Question about chilled water return temperature 6

[toast26]

I’ve just started looking into chillers, so my knowledge is pretty minimal. One thing in particular is bothering me. From what I read, reducing the chilled water return temperature will reduce the chiller load. I understand this, because if your temperature difference from the return and supply is less, then your tonnage will drop.

What confuses me is why this is important? It seems to me that the tonnage of the chiller really doesn’t matter, just the supply temperature, since that is what will doing the cooling in the cooling loop. If we have the same flow rate through the chiller, then the output temperature seems to be all that matters (or will it not be possible to keep the flow rate the same?).

Why I’m thinking this: I was wondering whether or not it would make sense to run a heat exchanger to cool the water before it enters the chiller.

 

[Drazen]

There can be some differences in trade's language, so simple definitions are necessary here.

Temperature can be lowered or raised, not reduced.

Increasing return temperature setting (if you can, that depends on your consumer's needs) will reduce chiller load.

Normally, return temperature is monitored and controlled as it gives you the picture of actual load at your consumers, while supply temperature is of secondary importance.

Reasons for inserting exchanger would be:
- to separate media for any reason
- to receive loads from many consumers with distinct size/temperature levels that could make problems to chiller (you don't want all that different temperatures to mix near chiller, and more problems can occur and partial load patterns)

If your only reason for idea of heat exchanger is to change temperature levels, mixer valve will do the job.

If you have large difference between consumer and supply circuits flows, headers with bypass arrangement are needed.

 

[ChrisConley]

You're actually describing a negative condition called 'low deltaT syndrome'. The chiller, rated at say 250 tons, is unable to deliver it's specified tonnage because the return water is coming back too close to the supply water temp.

In actual situations clients will have hundreds of tons of chiller plant capacity that they can't use because low deltaT. The largest culprit is usually 'overflowing' coils. The relationship between cooling output and chilled water flow is such that a large amount of 'overflow' will only marginally affect cooling output.

This can happen when three-way valves are used in a system, or when primary/secondary piping is used, or with dirty or improperly selected coils also with incorrectly sized control valves.

 

[MechEngNCPE]

What is the First Law of Thermodynamics, anyone?

Now, what type of heat exchanger are you talking about, do you have another cold source with extra capacity that you can use to cool your chwr? If so, is what is producing it more efficient than your chiller?

 

[Drazen]

Chris, it is interesting for me to learn that there is even term for that syndrom.

What you desribed I saw in some industrial plants where chiler is incorrectly specified - HVAC chiller is specified for process plant where consumers need temperature regime very close to chillers regime, and low delta t does not allow for modifying.

It took me almost year to convince one client to disconnect HVAC chiller and buy process chiller with about 1/4 of HVAC chiller capacity perfectly fit for that plant where beverage cooling had to maintain precise low temperature.

HVAC chiller worked on 6/11 minimum range, while process chiller was tailor-designed and specified for 4/9 degrees Celsius and all worked wll, with additional anti-freeze provisions.

 

[marcoh]

chiller cooling (kW) = 4.187 x flowrate (L/s) x delte T (oC)

If a chiller is selected for say 14oC entering temperature and 6oC leaving temperature, and the return water is actually 10oC (not 14 as per design) then the chiller will only need to generate half the kW to produce 6oC water, anymore kW and the suppy will go below 6oC.

So while you may have sufficient kW of chillers installed to meet the building demand, low building return water temperature may mean you cannot actually produce the kW required.

It is often an issue on larger systems due to incorrecly chosen chilled water coils, leakage in valves (especially if they get old), and the way some people put 3-way valves a the end of every run which just by-pass water and reduce the CHW return temperature

 

[toast26]

Thank you for all your responses. The way our system is set up is we have a free cooling heat exchanger that supplies chilled water during the winter (working with a cooling tower). Anytime the demand is too great or the ambient temperature is too high, we have electrical/gas powered chillers.

I was trying to see if there was a way to incorporate free cooling with the other chillers. The system is set up such that there is no way to use both at the same time to supply chilled water.

However, we can run the heat exchanger so it reduces the return water temp, such as from 55 to 50 F (effectively leading to a low deltaT syndrome).

I was initially thinking that lowering the entering temperature of the chiller may help increase the efficiency of the chiller, since the chiller would only need to lower the water temperature from 50-44 F rather than 55 to 44 F. I wasn’t particularly worried about reducing the load of the chiller, since we wouldn’t need anymore cooling.

I guess I failed to realize that reducing the load of the chiller will reduce the efficiency of the chiller, negating anything the heat exchanger does (hopefully what I just said is correct).

 

[ChrisConley]

Backing up now, the scenario you're talking about (free cooling HX) will reduce the load on your chiller. Low deltaT is a system problem, not a chiller problem. If your system is returning at 55F (which is design) then you are fine using a free cooling HX to reduce your chiller load.

This goes to NCPE's statement above, as long as the 'energy source' for your HX is lower than the chiller, this will reduce overall energy.

 

[MechEngNCPE]

Actually toast, part-loading of a typical HVAC chiller can give you better kw/ton, so higher efficiency. The efficiency goes way down below 25% part load, however.

 

[MechEngNCPE]

Just wanted to say well done guys, good points made, and +1 for all.