Hello Simge:,

Unfortunately, I don't think you'll ever get EP+ and eQUEST to agree without significant effort. At best, you can get specific aspects of the program to agree, but the overall results can be very different unless you do extrodinary things to the input files.

As part of RP1468 we developed a comparison simulation of the SAMP1E.INP file RUN3A from DOE-2.1e against eQUEST and EP+, which we could not get to agree very well. In fact, we could not even get eQUEST and DOE-2.1e to agree. The results can be found in the final report for 1468.

In addition we developed a comparison tool the "RASR" that automatically chooses the correct variables to compare, which was a major problem with the comparisons...what to compare against what...not even clear between DOE-2.1e and DOE-2.2.

Finally, the defaults all need to be "revealed" as well, since this can be one other areas where you think you're simulating the same thing...but don't know it because the defaults are set to X,Y,Z and you don't see it unless you have the program print all the defaults. In addition, you'll need to "reset" all the performance curves in one of the other programs to match.

In the long run, it may be easier to solve clmate change than it may be to get the different simulations engines to agree.

Jeff

8=! 8=) :=) 8=) ;=) 8=) 8=( 8=) 8=() 8=) 8=| 8=) :=') 8=) 8=?
Jeff S. Haberl, Ph.D.,P.E.inactive,FASHRAE,FIBPSA,......jhaberl at tamu.edu
Professor........................................................................Office Ph: 979-845-6507
Department of Architecture............................................Lab Ph:979-845-6065
Energy Systems Laboratory...........................................FAX: 979-862-2457
Texas A&M University...................................................77843-3581
College Station, Texas, USA, 77843..............................URL:www.esl.tamu.edu
8=/ 8=) :=) 8=) ;=) 8=) 8=() 8=) :=) 8=) 8=! 8=) 8=? 8=) 8=0

Simge,

I feel compelled to respond since you mentioned me so prominently in
your post :-)

I think you're bringing up a very germaine question, and indeed it's one
that in my opinion needs further study, i.e.,
how does EnergyPlus compare with eQUEST or other building energy
simulation programs in their modeling results?

The LBNL paper you cited is a summary paper from a SimBuild Conference
in 2008 that only skimmed the findings from a
year-long effort supported by the California Energy Commission. I've put
the full report on the Web, along with a later study
I did last year as part of an EnergyPlus Feasibility Study by AutoDesk
done for Southern California Edison that you might find
equally interesting:

http://www.whiteboxtechnologies.com/download_AutoDesk_EP_feasib_rpt.htm

The second study was under wraps for a long time, but I received
clearance from SCE to release it two months ago.
My part of the study is strictly the EnergyPlus to DOE-2.2 comparison
pp. 57 -98, and I have no opinions one way or
the other in respect to the rest of the study. My task was to take
several DEER prototypes (office, residential, and retail)
modeled in eQUEST/DOE-2.2 and convert them to EP 7.X. Compared to what
I saw in the 2007 study, the discrepancies
in heating energy consumption were even more striking, probably because
all the buildings were modeled in southern
California climates. I was getting consistently from a 5:1 to a 20:1
difference in heating energies (DOE-2 high, EP low).

Following up on Jeff's earlier comment, diagnosing the differences
between two simulation programs requires a lot of
time, experimentation, and often in-depth study of the source codes. In
the 2007 paper, I mentioned somewhere between
15 and 20 areas of modeling differences with significant effects on the
results.

It's been 7 years since the first study, and I've been disappointed by
the lack of progress. Some of it is understandable, since
speaking frankly, there are a very limited number of people with
sufficient knowledge and interest across two programs to make heads or
tails out of these comparisons. But it is a very important issue, not
just from technical curiosity, but because a lot of
money have been invested based on what these programs say, so if they
give significantly different results, policy makers would
like to know what's going on and ideally, what is the ground truth?

I don't want to sound like a modern-day Cassandra, and I don't have
enough resources or time to pursue this as a private
adventure. However, I hope that public institutions will see the need
and benefit, and support some serious work in this area.

I know that you (Simge) or I are not the only people who've wrestled
with this problem. I would like to hear the experiences of
others (not you, Jeff, I already know what you would say :-) ). Since
all the efforts I've heard of are going from eQUEST
to EnergyPlus, not the other way around, the domain knowledge is
probably higher on the eQUEST/DOE-2 side. Therefore,
it would be most valuable to get the EnergyPlus experts to chime in on
the nuances of EnergyPlus modeling of which we
may be lacking. For the same reason, you might consider posting your
message to the EnergyPlus_Support bulletin board
as well.

See, Simge, now you got me wasting a couple of hours on a Friday afternoon!

I'm not sure when I would have time to look at your input files, but if
I see anything I'll let you know.

Joe

Hi Joe,

Thank you for sharing this report with the list. It is very interesting to see a thorough comparison of EnergyPlus with DOE 2.2 results. I am an experienced user of eQUEST and I have been looking at EnergyPlus with some interest for future work.

Two things jumped out at me from your report, both of which you noted. One is that, as you point out on page 74, the heating load is relatively small in these climates. This would tend to amplify the effects of any discrepancies in how the heating loads are calculating and in how the plant responds to those heating loads. It would be very interesting to see the same comparison done in the same model in a different location with a much higher heating load. Would there still be a factor of 30 difference in the results in this case? The other is that it would help to see a comparison of the heating loads in the two models. Is the discrepancy coming in primarily in the calculation of the heating loads, the calculation of the energy required to meet those loads, or is it a combination of the two?

I have to say that the more alarming result appears to be the discrepancy in results for domestic hot water. I might have expected the domestic hot water load to be the same in both the DOE 2.2 and EnergyPlus models. Are the boiler models really so different as to entirely account for such a huge difference in natural gas consumption, or are there other loads showing up in the DOE 2.2 model that don?t appear in the EnergyPlus model?

With thanks and best wishes,
Dan

?
Daniel Knapp, PhD, P Phys, LEED? AP O+M
danielk at arborus.ca

Arborus Consulting
Energy Strategies for the Built Environment
www.arborus.ca
76 Chamberlain Avenue
Ottawa, ON, K1S 1V9
Phone: (613) 234-7178 ext. 113
Fax: (613) 234-0740