Crack method of infiltration not set up correctly?

4 posts / 0 new
Last post

I have another question about whether people have gotten wind speed to
have an effect on crack infiltration.

Below is an hourly report for a few hours of January 1st. I have a
building set up with each space's infiltration based on the crack
method, with an infiltration schedule of all 1.0s. Exterior walls are
set up with crack coefficients pretty high (leaky building) - the
coefficient is about 7.5, which corresponds to a 0.6" crack around the
entire perimeter of the wall (it's a leaky corrugated metal wall
construction).

This report is for a typical west (270) facing wall, about 45 feet by 8
ft tall. 70 feet below the neutral level. As expected, there is CFM
infiltration, which is great. The CFM values decrease as you go up the
building (12-story), which is perfect. Means the whole crack method vs
neutral height is working. However, there appears to be NO impact of
the infiltration based on wind speed. The CFM for this given wall is
entirely dependent on the OA DB temp, and does not change at all with
wind speed. See below. Highlighted areas show similar temp hours but
with varying wind speed. CFM for this 1st floor exterior wall doesn't
change at all with differing wind speeds.

I thought that the crack method was supposed to alter CFM based on wind
speed, but NOT alter based on difference in indoor and outdoor temp.

What am I doing wrong here? For a leaky building, wind speed SHOULD
have a greater impact on CFM than stack effect.

Var 4

Var 17

Var 19

Var 3

Var 4

Var 7

OA DB

Wind Spd

Wind direction

Outside air film U-value

Pressure diff due to wind velocity & stack effect

Crack method air flow for wall (cfm)

19

6

270

1.81016

0.109004

703.896

19

6

270

1.80501

0.109004

703.896

19

8

270

2.10433

0.109004

703.896

19

4

270

1.50129

0.109004

703.896

18

5

315

1.66397

0.111218

711.009

16

3

248

1.37812

0.115674

725.112

12

0

0

1.13297

0.1247

752.869

9

4

248

1.53612

0.13157

773.33

9

0

0

1.10755

0.13157

773.33

12

0

0

1.10909

0.1247

752.869

18

3

180

1.35996

0.111218

711.009

21

10

180

2.40378

0.104603

689.541

21

3

180

1.34907

0.104603

689.541

James Hansen, PE, LEED AP

James Hansen's picture
Offline
Joined: 2011-09-30
Reputation: 200

James,

Out of curiosity, how did you output the "pressure difference due to wind & stack effect"?

The Crack method doesn't include an explicit temperature correction term, ?T, but it does account for the stack effect (which is the pressure difference caused by air density/temperature gradient). But I couldn't find how this stack effect is calculated in the documentation, unless it uses the same equations as for the S-G (Sherman-Grimsrud) or ASHRAE-ENHANCED methods.

Table 32 in the documentation reports that for the Crack method, wind-speed correction depends also on the following SITE-PARAMETER keywords: TERRAIN-PAR1, TERRAIN-PAR2, WS-TERRAIN-PAR1, WS-TERRAIN-PAR2 and WS-HEIGHT. The method that references these parameters is the S-G infiltration method (see http://gundog.lbl.gov/dirpubs/10852_ShermanGrimsrud.pdf). Perhaps your site parameters are specified such that wind is not affecting the infiltration rate.

Nevertheless, if cfm = INF-COEF * dP0.8 * Area for a wall (help file under Envelope Components > Exterior wall), then for your case cfm = 7.5*.109004 (inches of water)^.8*(45*8) = 458 cfm, not 703.896 as calculated by eQuest below.

Maybe someone with more knowledge of the documentation can shed light on the calculations done by eQuest for infiltration. It would be interest to me as well. I am reviewing infiltration calculations/assumptions in several energy software.

Lisa

Ng, Lisa's picture
Offline
Joined: 2011-09-30
Reputation: 0

Lisa, that's one of the available parameters for an hourly report block.

As for the stack effect: each space has an input for the level above or below the neutral pressure height (NPH). That's how the stack effect is modeled in my project, and it appears to be working fine.

Jeff Hirsch emailed me back and basically said that:

1) Crack modeling in DOE2 is approximate at best, and I might be better off just doing a CFM / sq ft "with a wind and dT modifier with differential schedules multipliers based upon the HVAC schedule and door openings and elevator activity". Not entirely sure what that means - I think maybe Jeff is saying to actually make an infiltration schedule that mimics the wind speed, which would be terribly time consuming, but I guess more accurate. Or maybe the CFM / sq ft automatically has a wind and dT modifier, although the crack method is supposed to do that but it's not modeling things correctly right now. Jeff says that might be because the wind direction is not lining up with the outward normal direction of the wall, but I can't seem to get ANY variation in infiltration CFM no matter what the wind direction is.

2) Exfiltration at the top of a building is not modeled (which limits the accuracy of stack effect modeling in my opinion)

Also, as for the calc you reference: the CFM calc is different for U-value walls (it's more like DP^0.5). Still don't know if that adds up or not.

Thanks!

James Hansen, PE, LEED AP

James Hansen's picture
Offline
Joined: 2011-09-30
Reputation: 200

James,

As for Jeff's advice, it would be "terribly time consuming" and probably not too accurate either since what would be the values you would use for the "wind and dT modifier"? That is the current limitations of energy programs - not being able to appropriately determine pressures (and thus airflows) with changing indoor and outdoor conditions, which airflow models such as CONTAM and COMIS calculate.

Even with the dP^0.5, the numbers still don't add up.

Will you leave your simulation as is then?

Lisa

Ng, Lisa's picture
Offline
Joined: 2011-09-30
Reputation: 0