VFD Centrifugal chillers - suitable for locations with very little wet bulb changes 4

[HVAC68]

Are VFD Centrifugal chillers suitable and/or worth their investment and additional maintenance costs for locations where the wet bulb changes are very less ?

HVAC68

 

[DRWeig]

Is your internal load constant too? What sort of building is it?

It is not just the outdoor wet bulb that governs load. The sun, which varies from maximum at noon to zero at night, people, equipment, ventilation, and other factors play into the variability of chiller load.

In general, in the US, the investment is a very good one in a new building. Changing an existing chiller to a variable-speed chiller is questionable, though.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[KiwiMace]

I worked in Singapore in the early 2000's and we were advising against installing vfd chillers in hotels and institutional buildings at that time where the 24hr daily range was between 50-100% max and changed little over the year. We usually went with a quantity of big units running at peak efficiency point and a pony chiller for trim. However, the vfd option price step was large back then and it since come down substantially.

If you have an unoccupied period in your building operation, with a significant night setback, it could be worthwhile. Ultimately, it will come down to your client and the efficiency target. A vfd chiller will certainly save some energy and not all energy decisions are based on a short term ROI.

 

[urgross]

VFD chillers are stable, and I don't know of any additional maintenance costs. As noted above, you need to figure out whether you will be running part load, and how much of the time you will be running part load. As you have no latent load, that makes calculations (ASHRAE 90.1 Appendix G) much easier.

 

[EnergyProfessional]

I can't imagine an application where a variable speed compressor would not make sense. Even if load is nearly constat, are you really designing for 100% of the calculated load and get a model with the exact capacity? Probably not... and all building applications have variations, no matter what climate.

variable speed equipment is pretty standard nowadays, not much upcharge anyway.

 

[ozmosis]

Here's a useful article on a VFD chiller retrofit where the wet bulb delta is relatively small.

http://www.turbocor.com/literature/pdfs/casestudies/juvenile_hall.pdf

Disclaimer: I work for Danfoss, although not for the Turbocor group.

 

[HVAC68]

Thanks for the responses. This is a 24-hour application - a data centre to be precise. The internal loads are more or less constant. The external loads vary considerably thanks to the outside temperature variations.

2 schools of thought. One pitching in for VFD regardless of whatever the climate. Second one saying that if the wet bulb differences or delta is not significant, VFD is an additional piece of equipment that has to be maintained.

I am referring to a WATER-cooled centrifugal chiller !

HVAC68

 

[DRWeig]

Attachment is a pretty authoritative piece from an independent lab (HP).

Also, google up this phrase:

variable speed chiller data center

The first BUNCH of hits will be: a) the HP paper I've attached, and b) all the various chiller manufacturers' case studies, technical data, and propaganda.



Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[SAK9]

HVAC68,
There seems to be little merit in going for VFD chillers in your case.If my memory serves me correctly,the control algorithm for VFD chillers requires a change in lift for the VFD to effect a change in speed.A lift reduction happens eiher by lower condenser temperatures or resetting the chilled water temperatures higher.
Since your wet bulb is stable,there is not much variation in your condenser water temperature . Data centres use a higher than normal chilled water temperature leaving very little room for CHW temeperature reset either.With a fairly constant lift the VFD case is not on in your case.

 

[KiwiMace]

Yeah, I doubt it would be worthwhile.

 

[KiwiMace]

I just read that turbocor case study. Interesting enough but San Diego has a fairly typical wet bulb variation so I would hardly call it relevant. The San Diego 0.4% monthly design figures range 8 degC (14.4 degF) vs an equatorial country like Singapore where the range is 1.3 degC (2.3 degF).

 

[zuccus]

HVAC68,

Don't assume that just because it is a data center that the loads will stay constant. Depending on the applications/services being maintained on those servers/storage, the load can vary wildly or stay pure constant.

One data center processing facility I did had an almost constant load 24 hours/day, 365 days a year. It was processing large amounts of data that was staged.

Other ones I have done, especially if they are support internet apps or office applications, can vary as the demand comes online, as the system grows, etc.

Zuccus

 

[RobsVette]

Is this an upgrade or a new installation? if its an upgrade another thing to consider is the need to change our the motor (unless you have an open drive machine, like a York.)

If its a closed drive machine and you have to change the motor it could really be a pain. (VFDs usually need a motor with class F or H insulation) Plus if anything ever happens and the motor goes bad, you could stand to cause a serious problem with you refrigerant in the machine.

As previously stated if the load on the machine varies constatly then it is a good investment. Although it is important to remember that a chiller with a VSD is actually slowing down the motor to save energy, thus reducing the pressure output of the compressor (assuming its a centrifugal compressor) and resulting in teh requirement for raising the CHW temperature or lowering the CW temperature. The alternative to this is you need a compressor outfitted with a Varaible Geometry Diffuser to handle the low load high head conditions. So now you may end up with a new compressor as well.

It is important to consider all these items in your evaluation.

 

[HVAC68]

Thanks for the inputs. Was busy with a few things and hence couldn't log on earlier. This is a new installation. Another issue that has come up is the restart time required, since this is a critical data centre installation. Eventhough there is redundancy available in the form of a standby chiller, client still wants to know what would be the restart time for the chiller in the event of a power shutdown. Weighing the pros and cons of a VFD vs traditional starter from the point of view of restart time as well. Will update as I get more inputs.

HVAC68

 

[KiwiMace]

Starting the chiller with a vfd or not won't be a big factor in your restart time following power outage. What matters most is that when the chiller gets power back, the chiller controls initiates a sequence to ensure the oil system and compressor are all well before allowing a restart. This can be 15 mins. You can request a fast-start microprocessor control system as an option (~5 min I think), but there are warranty implications. Some owners go for fast start, others have chillers on independent power feeds and/or chilled water storage to maintain operational continuity while the chillers restart on their normal schedule.

 

[Waramanga]

Watch out for what condenser water temp the chiller manufacturer quotes their 'incredible' turn down. I have a chiller on a job that can turn down to 10% when the cond water temp is depressed, and only 70% when it isnt.

 

[urgross]

If wet bulb temperature is not an issue, and computer centers can use a higher (64*F here) supply temperature, have you thought about a chiller with evaporative cooling? Depending on size and loading, inlet guide vane control might do you as much good as VFD.