly new equipment and a new 3 phase service could come into view. The new chillers could come with vfd's and are much more efficient at part load.

If everything as is, is working correctly (breaker trips properly for being undersized), given the concerns about what they did not look at, at the time of the original install, it's possible they might want to start with a blank sheet of paper again and see what it is they really want to do (cooling load wise) and how best to go about it. Then they can take the next step. It is the operating costs that will really kill them over time.

 

[winnie]

I did that. 79-82A on supply conductors to RPC with all 3 units running

When you said:

RPC is feeding three 3 phase chillers. 1- 50A, 1- 40A, 1 30A. (actually running loads are approx. 35A, 25A and 18A respectively). Total connected running load when all three units are running is 79-82 amps per phase. There are no other loads on the RPC

Were the above currents measured between the RPC and the chillers or upstream of the RPC?

Are there any breakers between the RPC and the chillers?

What are the nominal current numbers from? (The 50, 40, 30) Do you have the max OCPD and min circuit ampacity numbers for the chillers?

If the running load is 79-82A as measured at the feed to the RPC, then I'd expect starting current to exceed 82A by quite a bit. It may be that starting current is tripping the breaker.

If the undersized SER is long enough then voltage drop during starting could increase starting duration, which would increase the chance of tripping.

Since you've already measured the running current on the single phase side of things, the IMHO the next step is some sort monitor that measures current and voltage vs time so you can see the values and evaluate starting events.

Jon

 

[collinkonow]

Zu

Generally RPC's have poor line to line Voltage matching. If you measure (load side) Voltage line to line , A - C, A - B, B - C, there's likely to be a large Voltage imbalance exceeding 1%, which causes a phase current imbalance. you would have to Amp clamp each phase for each load and likely there is a problem. Line to line Voltage imbalance for a three phase motor causes excessive heating in the motor.

Locked rotor starting current on the chillers could likely be an issue. Generally (old code rules), the supply conductors would be 125% but the upstream short circuit protection, the breaker, would be at 250%, to allow carrying motor starting current, applied to the largest motor for motor groups.

At 50 Amps three phase largest chiller, 125% and 250% is 62 Amps conductor size, 125 Amps breaker size.

Translating that through the RPC going from single phase to three phase at the same voltage (240 V ), kVA through the RPC stays the same neglecting efficiency losses, line current goes up by 1.732 at the same load single phase.

So the 62 A, 125 A, x 1.732 goes to 108.25 Amps supply conductor size (possibly the RPC rating also), 216 Amps RPC breaker size.

That's only for the one largest.

In the code rules I don't know if the 250% motor breaker allowance would be allowed upstream of the RPC. But most of the effect of LRC starting currents would be present at the upstream RPC breaker, with some mitigation by the impedance of the RPC itself.

So in sum, the RPC generated voltage imbalance could have an effect which is bad for the motor, including making it harder to start or increasing starting current time. Plus, the RPC breaker is not sized with an allowance to carry that motor starting current.

I don't know what the code would allow for the RPC breaker with motor starting load, with consultation with the manufacturer, though I have installed them myself (not that big).

That is if there is no other issue

An installation like that, operating and maintenance costs, the electric bill, is going to be far larger over time compared to the installed cost. I suspect here is an issue there, especially if the use is continuous year round commercial.

If you or they look at it from the point of view of total costs over the life of the install, possibly new equipment and a new 3 phase service could come into view. The new chillers could come with vfd's and are much more efficient at part load.

If everything as is, is working correctly (breaker trips properly for being undersized), given the concerns about what they did not look at, at the time of the original install, it's possible they might want to start with a blank sheet of paper again and see what it is they really want to do (cooling load wise) and how best to go about it. Then they can take the next step. It is the operating costs that will really kill them over time.

Thanks Dan. All great info. I'm going to continue to look into all of that.

 

[collinkonow]

When you said:


Were the above currents measured between the RPC and the chillers or upstream of the RPC?

Are there any breakers between the RPC and the chillers?

What are the nominal current numbers from? (The 50, 40, 30) Do you have the max OCPD and min circuit ampacity numbers for the chillers?

If the running load is 79-82A as measured at the feed to the RPC, then I'd expect starting current to exceed 82A by quite a bit. It may be that starting current is tripping the breaker.

If the undersized SER is long enough then voltage drop during starting could increase starting duration, which would increase the chance of tripping.

Since you've already measured the running current on the single phase side of things, the IMHO the next step is some sort monitor that measures current and voltage vs time so you can see the values and evaluate starting events.

Jon

1-Those were measure on the line side of the RPC because that's where I'm having the issue.

2-I don't have MCA and max OCPD handy but I'll be back there this week

3- I don't disagree with starting current being the issue.....but it ran fine for 3-4 years.

4-The run is under 100 feet

5- I think thats a good idea thanks!

 

[__dan]

You really need the RPC specs and instructions, load side kVA rating, load Amps, upstream breaker size. The vibe I am getting is it was probably sized tight correctly at the time but never for all three units simultaneous.

Does not matter if they could start and run ever, what matters is the numbers they worked with at the time of the original install were not calc'ed for all three at the same time.

That's where "it ran flawlessly", you want to know if that statement is true or false. The load calc would show it's false. It's a "flaw" if the supply is not big enough per code.

I'm getting 90 Amps line side for the #2 AL at 75C. Loaded to 80% for continuous type, heating, and cooling equipment, your max load as is, is 72 Amps singe phase or 72 x 240 = 17.2 kVA running. That's it. That's how big your supply is.

I suspect the RPC VA rating running continuous is in the same range.

At the load side it's 17.2 kVA = A x 240 x 1.732, is 41 Amps running load side 3 phase 240 V. So you may have been sized for one 50 A unit or the two smaller ones at the same time but nothing else. 41 Amps load side 3 phase is how big your supply is.

That's where looking at the customer cooling load, do they ever need to or intend to run all three at the same time.

There is a problem with the customer expectations if they say yes they do and want to, now fix it.

 

[GoldDigger]

The combination of RPC and motor is likely to have a significantly lower power factor than the motor alone, especially since the RPC must be oversized compared to a motor with the same current.
One result will be higher than expected current on the input side, resulting from the lower power factor.