[BLDG-SIM] Major disagreement on Chiller Performance Curves in DOE-2.1E and DOE-2.2

[Mike Tillou]

DOE 2.2 has a built in feature that allows you to enter raw data points and
create various types of performance curves.  I cannot comment on how it
works though, we have our own proprietary spreadsheet that we use to solve
for the coefficients.

My experience is the more points you use the less reliable the curve becomes
especially in the extreme areas of the curve (using a bi-quadratic).  I find
that I get better results creating a curve that is "tailored" to how a
chiller is going to operate using fewer points.  For example if I know the
chiller is going to be supplying 42-44F water I don't bother creating a
curve using 40-41F or 46-48F data.  

Although I have never done it I'm sure you could create a user expression to
dynamically change the chiller curves (ie. one set for 40-44 CHWT and one
for 44 - 48 CHWT) while the program is running (ie. based on a change in
CHWT or CWT).

I would be curious to hear if anyone has ever tried that.  

MMT


-----Original Message-----
From: Michael Wilson [mailto:mwilsonbc at yahoo.com] 
Sent: Thursday, October 16, 2003 4:50 PM
To: miket at etcgrp.com; BLDG-SIM at GARD.COM
Subject: Re: [BLDG-SIM] Major disagreement on Chiller Performance Curves in
DOE-2.1E and DOE-2.2


I believe that you need to request the various input
and capacity data holding two of the three parameters
(CHWT, CDWT, %load) constant, and varying the third.
Which is a lot of points for the manufacturer, but
they can do it. For recips this info is often
available in the catalog. However, once you've got the
data, its fairly difficult to turn it into DOE2
curves, as the curves need to be interpolated between
each other. ie. you'll get a capacity vs CDWT curve
for CHWT of 42,44,46,48, etc, but they're not exactly
the same. The easy way out is just to use the 44° one,
but its not as accurate. You don't have to do this
with PLR though.

I've attached a spreadsheet that does some
interpolation and calculation of curve co-efficients
if anyone wants to try it. It might need to be
adjusted for the data available though.

===========================
You received this e-mail because you are subscribed 
to the BLDG-SIM at GARD.COM mailing list.  To unsubscribe 
from this mailing list send a blank message to 
BLDG-SIM-UNSUBSCRIBE at GARD.COM