Baseline chiller unusual part load COP

5 posts / 0 new
Last post

Hi everyone,

I am working on a office building project (e-quest 3.65-7175) in which
baseline case is modelled following ASHRAE-90.1-2010 standard.

Two centrifugal type chillers of 690TR each are assigned to the chilled
water loop with full load COP value entered as 6.16 in chiller input tabs.

As I was analysing the hourly reports for chiller, it was observed that
chiller COP values (based on corrected EIR values in hourly report) are
quite high ranging from 11 to around 25 at part load values in the range of
50% to 0.1%.

This seems very unusual considering that as per chiller manufacturer part
load datapoints the part load COP values starts decreasing below 35-30%
part load; whereas using e-quest default chiller curves, the chiller COP
value seems to go on increasing even at lower part load values.

Has anyone observed this before, or am I going wrong somewhere?

Also due to above issue, when part load data points provided by chiller
manufacturer are entered in proposed model and default e-quest performance
curves are used in baseline model, expected energy savings under space
cooling consumption are not achieved during comparison.

I have attached screen-shots showing e-quest default curve coefficients
(EIR-FT, EIR,-fPLR & Cap-FT curves) for centrifugal chiller and a sample
manufacturer's part load details for reference.

Do I need to change the default curve coefficients from e-quest to get more
accurate results? If so, please kindly let me know the changes to be made
or the process for finding correct part load curve coefficients for chiller.

Thanks in advance.

Rushi Quest's picture
Joined: 2019-12-11
Reputation: 0

Yes ? you should develop curve fit parameters for your specific chillers, with the appropriate reset conditions, etc. eQUEST includes a curve-building function if you enter the performance data there.

Also see what the weighted average of the COP is against load, maybe it isn?t as far off as it seems at first with a few outliers, or it may be way off if the majority of hours are at part-load.

The idea that you should use a consistent set of curves is true though, using the default curves based on some historical piece of equipment vs the curves from a modern chiller today may not be consistent, it would be best if your chiller vendor can provide some data for a code-minimum unit to make curves for the baseline based on the baseline capacity, etc.


David S. Eldridge, Jr., P.E., BEMP, BEAP, HBDP, GGA

Direct: (847) 316-9224 | Mobile: (773) 490-5038

Grumman/Butkus Associates | 820 Davis Street, Suite 300 | Evanston, IL 60201
Energy Efficiency Consultants and Sustainable Design Engineers | Blog | Facebook | Twitter

David Eldridge's picture
Joined: 2012-05-08
Reputation: 1

Thanks David for taking out time to reply, will go through your suggestions
and follow up for any further queries.

On Thursday, March 25, 2021, David Eldridge

Rushi Quest's picture
Joined: 2019-12-11
Reputation: 0


David beat me to it, those are good resources. But you might get a headache.
The default curves unloading unrealistically well is problem we are aware of and have run into many times. It has a real impact; chillers don?t unload like the curves describe.
For history, I believe the default curves are based on data from the COMNET 2011 reference guide, but don?t think I?ve been able to get to the root source of what chiller data they?re supposed to be based on. These are also reproduced in the PNNL Guides for ASHRAE 90.1-2010 & 2016. If any one knows original source for data, would love to review it.

The path I?ve taken typically is to either 1-run default curves in Proposed & Baseline if we can?t get data needed from a rep. 2- get Proposed Curve information and if the chiller types are the same in Proposed & Baseline (Screw/ Centrifugal) use the same in both 3-if the rep will get us data for Proposed & Baseline, use that data.
The availability of generic, but realistic curves would greatly assist modelers.

daric adair

Daric Adair's picture
Joined: 2017-10-21
Reputation: 0

Hi David,

Thanks for all the suggestions.

I too was following path-1 (default curves in baseline and proposed) till
now, however sometimes, as you have mentioned the unrealistic unloading
issue results in low space cooling consumption (even lower than the
building lighting consumption which is quite unexpected).

Path 2 seems to be only possible, if chiller types are same in both
baseline and proposed case (which is not always the case).

I think path 3 may be followed where we need to get both baseline and
proposed data from manufacturer rep. The only problem that I see with this
is that it will be hard convincing them to fetch data for baseline chiller
(whose chiller type and capacity is different than what he has supplied for
the project).

As you suggested, availability of more realistic curves as per ASHRAE
standards will really be of great help to modelers. This will also lead to
having some constant baseline part load efficiencies to compare with for
various types & capacity of chillers.


On Friday, March 26, 2021, Daric Adair

Rushi Quest's picture
Joined: 2019-12-11
Reputation: 0