Flow Exponent Values and
Implications for Air Leakage
Testing
NBEC 2014
ROBIN URQUHART, MBSC, MA NRES, RDH BUILDING ENGINEERING
DR. RUSSELL RICHMAN, P.ENG, RYERSON UNIVERSITY
Outline
 
Introduction to air leakage testing
 
Relationship between flow and pressure
 
Case study building
 
Abnormal flow exponents
 
Data extrapolation to operating pressures
 
Conclusions/Implications
 
Further study
Air leakage testing
 
Air leakage = Uncontrolled air movement through the
building enclosure
 
 
Can account for up to 50% of a buildings heat loss/gain
Affects assembly durability
Stack
Wind
Fan induced pressure
 
Induced pressure differentials using fans
 
 
 
ASTM E779-10 (10-60 Pa)
 
 
multi-point
ASTM 1827-12 (50 Pa)
 
 
Mechanical
Blower door
single point
USACE (75 Pa)
 
single point
Why test?
 
Evaluate performance and determine retrofit options
 
 
 
Determine leakage rates for energy models
 
 
Single building
Building stock
4 Pa – 10 Pa
Code or other
specification compliance
 
LEED, Energuide, etc.
The Relationship between Flow and Pressure
Pressure (P) and Flow (Q)
Power Law Equation (Bernoulli):
!="​(∆%)↑( 50
10 Pa
Where,
Q = flow
C = flow coefficient
P = pressure
n = flow exponent (dimensionless).
Flow exponent (n)
Power Law Equation:
log Flow vs log Pressure
2.5
! = #$ + & 2.45
2.4
0.5
= Turbulent
2.35
log Flow
•  flow resistance varies with the square of velocity
2.3
2.25
1.02.2 = Laminar flow
2.15
•  flow resistance proportional to velocity
Slope = n > 0.5 and < 1.0
2.1
2.05
2
0.8
1
1.2
1.4
1.6
1.8
log Pressure
Series1
Linear (Series1)
2
Case study building
  13-story
MURB
  Enclosure
 
retrofit 2012
focus on air tightness
  Air
leakage tested
 
Pre-retrofit
 
Post-retrofit
Guarded-zone method
Suite 101
Suite–102Case
Suite
301
Suite 302
Suite 303
Flow exponent
values
study
building
Pre-press
Post-press
Change2
0.59
0.56
-0.03
0.56
0.67
+0.11
0.60
0.75
+0.15
0.68
0.85
+0.17
0.54
0.85
+0.31
Pre-depress
Post-depress
Change
0.92
0.79
-0.13
0.80
0.69
-0.11
0.50
0.60
+0.10
0.62
0.57
-0.05
0.61
0.37
-0.24
Suite 1101
0.52
0.71
+0.19
Suite 1102
0.49
0.56
+0.07
Suite 1103
0.53
0.64
+0.11
Suite 1301
0.74
0.77
+0.03
Suite 1302
0.52
0.72
+0.20
0.47
0.61
+0.14
0.46
0.43
-0.03
0.78
0.59
-0.19
0.57
0.57
0.00
0.63
0.47
-0.16
Pre-press
Post-press
Change
Pre-depress
Post-depress
Change
1
pre-press = pre-retrofit pressurization, post-press = post-retrofit depressurization, pre-depress = pre-retrofit depressurization,
post-depress = post-retrofit depressurization
2
positive change indicates a move towards laminar flow (smaller cracks), negative change indicates a move towards turbulent
flow (larger cracks)
Flow exponents outside of theoretical range
“…blower door tests occasionally yield exponents less than the
Bernoulli limit of 0.5. In fact, it is physically impossible for such
low exponents to occur.”
Sherman, Wilson and Kiel. 1986
So why do we get them?
Flow Exponents <0.5
Series 1: n=0.5-1.0
Q
P
logQ logP
800
10
2.9 1600
30
3.2 1.48 2200
50
2500
60
1 3.34 1.7 3.4 1.78 Series 2: n<0.5
n
r-sq
Q
P
0.63
0.999
800
10
1300 30
logQ logP
2.9 1 n
r-sq
0.42
0.999
3.11 1.48 1590 50
3.2 1.7 1710 60
3.23 1.78 Flow exponents <0.5
3.5
3.4
R² = 0.9999
Log Flow
3.3
3.2
3.1
R² = 0.9993
3
2.9
2.8
0.9
1.1
1.3
1.5
Log Pressure
n = 0.63
n = 0.42
1.7
1.9
Geometry of the Enclosure
 Higher
pressure differentials change
leakage path geometry
 Resulting
in abnormal flow exponent
(n) values.
10 Pa
50
 What effect on the data does a changing n
value cause?
Extrapolating to lower pressure differentials
40 L/s
Difference of ~35%
n = 0.40
n = 0.65
n = 0.50
n = 0.80
Enclosure change at high pressure
N = 0.68,
Rsq = 0.99
Q P 650 865 1100 1180 1250 1375 30 50 70 75 80 90 N = 0.62,
Rsq = 0.99
Q P 650 865 1100 1150 1200 1250 30 50 70 75 80 90 Flow
Flow
Enclosure change at higher pressures
1500
1400
1400
1350
1300
1300
1200
1250
1100
1200
1000
1150
900
1100
800
700
1050
600
1000
20
65
70 40 75
static enclosure
static enclosure
60
80
Pressure
Pressure
85 80 90
enclosure changes
enclosure changes
100
95
Adjusted values using n
250
1500
1400
1400
1200
200
1300
Flow
Flow
Flow
1000
1200
150
1100
800
1000
600
100
900
400
800
50
700
200
600
00
20 20
25 l/s
14% difference
20 340
static enclosure
nclosure
static eenclosure
static 40
60
4
Pressure
Pressure
Pressure
60
80
5
changing enclosure
changing enclosure
enclosure changes
80
1006
Conclusions
 
The enclosure may not remain static throughout the
test
 
The linear relationship may not be accurate at operating
pressure
 
Substituting n values can have magnified impacts at
lower/higher pressure differentials
 
R-squared values not perfect indicators
Implications for the industry
 
Energy models using extrapolated flow rates
may be inaccurate
 
Sizing mechanical systems
for predicted in operation
flow rates
 
Retrofit options
The Good
 
Multi-point testing still important
 
Order of magnitude for comparison purposes
 
Flow exponents tell us something, we might not know
exactly what it is
 
Standards at higher pressures can be used for
compliance purposes
Flow exponent corrections
In-test corrections
 
Check R-squared values during testing
 
Check n values at the termination of each test
Post-test corrections
 
Review data points and remove lowest pressure
differential point(s)
 
If not reconciled, single point test at 50 Pa
Further study
 
Sensitivity analysis: R-value threshold
 
The use of flow exponent values in forensics
 
Standardized leakage metric and pressure differential
for comparison purposes
 
Effect on flow exponent of enclosure geometry
changes
Questions
ROBIN URQUHART
[email protected]