Natural Ventilation

8 posts / 0 new
Last post

Hi All,
Does any one know how to model natural ventilation using cfd
tools like fluent anysis starccm+ etc and what should be the
boundary conditions applicable for natural ventilation in room.


sambhav tiwari's picture
Joined: 2011-09-30
Reputation: 0

Hi Sambhav,

For natural ventilation analysis, you may start with some field measurements
or wind tunnel test which can help you decide the boundary condition
precisely. Some key parameters, such as prevailing wind direction,
representative wind speed, air temperatures, surface temperatures, etc. are
all important components of the boundary condition. Meanwhile, building
geometry, building materials, surrounding buildings, topography, etc.
may influence the boundary condition significantly. You will have a holistic
consideration of all these variables and make reasonable simplification to
work out the boundary condition.



chen yu's picture
Joined: 2011-10-02
Reputation: 200


You should carry out a dynamic thermal model first to will allow you to
establish far more accurately (than just guessing) what the boundary
conditions for the specific time you are looking at are.


Eric Roberts's picture
Joined: 2011-10-02
Reputation: 0


It would be helpful to know more about what you want to achieve with
natural ventilation. In order of precedence:

1. How temperate is the climate? The climate needs to be fairly
temperate. Most of the UK will be ok for natural vent in many
commercial buildings. The same is not true for the South of France!
2. How heavily used is the building? Forget natural ventilation if a
building has dense occupancy or high IT loads.
3. If rules of thumb are followed (e.g. 5% of floor area reflected as
openable area, thermal mass, floor depth to ceiling height ratio,
glazing!) this is a good starting point for the design
4. A zonal (not CFD) model will allow you to assess the performance
of the design against weather data and standards such as ASHRAE
55. Zonal models are computationally cheaper than CFD and will
help consider strategies such as night time purge ventilation and
optimisation of thermal mass.
5. Finally, if you're doing something really unusual or bespoke (like
working for Norman Foster) CFD may be warranted. Boundary
conditions such as surface temperatures and flow through openings
can be extracted from a "snapshot" of the zonal modelling. You may
want to choose more than one design day to cover variations in
climate and use (e.g. still days when there is little wind to
drive ventilation).


Chris Yates

Chris Yates2's picture
Joined: 2011-10-02
Reputation: 0

Hi Shambav,

Both Eric and Chris offered very good points. I'd like to remind you that no
matter how your get the boundary condition, it must be reasonable. I know
some simulation programs can generate boundary condition through
dynamic thermal calculation or bulk air model, such IES VE or TAS. However,
you'd better verify the boundary condition based on your common sense and
experience. I did some natural ventilation related study before and
noticed the results from CFD simulation could be logical only when the
boundary condition is well established. That's why I highly recommend field
measurements or wind tunnel test as a starting point. What's the difference
between them and those dynamic calculation results:

1. Natural ventilation is quite micro-climate oriented. Surrounding
buildings, topography, landscape, etc will have very significant impact on
it. However, most dynamic calculation will use a standard weather file
generated from nearby airport or open sapce. As a result, a misleading
boundary condition could be generated.

2. Some variables, such as surface temperatures, may vary a lot between real
world and computational one. Most time I have a feeling that simplified
thermal engine could not trace the complicity of the real world perfectly
and underestimate the surface temperatures. I did mention it in one of my
recent post.



chen yu's picture
Joined: 2011-10-02
Reputation: 200

Hi Arpan,

IES VE's macroflo module may be useful here. If you go to Macroflo/Macroflo
opening types/Exposure types, you may find that there are couple of
scenarios pre-defined. They are sort of simplified way to estimate the
impact from surrounding buildings. Other than that, I doubt IES VE thermal
calculation modules can trace the real 3D model and generate very accurate
wind flow pattern accordingly. You know what, sometimes, when field
measurements or wind tunnel test are not possible, we will run CFD twice.
The first simulation will focus on the external CFD analysis with respect
to the designed building and its surroundings (include adjacent buildings).
The results from the first one will be a good reference for the second round
of internal CFD analysis.



chen yu's picture
Joined: 2011-10-02
Reputation: 200


Dr Atch Sreshthaputra did a good job with this for a Buddist temple in Thailand, which was published in Energy and Buildings. He combined CFD with DOE-2 and got some very interesting results. Problem is that it took 24 hrs to simulate 24 hrs on a fast PC. So he limited his analysis to a few days each year.

Jeff S. Haberl, Ph.D., P.E., FASHRAE

Jeff Haberl2's picture
Joined: 2011-10-02
Reputation: 0

Hi all,

The artical recommended by Jeff can be found at



chen yu's picture
Joined: 2011-10-02
Reputation: 200