To view the figures and tables associated with this article, please refer to the flipbook above.

Many practicing engineers (the authors included) have attempted to determine wind pressures on roof mounted screen walls and have wondered which provisions of the American Society of Civil Engineer’s Minimum Design Loads and Associated Criteria for Buildings and Other Structures (ASCE 7-22) should be used and whether they estimate wind pressures accurately. Wind pressures on these elements can be seemingly difficult to determine with any degree of accuracy and several code provisions within ASCE 7 may appear to provide conflicting information.

Screen walls are walls that project vertically from a roof structure, whose purpose is often to conceal mechanical, electrical, or plumbing (MEP) equipment from view. They can be in any configuration or geometry, and they do not have a roof over them. They generally are used for aesthetic reasons and are not a part of the primary building structure. The requirements for screen walls are almost always driven by architectural design. Parapets may also serve as screen walls, or screen walls may be parapets. The International Building Code (IBC) defines a parapet as “the part of any wall entirely above the roof line;” while in the context of wind engineering we often assume it is a continuation of the exterior wall above the roofline. For this article, rooftop units (RTUs) are defined as mechanical or building HVAC equipment supported by a building roof structure.

Figure 1 shows the mechanics of wind flow over a building with a flat roof without screen walls. When the wind flows up and over the building, it separates from the roof surface near the windward edge and creates separation vortices. Wind pressures in this zone are generally characterized with high uplift (suction). ASCE 7 provides a method to calculate the width of this zone as a function of the building height and width, called the “Width of pressure coefficient zone” represented by the symbol ‘a’. After a distance of ‘a’, this turbulent flow re-attaches to the flat roof and reloads a vertical surface with a windward pressure. Figure 2 shows the mechanics of wind flow over a flat roof with screen walls/parapets at both edges of the building. The windward screen wall/parapet is designed for the windward wall pressures on the front face and roof uplift pressures on the interior face. The leeward wall is designed for leeward wall pressures on the exterior face and windward wall pressures on the interior face. Figure 3 shows the local effects of wind flow on screen walls with wind flowing above and below the wall. Design wind pressures on screen walls will vary greatly, depending on the wind direction and their location on the roof. How practicing engineers can account for this variability is addressed later in this article.

The ASCE 7 standard is a valuable tool for estimating wind pressures on a variety of building and non-building structures. However, the standard does not specify which provisions should be used to calculate wind pressures on screen walls. Additional confusion may arise when deciding whether main wind force resisting system (MWFRS) or components and cladding (C&C) pressures should be used, and how they are calculated.

C&C wind pressures should be used for the design of all screen wall elements, roof supports, and connections. MWFRS pressures should be used for calculating wind story forces, base shears, and overturning moments used for the design of the lateral force-resisting system and foundations. Estimating design forces for rooftop equipment is often challenging during the early design phases because the locations, sizes, and layout of these elements may not be finalized yet. In this case, the engineer must apply judgement to make conservative estimates of locations and heights. Neglecting wind effects on these elements early in design can cause unconservative designs, leading to re-work or conflicts with architectural or MEP elements later in the design process.

The authors have heard of engineers using a variety of ASCE 7 methods to calculate MWFRS and C&C pressures applied to screen walls. These include the following, but in the authors’ opinion, not all of these are correct:

In order to demonstrate the wide range of external pressure coefficients (Cp/GCp) that can be obtained using the methods above, the maximum and minimum coefficients for each method are listed in Table 1. Possible coefficients range from 0.6 to 5.5. Parapet coefficients tend to be the highest, and MWFRS wall coefficients tend to be the lowest. The Chapter 29 rooftop structures provisions provide a fairly tight and central range of coefficients between 1.0 and 1.9.

ASCE 7 requires using the rooftop structures provisions of Chapter 29 to calculate wind pressures on roof mounted mechanical equipment for both MWFRS and C&C loads. However, the ASCE 7 standard does not have specific language addressing screen walls. While the commentary Section C29.4.1 offers valuable recommendations, it is not enforceable code, and building officials or peer reviewers may reasonably question the design approach. However, this is the best currently available approximation in ASCE 7, and this is the method that should be used if a wind tunnel procedure is not performed. This approach can be conservative for buildings that have screen walls on lower roofs of buildings because it is based on the mean roof height of the overall building and not the height of the screen wall itself.

The Chapter 29 rooftop structures procedure does not consider the location of the rooftop structure on the roof but instead provides a single pressure regardless of location which can be unconservative for screen walls located near the edge of a roof. The ASCE 7 commentary to Section 29.4.1 states that screen walls located close to a building edge should be designed for parapet pressures. However, it does not provide a method for determining the proximity to the roof edge where roof top equipment pressures stop and parapet pressures begin. The authors believe, based on our understanding of the research that is the basis for ASCE 7, that it is reasonable to design screen walls located in edge or corner roof zones for parapet pressures and screen walls located in interior roof zones for Chapter 29 rooftop structures and equipment pressures. When calculating MWFRS pressures on screen walls that are in edge or corner roof zones, the wind pressures should be calculated per both Chapter 27 parapet provisions and Chapter 29 rooftop structures provisions and the largest of the two calculated pressures should be used for design.

In order to better understand the variability of the pressures that can be obtained from the different methods in ASCE 7 and how they compare with wind tunnel testing, three real world examples of mid-rise buildings with roof mounted screen walls are explored here. A reputable wind engineering consultant performed the wind tunnel testing for each example. See Table 2 for external pressure coefficients and corresponding design wind pressures for each screen wall as determined by different methods from ASCE 7 and wind tunnel testing. The recommended procedures and coefficients are shown in the table notes. Some of the procedures in Table 2 are not recommended but values are provided for the sake of comparison. This data shows the significant variability of the pressures obtained using each different procedure. The ASCE 7 recommended C&C procedure produced higher wind pressures when compared with wind tunnel testing in every case except for at the Building 3 interior condition. Building 3 happened to be the only building that was classified as Exposure Category B. If it had been classified as Exposure Category C, then the wind pressures would have been in better agreement. This highlights the critical role of surface roughness and exposure category in the calculation of accurate wind pressures. The C&C parapet pressures from Chapter 30 yielded high pressures in all cases when compared with wind tunnel testing. The remaining methods yielded varying results that were sometimes higher and sometimes lower than the wind tunnel testing. This information represents only a small sample of buildings and is not intended to be comprehensive. Any conclusions drawn from this information should be corroborated with the code and the literature.

ASCE 7-22 does not currently provide specific code provisions for calculating wind pressures on screen walls. Wind tunnel testing or CFD performed by a wind engineering consultant will provide the most accurate and least conservative wind pressures for design. If those methods are not possible, the best available procedure for screen walls that are not located close to a roof edge is the Chapter 29 provisions for rooftop structures for MWFRS and C&C. The best available procedure for screen walls that are located close to a roof edge is the Chapter 30 parapet procedure for C&C. For MWFRS pressures on screen walls close to roof edges we recommend using either the Chapter 29 provisions for rooftop structures or the Chapter 27 parapet provisions for MWFRS, whichever results in the largest pressures.

Additional research and wind tunnel testing is needed to develop a wholistic method for calculating wind pressures on common configurations. The following is a list of enhancements to the existing method that should be considered.

Research can take a long time to be funded, completed, and incorporated into ASCE 7. In the meantime, practicing engineers can use the suggestions in this article as a guide to make educated decisions and use proper engineering judgement when designing buildings with screen walls. ■

Can shielding be considered when designing screen walls or RTUs in close proximity?

It can seem intuitive that if a RTU is located behind a screen wall, penthouse, or another large RTU that wind loads might be lower. However, ASCE 7 does not currently allow any reduction in wind pressure due to shielding by other structures or by topographic features because of the “lack of reliable analytical procedures for predicting the effects of shielding.” However, this does not preclude the determination of shielding effects and the corresponding reductions in velocity pressure by means of wind tunnel testing.

Research has shown that in certain conditions, an object upwind of another structure can create an effect called dynamic interference, which can cause increased wind pressures on the downwind object. In reality, structures have failed when a new structure was placed upwind. These are rare occurrences but should cause some pause when an engineer automatically assumes an upwind element shields a downwind element.

The most recent relevant research for wind loads on mechanical units shielded by screen walls (Morrison and Miller, 2017) was based on wind tunnel testing of 4-foot tall screen walls. The following is a short list of relevant findings from this report:

If a very large obstruction is in front of a mechanical unit, in theory, uplift loads on the unit will be increased and horizontal loads will be decreased. The roof wind uplift coefficients shown in Figure A indicate that air flow reattachment occurs at about 2h downwind from the edge of a roof, where ‘h’ is the mean roof height of the building. Therefore, it may be reasonable to assume that if the height of obstruction = ‘h’, then the wind is turbulent (i.e., not able to reload a vertical surface) for 2h down wind. Some theory supports the reduction of MWFRS for shielded RTUs downwind, but the research does not support any reduction in C&C wind loads.

Can wind loading on screen walls be reduced for louvered or perforated panels?

Wind pressures on perforated screen walls depend upon the screen wall geometry, the size/spacing of the perforations, and the wind speed. ASCE 7 commentary C29.4.1 states that in the absence of appropriate wind tunnel testing, the full wind load should be applied to porous screen walls. Wind flow takes the path of least resistance. At low wind speeds, wind may be able to easily pass through louvers and perforations. However, at design wind speeds the wind flow through the panel becomes constricted and the wind is more likely to flow around the full width/height of the wall, resulting in pressures similar to a solid panel. Recent research (Morrison and Miller, 2017) concluded that screen wall type/porosity did not significantly change wind loads.

Do increased wind pressures on existing building elements need to be considered when adding a roof top structure to an existing building?

Extending parapets or adding roof top screen walls and equipment to an existing building can create a wind sail effect that increases MWFRS and C&C loads. The increase in lateral load and overturning needs to be checked in accordance with the International Existing Building Code (IEBC). It can be beneficial to perform a wind tunnel test for existing building renovation projects that include new rooftop structures in order to possibly reduce wind pressures on these new elements and account for any shielding that may be present as mentioned previously. This may help minimize costly strengthening of existing building elements and the existing lateral system.

What Risk Category should be used for determining the basic wind speed for screen walls and RTUs?

The Risk Category used for the design of screen walls and RTUs should not be less than the Risk Category of the building that they are connected to. This is intuitive for RTUs because they are required for the building to function properly. Likewise, a screen wall blowing off of a building’s roof could damage other parts of the building or cause an obstruction that restricts the intended function of the building. Additionally, if the equipment provides necessary service to another building, the Risk Category should also consider the building being served. In some cases, the equipment may be designed for a higher risk category than the building supporting it. Additional guidance can be found in the ASCE 7 Chapter 29 commentary.

About the Authors

Todd Clapp, PE, SE, is a Senior Project Engineer at Martin/Martin Consulting Engineers and a member of the NCSEA Wind Committee.

John O’Brien, PE, SE, is a Senior Associate and Director of Structural Engineering at PES Structural Engineers. He serves as Chair of the NCSEA Wind Committee and Chair of the ASCE 7 Chapter 29 Task Committee for Wind Loads on Building Appurtenances and Other Structures.

References

Morrison, M. J., C. Miller. 2017. “Effects of shielding on the wind loads on roof-mounted equipment. ASHRAE Research Project Report 1692-RP. Peachtree Corners, GA: ASHRAE