Optimising parallel water chillers for energy saving

Hello All,
My colleague and I are currently tasked with examining the energy consumption of the whole of our site.

One thing that I am questioning is the (short) cycling of a pair of TRANE RTAB213 water chillers. Each unit has 2 compressors A and B. Both chillers are connected in parallel on the same water circuit, and both set for similar operating parameters.

These are: Setpoint 7deg C, Differential to start 1.1deg C. The only difference appears to be the 'design delta temperature setpoint' which is 8.5deg C on Chiller #1 and 7.5deg C on Chiller #2. Both are set to Setpoint: 'Local' and Mode: 'Auto'.

They are controlled (?) by a TRANE Sequencer, and the operation has been observed and recorded thus:
(Note: Compressor B runs in both chillers rather than compressor A due to the difference in run hours of each compressor) Return temperature reaches 8.1deg C, Chiller #1 starts runs at 90%FLA, 2 mins later Chiller #2 starts running at 50% when its return temperature reaches 8.3deg C, 3 mins later Chiller #1 stops as output water reaches 4.9deg. Chiller #2 continues to run (at 50%FLA) for up to 45mins!

Consequently Chiller#1(B) has low running hours but a higher number of starts than Chiller#2(B).

This is one set of readings:
Ch#1(A): 21152 starts, 12123 run hrs.
Ch#1(B): 17588 starts, 7237 run hrs.
Ch#2(A): 20114 starts, 30177 run hrs.
Ch#2(B): 12167 starts, 15497 run hrs.

It seems that the sequencer might be working in the fact that it starts the one with the lowest run hrs, but it appears that the fact that both start and #1 then shuts down to be wrong and wasteful of power, to say nothing of motor/compressor wear.

The system is presumably a 6 / 12 deg c design, so should the setpoint(s) be raised and perhaps a higher differential to start be set on Chiller #2 ?

Your ideas are welcome, especially from the TRANE guys who might have come across similar problems before.

2 questions.

What is the water cooling and what is the total water volume?

EDIT:
Sorry just thought about another thing, what is the system layout?
(2 primary circuits on one secondary? Mixing vessel or loop? Heat exchanger?)

Hi Viking,
Water is for (large) building airconditioning.

I'm afraid that I don't know what the volume of water is, but there are no buffer tanks at the chillers, and it is probably a loop with proportional valves on each air handler.

Its piped up in 12" / 300mm pipe at the chillers with both shell and tube evaporators piped in parallel.

My thoughts are that either you got a problem with the water flow (most likely) or the controls are stepping up too rapidly.

You will have to investigate the water system further, preferable sketching up a pipework diagram, to get an understanding of what water flows you got in the system and how much water the chillers got to work with.

Good luck.

Quote:

---

Originally Posted by Electrocoolman

These are: Setpoint 7deg C, Differential to start 1.1deg C. The only difference appears to be the 'design delta temperature setpoint' which is 8.5deg C on Chiller #1 and 7.5deg C on Chiller #2. Both are set to Setpoint: 'Local' and Mode: 'Auto'.

They are controlled (?) by a TRANE Sequencer, and the operation has been observed and recorded thus:
(Note: Compressor B runs in both chillers rather than compressor A due to the difference in run hours of each compressor) Return temperature reaches 8.1deg C, Chiller #1 starts runs at 90%FLA, 2 mins later Chiller #2 starts running at 50% when its return temperature reaches 8.3deg C, 3 mins later Chiller #1 stops as output water reaches 4.9deg. Chiller #2 continues to run (at 50%FLA) for up to 45mins!

---

I must be missing something here. If the entering water is 8.3C and the leaving water is 4.9C = 3.4K dT, then where is the 'design delta temperature setpoint' of 8.5 or even 7.5? What delta temp are they controlling and how are they controlling it?

Quote:

---

Originally Posted by The Viking

My thoughts are that either you got a problem with the water flow (most likely) or the controls are stepping up too rapidly.

You will have to investigate the water system further, preferable sketching up a pipework diagram, to get an understanding of what water flows you got in the system and how much water the chillers got to work with.

Good luck.

---

I agree. It seems almost as if the water stops flowing through chiller #1 as soon as chiller #2 starts up. How else could chiller #1 sense the 7C incoming water temp it takes to shut it down?

I'm thinking an accurate piping diagram would tell the story.

It's entirely possible that one of the chillers is intended to be held in reserve, shut down and valved off while the other chiller handles the load.

Hi Gary and Viking,

Thanks for your input.

Yes Gary, I was wondering what the 'design delta temperature setpoint' was and how it figured in the operation of the chillers.

It does seem to point to the thought that the load on the chillers is too little for both to be operating together.

Viking, I'll try and take some pictures of the pipework and post them.

Before I jump in and change anything (i.e. setpoints, differentials), I was wondering if any one else had any ideas.

I know that one or two of you are Trane guys.

Thanks, ECM

The chiller controllers (UCM or UCP) will take the design dT in the calculation to load/unload. If there is a sequencer installed on them, is this a duo-controller or is there a BCU?
If both chillers are enabled it is possible that the sensors are off. This can only be corrected by Trane. The design dT should none the less be set at the right number. That is the actual dT when the chiller is at full load. But don't look at the %RLA. The chiller will onl give 100% RLA when it's at full load and rinning in maximum OAT!