VRF Heat Recovery Calculations

4 posts / 0 new
Last post

I have created some heat recovery calculations for a VRF system following the Energy Trust of Oregon guidelines and I am getting some underwhelming savings.

The project is a 4-story office building, one of the floors is a basement, and there is approximately 9000 SF per floor. 1 VRF air cooled unit per floor except for the third, where there are two. There might not be enough load diversity between zones to generate much heat recovery. One strange thing is that I am getting negative elec input to heat/cool numbers exported from the CSV hourly reports. Has anybody encountered this before?

I have attached my excel file for reference (sorry for the size - not sure how to compress an excel file). Any feedback would be much appreciated. Thanks in advance for your help.

Travis Miller, PE
DC Engineering
O: 406-829-8828; x. 209
123 W. Spruce ST.
Missoula, MT 59802

Travis Miller's picture
Joined: 2011-09-30
Reputation: 0


I have experienced negative electric consumption in an energy model, but it was in Trane TRACE 700. (The solution from CDS was to use equipment curves from a manufacturer other than Trane!) Perhaps the equipment curve is to blame in your case too.

Keith Swartz, PE
Senior Energy Engineer | Seventhwave | Madison.Chicago.Minneapolis
(formerly Energy Center of Wisconsin)
608.210.7123 | www.seventhwave.org

Keith Swartz2's picture
Joined: 2015-04-29
Reputation: 0

If you are seeing negative heating energy there is a probably a problem with the performance curves.

If you are trying to model the Daikin system I recommend checking out this discussion on UnmetHours. The gist is that there is an line length error in the Daikin VRV curves for eQUEST that causes a negative heating energy. Fortunately it is easy to fix. See the answer from Julien Marrec:


?I've been trying to model a Daikin VRV system and trying to figure out why I was getting 100 times more space heating in the proposed (VRV) model compared to the baseline model (PTHP). After isolating the problem to the VRV, I found a conversation from March 2014 between a few folks (Byron Burns and Bill Bishop mostly - thank you both!), that basically explained why I was getting the issue: the .inp data from Daikin isn't importing correctly.

I just wanted to say that we're in July now, and this still hasn't been fixed (I downloaded the curve from Daikin a few days ago), so if you're having the problem, the fix proposed by Byron and Bill still works. It consists in to replacing the VRV Heating Curve with ?

COEFFICIENTS = ( -1.85025, 0.0620141, -0.000312822, 0.07145853,
-0.0003626328, -0.0007776424 )

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

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

Daniel Knapp's picture
Joined: 2011-09-30
Reputation: 0

Hi Travis,

I can understand your problem but the achieved negative cooling load can be
due to the less cooling requirement at that time.
The cooling system is deprived of that part by the heat recovery wheel as a
result of simulation.
The negative cooling load is being seen at the cooler part of year.
Now the heat recovery can use the hot air from the zone and return it back
to the zone and making the cooling system behave like this.
This is happening in only two zones of basement.
So due to the less heating load in those spaces this thing is happening.
Just try to find out those zones in basement and may be believe that
basements could have less cooling requirement.
The negative sign is just the opposite to that nature.
So if the cooling load is negative then it is believed that there is
heating load.If any more issue then may be there is some incorrect input in
the model. like thermostat set-point, heating, cooling or fan schedules.


Sharad Kumar2's picture
Joined: 2015-05-21
Reputation: 0