Re: [EnergyPlus_Support] Re: Using AFN for windows, but with default infiltration

I would want to ask. If you try to model a new building and take a value like 0.1 its very simple and you will have an answer that everybody knows. For me its.better to model the situation..you have to.assume plenty of things but you will have a "real'" value not the value of Mr everybody.

I think your model.its good. If you use afn and describe cracks why you wont have something good? Maybe the answer is going to.be different..but yat least you can know why.

Good luck.

On Oct 7, 2013 3:04 PM, "allinson_louis" <allinson_louis@xxxxxxxxxxx> wrote:

��

Thank you for this informative post. Plenty of food for thought. I was just wanting to ensure my model was 'equivalent' to other models I had seen (not using e+) which tend to model a standard 0.1ach as an assumption

which of course does not represent reality but apperas to be a typical assumption

--- In EnergyPlus_Support@xxxxxxxxxxxxxxx, "jeannieboef@..." <jeannieboef@...> wrote:
>
> Just saying a big "YES!" to Joe's comments.
>
> Extra points are:
> 1) You could pretend there was a blower door test and ASSUME all leaks
> in the closed possition are equally leaky per linear length. That
> means measuring out all your modelled exterior facing leaks (example

> top bottom sides of all modelled doors and windows). This allows you
> to calculate the "per meter length leakage coefficient at 50 Pa for
> example that is required for a desired air change for your facility by
> using the normal power-law equation (Q = C*dP^n). Note that you'll
> have to sum the total volume of the facility to convert the required
> ACH to m3/s for the power-law equation. After you have calculated C
> (assume C is the same at 1 Pa as at 50 Pa), you can place this same
> input on each exterior leak object as discussed. This is currently my
> prefered method for using AFN where a lot of assumptions must be made.
>
> 2) Data-driven approaches aren't wrong as such. The downside is that
> you rely heavily on the statistics you put to ground. For example, if
> you have statistical infiltration data for 100 builings of similar
> type and hvac to yours in your climate, then this is a viable option
> for you, if you can live with the granularity of the data. In my
> opinion this method should be used (normal infiltration objects) for
> most mechanically ventilated buildings (opening events are usually far
> less and at times where large temperature swings are unlikely). Where
> I start getting irritated is using data-driven approaches to only-
> window-ventilated buildings. The temperature swings and number of
> venting occurances are huge and seriously affect user comfort, which
> in turn effects the opening and closing of the windows --> heating of
> outdoor air and energy consumption. The minute-wise data just isn't
> there.
>
> 3) A third approach is possibly the best. A data-driven approach with
> stoastic modelling of the main factors which lead to when and for how
> long a widow stays open (Anderson)...these factors would be inputs
> into a formula which would open or close the window on a probability
> threshold coupled to a randomness effect. Inputs which would increase
> or decrease this probability are the usual:

> Temperature zone
> Temperature of outdoor air
> Time of year
> Solar
> Stuffiness of air possibly using CO2 as an indicator
> Number of people in the zone
> Number of people having entered the zone to stay in the zone from the
> last timestep
> Room air velocity or windvelocity
> Etc.
>
> 4) I've tried a bit of an advanced version of 1) (AFN + EMS control
> algorithm for x2 widow groups, based on monitoring comfort criteria)
> approach based mainly on PPD threshholds of comfort criteria including
> CO2. The only thing I kept constant were that the persons in my
> facility were evenly distributed (no human traffic modelling), but the
> densities were indeed scheduled. It worked quite well to identify
> problems with the design and possible solutions and demonstrated
> typical expected CO2 concentrations for my design. It also showed the
> problem that windows don't linearly modulate (even when split into two
> independant groups per zone)...this means that occupants close windows
> too quickly after the opening event in winter due to "cold draughts"
> or intense temperature swings in the space. My model required 1 minute
> timesteps to catch this effect. It also showed up the common problem
> of controlling a heating system which is based on an air temperature
> thermostat. I like this approach best, but would idealy prefer the PPM
> threshholds still to be affected by an element of randomness and to
> include the human traffic element (the most significant ones being the
> venting upon the entry of a new person who is likely to stay AND the
> venting of unoccupied rooms). My other significant assumption was
> imposing a limit on the opening event frequency (not closing). I found
> no data to base this assumption on.
>
> Hopefully there was some food for thought in my rambles.
>
>
> Mit freundlichen Gr��en- Sent from my iPhone (excuse the brevity)
>
> i. A.
> Jean Marais
> b.i.g. bechtold
> Tel. +49 30 6706662-23
>
> On 03.10.2013, at 19:21, Joe Huang <YJHuang@...>
> wrote:
>
> > I'm afraid you're missing the point.�� The original
> > question was how to model infiltration when the AFN is being used.
> > My point was that you cannot mix the infiltration object in
> > conjunction with the AFN.�� If you want to model
> > infiltration with the AFN, you have to define the leakage characteri
> > stics of the space for the AFN�� (this should probably be
> > done anyway for proper modeling - in fact, all these AFN programs su
> > ch as COMIS and AirNet were originally designed not for modeling nat
> > ural ventilation, but rather infiltration).��
> >
> > Sure, you can use a blower door test to get the leakage fraction,
> > but that has to be used as an input to the AFN, not to the
> > infiltration object.�� I've always been adverse
> > to modeling infiltration as a constant air change rate, because that
> > 's nonphysical and not what really happens.�� IMHO, it's
> > always much better to define the physical characteristics of the spa
> > ce (leakage-fraction, area of cracks, etc.) and then use a physical
> > model like the Sherman-Grimsrud or the AFN to then compute the hour-
> > to-hour infiltration.
> >
> > Joe
> > Joe Huang
> > White Box Technologies, Inc.
> > 346 Rheem Blvd., Suite 108D
> > Moraga CA 94556
> > yjhuang@...
> > http://weather.whiteboxtechnologies.com for simulation-ready weather
> > data
> > (o) (925)388-0265
> > (c) (510)928-2683
> > "building energy simulations at your fingertips"
> >
> > On 3/10/2013 9:25 AM, Oscar Hernandez wrote:
> >>
> >> There are many methods to calculate the real infiltration
> >> value...like the blower test.
> >>
> >> On Oct 3, 2013 6:06 PM, <josep.sole@...> wrote:
> >> ��
> >> In AFN models you need to assume a lot of other hypothesis it is no
> >> easier than assume the zone infiltration rate.
> >>
> >>
> >> ��
> >>
> >> Josep Sol��
> >> URSA Insulation S.A-
> >>
> >> Sustainibility & Technical Manager��
> >>
> >> M��vil +34 606 42 32 14
> >>
> >> www.ursa.es
> >>
> >> www.ursainsulation.com
> >>
> >> www.uralita.com
> >>
> >> ��<mime-attachment.gif>
> >> P URALITA / URSA se compromete con el ahorro de energ��a.
> >> Antes de imprimir este mensaje aseg��rese de que es necesar
> >> io.
> >>
> >> ��
> >>
> >> De: EnergyPlus_Support@xxxxxxxxxxxxxxx
> >> [mailto:EnergyPlus_Support@xxxxxxxxxxxxxxx] En nombre de Oscar
> >> Hernandez
> >> Enviado el: jueves, 03 de octubre de 2013 17:57
> >> Para: EnergyPlus_Support@xxxxxxxxxxxxxxx
> >> Asunto: Re: [EnergyPlus_Support] Using AFN for windows, but with
> >> default infiltration
> >>
> >> ��
> >>
> >> ��
> >>
> >> If you want to model an infiltration value use the zone
> >> infiltration. The AFN its to predict the infiltrattion value when
> >> you don"t know it.
> >>
> >> ��
> >>
> >> On 3 October 2013 17:42, Joe Huang
> >> <YJHuang@...> wrote:
> >>
> >> ��
> >>
> >> I don't think the infiltration objects can be used in conjunction
> >> with the AFN. That would violate the mass balance of the AFN,
> >> wouldn't it?
> >>
> >> Sent from my IPad
> >>
> >> ��
> >>
> >> Joe Huang
> >>
> >> ��
> >>
> >>
> >> On Oct 3, 2013, at 8:35 AM, Oscar Hernandez <eng.ohw@...>
> >> wrote:
> >>
> >> ��
> >>
> >> Hi,
> >>
> >> Use the zone infiltration.
> >>
> >> Oscar
> >>

> >> On Oct 3, 2013 2:48 PM, "allinson_louis"
> >> <allinson_louis@...> wrote:
> >>
> >> ��
> >>
> >> Hi All
> >>

> >> I am using AFN to model air flow through window openings that I am
> >> controlling using EMS
> >>
> >> However, when the windows are closed the only infiltration is
> >> through cracks and the infiltration is much less than 0.1ach. I
> >> want to model a 'default' 0.1ach when the windows are closed, as
> >> this is a typical value used in simulations.
> >>
> >> Is this possible?
> >>
> >> ��
> >>
> >
>