The Model for Smoke Control

CSE: What new technologies or trends are currently affecting the design of smoke-control and fire-detection systems? NADGIR: We have seen a tendency among end users to move away from the use of ionization detectors. However, this trend seems more pronounced in the detection and alarm business than in the fire-suppression business.

By Barbara Horwitz-Bennett, Contributing Editor April 1, 2003

CSE: What new technologies or trends are currently affecting the design of smoke-control and fire-detection systems?

NADGIR: We have seen a tendency among end users to move away from the use of ionization detectors. However, this trend seems more pronounced in the detection and alarm business than in the fire-suppression business.

In Europe and other parts of the world, carbon monoxide detection is also increasing in importance. Testing has proven that the burning of synthetic polymers, such as carpets and mattresses, generates colorless carbon monoxide earlier than the smoke.

BOWMAN: Computational fluid dynamics (CFD) modeling is becoming more commonplace in the design of smoke-control systems. In the past, many of the techniques used for determining the design criteria for smoke-control systems focused on either meeting certain air change rates, and more recently, preventing the smoke layer from descending to a certain height—often 6 to 10 ft. above the highest walking surface. These techniques reflected the analytical tools and calculations available at the time.

For example, commonly used zone-type fire models assume uniform smoke layer characteristics and a uniform smoke layer height. However, with the availability of powerful CFD modeling for desktop computers, it is possible to evaluate a smoke-control system on whether it can achieve the actual goal of maintaining tenability of the space through full evacuation, eliminating assumptions of uniformity, which is particularly useful in buildings with unique interior geometry.

CSE: Is increased use of CFD modeling a result of the influence of performance- based design in this area?

O’CONNER: Yes. Approaches and analysis methods using performance-based design (PBD) will likely have more impact on the design of fire detection and smoke-management systems. Rather than relying on the prescribed code requirements, PBD approaches will engender the consideration of what is to be achieved by these systems. This will take the form of goals and objectives relevant to a given building and occupancy and agreed to by the stakeholders such as the building owner, code authorities, architect, engineer and insurance interests. The PBD process allows for considerations of not only hazard and relative risks of credible fire scenarios, but factors related to other building design elements, design costs and long-term maintenance.

With PBD, we will see more emphasis on the intended performance of smoke- management and detection systems. Thus, fire and smoke movement models will play a key role in evaluating and demonstrating that a proposed smoke-management or fire-detection system works to fulfill the goals or objectives established by the stakeholders.

CSE: Speaking of code-related matters, how do most local or state codes stack up when it comes to smoke-control and fire-detection specifications?

BOWMAN: In my opinion, not well. For example, take the calculation methods in the Uniform Building Code—and subsequently, the International Building Code. The danger is that these prescribed methods can simplify the process of developing design criteria for smoke-control systems to the point where all credible scenarios may not be evaluated. In addition, the codes are written in a manner where the goal of the design is to keep the smoke layer at a certain height above the walking surface. This can cause issues in communication with the authority having jurisdiction (AHJ) when the design is based upon CFD modeling and not on the provided hand calculations or more simplistic zone-type fire models.

O’CONNER: Some architects and engineers believe that a few codes have gone too far in that they have taken a “more is better” approach. Often, there is a focus on the design of fancy and complex smoke-management and fire-detection systems without due consideration of the occupancy risks, building construction features, fire-suppression systems and occupant capabilities. With the greater use of PBD approaches, authorities and designers will come to a better understanding of the need for and purpose served by smoke-management and fire-detection/alarm systems.

Engineers and authorities should review the Society of Fire Protection Engineers SFPE Engineering Guide to Performance-Based Fire Protection for assistance in applying a performance-based approach. An additional resource will be a forthcoming guide oriented to code officials and developed by SFPE and the International Code Council (ICC).

CSE: Of course, a big piece of the smoke management puzzle is dealing with tricky specs such as for atriums. When designing for large open spaces, a common mistake is underestimating design-fire sizes by failing to take into account transient fuels. What other kinds of mistakes should engineers beware of when designing smoke- and fire-detection systems?

NADGIR: Many times, only ceiling spot-type smoke detectors are used in areas with high ceilings. This is generally not effective smoke detection, as the smoke will not always rise to the ceiling. In such a case, multi-level detection needs to be employed and oftentimes, air sampling detection or beam detection is useful.

BOWMAN: Primarily, the key mistake is to oversimplify the process. This can lead to leaving out critical design issues such as a full evaluation of the range of fire growth rates, toxicity of products of combustion, sizes and locations. It can also prevent a designer from fully evaluating critical system components such as detector location, as well as fan and make-up air locations. In addition, oversimplification can cause oversights in the development of simple design measures that would have a significant impact on the cost or ease of implementation of the system.

Unfortunately, while the codification of the smoke-control system design process provides an easier basis for review for the AHJ, important unique design elements can be overlooked or bypassed. Additionally, the codification invites designs by those who may not fully understand the complexity of a smoke-control system design from the standpoint of fire growth and behavior. These are obviously critical factors.

CSE: Could you be more specific?

BOWMAN: For example, the International Building Code permits the designer to use a 5,000-BTU design fire unless a “rational analysis” is provided. However, this analysis is not explicitly required, so a designer could rely on the 5,000-BTU criteria with little or no thought as to the fire growth rate or the toxicity of potential fuel loads. While the design process often delves into these issues regardless of how code requirements are written or interpreted, the biggest concern is where an oversimplified design is being reviewed by an AHJ with little or no experience with smoke-control systems.

O’CONNER: A most important step in smoke-management system design is planning for the acceptance testing. Design firms must be careful to consider and understand the acceptance criteria used by AHJs. Working through the sequence of operation of the smoke-control system in conjunction with other building systems and building features is important to assuring that various operational scenarios are accounted for and not overlooked. Simply looking at the initiation of smoke-control systems by associated smoke detectors and the operation of associated fans may not be adequate to determine if the smoke-control system will perform as intended. The concurrent operation of other detectors, fans or equipment in a building can play a role and may need to be considered.

Another smoke-management design issue that engineers need to address is provision of make-up air. More often than not, fans and air-handling units can be arranged to provide the exhaust needed, but the make-up air can be difficult to implement. Architects often place limits on the engineer regarding the location and size of inlet area for make-up air in a building. These limitations should be addressed early in the design to assure that adequate make-up air can be provided. Additionally, the designer needs to consider the effect of make-up air on the smoke layers. Make-up air introduced into the smoke layer will result in an expansion—and dilution—of the smoke layer and may require additional exhaust capacity to be provided by exhaust fans to limit the descent of the smoke layer.

CSE: Can you offer some advice as to how fire-protection engineers can best work with other building team members—particularly the architect and AHJs—to ensure that things go smoothly with these kinds of projects?

O’CONNER: Fire protection engineers are often called into projects after smoke-control acceptance tests fail, instead of in the early design stages when issues involving the incorporation of a smoke-control system can be identified. Also, many firms follow the prescriptive criteria included in building codes, which does not consider all aspects of smoke-control system design. Our recommendation is to involve a qualified fire-protection engineer early in the design process to analyze and coordinate the systems and building features that affect smoke development and movement. Additionally, the SFPE guide is a great document that will help one’s understanding of PBD process and considerations for a smooth project and issues related to smoke-management and fire-detection/alarm systems.

BOWMAN: As with every atypical design process, communication is the key. I would recommend using the structure in the SFPE guide to define the appropriate stakeholders, as well as a process that everyone could be comfortable with. Unfortunately, using this tool can have an impact on the design fees necessary to conduct the proper design. It then becomes a cost issue that may not be supported by the project. This leads toward oversimplification and reliance on the codified approach to smoke control design.

NADGIR: Fire-protection engineers should educate architects and AHJs as to new technologies and products. They should help ensure that the initial design of the building takes into account the different fire hazards present and suggest the appropriate fire protection solutions.

Oftentimes, the “released for installation” drawings are different from the “initial bid package” drawings. For suppression systems, this is a serious issue since the system design and agent concentration is based completely on the room volume. Changes to room dimensions should be communicated to all concerned at the earliest possible time.


Andrew Bowman, P.E., Principal, Gage-Babcock and Associates, Oak Brook, Ill.

Daniel O’Conner, Vice President of Engineering, Schirmer Engineering Corp., Chicago

Abhay Nadgir, Product Manager, Kidde-Fenwal, Ashland, Mass.

Code Watch: Telecom

Given our society’s reliance on telecommunications and the sharing and routing of information, we may see trends in National Fire Protection Association documents to address issues of protecting computer and telecommunications equipment for the purpose of assuring continuity of service. The concern to be addressed will be the non-thermal hazard of smoke due to relatively small fires in computer facilities and telecommunications environments.

This will be a challenge as the general need to control the temperature and humidity in a computer or telecommunications environment is currently a priority that can be upset by the injudicious use of smoke-management systems that respond too early or are not coordinated with an appropriate fire-detection scheme.