VSD Centrifugal Chillers

[SAK9]

Guys,

Has anyone had experience with variable speed drive centrifugal chillers? Do they have any significant advantages over normal chillers with inlet guide vane control?

 

[quark]

We had 500TR York millenium centrifugal chiller with VFD control(and vane control as well). We had two chillers without VFD just for comparision and York promised to give back VFD cost incase it didn't work.

At our load conditions and ambient conditions, the VFD chiller was working perfectly. I observed 5-10% savings.

Believe it or not : Goldbach's conjecture says any even number can be written as a sum of two prime numbers. Postulated in 16th century, nobody could, so far, prove it nor disproove it.

 

[VSSriram]

The extent of savings would depend on the load pattern - i.e. how much of part load requirement exists and the range of condenser entering temperatures which is primarily dependent on the ambient wet bulb temperature - i.e. the savings become more when the equipments operate more and more at part loads due to reduction in cooling load. If your cooling load is a constant one like a process requirement, then the savings would be due to reduction in ambient wet bulb temperature only.

A typical pay back period would be around 2 to 3 years.

 

[Rob1972]

Another thing to consider is the number of chillers in your plant. If you have a single chiller application, you have a lot of diversity in your load profile and the weather conditions are favorable (as indicated by VSSriram) then the variable speed chiller might be appropriate. However, if you have more then one chiller, then the VSD chiller may spend more hours at higher loads which may lengthen the payback period. A lot depends on the system design and site specific conditions so the best approach is to do an analysis.

 

[VSSriram]

Absolutely right Rob1972. Carrier has a very nice program, which can help you out on selecting the right type of system depending on the type of your application. It's called "Chiller System Optimizer". It's an inhouse software program and they don't give it outside. The more the accurate your input is, the more the accurate the output would be.

 

[simonho]

The benefit of VSD centrifugal chillers will largely depend on its utilization range. If your application does have varying loads and also low ambient temperatures, it will suit VSD chillers. Remember that VSD chillers are efficient at part load with low ambient temperatures (cool condenser water temperatures). If you don't get these conditions, then the VSD chiller(s) run at near full load and you don't get the great efficiencies the chiller manufacturers tell you.

In multiple chiller plants, the chillers are typically fully loaded so there's little advantage for VSDs. Even if you do get low loads and low ambient conditions, chances are if it is a commercial application, you might have an airside economy cycle which limits your chiller operation anyway. No matter how efficient your chillers can be, you cannot beat free cooling!

It all boils down to the return on your investment: would you pay that extra for a chiller to have a VSD or would you use the same money to buy an even more efficient chiller ? e.g would you buy a chiller with VSD with 0.6 kW/ton or a constant speed chiller with 0.5 kW/ton if they were priced the same?

The answer lies in comprehensive analysis/simulation programs, one like VSSriram mentioned. Trane also has one called the System Analyzer which does similar comparisons. The answer is it depends...

 

[psighi]

The vfd chillers were considered because of load diversity from day to night (24-hour operation). If you compare the IPLV's of the vfd vs. non-vfd chillers the efficiencies are quite significant. The efficieny curves go much lower at partial loads and stays lower down to 20% capacity.

Caution: I not only have two 750 Ton Carrier Centrifugals with vfd, but also variable primary pumping. Details need to be engineered properly for the vfd portion of the chiller to maintain optimization. Make sure you optimize your cooling towers for low ecwt, or potentially, the chiller will self protect and act like a non-vfd chiller at 60 Hz. Carrier seems to have licked this problem on the larger compressor sizes (>750T). I haven't seen the changes on the Trane or York...just precautions in operation.