chiller EIR curve

3 posts / 0 new
Last post

Hi,

I have a question about the chiller EIR curve in eQUEST. In the
attachment, there are two curves of chiller EIR. One is the kWH/Ton
curve from chiller manufacture. One the normalized EIR curve from eQUEST
default. These two curves showed quite big difference when the chiller
PLR drops to 30% or lower.

In my simulation, I used manufacturer data and this equation to
calculate the EIR, and then normalized the EIR to get the curve.

Cooling EIR = 0.941 * (1/(SEER/3.413))

Because about 20% of the chiller hours ran at less 20% of PLR, using two
different curves (calculated manufacturer curve and eQUEST default
curve) caused huge difference of chiller electricity consumption. So my
question is which curve is right for the chiller performance? Thank you
very much.

Xiaowei Xu

Xu, Xiaowei's picture
Offline
Joined: 2011-09-30
Reputation: 0

I have not spent much time researching the eQuest EIR curve origins, but I have spent years performing chiller ratings and analyzing centrifgual chiller part load curves under many circumstances.

In my opinion, the CVHE curve is much more representative of what the energy use will be for (1) chiller operating, than the eQuest curve. The nature of centrifugal chiller is, below 25% loading the kW/ton increases. Basically, system inefficiencies, which are relatively minor compared to power of compression at higher loads, are a bigger percentage of work input (kW) at low loads.

If you have mulitple chillers, more information is needed to make a determination of what your part load efficiency would be.

It would be good to get feedback from someone more familiar with eQuest EIR.

SEER is calculated differently than water cooled chiller performance, so that could be a source for differences as well.

I hope this helps,
Regards,

W. Peter Anderson's picture
Joined: 2011-09-30
Reputation: 0

Folks,
Don't be alarmed. The two curves in Xu's .pdf do not represent the same dependent value on the y-axis. The DOE-2 EIR f(PLR)curves essentially output kW (EIR(f)PLR x design kW) not kW/ton. The conversion to kW/ton has been discussed in previous posts, but, essentially, the "EIR"(kW) needs to be divided by the PLR (load, tons) to get kW/ton. The Trane curves probably include some condenser water reset, so a combination of DOE-2 curves would be needed to produce comparable outputs.

Fred Porter

Fred Porter's picture
Offline
Joined: 2011-09-30
Reputation: 0