Helping the A talk to the E about Air Leakage

Mike, Interesting topic. Typical spec on curtain wall is 0.06 cfm/sf at 75 pa (1.57 psf, or equivalent to 25 MPH wind) .  However, including doors, openings etc. most codes and standards have settled at .25 cfm/sf (some at 0.4 cfm/sf) 75 Pa Oregon per 2011 Oregon REACH Code is 0.25 cfm / ft² California per 2013 California Building Energy Efficiency Standards is 0.4 cfm / ft² Army Corps of Engineers per USACE Test Protocol for Building Envelope is 0.25 cfm / ft² However, if the modeler just use the above numbers, it may not be appropriate, since 25 mph wind is not a "typical" condition in majority of the cities in US, and also needs to account for building pressurization.  Using Chicago as an example, average wind speed in "Windy City" is about 10 mph, so the leakage rate (use pressure and flow laws), is about 0.1 CFM/sf (10 MPH wind) in average, not .25 CFM/sf (25 MPH wind), ignoring building pressurization. Tall buildings use 300 pa is conventional, which is 6.24 psf or 50 MPH wind.  But that number needs work.  When we did Burj Khalifa, the 30 years cycle wind is about 83 psf (yes, 13.3 times of 6.24 psf), and we wonder why there are stack effect air movement into the buildng and leakages.  Though compares a typical year with 50 years cycle wind is obviously not entirely fair, but it illustrate how the design criteria we se to test the curtain wall has room for improvment. Best,
Luke

I'll add this resource below to the discussion. Part of the issue is that the desired output values from leakage tests are not the same input values used in the analysis. For example, for window u value, the input is window u value. But for leakage testing, as Luke mentions below, the output value is in pressure, pascals, but the software inputs is often a leakage rate per area or volume at an assumed windspeed. And the windspeed varied depending on which program you use. This PNNL paper offers some guides for modeling. https://www.pnnl.gov/main/publications/external/technical_reports/PNNL-18898.pdf Adam McMillen, PE, CPHC, LEED AP BD+C Director of Sustainability [IMEG Corp.] IMEG Corp. 1100 Warrenville Road | Suite 400W | Naperville, IL 60563 (630) 527-2320 | phone (312) 852-1360 | mobile (630) 527-2321 | fax Adam.M.McMillen@imegcorp.com website | my linkedin | vCard | map | regional news [Twitter][Facebook][LinkedIn] Learn more about us and the IMEG story! This email may contain confidential and/or private information. If you received this email in error please delete and notify sender.

"What are we missing to make this conversation more consistently easier?"

More industry education in building physics/science. A lot (possibly most) engineers and modelers don't know much about air leakage testing, how it's conducted, what are appropriate values, or what that means for a load calc/model. I don't think it's something taught in most academic programs, so most of the knowledge a good one gets is through personal interest, reading up on white papers, attending conferences, and because of the willingness of people like you to help educate them through your work.

On the flip side, it's possible that many of those same engineers/modelers haven't learned about this stuff in part because most of their clients don't bring it up. The level of knowledge on the subject of air infiltration among architects is just as varied as the engineers'. We typically try to include air leakage performance criteria into the OPR, and you should see some of the pushback we get from architects for putting down on paper what is either already code or standard. I've countered with, "OK, what value do you suggest?" which usually turns into an explanation of all the reasons why there's no way to commit to *any* number. Usually at that point in the meeting, the Owner/Owner's Rep has stopped firing off emails and has gotten *really* interested in what's happening. 

In terms of making this conversation easier, you would probably see significant improvement on most projects by getting the architect and engineer to agree to make calculations based on code or the applicable standard, and then commit to testing that proves the building performs to those rates. That's where the biggest opportunity lies in my opinion. If I used the apartment in the white paper, here's what I mean, converting with the 0.112 factor (which isn't perfect either). 

• Engineer's Assumed Leakage: 18,963 cfm
• Actual Leakage (@0.4 cfm/sf): 2,370 cfm (about the same as CA Code, GSA P100, and some others)
• Army COE Std (@0.25 cfm/sf): 1,481 cfm
• PHIUS+ (@0.08 cfm/sf): 474 cfm

One could spend some time figuring how much effort is worth chasing out that leakage to get from code to higher standards, but I think it shows the biggest gain could be in building a whole team strategy about how to at least meet the applicable code/standard target.