Changes in design of chilled water system over the years

[RobsVette]

Hi guys, looking for some feedback on the design of temperature differential for chilled water systems. Here is what brought this up to my mind.

Lately, I have seen a lot of 15 degree chilled water systems on new building here in the NYC area. Previosly, most of the 15 degree systems I had seen where for cooling data centers and had a range of 45F to about 60F. Which is fine because there is not alot of humidity in a data center.

However, I have seen alot of chilled water systems designed for temperature ranges of 42 to 56 lately. And just recently I saw one designed for 42F to 60F, which is for a building with lots of laboratories.

The question would be as follows;

- Conventional thinking on design of chilled and hot water systems was that on hot water system you can tolerate a larger temp change because the water distribution temperature was so far away from the air temperature that you wouldnt notice much of a difference in average temp.

With chilled water system I always assumed that a 10 degree diff was used to keep the effective average coil temperature as low as possible to be able ensure that the air moving over it was cooled properly.

Also, most of these systems where I am seeing a high temp diff are variable flow primary pumping systems. I dont know if that makes a difference, but I figured I'd mention it. (on an unrelated topic, I personally feel that these larger temp differences in conjuction with variable flow primary can cause havok with chiller control systems since all of the evaporation of refrigerant happens in one spot, but I digress)

Here is what I am thinking on the design end.

I assume the point here is to make the chilled water system as energy efficient as possible so they start with the coils and increase the size of the coils to give the air more contact time and get it closer to the effective coil temp. From there the pumping system energy can be reduced because of using a higher temper differential.

My only thought here is having a chilled water setpoint on the chiller of only 42F does require a decent amount more energy than does a setpoint of 44F. But I assume when viewing the system as a whole this is less enegy intensive than the standard 10F differential.

If my understanding of anything here is incorrect please let me know where I am incorrect as this is the point of this post, to gain insight into the changes in designs for chilled water systems due to the use of high temp diffs and variable flow primary pumping.

If anything other information is needed please let me know.

Thanks in advance for any help.
Rob

 

[Drazen]

I am not familiar with 10 degrees difference in commercial HVAC applications, it is only special-purpose use when you have other constraints.

In my opinion it is not issue of designer's selections, but thermodynamic laws that guide chillers efficiency.

Lowering temperature is "unnatural" process in terms of thermodynamics laws, and therefore the more you are lowering temperature the more you are losing on useful energy. That can be seen through Carnot process analysis which seemingly goes beyond level of this posting.

It is sufficient to be aware that some things cannot be comprehended by "hands-on" measurements and reasoning. In short, without low temperature difference chillers could not reach SEER ratios they have now, and if you look at SEER for some special-purpose chillers you will notice large difference - that is why they are special-purpose, used when they are really needed according to engineering judgment.

 

[SAK9]

ARI conditions for chilled water (44/54 F) holds good for most comfort applications. When this range is increased keeping supply temperature constant, it increases the chiller LMTD and its output. But the higher return temperature causes coil LMTD to drop. Coils become deeper to provide the same output and increases fan energy consumption. You would need to do an energy analysis to see what is the best operating range for your system on a case to case basis. Things such as how big the chiller is, how big the pumps are, what percentage of total power the fans consume can affect the analysis results. A high delta T of 18 or 20 will help reduce the size of pumps, piping, valves, switchgear, cabling etc and help bring the cost down though this may be negated to some extent by the increased coil size.

For a good primary secondary system, the production flow must be always greater than distribution flow.This would mean keeping the distribution delta T higher than chiller delta T.But for the variable flow primary systems, there is no such issue as production and distribution delta T are the same and therefore you can keep the chiller delta T as high as economic operation would allow.

 

[Zesti]

I'm in Europe, so it's degrees Celsius for us...

Anyway, usual temperatures for chilled water in HVAC is 6/12, so 6 degrees difference.

We have looked at 8/14 or 10/16 but as it turns out the increase in chiller efficiency is lost on extra resistance (air) of the larger coils in the AHU's.

So, as far as energy-consumption of the total system is concerned, there is not much in it.

 

[urgross]

Just finishing up a variprime with distributed booster pumps, runs 42*F setpoint. The part load efficiency is great, which is how it runs most of the time. Finished the bypasses, valves, dP's and controls, so now we typically have at most one booster pump running. Nice on the elctric bill.

The temperature setpoint was chosen based on application. If dewpoint control is needed, lower temperatures are needed, and chillers were selected on that basis (and PLV). If it were for a computer center, I'd select a higher temperature, and maybe consider supplemental evaporative cooling. If application were for chilled beam, I'd stay above the dew point.

I traded in my Carnot cycle for a Honda.

 

[MechEngNCPE]

Don't forget about dehumidification, in order to get say 48F LAT off a coil (to achieve a space of 68db/50%RH, having chilled water at 42 requires a ten row coil for a job I'm doing now. I don't know if I could do it with 45 degree chw.

 

[urgross]

That's why we used the 42* F, a 100% OA OR. We need to keep the discharge dewpoint between 47-48* F for the weather conditions.