I've been tasked with finding a software package that can model our chilled water system. We have a district plant that distributes chilled water to many buildings on our campus. My specific task is to model the chilled water system of one specific building, then those results will determine if we continue looking for a different software package, or proceed to model the other buildings. Pretty much they want a computer model of the buildings chilled water system that can predict the chilled water flow requirements for each AHU & the total building as a function of the buildings chilled water supply temperature, outside air conditions, and air handling unit flow rates.
I'm looking at eQuest & EnergyPlus (as they're the only free options I've found so far to play with), and I'd really like some opinions/suggestions on how to proceed.
Thanks
Jason
Dear Jason,
Assuming that you've got the campus HVAC wired to a BMS, I'd consider
gathering data and doing a regression analysis using key parameters to
predict energy or load.
There are hundreds of variables which affect actual load, including poorly
calibrated sensors, leaking control valves, etc. which are not well
addressed by ANY modeling software. On the other hand, you already have the
actual usage data and can get concurrent weather data. even assuming that
you'll spend significant time tuning the regression, I think you'll be MANY
hours ahead and have a more accurate model from which to make improvements.
p.s., I'm no stats wizard. If you aren't either, there are many on a
typical University campus that would probably love a real problem to
solve....
James V. Dirkes II, P.E., LEED AP
Jason,
Based on your description, I don't think your best option is a
whole-building energy model such as eQuest or EnergyPlus. I believe your
best option is an empirical model. I suggest you review the ASHRAE 1050-RP
research, and see if the models presented there might be more applicable to
your task. They will also be much, much easier to learn and use. You can
either use the 1050-RP toolkit, or depending on your needs, you might find
it handier to create spreadsheet models. You can find some good examples of
the use of those approaches in the Texas A&M web library.
These approaches are typically described in the context of Measurement and
Verification, but they are applicable to many other energy analyses.
--Bill
Jason,
James Dirkes expressed what I was saying much more clearly. But we are
saying the same thing. He is correct in everything he said. Fortunately,
in most cases, you won't need a stats wizard. Study the regression
approaches used by Texas A&M and others, and provided in the ASHRAE 1050-RP
toolkit, and you may find your task to be fairly simple.
--Bill
I agree that whole-building energy models are not the way to go for chilled
water simulation. ASHRAE Research Project 141 developed a program for
simulating central chilled water systems. The original work was done by the
University of Illinois at Urbana-Champaign. Dr. James A. Corll converted
this mainframe program for use on PCs. He made extensive modifications to
the program to add features and correct some program bugs. His version,
WTRSYM, was a powerful tool for simulating chilled and hot water systems.
The program allows you to describe a chilled water piping network including
pipe size, length, and fittings. It models variable and constant speed
pumps, chillers, coils and coil loads, pressure relief valves, check valves,
and control valves including two-way, three-way, and wild coils.
Program output includes an echo of the program input, a summary of control
valve conditions including Cv and flow, a summary of air side and water side
coil conditions, and a list of pressure, flow rate, and temperature at every
node in the system. It also notes the load on each chiller and lists
overloaded coils.
There are several applications for the program. If the total system capacity
is less than the total of the peak design cooling loads, simulations may be
made to see if system inadequacies occur for different load combinations. A
second design application would model proposed changes or additions to an
existing system.
The program can also simulate problems which occur under certain load
conditions and may indicate the cause or solution of the problem. It may be
used to simulate corrective measures to a problem. The program will also
allow simulating the effects of turning off part of the equipment under part
load conditions for energy conservation.
There are a couple of drawbacks to the program. First, Dr. Corll did not
design much of a user interface. The input actually dates back to the
mainframe roots of the program. The input must be in a column oriented
format similar to punch cards in order to work. Second, it is very easy to
create a Chilled Water network that WTRSYM cannot solve. The program uses an
iterative algorithm that must converge to arrive at an answer. The
combination of these two things make WTRSYM a very difficult program to work
with. It does, however, give you results that make it all worth while.
I have use WTRSYM on several projects over the last 20 years and have never
been disappointed with the results. The program can give amazing insight
into a chilled water system, especially one that has been added on to for
the last 20 or 30 years.
Logo. Click here to open window on Elite's HSYM program.
lost contact with Dr. Corll several years ago. I am not sure if his program
is still available. Elite Software bought the rights to use his program as
the calculation engine for their HSYM program. Elite Software has added a
Windows user interface, which should be very helpful. It will not help with
the convergence problem. HSYM is available through Elite. If you are
interested, click the Elite Software Logo. Understand that this is a
qualified recommendation. I have no direct experience with Elite's version
of the program.
Mike Roberts