[Strawbale] (not) Re: Blower Door test set-up

[Mark Bigland-Pritchard]

OK, let's try to clear this one up.

Yes, Rob, you obtain the ACH50 value by testing on the actual house.  
Then you use that number to come up with a typical a.c.h. figure under 
actual conditions (not at 50Pa but at the pressure range which actually 
applies to the building).  In order to do so, you have to apply a 
multiplier which is not going to be any better than an educated guess.  
 From that, you can _predict_ the air infiltration heat loss from the 
building.  It is no more than a prediction.  Indeed it is a prediction 
whose accuracy is rarely tested - how often does anyone carry out 
_direct_ measurements of air infiltration rates from a building over an 
extended period of time?  (And the only method we have for doing so is 
the tracer gas method, which brings its own complications.)

The actual pressure differences across a building envelope - and 
therefore the actual air exchange rates - depend on a multiplicity of 
interrelated factors (site orientation, site sheltering, local wind 
conditions, locations in the building of principal leakage areas with 
respect to each other and to prevailing winds, potential for stack 
effect and hence vertical distance between principal leakage areas, and 
so on).  So the correct multiplier for a given building could be quite 
different from the one chosen for a particular class of building in 
simple software packages.  That is why I plucked a figure out of the air 
and talked about 30% errors - I think that would be an entirely 
plausible eventuality.  Though, as I note above, we may never know 
because it's so hard to get direct data.

Mark
(who sometimes feels like he is a question)




Rob Tom wrote:

>On Mon, 09 Apr 2007 17:25:00 -0400, Mark Bigland-Pritchard  
><mark at lowenergydesign.com> wrote:
>
>  
>
>> For getting reasonably accurate predictive
>>numbers for heat loss, I stick with my description of problematic.  What
>>factor do you divide the 50Pa ach value by to get the "actual" average?
>>- it depends on a number of factors which generally cannot be adequately
>>accounted for in the best thermal modelling suites let alone in a simple
>>assessment tool (like HOT2000, or the British SAP procedure, for
>>example).  Having a maybe 30% error in your air exchange heat loss can
>>make a big difference to design of a near-net-zero dwelling.  And it's
>>not insignificant even for say a standard sb building in mid-Canada when
>>you come to assessing the financial viability of HRVs or solar preheat,
>>maybe even for sizing the heating system.
>>    
>>
>
>Sorry Mark, I'm confused. (It seems to happen a lot to me these days, so  
>nothing new there).
>
>Blower door test ACH50 numbers are something that are acquired after the  
>house is built.
>
>Predictive heat loss numbers (as a result of air changes whether natural  
>or forced) for the purposes of design (ie at the design stage where the  
>house is not yet existing) would, as a minimum, be based on the minimum  
>air changes as prescribed by ASHRAE or if one felt that the  
>ASHRAE-prescribed air change rate was too low to be healthy, a  
>mechanically-induced higher target rate.
>
>Once the house is built, testing would confirm that the house is either  
>tight enough or not tight enough to perform as per the design specs.
>
>ie No house that is built to be air-tight should have a natural leakage  
>rate that is anywhere close to being as high as the ASHRAE minimum air  
>change rate.
>
>And any house that aspires to be near net-zero energy would have to be  
>built air-tight since it would necessarily be superinsulated and if  
>super-insulated would need to be air-tight or else it would experience  
>moisture problems and/or rot in the envelope materials. (Or least that  
>would be the case for a house built in Canada. Maybe not for one built in  
>the desert.)
>
>Given the above, I don't follow how there could be the possibility of a  
>30% error in heat loss due to air exchanges (forced or natural) nor any  
>major errors in sizing of a heating system.
>
>In all likelihood, I think that the Code-required automated heating system  
>would be oversized simply because the units avaialable on the market are  
>designed/made for conventional homes with conventional heat load  
>requirements and the capacity of even the smallest unit made will likely  
>be more than is required (assuming the SB house is designed & built to be  
>an energy-efficient/low-energy structure.
>
>So more often than not what one ends up doing is installing an electric  
>furnace or electric baseboard units (to satisfy (Canadian) Code  
>requirements, the easiest/simplest/cheapest route) and which seldom get  
>used if at all and a woodstove is fired up for those occasions when  
>auxilliary heat is necessary.
>
>Fortunately, they're relatively cheap (ie big toasters),easy to install  
>and take up minimal space. (ie the space under the stairs usually will  
>suffice for a whole-house capacity electric furnace and the ductwork would  
>be there anyway for the ventilation system that is necessary to ensure the  
>delivery of fresh air to every liveable space (again required for Canadian  
>homes, don't know about elsewhere. Perhaps in David's Texan (?) locale, a  
>exhaust-only/passive inlet ventilation system is feasible and permitted.)
>
>If anything, tools like HOT2000 (don't know anything about the Brit SAP)  
>are conservative so if there is any error, it is typically on the safe  
>side ("safe" from a homeowner PoV). ie The actual house performance will  
>be better than predicted if it was built to be as tight or tighter than  
>designed (as confirmed by testing.)
>
>I just noticed that there is no "question mark" (in quotes so as not to be  
>confused with Mark, who is not a question) anywhere in the above but this  
>whole posting  should read as one big "question mark".
>
>  
>