reliability of Chillers.idf entries?

asked 2017-12-07 11:16:03 -0600

Matt Koch
1998 ●3 ●17

updated 2017-12-07 12:49:07 -0600

10147 ●1 ●19 ●26 https://www.nrel.gov/b...

I was looking through various chillers around 425 ton in the EnergyPlus-8.7.0\DataSets\Chillers.idf file (Chiller:Electric:EIR):

Trane CVHE 422 ton VSD York YT 425 Ton VSD Carrier 19XR 443 ton VSD Carrier 19XR 465 ton Vanes McQuay WSC 432 ton Vanes

All of these have their Optimum Part Load Ratio set to 1.0. Even for the two Vanes chillers, I don't believe this is true, and certainly not for the VSD chillers. When plotting EIR = EIRFT*EIRPLR/PLR/COPref vs. PLR, the Vanes chillers have their optimum PLR (point of best COP according to EnergyPlus I/O reference) near 75% and the VSD chillers have their optimum PLR near 40%.

In addition, most of these have non-standard Entering Condenser Water Temperatures.

Just curious, how useful/trustworthy are these chiller curves, or is the optimum PLR usually the only major discrepancy in these?

edit retag flag offensive close merge delete

Comments

@Matt Koch, do you know how is the Optimum Part Load Ratio used in the simulation? I quickly looked at the engineering manual (for the chiller model) and didn't see a mention of it, perhaps I missed it.

Jeremy ( 2017-12-08 12:04:11 -0600 )edit

I do not know and just assume that if it is asked for, it is being used? I could see this being of use in the Optimal Load Distribution Scheme, for example. But maybe you are correct, and this is just a dummy parameter and thus set to one for simplicity.

Matt Koch ( 2017-12-08 12:46:27 -0600 )edit

You are probably correct, the Optimal load distribution scheme seems like an obvious application!

Jeremy ( 2017-12-08 13:00:37 -0600 )edit

add a comment

answered 2017-12-09 10:20:42 -0600

Matt Koch
1998 ●3 ●17

I just did a quick and dirty experiment with two water-cooled parallel chillers in a loop set to the Optimal load distribution scheme and connected to an old load profile. When I left the chillers' Optimum PLR at 0.3, I got 2,160.14 GJ for their electricity consumption, when I changed their Optimum PLR to 1.0, I got 2,143.47 GJ. So, even though I would have expected a higher rather then lower energy consumption by "de-tuning" the loop, at least this seems to prove that the Optimum PLR is being used in the calculations and therefore probably out to be set to a realistic value.