House seems much tighter than ACH50 suggests?
I have an older “somewhat” airtight house that tested at ACH50 = 8 and has no mechanical ventilation as it shouldn’t really be needed.
However due to persistent smells and stuffiness, I installed a couple of CO2 data loggers. These show that CO2 in living areas is generally 1000-2000ppm, and peaks at 3000ppm in bedrooms even with their doors ajar at night.
More tellingly when the house becomes unoccupied, it takes 2-3 days for CO2 levels to decay back to outside levels.
Using the CO2 decay curve when unoccupied seemed like a simple way to calculate ACHnat via the gas concentration curve, Ct = Cair + (C0 – Cair) * e ^-nt. Using this approach over a number of tests in mild spring weather I consistently calculated ACHnat = 0.1, well below what the ACH50 divided by 20 rule of thumb would suggest.
So a couple of questions I was hoping to get some help on 🙂
1. What does ACHnat this far from (ACH50 dived by 20) suggest? Do some leakage paths only show up under pressure?
2. Is using the CO2 decay curve a reasonable way to calculate ACHnat? I couldn’t find any reference to it on GBA.
thanks,
Rowan
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
The ACH50PA to AchNat factor is really a giant fudge factor. There is no way to accurately calculate AchNat from a blower test as a lot depends on where the holes are. For example a house in heating climate with all the leaks near the basement and no leaks in the roof would have very low AchNat.
For your specific test, outside temperature also matters a lot. The less temperature delta between the house and outdoors the lower stack pressure thus lower AchNat. The fact you did your test on a mild spring day would definitely reduce the leakage you see. I would also run the furnace fan during a test to make sure there is good mixing of indoor air.
Yeah the CO2 loggers definitely confirm leakage is worse in winter than in spring.
If there is a high variability in ACHnat then why do we put so much stock in ACH50? I guess in my case the problem is it's right on the edge... if it was lower then ventilation would have been clearly required, if it was higher then there wouldn't be issues.
Presumably with the CO2 results and ACHnat I am seeing, I should be planning to retrofit some mechanical ventilation? If I am targeting ACH = 0.4, that's actually going to need a substantial system if my measured ACHnat = 0.1 is accurate.
"If there is a high variability in ACHnat then why do we put so much stock in ACH50? "
Because it's something you can measure. And while ACH50 may do a poor job of predicting ACHnat, in general houses with lower ACH50 are going to have lower ACHnat and better energy efficiency.
So I guess another way to put it is that ACH50 is just an indicator, it should never be used directly for ventilation or energy efficiency calculations.
e.g. in my border line case it tells me "This house is probably tight enough to need mechanical ventilation". But the rate of ventilation will be based on volume and occupancy rates, not the ACH50 value itself.
Yes.
Mechanical ventilation is required in every house for exactly this reason. A leaky house may average the required number of air changes, but unfortunately you need to breathe 24/7/365.
> if my measured ACHnat = 0.1 is accurate
It's not. Sometimes it will be higher and sometimes it will be lower.
do you have data on summer ACHNAT vs winter ACHNAT
Ive seen some data that show that infiltration rate is SIGNIFICANTLY more in the winter vs summer
Estimated ACHnat is an average, useful for energy use modeling (and nothing else). Low wind days will also cause it to be much less.
For anyone wanting to use CO2 decay in an unoccupied house to calculate ACH:
ACH = (1÷5) × ln((800−400)÷(600−400))
Where 5 is the hours between readings of 800 then 600 ppm and 400 is the outdoor ppm.
Thank you for the explanation!
You have to think about how ACHnat is used. It's used in the Manual J to calculate heat and cooling loads due to infiltration. Manual J just cares about maximum loads, worst case scenarios. So it cares about infiltration on the design day -- when you're at the first and 99th percentile of temperature.
ACHnat is going to be driven by two things -- temperature difference and pressure difference (known colloquially as "windiness."). On a calm day with the temperature near room temperature there is going to be very little air exchange with the outdoors unless you open some windows. But the guy designing the HVAC is thinking about that night in February when it's clear and cold and the wind is howling. That's the ACHnat he's trying to estimate.
Also, a word about the difference between engineering and science. In science, you try to come up with models for the world that are accurate. In engineering you try to come up with models that are useful. The question of the estimate of ACHnat is not whether it does a good job of predicting actual air infiltration. The question is whether it leads to sizing equipment that is appropriate for the job it is asked to do.
Estimated ACHnat that is commonly referred to is an annual average. As it should, Manual J (heating) uses a higher number that is based on a higher than average 15+ MPH wind. Ie, ACHnat should not be used for design day load calculations.
https://higherlogicdownload.s3.amazonaws.com/ACCA/c6b38bda-2e04-4f93-bd51-7a80525ad936/UploadedImages/Infiltration%20per%20Blower%20Door%20Test%20Oct2016.pdf