Final Report

Clearance Distance for Gas

Appliance Sidewall Venting

To:

Fire Protection Research

Foundation

December 5, 2007

Final Report
for

Clearance Distance for Gas Appliance Sidewall Venting

December 5, 2007
Prepared for:

The Fire Protection Research Foundation

1 Batterymarch Park
Quincy, MA 02169

Prepared by Battelle:

Stephen Ricci
Sherwood Talbert
Darrell Paul

Battelle Memorial Institute

Applied Energy Systems
505 King Avenue
Columbus, Ohio 43201

TABLE OF CONTENTS
1.0 INTRODUCTION .................................................................................................................... 1
2.0 APPROACH ............................................................................................................................. 1
3.0 LITERATURE REVIEW ......................................................................................................... 2
4.0 CFD ANALYSIS ...................................................................................................................... 2
4.1 Model Setup and Boundary Conditions.............................................................................. 4
4.2 Method of Analyzing the Results ....................................................................................... 5
4.3 Discussion of Results of Cases Initially Agreed Upon....................................................... 6
4.3.1 Effect of House Separation Distance ......................................................................... 7
4.3.2 Effect of Outdoor Temperature.................................................................................. 7
4.3.3 Effect of Vent Terminal Design................................................................................. 8
4.3.4 Effect of Wind Speed and Direction.......................................................................... 8
4.4 Discussion of Results of Two Final Cases.......................................................................... 9
5.0 RECOMMENDATIONS........................................................................................................ 14
6.0 REFERENCES ....................................................................................................................... 15
APPENDIX A: Final Presentation Delivered to Technical Review Committee Sept 6, 2007

1.0 Introduction
The objective of this work was to assess the effects of sidewall venting of flue gas from a highefficiency, natural gas appliance in the presence of a neighboring home. The National Fire
Protection Association (NFPA) and the American Gas Association (AGA) cosponsored this
study as an initial attempt to identify basic criteria related to the dispersion of sidewall-vented
flue gas between neighboring houses. For this study, Battelle proposed to perform a review of
the literature for information on issues related to sidewall venting and a computational fluid
dynamics (CFD) modeling study to investigate the effects of important parameters, such as the
spacing between houses, wind direction, and vent terminal design.

2.0 Approach
A technical review committee, comprised of representatives of NFPA member companies and
AGA, was assembled to guide the project and evaluate the results. Initial discussions between
Battelle and the review committee were held to establish consensus on a detailed scope of work
for this study, taking into account the funds available and the desired outcomes. With respect to
the CFD study, there is a vast array of parameters that are, or could be, important in determining
how flue gas will disperse in the real world and what the limits might be for safe and problemfree venting. The input rating, excess air, and efficiency of the appliance determine the flow
rate, temperature, and composition of the flue gas. The design and location of the sidewall vent
terminal has an important influence on how the gas will disperse. The spacing between the
houses is critical, as one can reasonably expect that there is some separation distance below
which dispersion would be restricted. Wind speed and direction are also important, as well as
the outdoor ambient temperature. The topology in the vicinity of the houses, and the presence of
barriers such as nearby buildings, trees, and hills could also have a substantial impact on the
dispersion of flue gas. To investigate the effects of all of these important parameters would
involve an exhaustive and costly modeling study and would require more funding than was
available for this work. The fundamental objective of this study was to take a first step toward
understanding the physics and to gather insight as to the factors that may be important in
generating guidelines or codes for sidewall venting.
It was decided in those initial discussions to select a typical Category IV appliance with a 100
kBtu/hr input rating operating with 40 percent excess combustion air. Such an appliance,
assuming roughly 92 percent efficiency, would produce a flue gas with a flow rate of 23 acfm
and a temperature of 115 F. Two identical houses, each 36 ft wide, 25 ft on the sides, 20 ft high
to the bottom of the roof, and with a peak height of 28 ft were used. Two house separation
distances, 5 ft and 10 ft, were decided upon. The vent would be located one foot from the
ground, centered horizontally, and protrude 1 foot from the face of the house with the vent. The
vent would be 2 inches in diameter and have two terminal designs, one with no terminal
(straight vent) and one with a 6-inch diameter plate facing the vent outlet to force a radial
dispersion of flue gas as it exits the vent (radial vent). Winter (0 F) and summer (75 F) outdoor
temperatures would be investigated, with no wind and a 7 mph wind blowing perpendicular to
the houses and parallel (between) the houses. This agreed-upon set of parameters resulted in the
1

eleven initial CFD cases. The definition of two additional cases was held in reserve to allow for
investigation of questions or opportunities that might arise from the results of the initiallydefined set of cases. Table 2-1 lists all of the CFD cases run in this study. Cases 12 and 13,
which are highlighted in Table 2-1, are the two additional cases that were defined after
completion of the first eleven. The rationale for selecting those cases is discussed later in the
report.

Case
Number
1
2
3
4
5
6
7
8
9
10
11
12
13

Table 2-1. Parameter Matrix for CFD Modeling Cases

House
Outdoor
Wind
Vent Type Separation (ft) Temperature (F) Direction
Radial
5
0
NA
Radial
5
75
NA
Radial
10
0
NA
Radial
10
75
NA
2 Straight
5
0
NA
2 Straight
5
75
NA
2 Straight
10
0
NA
2 Straight
10
75
NA
2 Straight
5
75
East
2 Straight
5
75
North
2 Straight
5
75
South
45o Down
5
0
NA
2 Straight
5
0
NA

Wind Speed
(mph)
0
0
0
0
0
0
0
0
7
7
7
0
0

3.0 Literature Review

A sound technical basis for determining the minimum clearances for sidewall venting to adjacent
buildings is necessary to affect a uniform code change that can be accepted by industry and
adopted by code authorities. A literature search and review was conducted using the Dialog
Database to determine if any of the past work on venting of gas appliances could be used to form
the basis for decision-making regarding the separation distance between buildings in the
presence of a sidewall-vented appliance. Only a few relevant papers were found and these are
summarized below.
From 1988 to 1994 the Gas Research Institute (GRI) funded research into the venting of gas
appliances (Ref 1). The objective of this program was to develop guidelines for the practical and
safe venting of flue gases from mid- and high-efficiency gas-fired appliances. In 1992, the GRI
venting program conducted a study to examine means to sidewall vent gas appliances in
multistory buildings (Ref. 2). As part of this work, the issue of sidewall venting between closely
spaced buildings was identified. Several European vent configurations were studied which
allowed sidewall vented appliances to be discharged into a manifold that ran vertically up the
side of the building and exited above the roofline. The simplest design incorporated a vertical
manifold with two openings, one at the top and one at the bottom. Sidewall vents could be
discharged into the manifold, which would cause an up flow due to the buoyancy of the warm
vent gases, while maintaining nearly atmospheric pressure in the vent due to the bottom opening.

Several variations of this concept are currently in use in Europe, including a U-shaped vent with
both openings above the roof. Vent gases are discharged into one leg of the manifold, and the
other leg is used to supply combustion air to the appliance. These configurations, allow sidewall
venting to take place between buildings that are right next to each other because the manifold
discharges the vent gases above the roofline.
Between 1999 and 2002, Battelle examined the minimum clearance distances between sidewall
vented gas appliances and adjacent building openings along the same wall as the sidewall vent
(Refs. 3 and 4). Dr. James Reuther used analytical expressions for gas jet dispersion to estimate
dilution factors for vent gases being discharged horizontally from the side of a building. His
results show that for a straight out discharge of a sidewall vent, the minimum separation distance
could be less than one foot between the vent outlet and any of the same building openings, and
still provides adequate dilution to the vent gases to alleviate safety concerns for the occupants of
the building. However, his study did not include the effects of wind on the vent discharge, the
effect of different types of vent terminals, or the influence of adjacent buildings or structures on
plume dispersion.
In 2006 and 2007, Battelle conducted a study for the Gas Appliance Manufacturers Association
on the minimum separation distance between sidewall vents and adjacent building openings in
which wind effects and different types of vent terminals were examined. This study used
Computational Fluid Dynamics (CFD) models to compute plume dispersion and dilution factors
which include the effects of different wind speeds and directions, different types of vent
terminals, different input ratings of appliances, and vent terminal height above grade. This study
showed that the minimum separation distances between the vent discharge and the same wall
openings currently in the NFGC could probably be reduced if the industry could agree on a
reasonable dilution factor for the vent gases. The GAMA study did not calculate dilution factors
for adjacent buildings located in the vicinity of the sidewall vent.

4.0 CFD Analysis

Computational Fluid Dynamics (CFD) simulations were used in this project to estimate the
dispersion of flue gas from high-efficiency, sidewall-vented appliances. In CFD, a graphical
representation of a fluid-dynamics problem is generated and the fluid space in the domain is
discretized using a computational mesh. CFD software is used to numerically solve the
differential equations of fluid-flow, energy, and species transport, three-dimensionally, on the
computational mesh given conditions specified by the user on the boundaries of the fluid
domain. If done correctly, the result is a reasonably accurate calculation of fluid velocity,
pressure, temperature, and concentration on each node of the mesh. The solution can be
examined and analyzed either graphically or with numerical calculations to determine the
conditions anywhere in the domain. FLUENT, a commercial CFD software package now
marketed by Ansys, Inc., was used for these analyses.

The results of the CFD analyses were presented to the technical review committee via a livemeeting presentation on September 6, 2007. The presentation that was used for that live meeting
is incorporated into this report as Appendix A. The discussion here will focus on the overall
effects noted from the analysis results and will refer to the figures and plots in the presentation
when appropriate. The reader is referred to Appendix A for explanation and discussion of caseby-case results.
Note: In the discussion that follows, House 1 refers to the house with the vent and House 2
refers to the neighboring house.

4.1 Model Setup and Boundary Conditions

Slide 4 in Appendix A shows a diagram of the model used in the simulations without wind. For
the cases without wind, the two houses are enclosed in a hemispherical dome, centered on the
ground halfway between the houses, with a radius of 400 ft. The surface of the dome represents
the sky, far from the houses, and is set as an atmospheric pressure boundary in the model. A
pressure boundary in CFD allows fluid to flow into and out of the domain such that the pressure
everywhere on the boundary is held constant at a specified value (atmospheric pressure in this
case). The bottom face of the dome represents the ground and is modeled as a wall boundary
where the air velocity is fixed at zero. The two houses are visible in the center of the model.
House 1 (with the vent) is on the left (south) side and House 2 (the neighboring house) is on the
right (north) side. The houses are separated by a distance of 5 feet in the diagram shown on
Slide 4.
Slide 5 in Appendix A shows a diagram of the model used in the simulations with wind. For the
cases with wind, the two houses are enclosed in a rectangular box that is 800 ft x 800 ft
horizontally and 400 ft high. The bottom face of the box represents the ground and is modeled
as a wall boundary where the air velocity is fixed at zero. The top face of the box is modeled as
a free-stream boundary far from all walls and obstacles where it is assumed that the velocity is
not changing (zero shear stress). There is no friction or loss of mechanical energy (pressure) on
a free-stream boundary. The side walls are used to generate a wind with a specified velocity at
the outdoor ambient temperature. To generate a south wind, for example, the left side of the box
is assigned a wind velocity where air will enter the domain uniformly. The opposing right face is
assigned atmospheric pressure. The two sides of the box that run parallel with the wind direction
(east and west, or front and back faces) are specified as free-stream boundaries, similar to the top
face of the box.
Slide 6 in Appendix A shows diagrams of four vent designs that were modeled in this study. In
each diagram the flow direction is from left to right. The straight vent (top left) and radial vent
(top right) were the designs decided upon in the initial meetings on this project. The 45 degree
down (bottom left) and three-inch straight (bottom right) vents were decided upon after
completion of the eleven cases shown in Table 2-1. The rationale for selecting the conditions
and vent configurations for the two final cases (Cases 12 and 13) is discussed later in this report.

The boundary condition set at the vent outlet, where the venting gas enters the fluid domain,
depends on the appliance being modeled. The input rating, excess combustion air, and efficiency
of the appliance determine the flow rate, composition, and temperature at the outlet of the vent.
In this study, the appliance was fixed at 100 kBtu/hr, 40 percent excess air, and 92 percent
efficiency, resulting in a constant flow rate and temperature of 23 acfm and 115 F, respectively.
The velocity of flue gas through a two-inch vent, at a volumetric flow rate of 23 acfm, is 6.1 m/s
(20 ft/s).
The concentration of carbon dioxide and water vapor, in the flue gas are certainly important
variables in these analyses and can be computed from combustion calculations using the fuel
composition, excess air, and their associated temperatures. There is however, a degree of
complexity to modeling the flue gas as a multi-component gas in the CFD simulation. In order
to track the concentration of each component of the flue gas through the fluid domain
surrounding the house, the multi-component diffusion coefficient of each component would need
to be known or calculated. We have found from experience that the uncertainties in performing
the multi-component calculations do not warrant the added complexity and computational cost.
In similar analyses on past projects, we have chosen to model the flue gas as a single component,
which well call flue gas, with a representative diffusion coefficient in air and an ideal-gas
density based on the temperature. The mass fraction of flue gas, by definition, has a value of one
at the vent outlet. By tracking the concentration of flue gas in the domain surrounding the house,
the model computes the degree to which the flue gas is diluted by the outside air at every point in
the domain. The concentration of a given component, such as carbon dioxide, at every point in
the domain can then be estimated with reasonable accuracy by assuming a concentration at the
vent outlet and multiplying by the normalized concentration from the CFD calculation. In other
words, modeling the flue gas as a single-component gas with a maximum concentration at the
vent outlet, which is defined as one, normalizes the concentration profile in the domain. It
assumes that the concentration of any component of the flue gas is dispersed and diluted by the
same amount as the flue gas itself.

4.2 Method of Analyzing the Results

A parametric study was conducted using CFD simulations to determine the concentration profile
of flue gas everywhere in the fluid domain, with emphasis on the region surrounding the vent
and between the two houses. In each of the cases, the steady-state concentration profile of flue
gas was computed on the face of House 1 (with the vent) and the face of House 2 (neighboring
house) opposite the vent. The objective of the simulations was to determine the maximum
concentration of flue gas (or carbon monoxide) on the surface of each house. The point of
maximum concentration is also the point of minimum dilution, which is defined as the reciprocal
of the maximum normalized concentration. It measures the degree to which the flue gas has
been diluted in the worst case position on each house (i.e. a minimum dilution of 1000 means
that the flue gas has been diluted by a factor of 1000 in the worst case location). The point of
minimum dilution on House 1 was determined at distances greater than four feet from the vent
because existing guidelines state that doors or windows should be located at least this distance
from the vent location. The point of minimum dilution on House 2 could occur anywhere. No
assumptions about the location of doors or windows on either house were made.

The contour plots in the presentation shown in Appendix A show the normalized concentration
of flue gas. The plots are scaled by color, from blue to red, with blue representing the lowest
concentration and red representing the highest concentration. The scale is from a normalized
concentration (or mole fraction) of 0.001 (dilution factor 1000 on the blue end) to 0.2 (dilution
factor 5 on the red end), and is the same in all of the plots. Setting low and high limits for the
contour plots allows for a view of the shape and trajectory of the plumes. There is no color in
the plots where the concentration is outside the limits. In other words, at positions outside the
edge of the plumes shown in the contour plots, the normalized concentration is below 0.001 and
the dilution factor is greater than 1000.

4.3 Discussion of Results of Cases Initially Agreed Upon

The value and location of minimum dilution (maximum concentration) on the face of each house
was recorded for each of the CFD cases run in this study. Table 4-1 lists the cases, their input
parameters, and the results for Houses 1 and 2. The two cases that are highlighted at the bottom
of Table 4-1 are the two additional cases that were decided upon after completion of the initial
eleven cases. The rationale for selecting those cases is discussed in section 4.4 of this report.
The X and Y coordinates for the location of maximum concentration are given in feet, with the
origin at the base of the house vertically and centered horizontally. The X coordinate measures
how far horizontally the location of maximum concentration is from the center of the house,
while the Y coordinate measures how far the location of maximum concentration is from the
ground.
With the exception of the case with an east wind of 7 mph (Case 9), which is the direction that
runs between the houses, the horizontal location of maximum concentration is within 4 feet from
the center of either house. Unless deflected horizontally by wind, the plume generally rises
directly in plane with the vent. The vertical location is affected most by the outdoor temperature
(or more specifically, the difference in temperature between the flue gas and ambient) and house
separation distance. Temperature difference and separation distance influence the vertical
location of maximum concentration because of the competing rates of flue-gas convection and
buoyancy. When the outdoor temperature is low, the flue gas will be more buoyant and rise
more rapidly relative to the rate that it is forced horizontally by the vent. The opposite is true
when the outdoor temperature is warmer. When the houses are separated by a greater distance,
the plume has more time to rise before impacting the neighboring house.
The presentation in Appendix A provides a case-by-case discussion, with graphical
representations of plumes, for the CFD analysis. The effects of parameters varied in the first
eleven cases are discussed in the sections that follow. Reference to case numbers will be made
when appropriate so that the reader may refer to Appendix A for additional information.

Table 4-1. Matrix of Results for All CFD Cases

Parameters

Case
1
2
3
4
5
6
7
8
9
10
11
12
13

Vent
Radial
Radial
Radial
Radial
2" Straight
2" Straight
2" Straight
2" Straight
2" Straight
2" Straight
2" Straight
45o Down
3" Straight

Bldg.
Sep.
(ft)
5
5
10
10
5
5
10
10
5
5
5
5
5

Results for House 1

Results for House 2

Max
Max
Max
Max
Wind Amb
Min
Conc X Conc Y
Min
Wind Speed Temp Conc X Conc Y
(F) Pos (ft) Pos (ft) Max Conc Dilution Pos (ft) Pos (ft) Max Conc Dilution
(mph)
Dir
NA
0
0
0.15
7.25 8.00E-02
13
-0.05
19.50 1.15E-03
869
NA
0
75
-0.08
7.40 1.17E-01
9
0.45
19.73 1.62E-03
616
NA
0
0
-0.15
7.25 7.55E-02
13
-0.49
17.76 3.29E-06 304,300
NA
0
75
-0.15
7.25 1.04E-01
10
0.03
14.99 6.40E-06 156,200
NA
0
0
2.01
19.63 1.61E-03
622
-0.46
1.53 8.46E-02
12
NA
0
75
4.39
19.74 4.55E-03
220
0.32
1.07 9.91E-02
10
NA
0
0
-0.17
19.57 2.41E-04
4,145
0.15
10.14 2.05E-02
49
NA
0
75
0.59
7.20 3.18E-05
31,420
-0.06
2.30 5.93E-02
17
East
7
75
17.85
0.82 6.68E-03
150
17.39
2.08 9.40E-04
1,064
North
7
75
-2.28
7.48 2.63E-02
38
-2.38
1.51 6.60E-02
15
South
7
75
3.66
17.48 8.57E-03
117
0.32
1.07 7.20E-02
14
NA
0
0
-0.50
19.82 1.77E-03
564
0.49
0.28 1.20E-01
8
NA
0
0
0.97
19.59 1.23E-03
813
0.31
3.53 8.48E-02
12

4.3.1 Effect of House Separation Distance

Figures 4-1 and 4-2 show plots of minimum dilution factor versus house separation distance,
without wind, on Houses 1 and 2, respectively. In Figures 4-1 and 4-2, dotted lines refer to the
radial vent cases and solid lines refer to the 2-inch straight vent. Blue lines with circular symbols
refer to an outdoor temperature of 0 F, while red lines with triangular symbols refer to an outdoor
temperature of 75 F.
In Figure 4-1, there is very little effect of separation distance on the House 1 minimum dilution
factor with a radial vent (dotted lines). The radial vent design traps the flue gas nearer to House
1 regardless of house separation distance. There is an effect of separation distance on the
dilution factor for House 1 with the 2-inch straight vent (solid lines) because the straight vent
design throws the flue gas away from House 1 toward House 2. When the separation distance is
higher, the gas is allowed to disperse away from House 1 and is more diluted near House 1.
Regardless of separation distance, however, the flue gas is diluted by more than a factor of 100
before reaching the surface of House 1.
Figure 4-2 is a similar plot of minimum dilution factor on House 2 (the neighboring house). The
results are virtually opposite those in Figure 4-1 for House 1. Because the radial vent traps the
flue gas nearer to House 1, the dilution factors are much higher, and well above the threshold, on
House 2. Because the straight vent throws the flue gas toward House 2, the dilution factors are
much lower on House 2. The case with a 2-inch straight vent, an outdoor temperature of 0 F, and
a house separation distance of 10 feet (Case 7) is the only straight vent case with no wind that
meets the threshold requirement for the neighboring house. In that case, the rate of buoyancy
(rise) is just sufficient to allow the flue gas to disperse and be diluted by a factor of more than 44
before reaching House 2.

4.3.2 Effect of Outdoor Temperature

Figures 4-3 and 4-4 show plots of minimum dilution factor versus outdoor temperature, without
wind, for Houses 1 and 2, respectively. As discussed previously, outdoor temperature
determines the rate of buoyancy or rise. When the outdoor temperature is cooler, the flue gas
will rise more rapidly. Both plots show that temperature has a relatively small effect on dilution
at the surface of either house. All of the plots thus far indicate that the dilution factors are
governed more by vent design than by temperature. The effect of vent design is discussed in the
next section.

4.3.3 Effect of Vent Terminal Design

Figures 4-5 and 4-6 show plots of minimum dilution factor versus vent terminal design, without
wind, on Houses 1 and 2, respectively, for the two vent designs considered in the initial set of
cases (radial vent and 2 straight vent). In these plots, the dotted lines are for cases at a house
separation distance of 5 feet and the solid lines represent cases with a separation distance of 10
feet. These plots clearly show that in the absence of wind, the dilution factor is most dependent
on vent terminal design. The radial vent clearly traps the flue gas nearer to House 1 (low
dilution factor) and protects House 2 (high dilution factor). The 2-inch straight vent produces the
opposite effect. It throws the flue gas toward House 2 (low dilution factor) and protects House 1
(high dilution factor).

4.3.4 Effect of Wind Speed and Direction

Two wind cases, both at a wind speed of 7 mph, were initially selected for analysis in this study,
one perpendicular to the houses and one parallel, or between the two houses. After some
discussion among members of the technical review committee, it was decided to run the
perpendicular wind cases in both directions (i.e. wind from behind House 1 and wind from
behind House 2) because it was not clear the effects would be the same. When the wind blows
over a building, there is a zone of recirculation (recirculation shadow) on the downwind side of
the building. There is turbulence on the downwind side that forces the air downward in a
circular fashion and back toward the downstream face of the building. It was not clear that the
effect of a shadow cast by House 1 in the presence of a south wind (from behind House 1) would
be the same as a shadow cast by House 2 in the presence of a north wind (from behind House 2).
Figure 4-7 is a bar chart that shows the dilution factor on both House 1 (blue bars) and House 2
(red bars) in each of the three wind cases. In each of the cases, the vent is a 2-inch straight vent,
the house separation distance is 5 feet, and the outdoor temperature is 75 F. Case 6, which is the
similar case with no wind, is included in the plot for comparison. The east wind, which is the
direction between the houses and parallel to them, carries the flue gas away and dilutes the space
between the houses considerably.
The effects of wind perpendicular to the houses are similar in each direction. As was shown
earlier in this report, and as is evident in the bars for no wind in Figure 4-1, the straight vent
tends to carry the flue gas away from House 1 and toward House 2, resulting in higher dilution
factors on House 1 (blue bar) and lower dilution factors on House 2 (red bar). The bars for the

north and south wind cases are similar, but the dilution factor on House 1 is decreased compared
to the case with no wind, and the dilution factor on House 2 is slightly increased compared to the
case with no wind. Generally speaking, the turbulence that is created between the houses as a
result of wind perpendicular to the houses tends to mix and disperse the flue gas. As a result, it
reaches the surface of House 1 at a slightly higher concentration, and House 2 at a lower
concentration, than it would without wind.

4.4 Discussion of Results for Two Final Cases

The definition of two final cases was held in reserve pending the insight that might be gained
from the analysis of the results of the eleven cases initially defined. From those initial cases, it
was determined that wind certainly can have a profound effect, and helpful in most cases, on flue
gas dispersion. Wind helps disperse and sweep the flue gas away, particularly in this application
where the vent is located between two buildings and the wind is forced to approach the vent at an
angle. Outdoor temperature had little effect, and no real determination could be made as to a
limit for house separation distance for safe venting. Vent design emerged as the dominant factor.
It is also a factor over which code bodies, manufacturers, and installers have some control.
Regardless of house separation distance, the radial vent design protects the neighboring house
and traps the flue gas near the house with the vent, while the 2-inch straight vent sends the flue
gas away from House 1 toward House 2. The two final cases were designed to begin answering
whether it is possible to achieve sufficient dilution on both houses at a separation distance of 5
feet with different vent terminal designs. It was further decided to run the two cases with no
wind and at a winter outdoor temperature of 0 F.
The two vent terminal designs selected for the final cases are shown on Slide 6 of Appendix A.
The 45-degree down vent has a nominal diameter of two inches and directs the flue gas
downward at a 45 degree angle. It was selected because the interaction of the flue gas with the
ground could create turbulence and allow the gas to disperse more before reaching House 2. The
3-inch straight vent design was chosen to decrease the momentum of the flue gas as it exits the
vent. Increasing the diameter from 2 inches to 3 inches decreases the flue-gas velocity by a
factor of 55 percent and could perhaps allow the gas to disperse more before reaching House 2.
Figure 4-8 shows a bar chart of minimum dilution factors on House 1 (blue bars) and House 2
(red bars) for the four vent designs modeled in this study. In all of the cases there is no wind, the
house separation distance is 5 feet, and the outdoor temperature is 0 F. Both the 45-degree down
vent and the 3-inch straight vent produce results similar to the 2-inch straight vent. The 45degree down vent sends the flue gas along the ground, but it still reaches House 2 at a
concentration similar to the 2-inch straight vent. Similarly, the 3-inch straight vent does not slow
the flue gas down enough to lower the concentration as it reaches House 2.

Minimum Dilution on House 1

Effect of House Separation Distance
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
4

10

11

House Separation Distance (ft)

Radial Vent @ T = 0 F

2" Straight Vent @ T = 0 F

Radial Vent @ T = 75 F

2" Straight Vent @ t = 75 F

Minimum Dilution Threshold

Figure 4-1. Effect of house separation distance on the minimum dilution factor for House 1.

Minimum Dilution on House 2

Effect of House Separation Distance
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
4

10

11

House Separation Distance (ft)

Radial Vent @ T = 0 F

2" Straight Vent @ T = 0 F

Radial Vent @ T = 75

2" Straight Vent @ T = 75 F

Minimum Dilution Threshold

Figure 4-2. Effect of house separation distance on the minimum dilution factor for House 2.

10

Minimum Dilution on House 1

Effect of Outdoor Temperature
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
0

15

30

45

60

75

Outdoor Temperature (F)

Radial Vent @ Sep = 5 ft

2" Straight Vent @ Sep = 5 ft

Radial Vent @ Sep = 10 ft

2" Straight Vent @ Sep = 10 ft

Minimum Dilution Threshold

Figure 4-3. Effect of outdoor temperature on the minimum dilution factor for House 1.

Minimum Dilution on House 2

Effect of Outdoor Temperature
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
0

15

30

45

60

75

Outdoor Temperature (F)

Radial Vent @ Sep = 5 ft
Radial Vent @ Sep = 10 ft

2" Straight Vent @ Sep = 5 ft

2" Straight Vent @ Sep = 10 ft

Minimum Dilution Threshold

Figure 4-4. Effect of outdoor temperature on the minimum dilution factor for House 2.

11

Minimum Dilution on House 1

Effect of Vent Terminal Design
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
Radial

2" Straight
Vent Type

Sep = 5 ft @ T = 0 F

Sep = 10 ft @ T = 0 F

Sep = 5 ft @ T = 75 F

Sep = 10 ft @ T = 75 F

Minimum Dilution Threshold

Figure 4-5. Effect of vent terminal design on the minimum dilution factor for House 1.

Minimum Dilution on House 2

Effect of Vent Terminal Design
1.E+06

log(Dmin)

1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
Radial

2" Straight
Vent Type

Sep = 5 ft @ T = 0 F

Sep = 10 ft @ T = 0 F

Sep = 5 ft @ T = 75 F

Sep = 10 ft @ T = 75 F

Minimum Dilution Threshold

Figure 4-6. Effect of vent terminal design on the minimum dilution factor for House 2.

12

1.E+06

1.E+05

Effect of 7 mph Wind

2" Straight Vent - Bldg Sep 5 ft - Amb T 75 F

House 1 (vent)
House 2

log(Dmin)

1.E+04

1.E+03

1.E+02

1.E+01

1.E+00
East Wind

North Wind

South Wind

No Wind

Figure 4-7. Effect of wind on the minimum dilution factor for both houses.

13

1.E+06

1.E+05

Effect of Vent Terminal Design with Two Additional Cases

Bldg Sep 5 ft - Amb T 0 F

House 1 (vent)
House 2

log(Dmin)

1.E+04

1.E+03

1.E+02

1.E+01

1.E+00
Radial Vent

2" Straight

45 Down

3" Straight

Figure 4-8. Effect of vent terminal design on the minimum dilution factor for both houses.

5.0 Recommendations
This work takes the first steps toward understanding the physics and identifying the important
factors in sidewall venting of gas appliances between buildings. It provides an initial foundation
for building a sound technical basis for establishing codes and guidelines. Fluid dynamics
modeling by itself is rarely sufficient for answering important real-world questions, particularly
when those questions relate to safety, potential damage to property, or the success or failure of
commercial products. To move forward, it will be important to establish the validity of these
results, and to extend the modeling analyses to more real-world conditions.
There are a number of ways that one might consider to accomplish the necessary validation,
extend the modeling, and move toward a more detailed, comprehensive and technically sound
justification for sidewall venting guidelines. It is possible to conduct experimental studies using
a controlled environment, an appliance or set of appliances, and by constructing walls and/or
inexpensive buildings. It would be possible then to visualize the plumes under various
conditions and measure plume dispersion using tracer gas. CFD models of the experiments
could be conducted in parallel to validate the models and refine the modeling procedures. An
experimental study such as this, while valuable, could be costly.
Another approach would be for the industry participants to note field cases where sidewall
venting between houses resulted in problems such as ice buildup or infiltration of vent gases. If
the industry participants were to note the conditions in the problem cases, such as appliance
details, vent design, house separation, outdoor temperature, etc., then Battelle could use that
14

information to construct and validate fluid-dynamics models. It would probably require a

heating season to accumulate problem cases of interest, after which a set of them could be
selected for analysis by Battelle. After validation, the modeling could then be used with
confidence to answer questions related to conditions for which no direct data are available.
While there would be lag time in generating the needed results, the cost of this approach is likely
to be far lower than a detailed experimental study.
The results of this limited study do show that vent design is important for safely venting flue gas
between buildings. This study showed that the radial vent design could possibly cause problems
for the house with the vent, while straight vent designs could cause problems for a neighboring
house that is less than ten feet away. It seems reasonable that there are alternative vent-terminal
designs between these limits that could prevent problems on both houses. There is an
opportunity for industry to meet this need by designing new vent terminals that will disperse and
dilute sidewall-vented flue gas between buildings.

6.0 References
1. Rutz, A. L. ; Paul, D. D. ; DeWerth, D. W. ; Borgeson, R. A.; GRI's Venting Research
Program: Activities at Battelle and A.G.A. Laboratories (1988-1994), Final Report No.:
GRI-94/0371, Gas Research Institute, October 1994.
2. Barrett, R. E.; Venting Gas-Fired Appliances in Multistory Buildings: A Review of the State
of the Art, Gas Research Institute Report No. GRI-92/0529, June 1992.
3. Reuther, J.; Leslie, N.; Hemphill, R.; Justification for Sidewall Vent -Terminal Locations
for Gas Appliances, 50th Proceeding of the International Appliance Technology Conference,
P: 231-242, IATC, 1999.
4. Reuther, J. J.; Hemphill, R.; Vent - terminal locations as related to air infiltration and indoor
air quality, American Society of Heating, Refrigerating and Air-Conditioning, ASHRAE
Transactions, v 108, n 1, p 563-571, 2002.
5. Paul, D.; Ricci, S.; Talbert, S.; and Reuther, J.; Minimum Clearances for Sidewall Venting
of Gas-Fired Appliances, Gas Appliance Manufacturers Association, Final Report prepared by
Battelle, April 2007.

15

Appendix A
Presentation of CFD Results Delivered to the Technical Review Committee

16

NFPA Sidewall Venting Study

CFD Results
September 6, 2007

BUSINESS SENSITIVE

Overview of CFD Modeling

Appliance is 100 kBtu/hr, 92% efficiency, 40% excess air. Flue gas
temperature is 115 F and flow rate is 23 acfm.

Vent is 2-inches ID, 1 ft from ground, and protrudes 1 ft from house.

Flue gas is an incompressible ideal gas with air properties. Flue gas
concentration in the ambient air is tracked on a normalized scale (mole
fraction = 1 at vent outlet).

Ambient temperature is either 0 F or 75 F.

House separation is either 5 ft or 10 ft
Calculations are steady-state.
In the pages that follow, House 1 is the house with the vent and House 2
is the neighboring house.

Maximum concentration (minimum dilution) is computed on the wall of

House 2 facing the vent on House 1.

Maximum concentration is also calculated on the wall with the vent on

House 1, at a distance of at least 4 ft from the vent.
BUSINESS SENSITIVE

17

13 Cases Run
Vent Type

House
Separation (ft)

Radial
Radial
Radial
Radial
Straight
Straight
Straight
Straight
Straight
Straight
Straight
45 Down
3 Straight

5
5
10
10
5
5
10
10
5
5
5
5
5

Outdoor
Temp. (F)
0
75
0
75
0
75
0
75
75
75
75
0
0
BUSINESS SENSITIVE

18

Wind Direction Wind Speed

(mph)
NA
NA
NA
NA
NA
NA
NA
NA
East
North
South
NA
NA

0
0
0
0
0
0
0
0
7
7
7
0
0
3

19

20

21

22

23

24

25

26

27

28

Conclusions
General effect of vent terminal design
Radial vent design protects House 2 and keeps the plume near House 1.
Straight vent design protects House 1 and transports the plume to House 2.

Worst case dilution factors in this study were on the order of 10.
Any wind increases dilution and dispersion of the plume. The effect of
wind normal to the two houses in both directions is similar. Wind parallel
to the houses sweeps the flue gas away.
The influence of wind at a building separation of 10 ft was not modeled in
this study.
There is an opportunity to design a new vent terminal to achieve sufficient
plume dispersion, perhaps even at building separations of 5 feet.
Experimental studies to visualize plumes, measure dilution, and validate
models could be useful for engineering practical strategies for sidewall
venting between buildings.

BUSINESS SENSITIVE

29

25