Simulation time explodes with AirLoopHVAC:UnitarySystem

asked 2016-07-19 09:29:22 -0600

updated 2016-07-20 09:47:17 -0600

I just added an AirLoopHVAC with an Outdoor Air System and an AirLoopHVAC:UnitarySystem .

The AirLoopHVAC:UnitarySystem itself contains a Coil:Cooling:WaterToAirHeatPump:EquationFit, aCoil:Heating:WaterToAirHeatPump:EquationFit and a supplemental heating Coil:Heating:Water as well as a Fan:VariableVolume, and it is set to Control Type Load with one of the two zones it serves (a lobby and a vestibule) as a control zone.

My simulation time completely bloated. It went 08min50 to 1h35min (and 15 minutes later it's still writing the SQL file...), and I'm getting convergence issues during the warmup. I didn't change anything else.

Is there anything I can do to avoid having such a long simulation time?

Any workarounds I could use such as wrapping a single coil in a dedicted unitary system set to Setpoint or something? This is making my model completely useless.

edit retag flag offensive close merge delete

add a comment

2 Answers

Sort by » oldest newest most voted

answered 2016-07-19 09:40:01 -0600

Adam Hilton

3465 ●1 ●15

updated 2016-07-19 09:40:25 -0600

I've had success using separate unitary systems set to control to setpoint for each WaterToAirHeatPump coil. I put all other components (fan, supplemental coil, etc.) just in the air loop itself and it seems to work okay. Although now you're going to cause me to go back and double check it's working as I expected. 3 air loop 40 TZ model at 4 timesteps/hr solves in <10 minutes.

edit flag offensive delete link

Comments

I'm definitely going to try doing that too... Anytime I have to work with WaterToAir coils I just weep in advance... ZoneHVAC:WaterToAirHeatPump is definitely fine by me (though akward to work with), but man do I dislike the concept of AirLoopHVAC:UnitarySystem. It's like an airloop within an airloop...

Julien Marrec ( 2016-07-19 09:42:14 -0600 )edit

Haha, ya, the requirement of a unitary object in the workflow just doesn't make sense to me. All other Coil:Cooling and Coil:Heating can be be directly in an air loop, I don't understand why WaterToAirHeatPump has to be different.

Adam Hilton ( 2016-07-19 09:49:43 -0600 )edit

We should open a user voice request I think... "Allow placement of WaterToAir coils directly on AirLoopHVACs" maybe.

Julien Marrec ( 2016-07-19 11:13:27 -0600 )edit

@Adam Hilton, I have been struggling to implement this, particularly the setup you have outlined to be able to use the Setpoint control type (2 unitary systems representing heating and cooling coils). The heating Unitary component is working as expected, but the cooling component is not (cooling coil is not activating at all).

I deleted all of the other components in the Unitary systems (Fan, other coils) and re-assigned the nodes. I am using Scheduled SPMs and a condenser loop with a Min/Max deadband control (70F/90F). Any help/insight would be greatly appreciated! Thanks

mhudson ( 2023-01-04 15:13:12 -0600 )edit

add a comment

answered 2016-07-19 11:13:00 -0600

Julien Marrec

29706 ●94 ●76 http://www.effibem.com/

So I've tried multiple ways and wrote down the simulation times for each.

Without this AirLoopHVAC => 8min50s

With the AirLoopHVAC, single Unitary => 1h35min

Everything wrapped into a single AirLoopHVAC:Unitary System set control type Load.

Single Unitary

With the AirLoopHVAC, separate unitary for each coil => 9min 04s

The Coil:Cooling:WaterToAirHeatPumpEquationFit andCoil:Heating:WaterToAirHeatPumpEquationFit wrapped into seperate AirLoopHVAC:Unitary System set control type Setpoint and a SPM placed directly on the outlet: 9min 04s**

Separate Unitary

Conclusion

It seems that using the UnitarySystem as a simple wrapper for each coil is much more computationally interesting. I really wonder why it isn't possible to just place these WaterToAir coils directly on the AirLoopHVAC... Maybe a user-voice request is in order, unless we're missing something here?

edit flag offensive delete link

Comments

Send me the UnitarySystem file that runs for over an hour. It's not the UnitarySystem, it's the implementation of the WSHP coils within the UnitarySystem. Somethings wrong. You never said if it controls correctly or not. All coils need a controller. The water coil has a controller, the other coil types do not so therefore need a parent object.

rraustad ( 2017-03-16 11:57:08 -0600 )edit

@rraustad I believe we've just run into the same issue as the OP referenced in the thread; significantly increased simulation time (~12X) when trying to wrap a Coil:Cooling:WaterToAirHeatPump:EquationFit, , Coil:Heating:WaterToAirHeatPump:EquationFit, Coil:Heating:Electric, and Fan:ConstantVolume inside of a AirLoopHVAC:Unitary System with the control type as Setpoint

Would you mind giving an update or perhaps a recommended way for modeling water-to-air heat pump systems with E+? @julien's solution ok? What can we provide that will help pinpoint the issue, assuming it is still unclear

NickC ( 2017-06-21 13:33:04 -0600 )edit

@rraustad, sorry for not sending this (the request came with a bit of a delay to my defense). I've shared the OS model with Mark Adams that requested it on Slack.

Julien Marrec ( 2017-06-26 03:49:15 -0600 )edit

Any progress on this issue? The unitary system does not provide adequate enough detailing when design 100% DOAS units on the geothermal systems I design. I would love to be able to place the coils individually.

jmoody ( 2018-02-09 12:57:28 -0600 )edit

@Julien Marrec , I have been struggling to implement this, particularly the setup you have outlined above to be able to use the Setpoint control type (2 unitary systems representing heating and cooling coils). The heating Unitary component is working as expected, but the cooling component is not (cooling coil is not activating at all).

mhudson ( 2023-01-04 15:11:44 -0600 )edit

add a comment