[TRNSYS-users] Greenhouse

Tue Jan 20 01:20:52 PST 2009

Rahma Guechchati and Azelarabe Dkaki,

I have previously asked questions about TRNSYS and greenhouses to the
forum. Below you can find the answers I got.

Kind regards,

Knut Erik Enerstvedt

Dear all,

I wish to use Type 56 to model a greenhouse, and would like to know if
there is an appropriate building template available from somewhere.

Best regards,
Knut Erik Enerstvedt

Knut-Erik,
 Not that I am aware of - when we have had to model greenhouses, they have
just been normal Type56 zones with a lot of windows. I have even had pretty
good luck getting Window5 to come up with window properties for some of the
glazing materials that are used in greenhouses but not in buildings
(celular plastic panels, etc.). There are some problems with treating the
greenhouse as a typical Type56 zone in that Type56 does not account for
beam radiation that passes right through a zone as is the case when you get
a large fraction of windows. I asked a similar question on this forum some
time ago and got some useful help.... the thread is attached below:

>From Pierre HOLLMULLER
I confirm that our monitoring data (18 month in 5 minute step, also
available in 1 hour step) would be available for testing of such a
greenhouse model. The data concerns 3 identical agricultural greenhouses,
of which 2 were equipped with daily heat storage systems (water tank and
air/soil heat exchanger).

For the standard greenhouse (without storage), the data concerns :
- solar radiation
- ambient temperature and humidity
- internal temperature and humidity at 1 and 2 m height
- temperature of internal shading device
- soil temperature at 40 and 70 cm depth
- auxiliary heating
- humidification system
- opening of windows and activation of shading device

A detailed report can be found on:
http://www.unige.ch/cuepe/html/biblio/detail.php?id=20

It is in French, but going quickly through the figures will give you an
idea of what was done and what is available (see in particular annex 1)

>From Ian BEAUSOLEIL-MORRISON
I suggest you look at some of the work published by Jan Pieters et al. of
Ghent University (Belgium).  They presented an interesting paper at the
System Simulation in Buildings conference last December, but I am not sure
whether there are proceedings available yet.

>From Michaël KUMMERT:
You may find some information in a paper that was recently presented at the
"Systems Simulation in Buildings" conference:

E. Mashonjowa, J. G.  Pieters, J. R. Milford1, F. Ronsse, S. Denys, R.
Lemeur - THE USE OF A TRNSYS BASED DYNAMIC GREENHOUSE CLIMATE  MODEL TO
SIMULATE THE MICROCLIMATE IN A TYPICAL ZIMBABWEAN GREENHOUSE  in proc. of
the 7th International Conference on System Simulation in Buildings
(SSB), Liège, Belgium, December 11-13.

The authors use the so-called "Gembloux Dynamic Greenhouse Climate Model
(GDGCM)" that was originally written for TRNSYS 13 and was adapted and
validated in TRNSYS 16. Another reference is:

Pieters J. G. and Deltour J. M. 1997. Performances of greenhouses with the
presence of condensation on cladding. J. Agric. Engng. Research. vol. 68:
p. 125-137.

Kind regards,
 David

> Did your greenhouse consist of just a single Type 56 zone, or several?
> Maybe two zones: One for the main compartment and one for the top
> compartment?
>
our model had two zones - one below and one above. In our greenhouse,
there was a controlled horizontal shade that got pulled across just
above the plants if they got too hot. We used CONTAM coupled with TRNSYS
to determine the air exchange between the two (CONTAM is quite similar
to COMIS)
> How did you define the walls? I have tried to define very thin walls with
> huge windows, but have problems with the timebase because the walls are
so
> lightweight. Do you think I can use a massless wall?
>
I think you could probably just ignore the walls completely unless they
make a significant thermal bridge. If they do, then I would just treat
them as massless.
> Also, did you use one huge window for each wall or many smaller windows?
>
I treated it as one big window because that is how the greenhouse was
set up in my case. If the actual greenhouse is a lot of windows, then
thermally, there isn't much difference as long as you take care to set
the framing factor correctly in the window Typemanager.

Kind regards,
 David

knut.erik.enerstvedt at niva.no wrote:
> Dear TRNSYS users,
>
> I am trying to model a greenhouse with a thermal screen using Type 56.
The
> greenhouse consists of two zones, one above and one below, with the
screen
> as a horizontal shade between the zones. My plan to model the screen is
to
> define an ADJACENT window between the upper and lower zone, choose a
window
> with no glazing (non-existant or open) and have an internal shading
device
> controlled by an input variable. I will also use CONTAM to calculate the
> air exchange between the two zones.
>
> Can anyone tell me if this approach seems reasonable or not?
>
> I also need to know the amount of radiation being transmitted through all
> windows of the lower zone (including the virtual screen window) to
> determine the radiation level just above the plants. Does anyone know the
> proper NType for this?
>
> All help is higly appreciated.
>
>
> Kind regards
> Knut Erik Enerstvedt

Knut Erik,
 I think your approach is good. The one thing that you have to keep in
mind is that with CONTAM (or COMIS / TRNFlow), the flow through a
horizontal opening is uni-directional because it is driven by a pressure
difference and a temperature difference. Those flow paths do not have an
ability to show an updraft in the middle of a zone and a downdraft at
the perimeter. This can lead to a non-convergent solution in CONTAM.
Sometimes you have to artificially restrict the effective size of the
horizontal opening to prevent too large a volume of air from going from
one zone to the other. I have talked to the CONTAM developers about the
problem and how best to try and solve it but we have never come up with
a very good verdict.
Kind regards,
 David

PS: hopefully, Transsolar will answer the question about the NTYPE...
those confuse me terribly and I don't want to tell you something that is
incorrect!

David,

Thanks again for a thorough answer.

I have some more questions related to the same topic:

I wish to model a greenhouse with sloped roof and more than one top. In
Type 56, will adding one wall tilted in the first direction (upward slope)
and a second wall tilted in the other direction (downward slope) have the)
same effect as adding several walls in both directions, as long as the
surface area matches?

I usually set geosurf to zero for all surfaces to keep things simple. Do
you think this is acceptable for a greenhouse?

What do you think the thernal capacitance should be in a greenhouse (upper
and lower zones)?

Thanks again and kind regards,
Knut Erik

Knut Erik,

  I am not 100% sure that I follow your question about slopes - do you have
a cross section of the greenhouse that you can send me so I can see?
Usually if you have an upward slope and a downward slope, the two surfaces
are defined separately with the same slope but with opposite azimuth
angles.

  I left all the geosurf values set to 0 in my case. My thinking was that
if there were no plants, then it would be more appropriate to set geosurf
to 1 for the floor and 0 for everything else. However, the plants are going
to absorb most of the radiation and turn it into growth so there won't be
much direct radiation hitting the floor and allowing it to do so would
mislead the model.

  For the capacitance, I think I set the lower zone to 10 times the default
value and the upper zone to 5 times the default value (I didn't have any
particular justification for my choice though)...

Kind regards,
david

             trnsys-users-requ                                             
             est at cae.wisc.edu                                              
                                                                        To 
             17.01.2009 19:03          trnsys-users at cae.wisc.edu           
                                                                        cc 

             Please respond to                                     Subject 
             trnsys-users at cae.         TRNSYS-users Digest, Vol 49, Issue  
                 wisc.edu              25                                  

----- Message from "Rahma GUECHCHATI" <rguechchati at gmail.com> on Sat, 17
Jan 2009 10:51:26 +0000 -----

            To: trnsys-users at cae.wisc.edu                                  

       Subject: [TRNSYS-users] Fwd: Greenhouse                             

Hello;

I want to make a simulation of a  greenhouse by using software TRNSYS.

Can i use type 56 to carry out the energy assessment of a greenhouse and
which are the necessary parameters  to take into account.

Thank you,

Kind regards,

Rahma

----- Message from Azelarabe Dkaki <d_azelarabe at yahoo.fr> on Mon, 19 Jan
2009 10:07:58 +0000 (GMT) -----

              To: TRNSYS-users at cae.wisc.edu                                

         Subject: [TRNSYS-users] (no subject)                              

Hello TRNSYS-users,
currently I work on the greenhouse by using the software of simulation
TRNSYS, can you give me some refer which can help me.
It's the first time that I work with this software.
Best regards.

----------------------------------------------------------------------------------------------------------------------
NIVAs hovedkontor har flyttet til nye lokaler i CIENS - Forskningssenter
for miljø og samfunn; Gaustadalléen 21, 0349 Oslo. Meld deg på vårt
nyhetsbrev på www.niva.no