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

Who uses ZoneMixing?

asked 2021-06-22 14:45:13 -0500

Keigo's avatar

updated 2021-07-11 16:10:21 -0500

I did case studies with a very simple model in order to check the function of ZoneMixing and ZoneCrossMixing.

Two zones (ZoneA and ZoneB) are adjacent to each other through an air wall. They have different cooling setpoint: 30°C for ZoneA and 24°C for ZoneB. Cooling system is HVACTemplate:Zone:IdealLoadAirSystem. I simulated 7 cases with different conditions. The table below summarises each condition and simulation result (annual cooling end use). "AtoB" means that ZoneA is the "source" zone and ZoneB is the "receiving" zone. The same Design Flow Rate was specified inZoneMixing and ZoneCrossMixing, and the schedule was always on.

image description

Case2 and Case4 resulted in lower cooling end use than Case1(No mixing), which means that the cool air in ZoneB was distributed to ZoneA and helped to cool ZoneA, but no additional cooling energy was used in the HVAC system in ZoneB.

On the other hand, Case3 has mush higher cooling end use than Case1(No mixing), which means that the hot air in ZoneA was distributed to ZoneB and the HVAC system in ZoneB required much cooling energy.

These results are consistent with what is described in I/O Reference and EngineeringReference. i.e.,

image description

But then, who would use ZoneMixing? It generates cool/hot air from nothing and completely destroys energy balance and energy simulation.

My case study model has an extreme temperature difference between two zones, but even if the temperature difference is very small, the difference causes inaccuracy of energy simulation.

ZoneCrossMixing makes sense. To me, ZoneMixing looks like a degraded version of ZoneCrossMixing. Is there any case where ZoneMixing would be more useful than ZoneCrossMixing?


This post script illustrates my interpretation of @Aaron Boranian 's answer (2 examples of using ZoneMixing). Could you please correct me if I'm wrong?

1. Atrium

What the purpose of this modelling? - I guess it is to estimate how hot the upper part of the atrium will be in summer. However, I think the upper part of the atrium will be Cooled because the air with the same temparature as the lower part of the atrium that is air-conditioned is supplied to the upper part. It is an opposite phenomenon of what happens in actual atriums. How can we solve this problem? I think ZoneMixing is not suitable for modelling the upward airflow due to buoyancy.

Or we may want to know how cold the lower part of the atrium will be in winter, but I think we can't model it because the lower part of the atrium("source" zone) is not affected at all by ZoneMixing. In this case, I think the atrium should not be divided into lower part and upper part, it should be modelled as a single zone.

image description

2. Office and Corridor

Again, what is the purpose of this modelling? If the corridor is not air-conditioned, normally we don't care about the air temperature of the corridor. If the corridor is air-conditioned, the "spilling over" air will ... (more)

edit retag flag offensive close merge delete

3 Answers

Sort by » oldest newest most voted

answered 2021-06-23 08:36:07 -0500

updated 2021-06-24 19:28:04 -0500

As you've noted, ZoneMixing is for one-way flow of air in between zones. A few situations where you want to use this are:

  1. Horizontal openings. ZoneMixing would represent the upward flow of air due to buoyancy, say in stacked zones that represent an atrium. You can read more details here.
  2. Pressurization or depressurization. ZoneMixing would represent the intentional pressurization of supply air entering one zone, then "spilling over" into an adjacent zone (supply into an office and spill into an adjacent corridor) or the intentional depressurization of exhaust air leaving one zone and "pulling" air from an adjacent zone (exhaust from a restroom that pulls air from an adjacent corridor). You can read more details here.

Note that either mixing object you've mentioned is a "bulk airflow" added by the user -- EnergyPlus is "blindly" assigning an air flow to the model that may not be realistic. If you want to model internal air flow accurately, it would be better to use the Airflow Network feature where EnergyPlus calculates a proper air flow that should occur due to a pressure difference. Some drawbacks of this are there is overhead with defining detailed inputs of AFN objects, and it does increase simulation time.



If you're striving for the "perfect model" (if such a thing exists :) ), then AFN objects should be used to model airflow loads instead of "bulk airflow" objects like ZoneMixing.

Full Response

As you mentioned from the documentation, "bulk airflow" objects are really just "zone energy balance accounting" for the receiving zone only. They tell EnergyPlus that air moves within the model and there is a resulting sensible and latent heat exchange where the air flow ends up. They do not enforce energy balance where the air comes from (outdoors or the source zone).

I agree that this does not match reality.

Why would we use "bulk airflow" objects in general?

Because the alternative is the AFN objects, which comes with extra overhead of assigning very detailed inputs and longer simulation times for EnergyPlus to calculate air flows due to pressure differences. Not all energy modelers want to deal with that, and adding "bulk airflow" objects to models is better than doing nothing to account for air flow loads.

Why would we use the ZoneMixing object specifically?

Because users want to account for the impact of air moving between zones in one direction.

  1. I should have clarified that the ZoneMixing object has simple control inputs for the user to specify limits on when the mixing air flow occurs. For the atrium example, that would be if the lower zone became hotter than the upper zone (air would flow upwards). In reality, you are correct that over time the hot air would naturally collect at the top of the atrium. However, EnergyPlus doesn't calculate this kind of fluid flow phenomena like a CFD tool would.

  2. "If the corridor is not air-conditioned, normally we don't care about the air temperature of the corridor." While ...

edit flag offensive delete link more


Thank you for your answer! However, even in the two examples you've listed, I still don't see any point in using ZoneMixing. I added my understanding with some sketches to my question. Could you please point it out if I'm wrong?

Keigo's avatar Keigo  ( 2021-06-23 23:09:01 -0500 )edit

@Keigo I've updated my answer above.

Aaron Boranian's avatar Aaron Boranian  ( 2021-06-24 19:28:22 -0500 )edit

@Aaron Boranian Thank you for your update! Your detailed explanation made sense to me.

Keigo's avatar Keigo  ( 2021-06-24 20:10:52 -0500 )edit

@Keigo you may find this post interesting, an OpenStudio developer asked EnergyPlus users what situations they use ZoneMixing and ZoneCrossMixing for.

Aaron Boranian's avatar Aaron Boranian  ( 2021-07-01 15:25:07 -0500 )edit

answered 2023-03-29 02:47:45 -0500

Keigo's avatar

ZoneMixing helped me for the first time when I modelled restaurant and kithcen without using Airflow Network model as shown below.

image description

Some air is transferred from Restaurant (seating area) to Kitchen. The transfered air is one-way, and it helps to cool kitchen a little. In this case, ZoneMixing is suitable rather than ZoneCrossMixing.

edit flag offensive delete link more

answered 2021-06-30 03:48:37 -0500

updated 2021-06-30 03:49:52 -0500

I think another good reason to use ZoneMixing is when you've airflows involving more than 2 zones and isn't perfectly symetric between each pair of zones but you still don't want to bother with a detailed AFN.

For example, you've 2 rooms adjacent to each other and connected to the same corridor. Air flow from room 1 to room2 (e.g. you have a fan in the door between them), then from room 2 to the corridor and comes back to room 1 from the corridor. You end up with a "swirl" of air between the 3 rooms that ZoneCrossMixing simply cannot depict as it enforce balance between the source and recieving zone.

edit flag offensive delete link more


Thank you for your answer. I/O Reference says that "ZoneCrossMixing can be entered once (in one of the mixing zones), twice (once for each zone), or multiple times (exchanging with more than one zone)." For your example, my understanding is that three ZoneCrossMixing objects (from room1 to room2, from room2 to corridor, and from corridor to room1) can be set. Do you think it is correct?

Keigo's avatar Keigo  ( 2021-06-30 05:23:42 -0500 )edit

I see what you mean. My understanding for that part is that if you enter multiple ZoneCrossMixing for a single zone if it will enforce 2 way exchanges between that zone and each zone you specified a cross mixing with. In the example I feel that if we specify ZoneCrossMixing between each room I think we'll end up with a 2 way air flow between each zone. Room 1 will send and recieve & send air both from & to room 2 & the corridor whereas it shoud only recieve some from the corridor and send some to room2. But maybe I misunderstood something

Tanguy Timmermans's avatar Tanguy Timmermans  ( 2021-06-30 05:44:17 -0500 )edit

OK, maybe I got what you want to say. Yes, one-way airflow can be modelled only by ZoneMixing. But ZoneMixing doesn't conserve energy...

Ideally, when some air volume (V m3/s) is moved from Source zone to Receiving zone, the enthalpy that the air volume has should be moved from Source zone to Receiving zone as well. I wish a future update to EnergyPlus would add such a feature for one-way zone mixing.

Keigo's avatar Keigo  ( 2021-06-30 06:06:36 -0500 )edit

Energy might not perfectly conserved. But if you balance the flows properly between the zone my gut feeling is that the overall energy will be conserved. Imagin Tr1 > Tr2 > Tcor in our example I'll "inject" heat in room 2 (without removing it from room1). Then heat is injected in the corridor (again without removing it room 1). Finally heat is extracted from room 1 (without adding it to the corridor). Individually I don't conserve energy, but overall I've 2 "injection" and 1 extraction and the sum is (I think) 0

Tanguy Timmermans's avatar Tanguy Timmermans  ( 2021-06-30 07:11:23 -0500 )edit

Injected in R2: Er2 = VrhoCp (Tr1 - Tr2) Injected in Cor: Ecor = VrhoCp (Tr2-Tcor) "Injected" in R1 (negative as Tr1>Tcor) as: Er1 = VrhoCp (Tcor-Tr1) And overall I've Er2 + Ecor + Er1 = 0

Tanguy Timmermans's avatar Tanguy Timmermans  ( 2021-06-30 07:14:12 -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: 2021-06-22 14:45:13 -0500

Seen: 403 times

Last updated: Mar 29 '23