Question-and-Answer Resource for the Building Energy Modeling Community
Get started with the Help page
Ask Your Question

VRF & DOAS - complete model?

asked 2019-01-17 18:37:40 -0500

Matt Koch's avatar

updated 2019-01-18 10:04:28 -0500

I am just curious. There are a lot of exchanges on VRF with DOAS for OpenStudio/EnergyPlus here. I have spent some time with that subject myself, including trying to use the VRF with DOAS measure. To this day, I have not been able to demonstrate the commonly alleged advantage of a VRF with DOAS system over a RTU system. This includes using the default OpenStudio VRF selection and the LG components, as well as homegrown Daikin components (based on Daikin's Trace-700 library).

In fact, I am having to fall back on Trace-700 - and its Daikin library - every time I have to do a LEED project, which is frustrating, given that OpenStudio/EnergyPlus ought to be eminently capable of handling this seemingly simple task? I have gotten some comments on earlier questions I posed, but not really actionable guidance.

Has anyone at all gotten VRF with DOAS to work under OpenStudio/EnergyPlus while seeing BOTH a roughly 25% reduction in demand and usage AND acceptable unmet hours (< 300 hours/year)? If so, would you be willing to share your model, so I could learn from it? I know there is a large number of Vulcan death grips possible in VRF with DOAS, and it is quite likely I am not applying them in the right combination (though I think I have been through a good number - if not all - of them). Thanks!

edit retag flag offensive close merge delete

2 Answers

Sort by ยป oldest newest most voted

answered 2019-01-18 12:51:05 -0500

The measure labeled 'VRF + DOAS' in BCL has high static pressure on the zone terminal units and some odd curves for main VRF condenser.

Try the VRF+DOAS system option in this measure. This should have acceptable unmet hours in all climates except very cold climates (ASHRAE CZ 7+), as the default system and curve set isn't set up to handle very cold outdoor air temperatures without a backup heating system.

VRF systems have improved rapidly in the last few years, and we need to update our curves, especially for low-temperature operation. No matter which VRF measure you select on BCL, I would recommend updating the VRF condenser curves to match the specific product you are using in your project.

edit flag offensive delete link more


Thank you kindly, mdahlhausen! I created a mock-up model and did a VRF-DOAS on the 1st version per the measure you suggested, then did a standard RTU-PFPB on the 2nd version. With all default information, notably default efficiencies, I get the VRF-DOAS to be better than the RTU-PFPB by about 14.5% and for it to have 83 UMH vs. 33 UMH. Not exactly where I need it to be, but definitely a step in the right direction. My mock-up model has two stories and only three zones per story, none of them truly core vs. perimeter either, so it needs a little more work more work.

Matt Koch's avatar Matt Koch  ( 2019-01-18 21:55:11 -0500 )edit

what climate zone? Also check the HVAC scheduling - if you have a 5% or less occupancy rate, the default HVAC schedule might be turning off the system when you have few people in the zone, causing occupied unmet hours.

mdahlhausen's avatar mdahlhausen  ( 2019-01-21 10:47:48 -0500 )edit

So, is that to say that:

1) the FSEC document shows "the right way" to do VRF (even in OpenStudio)

2) the "VRF+DOAS system option in this measure" reference by mdahlhausen above does NOT do it the FSEC way?

I think knowing the answers here would help better understand why OpenStudio may be so far off currently?

Matt Koch's avatar Matt Koch  ( 2019-02-06 19:10:16 -0500 )edit

The FSEC document explains how to make VRF curves from manufacturer data. The measure also uses curves derived from manufacturer data, though not the same equipment. In both cases, the curves are based on equipment >5 years old, and modern equipment is more efficient and capable of going down to low temperatures. That means if you are designing a system, make and use curves for the equipment you select. If modeling a existing system, it still helps to make curves for that equipment too. The default curves in OpenStudio are much older and give poor performance; don't use default curves.

mdahlhausen's avatar mdahlhausen  ( 2019-02-07 11:33:48 -0500 )edit

answered 2019-02-01 14:14:16 -0500

lrebello's avatar


Which is the best way to generate those VRF condenser curves mentioned on your answer?

I am trying to generate for plenty equipments to provide to all but I am having trouble on understanding the procedure of generating these coefficients.


edit flag offensive delete link more



There is a new object that allows you to use a performance map of manufacturer data instead of translating that data to curve coefficients. Or, if you have the data and want to generate a curve, fit a regression to it in excel or some related software. See the EnergyPlus reference manual to understand the equations. Either way, you need performance data for the manufacturer.

mdahlhausen's avatar mdahlhausen  ( 2019-02-01 15:06:31 -0500 )edit


For this project I am going to try to use the performance map.

Whenever I get some time to kill, I will work on generate those curves for different manufactors.

Thanks for your answer @mdahlhausen

lrebello's avatar lrebello  ( 2019-02-04 09:53:37 -0500 )edit

@mdahlhausen what is this performance map object you speak of?

TomB's avatar TomB  ( 2019-02-05 04:47:25 -0500 )edit

Have you read this document?

rraustad's avatar rraustad  ( 2019-02-05 07:38:49 -0500 )edit
mdahlhausen's avatar mdahlhausen  ( 2019-02-05 10:27:59 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Training Workshops


Question Tools



Asked: 2019-01-17 18:37:40 -0500

Seen: 895 times

Last updated: Feb 01 '19