Professional Documents
Culture Documents
Year7 Section3 RoofTopEquip RCMPY7
Year7 Section3 RoofTopEquip RCMPY7
Year7 Section3 RoofTopEquip RCMPY7
Prepared by:
In Partnership with:
The International Hurricane Research Center
Florida International University
August 2007
TABLE OF CONTENTS
CHAPTER PAGE
EXECUTIVE SUMMARY…………………………………………………………………………………. 3
7. FUTURE WORK................................................................................................................................ 27
REFERENCES ............................................................................................................................................. 28
2
ROOFTOP EQUIPMENT WIND LOAD AND ITS MITIGATION FOR
BUILDINGS IN HURRICANE PRONE REGIONS
Executive Summary
exhaust hoods, HVAC units, and communications equipment that are typically mounted on the
roof of structures. This type of equipment is subjected to wind loads that must be considered
when designing the anchorage connection between the particular component and the roof.
Damage reconnaissance studies conducted during the 2004 and 2005 hurricane seasons
witnessed widespread rooftop equipment failures. The aim of this research is to develop
mitigation techniques that will reduce the wind loading on mechanical rooftop equipment and the
forces transferred to the roof supporting structure by the equipment through the use of
1. Problem Statement
Rooftop equipment is subjected to high wind loads during extreme wind events such as
hurricanes and a structural system is needed to resist and transfer the loads. The equipment,
structural system, and anchorage connections to the roof members need to be carefully designed
to prevent failure during sever storms. Wind-induced failure of rooftop equipment during a
hurricane may result in large openings in the roof that will allow water to penetrate into the
building, puncturing of the roof membrane, again allowing water infiltration, and detached
rooftop equipment can pose considerable threats as windborne debris [Reinhold, 2006].
Secondary effects associated with rooftop equipment failure may lead to extended delays in
restoring occupancy and function of the building because significant drying out time may be
3
needed or difficulties may arise from exposed electrical and plumbing connections resulting from
such failures [Reinhold, 2006]. With significant roofing damage and secondary water damage
occurring related to the poor performance of rooftop equipment during extreme wind events,
high maintenance and costly repair works (Fig. 2) are often needed.
Figure 2. Repair of Rooftop AC Units and Structural Framing System after Damages from
Hurricane Wilma in 2005
Damage reconnaissance studies conducted during the 2004 and 2005 hurricane seasons
witnessed widespread rooftop equipment failures (Fig. 3). In the aftermath of the 2004
4
hurricanes Charley, Frances, Ivan, and Jeanne, the Federal Emergency Management Agency
building performance and damage caused by the storms. Rooftop mechanical and electrical
equipment was listed by the MATs as one of the key modes of observed building failure, causing
millions of dollars worth of damage and even impeding recovery efforts by impacting critical
and essential facilities [FEMA 490, 2005]. Similar results were found following Hurricane
Katrina (2005); the FEMA MAT identified wind impacts to rooftop equipment as one area
requiring “additional attention” from designers, architects, and contractors [FEMA 549, 2006].
Following hurricanes Katrina and Rita (2005), the National Institute for Standards and
Technology (NIST) also recognized the impact of rooftop equipment failure on major buildings
and windborne debris damage caused by rooftop equipment detachment in all locations of
Image Source: FEMA 489 “Mitigation Image Source: FEMA 549 “Mitigation
Assessment Team Report: Hurricane Ivan in Assessment Team Report: Hurricane
Alabama and Florida” August 2005. Chapter Katrina in the Gulf Coast.” July 2006.
5, Pg 5-61. Chapter 5, Pg 5-80.
5
Image Source: FEMA 489 “Mitigation Image Source: FEMA 488 “Mitigation
Assessment Team Report: Hurricane Ivan Assessment Team Report: Hurricane
in Alabama and Florida” August 2005. Charley in Florida” April 2005. Chapter 5,
Chapter 5, Pg 5-58. Pg 5-64.
Image Source: FEMA 489 “Mitigation Image Source: FEMA 489 “Mitigation
Assessment Team Report: Hurricane Ivan Assessment Team Report: Hurricane Ivan in
in Alabama and Florida” August 2005. Alabama and Florida” August 2005. Chapter
Chapter 5, Pg 5-60. 5, Pg 5-59.
2. Research Background
There is very limited research available to provide designers with guidance about the
wind-induced forces exerted on rooftop equipment. ASCE 7 Standard is the source adopted by
all the latest national model codes and standards for wind loading information [Reinhold, 2006].
6
In the 2002 and 2005 editions of ASCE 7, the force applied to rooftop equipment is computed
F = q z GC f A f (lb) (1)
where qz is the velocity pressure evaluated at height z of the centroid of area Af using the
appropriate exposure category, G is the gust-effect factor, Cf is the force coefficient, and Af is the
projected area normal to the wind except where Cf is specified for the actual surface area [ASCE
ASCE 7-02 was the first edition of the reference standard that sought to address the
proper design loads for wind-sensitive rooftop systems. The provisions of ASCE 7-02 require
the use of Figure 6-19 to determine the value for the force coefficient, Cf, to be used in Eq. 1.
Based on the ASCE 7-02 commentary, the ASCE 7 committee was vague about providing
guidance for dealing with the increased loads exerted on the equipment because there is no basis
to make a recommendation [ASCE 7-02]. Additionally, uplift forces on the rooftop equipment
ASCE 7-05 commentary mentions that because of the small size of the rooftop equipment
in comparison to the building, it is expected that the wind force will be higher than that predicted
by ASCE 7-02 due to higher correlation of pressures across the structure surface, higher
turbulence on the building roof, and accelerated wind speed on the roof. ASCE 7-05 commentary
also mentions that research [Hosoya et al., 2001] has shown high uplifts on the top of the rooftop
air conditioning units, although the net uplift on the units was not measured. The consensus of
the committee is that uplift forces may be a significant fraction of the horizontal force. Hence
7
ASCE 7-05 uses Figure 6-21, which is same as Figure 6-19 in ASCE 7-02, but specific
changes were added by providing a section that dealt with wind loads on structures and
equipment for low-rise buildings, that is, buildings less than or equal to 60 ft [ASCE 7-05,
Reinhold, 2006]. These changes were based on the results of a wind tunnel study conducted by
Hosoya et al. [2001], which modeled a 4x4x4 ft air conditioning unit mounted at three different
locations on the Texas Tech University (TTU) field site building. From the ASCE 7-05
commentary, the revised methodology requires a new factor ranging from 1.0 to 1.9, depending
on the size of the equipment, to be multiplied with the force determined using Eq. 1. Thus the
force should be increased by a factor of 1.9 for units with area Af less than 0.1Bh (10% of the
building area). Because the multiplier is expected to approach 1.0 as Af approaches that of the
building (Bh), a linear interpolation is included as a way to avoid a step function in load if the
designer wants to treat other sizes. The research only treated one value of Af (0.4Bh).
In nearly all cases, the area of the equipment will be less than 10% of Bh [Reinhold,
2006], meaning that a force factor of 1.9 will almost always be used. The implication is that
forces calculated for rooftop equipment on low-rise structures will nearly double when using the
ASCE 7-05 methodology versus the ASCE 7-02 methodology. Due to the limited research and a
significant level of uncertainty, experts in the wind engineering field recommend further
increasing the forces determined using ASCE 7-05 with a factor of safety of 2 for regular
structures and a factor of safety of 3 for critical and essential facilities [Barista, 2007; Reinhold,
2006].
3. Research Objectives
As the latest edition of the ASCE 7 standard is adopted by the national building codes,
the load increase from ASCE 7-02 to ASCE 7-05 translates to significantly higher costs in the
8
design and construction of framing and anchorage for rooftop equipment for new buildings.
Furthermore, existing buildings requiring new rooftop equipment may need retrofits to increase
the strength of existing roof members to withstand higher wind loads transferred by the
equipment and the structural framing. Better and cost-effective alternatives are to reduce the
wind loading on the equipment itself by aerodynamic retrofitting or to dampen wind loading
transferred to the roof structure by structural retrofitting. Potential retrofits may include, but are
not limited to, aerodynamic edge shapes, wind screens, and elastomeric dampers.
The primary focus of this research is to develop a baseline for cost-effective techniques to
mitigate wind loading on rooftop equipment, lowering the cost of framing and anchorage,
reducing the chances of rooftop equipment failure, and saving potential losses associated with
roof damage and water infiltration. The research objectives are stated as:
To achieve these goals, the scope of work for this project consists of the following tasks:
with the current Wall of Wind windstorm simulator setup in order to establish the baseline loads
9
The broader impact of this proposed research work is to make the structural community
aware of the impact from the higher wind loads and also suggest a path for development of
efficient techniques to mitigate these loads cost effectively. With improved load determination,
more awareness during design, and proper mitigation techniques, damage related to rooftop
Reduced scale model testing on rooftop equipments is not feasible due to the small size
of the units. Thus wind tunnel testing would not be appropriate for this purpose. To overcome
wind tunnel constraints, the current study aims to test rooftop equipment subjected to design-
level hurricane force winds at full scale, using the Wall of Wind windstorm simulator (Fig. 4a)
operated at Florida International University (FIU). For this testing, typical air conditioning
condenser units are mounted on the roof of a 10x10x10 ft building model (Fig. 4b). Consistent
with general practice the rooftop units are mounted to a steel frame and the steel frame is
anchored to the roof supporting members of the test building by six legs. Both the building and
the rooftop equipment are fully engulfed by the wind flow generated by the Wall of Wind.
Based on the findings of Hosoya et al. [2001] and the failure observations reported by FEMA
[2005, 2006], NIST [2006], and others, the wind-induced effects of interest for the rooftop
equipment are the shear force, the axial force, and the overturning moment.
Instrumentation of the rooftop equipment consists of six Omega LC402 force transducers
(Fig. 5) to measure the axial forces, and six Omega LC101 force transducers (Fig. 5) to measure
the shear forces; from the equipment’s geometry and the measured forces, the overturning
moment can be determined. The Wall of Wind data acquisition system records loading time
10
Figure 4. (a) 6-fan Wall of Wind at FIU, (b) One-story building model
WIND LC 402
2 1
4 3
6 5
LC 101
11
The steel support frame is mounted at two different locations on the building structure: at
the windward edge of the building, and with the centerline of the equipment at a distance of 5’-
0” away from the roof edge. The rooftop equipment are mounted perpendicular to the wind flow.
Three AC units are placed on the steel supporting frame and the tests comprised of placing the
frame with the AC units at two different locations (#1 and #2) as illustrated in Fig. 6.
5’–0”
1 1’
Building
Structure
Wall of Wind
Flow Field 2 2’
3 3’
Frame Frame
Centerline Centerline
Location #1 Location #2
12
A control test without any retrofit technique was performed for each setup to establish the
baseline forces exerted on the rooftop equipment by running the Wall of Wind hurricane
simulator at 4400 rpm and 4000 rpm for the top and bottom engines respectively. Once the
baseline forces were known, the effectiveness of mitigation techniques was sought. For this
project the mitigation technique selected was the provision of a perimeter wind screen around the
AC units to alleviate the wind loading on the units and the structural frame. The screen was
fabricated from metal grating with a porosity of approximately 50%. The screen was installed
around the rooftop equipments and the test assembly was subjected to wind loading using the
Wall of Wind engine RPM profile and time duration identical to that of the baseline
measurement tests. Figure 7 shows test configurations with and without the wind screen.
Force time histories for shear and axial loading on the bottom of each leg of the structural
frame were recorded for the wind screen installed condition and then compared to the baseline
measurements to determine the effectiveness of the retrofit technique for reducing the loads
exerted on the rooftop equipment and transferred to the roof supporting structure.
13
a b
c d
e f
Figure 7. Rooftop Equipment Set-up: (a) Middle of Roof without Screen, (b) Middle of Roof
with Screen, (c) Edge of Roof without Screen, (d) Edge of Roof with Screen, (e) Edge of Roof
without Screen (Side View), (f) Edge of Roof with Screen (Side View).
14
5. Results and Discussion
Wind loads on the rooftop equipment are transferred to the structural frame with pin-
connected legs in the form of shear and axial forces. With the given engine RPM profiles used
for this project the Wall of Wind fans generated a maximum wind speed of 120 mph which is
equivalent to a Category 3 hurricane and the forces generated at the bottom of each leg of the
frame were recorded with and without the wind screen retrofit technique. Figure 8 shows
selected loading time histories and Figs. 9 and 10 show the comparison of peak wind induced
loads for all the 6 legs (shown in Fig. 5) with and without screen for the two testing locations.
.
Middle of Roof
Leg 1
Axial Force Time History
0
0 20 40 60 80 100 120 140 160 180 200
-50
-100
-150
Force (lbs)
-200
Without Windscreen
-250 Windscreen
-300
-350
-400
-450
Time (sec) a
Figure 8. (a) Axial Compressive Loading Time History, (b) Axial Tensile Loading Time History,
(c) Shear Loading Time History (Cont.)
15
Middle of Roof
Leg 4
Axial Load Time History
400
350
300
250
Force (lbs)
200
150
Without Windscreen
Windscreen
100
50
0
0 20 40 60 80 100 120 140 160 180 200
-50
-100
Time (sec)
b
Middle of Roof
Leg 1
Wind-Induced Shear Force Time History
60
50
40
30
Force (lbs)
20
Without Windscreen
10 Windscreen
0
0 20 40 60 80 100 120 140 160 180 200
-10
-20
-30
Time (sec) c
Figure 8. (a) Axial Compressive Loading Time History, (b) Axial Tensile Loading Time History,
(c) Shear Loading Time History
16
Middle of Roof
Peak Axial Loads
600
400
200
Force (lbs)
Without Windscreen
0
1 2 3 4 5 6 Windscreen
-200
-400
-600
Leg
Middle of Roof
Peak Shear Loads
60
50
40
Force (lbs)
30 Without Windscreen
Windscreen
20
10
0
1 2 3 4 5 6
Leg
Figure 9. (a) Peak Axial Loading -- AC Units at Middle of the Roof, (b) Peak Shear Loading --
AC Units at Middle of the Roof
17
Edge of Roof
Peak Axial Loads
600
400
200
Force (lbs)
Without Windscreen
0
1 2 3 4 5 6 Windscreen
-200
-400
-600
Leg
Edge of Roof
Peak Shear Loads
80
70
60
50
Force (lbs)
Without Windscreen
40
Windscreen
30
20
10
0
1 2 3 4 5 6
Leg
Figure 10. (a) Peak Axial Loading -- AC Units at Edge of the Roof, (b) Peak Shear Loading --
AC Units at Edge of the Roof
18
A comparison between the peak loading with and without the wind screen is presented in
Table 1. Percentage reductions in peak loadings are also shown. Signification reductions have
been achieved with the wind screen which shows the effectiveness of the screen as a retrofit
method to reduce damage of rooftop equipment and secondary losses due to roof membrane
damage, water infiltration, debris generation (dislodged units) during hurricane events. Wall of
Wind full scale testing also showed that the wind screen and its connections didn’t have damage
Table 1:
Middle of Roof
Peak Axial Load (lbs) Peak Shear Load (lbs)
Percent Percent
Leg Without Without
Windscreen Reduction Windscreen Reduction
Windscreen Windscreen
1 -414.01 -204.21 50.7% 56.06 28.88 48.5%
2 233.48 82.79 64.5% 42.64 24.99 41.4%
3 -540.29 -332.43 38.5% 41.25 27.50 33.3%
4 370.38 112.51 69.6% 31.39 24.52 21.9%
5 -546.36 -265.53 51.4% 27.02 20.17 25.4%
6 323.36 73.84 77.2% 43.64 21.04 51.8%
Average: 58.6% Average: 37.0%
Edge of Roof
Peak Axial Load (lbs) Peak Shear Load (lbs)
Percent Percent
Leg Without Without
Windscreen Reduction Windscreen Reduction
Windscreen Windscreen
1 -360.09 -186.56 48.2% 39.79 30.08 24.4%
2 179.66 23.10 87.1% 32.23 27.33 15.2%
3 -565.78 -364.79 35.5% 74.30 44.83 39.7%
4 359.59 133.10 63.0% 39.98 31.15 22.1%
5 -466.12 -247.91 46.8% 34.91 28.06 19.6%
6 272.48 44.49 83.7% 45.79 25.15 45.1%
Average: 60.7% Average: 27.7%
6. Cost-Benefit Analysis
Cost and benefit analysis was performed to evaluate the economic effectiveness of the
proposed retrofit technique by installation of a wind screen to reduce the rooftop equipment
19
loading. Table 2 shows the cost of the AC units (cost is based on new AC units), structural
frame, labor and also the material and labor cost for the wind screen. The wind screen cost was
evaluated as approx. 20% of the cost of the rooftop equipments and the structural framing.
Table 2:
Labor
Installation of Frame and Equipment $1,500.00
Misc. $700.00
TOTAL $7,100.00
Windscreen
Materials
Steel and Expanded Metal $310.00
Labor
16 Hours @ $70 per hour for fabrication $1,120.00
Installation
TOTAL $1,430.00
For the benefit analysis, a numerical example has been presented to show the pressure
speed taken as 90 mph), Category 3 (sustained wind speed taken as 115 mph), Category 4
(sustained wind speed taken as 140 mph) hurricanes. Tables 3, 4, 5 show the conversions of 1-
min sustained wind speeds (as considered in Saffir Simpson Hurricane Scale) to 3-sec gust wind
speeds as considered by ASCE 7-02 and ASCE 7-05. The converted 3-sec gust wind speeds are
then used to calculate rooftop equipment pressure loading as per ASCE 7-05 as shown in Tables
6, 7, 8.
20
Table 3: Conversion of moderate Category 1 sustained wind speed to 3s gust wind speed
z0 (ft) α zg (ft)
21
Table 4: Conversion of moderate Category 3 sustained wind speed to 3s gust wind speed
z0 (ft) α zg (ft)
22
Table 5: Conversion of moderate Category 4 sustained wind speed to 3s gust wind speed
z0 (ft) α zg (ft)
23
Table 6: Calculation of Rooftop Equipment Load for moderate Category 1 hurricane
Chimneys, Tanks, Rooftop Equipment, & Similar Structures; Fig. 6-21 in ASCE 7-05
Height Kz q p Height Kz q p
ft "B" psf psf ft "C" psf psf
24
Table 7: Calculation of Rooftop Equipment Load for moderate Category 3 hurricane
Chimneys, Tanks, Rooftop Equipment, & Similar Structures; Fig. 6-21 in ASCE 7-05
Height Kz q p Height Kz q p
ft "B" psf psf ft "C" psf psf
25
Table 8: Calculation of Rooftop Equipment Load for moderate Category 4 hurricane
Chimneys, Tanks, Rooftop Equipment, & Similar Structures; Fig. 6-21 in ASCE 7-05
Height Kz q p Height Kz q p
ft "B" psf psf ft "C" psf psf
26
Based on the above results, the difference in wind loadings on rooftop equipment (on a
single-story building located in Exposure C open terrain) for a moderate Category 4 hurricane
and a moderate Category 1 hurricane is approximately 58%. The difference in wind loadings on
rooftop equipment for a moderate Category 3 hurricane and a moderate Category 1 hurricane is
approximately 38%. From the experimental results as presented in Table 1, the wind screen has a
potential of reducing the effect of wind loading by two to three levels of hurricanes (i.e., Cat 3 to
Cat 1 or Cat 4 to Cat 1 on Saffir Simpson Scale) and thus can prove extremely helpful to reduce
rooftop equipment damage. The wind screen retrofit technique may cost approx. 20% of the
rooftop equipments such retrofit technique will have other advantages such as (i) prevention of
roof damage, (ii) elimination of water infiltration thus preventing losses to building contents,
mold growth, dry wall saturation, (iii) prevention of windborne debris that may result from
7. Future Work
The future work on damage mitigation of rooftop equipment will include: (1) study of
uplift pressure on rooftop equipment by using Setra 265 pressure transducers installed on the
roof and the rooftop equipment to measure the pressure distribution, (2) use of different types of
wind screens (parametric studies: porosity, size, distance from AC units), (3) use of Elastomeric
damping devices to mitigate wind load effects and performing a cost-benefit analysis, (4) testing
under wind-driven rain injected into the flow field to reveal whether or not the presence of water
causes an increase in the loadings, (5) comparison of mean wind loads on rooftop equipment
obtained from Wall of Wind testing and design wind loads calculated using ASCE 7-05.
27
References
1. American Society of Civil Engineers, ASCE 7-02 (2002), Minimum Design Loads for
Buildings and Other Structures.
2. American Society of Civil Engineers, ASCE 7-05 (2005), Minimum Design Loads for
Buildings and Other Structures.
6. Hosoya, N., Cermak, J. E., Steele C. (2001), “A Wind Tunnel Study of a Cubic Rooftop AC
Unit on a Low Building,” Americas Conference on Wind Engineering, American
Association for Wind Engineering.
8. Reinhold, T. A. (2006), “Wind Loads and Anchorage Requirements for Rooftop Equipment,”
ASHRAE Journal, Vol.48, No. 3, p. 36-43.
28