It has been a windy week out here in Phoenix, AZ this past week.

I have a few tricks that I have used that works only sometimes. I use an empty plastic water bottle and put my reference to outside hose inside and try to shield it as much as possible. This has not worked so well the past few days where we have had gusty winds over 20+ mph.

I am open for some input on this, as my calendar is now chaos from all of the rescheduling.

Tags: Basline, Blower, Door, Testing, Wind

Views: 526

Reply to This

Replies to This Discussion

I don't have any magic to cancel out the wind, however, to help your schedule you can consider going ahead with the audit and return on a calm day to do the BD testing.  Distance and other factors will determine when this is an option, but you are now seeing the chaos that repeated rescheduling can cause. 

I'm in cold country and there are many cases where I need the cold or cold with no sun to give me the infrared images I want.  When mother nature is not co-operating I tell the home owner that I may have to come by at 4 AM to do some testing.  Often the wind calms down after dark as well.  There are many services that are necessarily inconvenient and we should not hesitate to tell them, an energy audit can be one of them.

Glad to hear you are busy.


Thanks Bud. 

When you see your baseline jump around like a grasshopper on a hotplate it might not be a good idea to test.

I listened to an online seminar once where it was suggested during windy conditions to multipoint test in both directions with no time averaging and results will be close.

I am not sure the risk outweighs the reward or visa versa. That would be a judgment call you need to make.

I like that grasshopper analogy, thanks for the feedback.

Multipoint testing will improve precision, mostly because (after you get experience with doing dozens of them) you can throw out data that doesn't fit the curve. It is probably not worth all the grief if you are testing a SF house, and a number that's +/- 5% is good enough. But I do it regularly when doing multifamily tests on large apartment buildings. It is true -- if you need more accuracy, testing both ways will help.

If you are running a single fan, a longer time average will help improve precision. It is sort of a low-tech way of doing a multipoint test. I have just set my DG-700 to "long" and let it run for several minutes to get a number that I trusted a little more.

The fastest way to get better precision and better accuracy might be to set yourself up where you can see a flag, or leaves, or something that will act like a wind sock. Then grab your data when it shows you that the wind is momentarily calm. That's because wind always makes a house seem a little tighter than it really is. More of the shell is subject to negative external pressure than to positive pressure, when a wind is blowing around a generally unshielded house. This is why testing both ways (de-pressurize AND pressurize) gets more accurate results -- it averages out the Bernoulli effects across the house shell induced by a wind.

That was extremely helpful, thank you. I would do this method if time would allow, however I am usually performing three to four audits a day on average a 2000 sq ft home, I have rescheduled as needed but not without a valiant effort. I have tried the opposite side of home (opposite from wind direction) and had better results. I have also been doing all of the other tasks other than blower door testing and it has helped out a lot. Thanks again Don.

Colin at Retrotec has told me that you can't get any more accurate than a 120 second baseline and a 120 second time average blower door test, multi-point included.

However, bear in mind that the ASTM standard states 20 MPH as the upper limit, and RESNET has rules about using standard deviation tables to determine accuracy.   So it depends on who you are reporting to - remember that we still use a 1989 version 62.2 that has not been supported for over 20 years.

Thanks Ed

I use a T on my outside hose and run hoses from each side of the T around either side of the house. 

I did a bunch of row houses recently where I couldn't use this trick and found that just extending my outside hose well away from the house worked pretty well too.

The first method averages out the pressures on two sides of the house.  Really to take this method to its full extent, I should be averaging all four sides but I find two sides works pretty well except on the windiest days.

The idea of the second method is to try to get away from the eddy effects of the house -i.e. high pressure on the windward side and low pressure on the leeward side.

I have found that trying to shield the end of the hose in a pop bottle or mailbox doesn't work either.  I suspect that beer or whiskey bottles will not work either, but I haven't tried.

I found this post thru a Google search. I was already a member and have posted a few times.

I am a mechanical engineer who has been doing blower door tests for decades. I only do 10 to 15 BD test per year (I am not doing energy audits or such).I am currently an expert witness in a court case. The one day the court ordered the owner to allow access to test the house was WINDY (steady winds 15 - 20 with gusts to 35 mph). I had to do the testing and deal with the wind effects. I feel like a compensated enough that the overall test results could be repeated n a still day even though there is clearly scatter in the in the data.

But of course opposing counsel is trying to say no part of the testing, much less the conclusions are valid. So in addition to methods for compensating for wind induced pressure fluctuations, I want to know if anyone has tested a building on a windy day and repeated the testing on a still day? How did the results compare? Any published references you can point me to?

Just for reference, I did the following to try to compensate for the wind - 

1) chose location for blower door that was shielded from the wind (down wind side of building, inside of garage with the two car garage door open

2) shielded the pressure reference hose end in a bottle

3) ran a multipoint test per CGSB

4) used Tectite software that takes 100 individual reading and averaged them for each test point

5) loosened the pressure fluctuations checks as little as possible until the test would run completion.

The results look decent, but there is scatter. the points do not fall right on the best fit line the software generates

Any comments would be appreciated. Even comments that say "You screwed up and here is why".

You did a pretty good job -- there are a few tricks of the trade that might have improved precision AND accuracy a little. The problem is that wind not creates a positive exterior pressure (lower delta-p across the shell) on only part of the building. The leeward side of the building and the windward half of a flat roof actually see negative wind pressure (the delta-p across the shell increases as the wind gets stronger).

The most accurate method is to test in both directions (depressurize AND pressurize) and average the results. Then the wind effects cancel each other out. However, if the building systems have dampers, you have to seal during the pressurization phase of your two-way test. This may not be so easy...

Another common (and simpler) technique, especially when doing a single-point test, is to run outdoor reference hoses to either windward and leeward sides, or even to all four sides of the  building, then manifold them together to make one averaged pressure reference where the differences cancel each other out.

I have also had success getting more precise data (and probably more accurate, as well) by letting TECLog run for long periods of time, then selecting the data collected when wind is at its lowest speed. (I used two reference channels on a multi-channel display, so I could see when the wind pressures on the building were lowest.) When I have done that (playing with several hours of data gathered on gusty days) I have come generally to believe that wind effects probably change the measure blower door number by 10% or less. I don't own an anemometer, and I have never done a detailed statistical analysis, so I can't prove that. But that's what I have concluded by observation.

I am not aware of any solid research that has quantified the effect of wind on measured infiltration in any meaningful way. If you contact me offline, I may be able to point you toward some people that may be more in touch with research-oriented testing work.

One thing that can be said with some certainty -- wind does bias a blower door test in one direction -- it makes a building seem/test tighter than it really is. Your test data is probably pretty accurate, but if there is error, the building is actually looser than your measured number. (A higher proportion of the building shell sees delta-p decrease when the wind blows, compared to the shell area seeing an increase in delta-p.)

So, if your job is to demonstrate construction defects that resulted in a leaky building, and measured number you have now is high enough to support that judgment, the attorney opposing you is wasting their time. The actual infiltration value is probably 5-10 percent higher than you measured, and is certainly not lower. 

If you were hired because your client alleges the building is "too tight" -- then your test value is probably not quite as low as you measured. But I am not aware of any way you can say with certainty just much more leakage there is than what you measured.  

Q:  When is the wind too much that you can't do a Blower Door test?

A:  When it keeps blowing your door set up out of the doorway.

A2:  When it is below 0° F. and the homeowner has  an issue with cooling the house off too much.

Otherwise you control for wind.  

I was having a problem on a commercial audit several years ago. I found a piece of PCV that had been romoved and piled up for the trash.  8 inches long with an elbow.  I used this and the readings settled down. So I took it home.  Put a plug in the long end with a 1/8 inch tube in the plug.  I covered the wood plug with epoxy.  I use this and it solves lots of problems.  Locate the PVC around the corner and put the open end close to the building.  I have had to go 180-° from the door with the end of the tube.  Sometimes it is easier to move the set up to another door.

I always try to do a Multipoint.  Lots of good info here.  In 7 years, I've done about 900 blower door tests. Picking and choosing your data points doesn't change much in the way of results.  Your tectite line may look better, but the important numbers don't change much.

1.  Correlation Coefficient: indicates the relative accuracy internal to the data points measured for the single test.

2.  Flow Exponent: indicates the relative size and shape of the hole(s).  Typically these range from 0.55 to 0.75.     Lower numbers indicate relatively larger holes that are more rectangular.  Higher numbers indicate relatively smaller holes that are more thin and long. (cracks).

The tightest home I've tested was gut rehab to PHI standards.  I did a post penetration test.  0.26 ACH @ 50.  I used the duct blaster fan. ELA of 5.6 inches Exponent of 0.84.  Volume was 23XXX cubic feet.  We found 1 4 foot crack along a top plate and two windows with a small crack, along the channel between the frame and the IGU.  (The windows were replaced by the manufacturer.

The Flow exponent indicates the slope of the line. So think of the change in CFM when you compare a reading at 45 to one at 50pa.  If that is 25 cubic feet or 100 cubic feet.   The same idea when you get your number and then open a window and see how much it changes compared to how much you opened the window.

3.  If you a good multipoint test, and you went off with a smoker, and IR and your biometrics to find the leaks, record what you found.  If you found most of the leaks, you can verify your actual findings to the actual test.  If this check doesn't check out, then you must find why it doesn't.

Note:  ASTM requires a 12° F ∆T with a blower door running to find air leaks.  To evaluate insulation quality, ASTM requires a 20° F ∆T over the past 12 hours and a few other things.

Advice;  Practice, protect the end of your tube, practice, do multipoint tests, practice, compare the results, practice.

You are not doing multipoint tests to provide info to your client.  You are primarily doing them to educate yourself. They have the nice side benefit of providing some great reporting of that info.  BTW:  I use the graph regularly.  I very seldom use the other reports.  I generate a detailed test for my own records.; I use the data in report formats that I have developed.  

FYI:  Most of the above info is presented in the Manual of my Blower Door and Manometer.


Forum Discussions

Electro Magnetic Fields (EMF)

Started by Barry L NewDelman in General Forum. Last reply by Brad Cook 1 hour ago. 4 Replies

When is it too windy to use blower door?

Started by Brandon Walton in General Forum. Last reply by Kent Browning 7 hours ago. 15 Replies

Affordable Housing Funding

Started by SUZETT MOFFITT in General Forum. Last reply by John Nicholas 13 hours ago. 5 Replies

Latest Activity

Brad Cook replied to Barry L NewDelman's discussion Electro Magnetic Fields (EMF)
"EMFs are everywhere and in varying strengths and frequencies. In my experience a small percentage…"
1 hour ago
Victor Rols Lindsay commented on A. Tamasin Sterner's photo

improper venting

"whiskey tango foxtrot. This is why we all get ridiculed. Less knowledgeable people  abound.…"
3 hours ago
Victor Rols Lindsay commented on Eric Aune's photo


"I guess a picture IS worth a thousand words"
3 hours ago
Victor Rols Lindsay commented on Michael Dunseith's photo

Proof is possible Tour - Corbett Lungsford

"So... what did you find? This will be fascinating, with some surprises?"
3 hours ago


  • Add Photos
  • View All

© 2017   Created by Lawrence Berkeley National Laboratory.   Powered by

Badges  |  Report an Issue  |  Terms of Service