VRV Special case: is it worth modeling?

9 posts / 0 new
Last post

Dear all,

My school has some 200 zones including classrooms, storage rooms,

The HVAC system is a VAV one, with FCU for 17 offices/storage rooms.

In the new updated design, 4 of these FCUs were replaced with a VRV
system. I am about to model that:

Do you think that these 4 small zones (electrical room (x 2), UPS room,
telephone room) will significantly change the Space Cooling results (not
that no heating is required)? I mean, these zones had FCUs and were
conditioned. All that changed is that a VRV (COP=10) replaced the
chiller (COP=5.4) for only 4 zones!

I would like to avoid the complexity of modeling VRVs if possible!

Many thanks,

Omar Katanani

Omar Katanani's picture
Joined: 2011-09-30
Reputation: 0


The only thing that would save you a significant amount of energy in my
opinion would have been scheduling large VAV air handling and plant
equipment that was originally serving them. I had a project where I took
data closets off of the VAV units and put VRV units in the closets so
that the VAV units no longer had to run 24/7. Scheduling these large VAV
units had a significant savings impact.

In your case, it sounds like you may have something similar but I would
need clarification. You said that your closets (that you want to convert
to VRV) are currently FCU. I assume that these FCUs are served by
chilled water. By switching these closets over to VRV, are you able to
now schedule the chiller off? If yes, I would assume that significant
savings would result by not running a chiller (and pumping) 24/7, even
with a high unloading capability and variable-speed pumping. (If you had
low unloading capability and constant pumping, more savings would

At face value, simply changing the efficiency of the cooling equipment
isn't enough to evaluate energy savings. If, however, you can
significantly change the schedule of plant equipment because it is
isolated from zones that require 24/7 operation, savings could be
significant and I would certainly recommend a model.

Sidenote: 10 COP seems very, very high for even the best VRV equipment.
I would check that value to ensure accuracy...

Best luck,

Gary Schrader

Gary.Schrader at buildings.schneider-electric.com's picture
Joined: 2011-09-30
Reputation: 0

For the most part, I agree with Gary. I believe that you will also find
significant savings from part-load performance from the VRV system. VRV
systems operate more efficiency at part-load, especially when compares to a
chiller. I just finished modeling a VRV system, and with the newest version of
eQuest, it is not that difficult. Simply use the PVVT system, and set the
compressor to variable speed. You will need a low-speed curve for both heating
and cooling. I took those curves from the ResidentialPVVT system.

I also agree that a COP of 10 seems unreasonable, but I assumed that you meant
an EER of 10


Eddie Corwin's picture
Joined: 2011-09-30
Reputation: 0

For the most part, I agree with Eddie. It is pretty easy to model a VRV
system in eQUEST but you do need to input curve data related to your
equipment if you want to accurately model them.


cmg750's picture
Joined: 2010-10-05
Reputation: 0

Omar, attached is an example .inp file (when opening, make sure you are viewing
INP files, and not PD2 files only)

You need to change your system to PVVT, then go throughput an setup the proper
inputs (fan power, schedules, condenser type...). There is one issue that I ran
into, and that is the following:

On the first tab, labeled "Basics", there is an option that says "WL/GS Heat
Pump" I selected YES (because selecting NO gave me unrealistically high
results), even though it is an air-to-air heat pump. After reading through the
literature, this does seem to be a proper selection, as long as
your condenser is set to air-cooled. After that, just make sure you have the
proper curves (Daikin can provide you curves for their unit).

Let me know if you run into more problems, but don't be afraid to dive right in.


Eddie Corwin's picture
Joined: 2011-09-30
Reputation: 0


You should not mix air source heat pump data entries with WS/GS heat pump
entries. eQUEST will let you do many things you shouldn't and this would
fall in that category. Hard to say what the impact is but safe to assume
it's not good. If you disagree with the output look elsewhere.


cmg750's picture
Joined: 2010-10-05
Reputation: 0

I know that it id not a good idea to do without looking into it more, but after
reading through the literature, specifically the new features document, I felt
ok doing it. The document specifically lists an air cooled condenser as being an
acceptable option. I was, and still am skeptical of it, but
the documentation clearly states that it is a valid option.

Eddie Corwin's picture
Joined: 2011-09-30
Reputation: 0

I agree with Gary I have never see that high COP for the VRV units. It can
be EER value. Please check that value again to avoid any mistake.

Good luck

Y?k.M?h.Berke ?elikel

Berke's picture
Joined: 2010-11-02
Reputation: 0

Yes yes, it is the EER! That was a mistake of communication.

Anyways, there is no place to enter the COP for the VRV system when modeling it by a PVVT system. All that matters is the performance curves, right?



Omar Katanani's picture
Joined: 2011-09-30
Reputation: 0