Chiller Entering

[SLBforever]

Where I work, we have an HVAC sytem with a single air cooled Chiller.
The Chiller as two identical COPLAND reciprocating compressors.
The Chiller capacity is controlled via the demand side.
We can only establish a set point in the entering water not in the leaving water.
With this strategy, don't we lose a certain degree of control?

Thanks.

SLB

 

[lilliput1]

Yes you should control the leaving water temperature. Move your sensor controling the setpoint to the leaving side.

 

[SLBforever]

lilliput1:
Thank you for your tip.

More details about the Chiller:
The Chiller has a controler with a interface were you can visualize several parameters: ECWT, LCWT ...
You can define a set point in the LCWT and for sure in the ECWT. However I don't know the code number witch define the parameter(ECWT or LCWT).
I don´t have the controler's manual.
The chiller is a CIAT(brand),IAB 600.2 (model).The controller is a MRS3-4.1.
This brand is from France, so probably it only sells here in Europe.
Does anyone knows this machine?


SLB
Lisbon, Portugal

 

[friartuck]

I asked the same question of a chiller manufacturer and they said that the best way to control the chiller is on a return sensor rather than a flow detector. I must admit, I have since forgot the reason why.

Perhaps there is soneone out there that can explain this oddity.

Friar Tuck of Sherwood

 

[lilliput1]

If your chilled water supply is higher than the design temperature, you would loose capacity specialy capacity for latent loads. If have multiple coils you would want to make sure you would be providing enough capacity to the coil with the maximum requirement. You would loose this ability if you control for the return.

 

[SLBforever]

Again, thank you all for your help.

At present the ECWT as a 9ºC set point.
The first band value is 10ºC and the second is 11.5ºC, i.e.when ECWT reaches 11.5ºC the second compressor(lets call it this way), starts and we have 2 compressors working.
After a while the ECWT, falls and at 10ºC the first compressor stops.
At first the ECWT falls further to 9.7ºC, and then starts to raise and the process repeats.
If the ECWT continues to fall(low demand)and reaches 9ºC, the second compressor stops and there are no compessors working until the ECWT reaches 10ºC again.
With this process, the LCWT varies from about 6.9ºC to about 9.5ºC.
Today, I personaly monitored the chiller performance.
We has a hot day with a maximum outside air temperature of 33ºC degrees(a week ago we had 43ºC).
We had three periods of about 40 minutes each with 2 compressors on.
The loads are a AHU for outside air and 40 fancoils.
The people on the offices in the west facade of the building, keep complaining about the heat.
However I thing we have Chiller capacity to deal with the load, otherwise the should have more periods with 2 compressors on.I thing West facade fancoils don't have enough power.
But I still think we should control the LCWT not the ECWT.

SLB,

 

[AEBuck]

Perhaps I too suffer from friartuck's malady, but I recall that older chiller control packages utilized return-side control because this temperature is much more stable than the supply-side. Nowadays, since modulation capability is more refined and control algorithms are more sophisticated, supply-side control is much more common, and typically preferred per lilliput1's description. Any factory engineer's in the crowd who care to confirm this speculation?

Does your building have a DDC system throughout? Do you employ an ECWT set point reset strategy to help satisfy the most demanding zone? I'm guessing there are 3-way valves at all coils. Is there a balancing valve in the bypass leg? Is it set correctly? Have you checked water flow rates on the west side during times of complaints? Is this facade at the end of the supply main? If I interpret your temperature measurements correctly, it seems you have ample overall flow. In our region we generally see and use a higher water temperature difference - anywhere from 10-to-14F - while yours is around 8F. Do you have original design documents? How due current conditions compare? Based upon your limited observations, I would agree that you seem have have ample chiller capacity, but this is subject to change if your chiller can't "see" all of the loads.

 

[HVAC68]

I am not sure whether there is "ample" chiller capacity, since the ambient temperature was only 33 deg C as against what was 43 deg C a week ago.

HVAC68

 

[AEBuck]

Thank you, I'll probably stand corrected on that item.

 

[SLBforever]

As far as I know, there is no possibility to implement an automatic ECWT reset strategy.
The possibility of a set point reset strategy is not implemented in the computer witch manage the AC system, but I’ll investigate if it can be implemented.
We have to go the chiller and manually reset the ECWT in the controller interface.
I’ll find out if is there any drawings of the building but I don’t think so.
The building as two floors: in the first floor there are 12 fancoils plus an AHU, and 28 on the second floor.
Each fancoil unit has a 3 velocity fan, a tree way modulating valve and a electronic controller linked to the computer witch manage the AC system.
I didn´t checked the water flow rates, but I was told that several modulating valves have been replaced.
I am almost certain, that the fancoils are all of the same size, witch if it is so, is an absurd: west facade units should have more cooling power, that’s for sure.
So I still think that west facade units are undersized

One other thing: I think the designer could have chosen a better chiller, because the partial load control strategy with this chiller is very poor: we have two on/off compressors.

Wouldn’t a small centrifugal chiller had been a better choice?


SLB

 

[absrbrtek]

Just from experience, your going to see more compressor cycling using LCWT then ECWT when using recip compressors on chillers. You do not have enough steps of cooling to control it tightly enough not to short cycle. LCWT is mostly used for centrifigal, screw and absorption chillers that have a more finite load control.

 

[sterl]

I agree with 'Tek on the ECWT vs LCWT...Its essentially a capacity balance routine where the big, integral steps don't lend themselves to low capacitance systems, distributiive control type systems...Unless your compressors are equipped with unloaders and some subtle sequencing controls.

With all those satellite zone controls, it would seem that pumped CW flow would be of large concern....Is the West zone a long ways from the chiller package? Are all the Units and Valve packages similar in size? Is the AHU controlled by pressure and humidity as well as temperature or just by temp?

 

[briand2]

With reciprocating chillers, unless you are having real problems with maintaining conditions in the areas served, just stick with control based on return temperature. Control on flow temperature is, whilst being more accurate in some circumstances, will lead to less stability. If at lower loads your system is happy with slightly higher (but stable) flow temperatures, then return temperature control is fine. An exception is if you're using chillers with infinitely variable control, such as centrifugal chillers, in which controlling the flow temperature is normal.

 

[MedicineEng]

SLB:

Some of the points that I will raise were already said but I think htat it is worthwhile to come to it:

Why don't you check your water network balancing instead of concentrating in the chiller control? Maybe your problem is that you don't have sufficient water arriving to your West wing of the building. You can make some experiences closing coils in other areas and see if the situation improves. Now that is vacation time in Portugal, maybe you can perform these tests without having your colleagues calling you every 5 minutes.

Other thing that you should also consider is the insulation, if the distance is big and having high ambient temperatures, you might be loosing some cooling capacity here.

Consider also the cleaning of the coils. Are they being cleaned frequently? If you have a problem of balancing, your West coils will be more prone to scaling.
Most of this checking you can make it with a infrared thermometer. If you don't have one, there are models relatively cheap (around 100 Euro or less).

Hope that helps.

Cumprimentos de um Português na China

Paulo

 

[SLBforever]

Paulo:
Boy, you're so far from "home"!
Como é que é para um português, viver na China?
I'm gonna try to follow your tips about the AC instalation.
You may very well be right.

Cumps para ti de

Vítor,
Lisboa, Portugal.