Smoke control design considerations

Learn from this overview of NFPA 92: Standard for Smoke Control Systems, and how fire protection engineers should apply it in design.

By Erik Anderson, PE; Shaun Wrightson, PE; Nicholas Sealover, PE; Koffel Associates May 20, 2016

Learning Objectives:

  • Illustrate NFPA 92: Standard for Smoke Control Systems and its basic guidelines.
  • Compare the various smoke control terminology and design parameters.
  • Recall the various equations and calculations when designing smoke control systems. 

If your fire protection engineering firm has been tasked with performing a smoke control system design, there are several design decisions and considerations that need to be addressed along the way. The primary document that deals with smoke control systems is NFPA 92: Standard for Smoke Control Systems. The latest edition of this standard, 2015, was issued by the NFPA Standards Council on Nov. 11, 2014. While some jurisdictions may directly mandate compliance with NFPA 92 via local codes or amendments, many jurisdictions reference NFPA 92 indirectly by mandating compliance with the International Building Code (IBC) or NFPA 101: Life Safety Code.

Several jurisdictions, including the state of Maryland, already have adopted the latest edition of the IBC (2015 edition), which references the 2015 edition of NFPA 92 for the design of opposed airflow and smoke exhaust systems (see Sections 909.7 and 909.8, respectively). The 2015 IBC also contains additional design criteria for these system types. Note that the codes do not always require compliance with NFPA 92 whenever a smoke control system is required. For example, for smokeproof enclosures (e.g., pressurized stairs and elevator hoistways), the IBC has self-contained criteria and does not reference NFPA 92. Please note that all references within this article are based on the 2015 editions of NFPA 92, NFPA 101, and the IBC.

Smoke control systems

The NFPA 92-2012 created a new hierarchy of terminology, which has not been changed for the 2015 edition. The term "smoke control system" is now used as a broad classification to include two subclassifications, or methods, of smoke control: smoke management and smoke containment (see Figure 1).

Smoke containment systems include stairwell, elevator, vestibule, and refuge-area pressurization systems, as well as zoned smoke control systems. These system types are referred to as "design approaches" within the context of NFPA 92 (Chapter 4). Smoke containment systems generally involve using fans to either inject air into protected areas of a building or exhaust air from fire areas to create pressure differences with respect to adjacent areas, thus "containing" the smoke outside of the protected area.

Building codes generally dictate when a smoke containment system is required, or in many cases, offer a smoke containment system as an alternative design feature. For example, the IBC requires all interior exit stairways serving floors more than 75 ft above the lowest level of fire department vehicle access to be designed as smokeproof enclosures. The code also permits stairwell pressurization systems to be used as an alternative to providing an open exterior balcony or ventilated vestibule (see Sections 403.5.4 and 909.20).

Smoke-management systems involve those used to manage smoke within large-volume spaces, such as atriums or smoke-protected assembly seating. Approaches for smoke management permitted by NFPA 92 include:

  • Mechanical smoke exhaust
  • Natural smoke filling (simply allowing smoke to fill a large void space above)
  • Opposed airflow
  • Gravity smoke venting (providing pathways for the smoke to naturally leave the space).

Choosing a smoke control system

Building owners and designers are often quick to jump to a decision by providing a mechanical smoke exhaust system without considering other potential design alternatives that may be easier and less expensive.

Once the type of system is selected, the designer must determine the applicable design criteria. In many cases, the design criteria for smoke control systems are the same in both the IBC and NFPA 92; however, this is not always the case, particularly with stairway or elevator hoistway pressurization systems. For example, Table 1 illustrates some of the differences between the IBC and NFPA 92 that must be taken into account when designing a stair pressurization system where both documents are applicable.

When evaluating the maximum pressure difference between a stairway and the interior of a building, the IBC specifically limits the design to 0.35 in. wc, whereas NFPA 92 relies more directly on door-opening forces. It should be noted, however, that for a standard-sized door, the maximum design pressure difference stated in the IBC can be shown to cause door-opening forces roughly equal to the maximums permitted by NFPA 92.

The equation for resolving door-opening force, given the design pressure difference, is given in the IBC equation 9-1:

F = Fdc + K(WADP)/2(W-d)


A = Door area, square feet (square meters)

d = Distance from door handle to latch edge of door, feet (meters)

F = Total door-opening force, pounds (Newtons)

Fdc = Force required to overcome closing device, pounds (Newtons)

K = Coefficient 5.2 (1.0)

W = Door width, feet (meters)

DP = Design pressure difference, inches of water column (Pascals).

Assuming it is a 3×7-ft door, with a 10-lb self-closer, a 0.35-in.-wc pressure difference causes about a 30-lb opening force. It should be noted that the force to overcome the door-closing device varies and can affect the maximum pressure difference some. Therefore, the 0.35-in.-wc maximum pressure difference should be considered a guide.

Calculating pressure and airflows

To calculate pressure differences and airflows, the codes and standards allow the designer flexibility in determining the most appropriate calculation methodology for a particular system. While algebraic calculations and/or spreadsheets can be used to design and analyze smoke containment systems, the Multizone Airflow and Contaminant Transport Analysis Software (CONTAM), published by the National Institute of Standards and Technology (NIST), is a viable solution. This software can be used to calculate the expected pressure difference across an opening such as a stair or elevator door based on leakage, mechanical injection, and exhaust airflow rates into and out of the stair and adjacent spaces. In addition to open-airflow pathways like open doors and windows, objects such as walls, floors, closed doors, and other elements may have varying leakage rates depending on their age, construction type, condition, undercut, side gap, etc. Airflow rates and pressure differences across all of these elements can be modeled and calculated in CONTAM.

The impacts of weather, stack effect, HVAC systems, locations of injection points, and other variables should also be analyzed and documented as required by the applicable code or standard as part of the rational analysis for the smoke control system. CONTAM can especially be useful for taller buildings, which are more susceptible to stack effect, and/or buildings with multiple smoke control systems, which may operate simultaneously and create complex interrelationships.

For example, if a stairwell and/or elevator hoistway protected by a pressurization system has a door that opens into an atrium provided with a mechanical smoke exhaust system, the atrium exhaust rate can have a significant impact on the pressure difference across the stair door, hence, the door-opening force and required fan size. Software programs such as CONTAM also are useful for performing sensitivity analyses to determine which variables have the greatest impact on the design. For example, the designer can do multiple trials to get a handle on the stack effect on a stair or hoistway pressurization system based on various outside and inside temperatures.

Determining the design number of doors open

One important consideration in any stair pressurization system design is the "design number of doors open"; that is, how many doors are anticipated to be open at any one point for a reasonable amount of time. Generally, the determination of the design number of doors open is the responsibility of the designer. He/she must consider the use of the building, egress configuration, occupant loads, and any other characteristics that may impact occupant movement.

NFPA 92 states that the pressure-difference calculations must take the design number of doors to be opened simultaneously into account (see Section This means that the minimum pressure-difference requirements listed in NFPA 92 Table must be maintained with the design number of doors open, which can include both interior and exterior doors.

The design number of doors open for NFPA 92 compliance is left to the discretion of the designer. In contrast, the 2015 IBC minimum and maximum pressure differences—0.1 in. wc and 0.35 in. wc, respectively—are required to occur with all interior doors closed. This allows the potential for leaving an exterior door open. Previously, the 2012 edition of the IBC required these pressure differences to be maintained with all doors closed.


NFPA 92 contains additional criteria in the testing chapter of the standard. During testing, pressure-differential measurements must be recorded with all interior doors closed, and any exterior doors that would "normally be open during evacuation" must be open (see Section 8.4.6). The design number of doors open and number of exterior doors expected to be open during evacuation can be somewhat subjective, but can have a huge impact on system design and fan sizing. Adding an open door, especially to the exterior, can potentially double or triple the required fan size.

It is important to discuss these variables with the authority having jurisdiction (AHJ) and reach a consensus early in the design process. Also, summarizing the testing procedures in the design documents for approval may avoid issues during the acceptance testing. As an added precaution, it is generally recommended to provide relief dampers to ensure door-opening forces are not exceeded during testing or during a real-world scenario. Relief dampers are used to overcome the potential of excessive pressure build-up when a door is opened and then closed.

In addition to stair pressurization, one of the most common design applications for NFPA 92 and the smoke control requirements in the IBC are large-volume spaces, i.e., atriums. When dealing with the objective of maintaining tenable conditions for occupant egress in an atrium, the most common design solution is mechanical smoke exhaust. Large exhaust fans are installed to extract smoke from the upper part or ceiling of the atrium. Either mechanical fans or openings (i.e., automatic-opening doors/louvers) to the exterior of the building, located in the lower section of the atrium, provide clean make-up air and avoid negatively pressurizing the space.

This approach to smoke control also is frequently found in large assembly spaces, such as arenas and theaters. NFPA 101 and the IBC contain provisions for "smoke-protected seating," which take advantage of large-volume smoke exhaust systems to maintain tenable conditions for extended periods of time, allowing for more leniency in the amount of egress provided for these spaces. Although these spaces are not atriums, the smoke-management systems protecting them are similar to those found in atriums.

Computer simulations

There are three methods of analysis permitted by NFPA 92 for the design of smoke-management systems: algebraic calculations, computer simulations, and physical modeling.

The algebraic calculations provided within NFPA 92 Chapter 5 are useful for very basic atrium geometries and design scenarios. When the atrium geometry is complex or there are many variables at play (i.e., opening/closing doors, large balconies, multiple interconnected atriums), the capability of algebraic calculations becomes too conservative and computer simulations or physical modeling become more desirable. Physical modeling of atrium designs (either scale or full-size) can be tedious and expensive because changes to the design of the atrium can necessitate rebuilding the model and repetition of the analysis. For this reason, many design professionals take advantage of the convenience and relative accuracy of computer simulations.

Most computer simulations used for smoke or fire modeling fall into one of two categories—zone fire models or computational fluid dynamics (CFD) models. A zone fire model is an analysis of simple geometries by dividing a space into an upper (smoke-filled) and lower (clean-air) layer, and evaluating temperature, smoke concentration, and other properties in each layer. A CFD model is a more detailed analysis that divides the space into small, 3-D computational grid cells and tracks the movement of heat and smoke through those cells. CFD models are more complicated to program and run, and also require more computational power than zone fire models, but the ability to visualize smoke and fire movement are very useful for smoke control system designs.

Designing for large-volume spaces

When designing smoke-management systems for large-volume spaces like atriums, most of the NFPA 92 and IBC requirements are very similar, with the exception of the required duration of system operation. NFPA 92 simply requires any system to operate for at least the required safe egress time (RSET) based on a timed-egress analysis. For instance, if calculations estimate it will take 8.5 minutes to clear an atrium, the system must be operational for at least 8.5 minutes.

The IBC, on the other hand, requires the duration to be either 20 minutes or 1.5 times the calculated RSET, whichever is greater (see IBC 909.4.6[F]).

Specifying temperature-rated equipment

In addition to sizing fans and associated ductwork properly for smoke-management systems, the designer also must specify appropriate temperature-rated equipment (i.e., fans, dampers, ductwork) that will be in contact with the smoke-plume/hot upper layer. Both NFPA 92 and the IBC contain the same equation for determining the temperature rating of smoke exhaust equipment. IBC equation 9-3 is:

Ts = (Q/mc) + Ta


Ts = Smoke temperature, Fahrenheit (Kelvin)

Ta = Ambient temperature, Fahrenheit (Kelvin)

c = Specific heat of smoke at smoke-layer temperature, Btus per pound Fahrenheit (kiloJoules per kilogram degree Kelvin)

m = Exhaust rate, pounds per second (kilograms per second)

Qc = Convective heat output of fire, Btus per second (kilowatt).

IBC equation 9-3 provides a conservative temperature and does not take into account any reduced temperatures due to dilution from cooler air. IBC section 909.10.1 allows for a reduced smoke temperature when adequate dilution air is provided.

Furthermore, the associated ductwork for a smoke-management system must be capable of withstanding the probable pressures to which they will be exposed when the system is operating. Ductwork must be leak-tested to 1.5 times the maximum design pressure in accordance with nationally accepted practices. The measured leakage cannot exceed 5% of the design flow (see IBC 909.10.2[F])

Control systems

Many of the smoke control design considerations discussed up to this point are performance-based in nature. The design criteria are given in the codes, but the designer has flexibility in how to meet the criteria. There is less flexibility with the system controls. NFPA 92 and the IBC contain specific criteria for the subsystems that control the overall smoke control system.

Control systems must be listed in accordance with ANSI/UL 864, Standard for Control Units and Accessories for Fire Alarm Systems, category UUKL. UL 864 is the fire-alarm-system test standard, but the UUKL category is specifically for equipment used in smoke control systems. It may be tempting for a designer to use the building automation system (BAS) to control the smoke control system because the controls would already be in place and a "smoke control mode" can be programmed in. This can be done, however, not all vendors can provide a BAS that is UUKL-listed for smoke control.

Simplicity should be the goal of a control system design. A single system should be used to control the various smoke control functions. The more complex a system is, the less likely it is to operate efficiently and to be properly tested and maintained.

Smoke control systems must be activated automatically. Activation is generally initiated in response to a smoke detector or a water-flow-switch activation. Special design consideration is warranted where smoke stratification can occur, such as within tall atrium spaces. Upward-facing beam-type smoke detectors or detection at multiple elevations within the space can be used.

A firefighters’ smoke control station (FSCS) is required for all smoke control systems per NFPA 92 and the IBC. The FSCS provides manual control, status indicators, and fault conditions for the system. Manual controls should be clearly marked and should show the graphic location and function served via diagrams or notations on the FSCS. Means of verifying correct operation of components upon activation must be provided. This includes positive confirmation for the operation of fans, any fault conditions, and manual overrides. Failure to receive or maintain positive confirmation of operation must provide an off-normal indication within 200 seconds.

NFPA 92 contains other time limitations on how quickly the system must react. For example, smoke control mode must be initiated within 10 seconds after an activation command is received. For smoke containment systems, the fans must operate within 60 seconds. Completion of damper travel must occur within 75 seconds. For smoke-management systems, full operational mode must be achieved before conditions exceed design smoke conditions.

Smoke dampers used as part of the smoke control system must be ANSI/UL 555S: Standard for Smoke Dampers-listed. This standard ensures that the dampers can withstand potentially elevated temperatures and higher pressures, and will allow minimal leakage rates through the damper.

Once the smoke control system is designed and installed, the system is tested against their design criteria. Therefore, the designer needs to stay involved through this commissioning process. There are three types of smoke control testing:

  • Component testing of each component or subsystem (e.g., fire alarm and detection systems, dampers, fans, controls, standby power)
  • Acceptance testing of the fully integrated system during system commissioning
  • Periodic testing performed over the life of the system.

Periodic testing should be performed at least twice a year for a dedicated smoke control system and annually if the system is integrated with the building’s HVAC system. Lack of maintenance testing and/or poor recording practices can lead to deficiency citations from the AHJ.

When documenting the testing and commissioning procedures, the designer should discourage the use of smoke bombs or similar "real" smoke tests. Visible smoke demonstrations are not required by the codes and can provide arbitrary results because they do not provide the heat, buoyancy, and air entrainment that real fire can produce. However, if the AHJ is adamant that some form of visible smoke must be incorporated into the acceptance testing, these means and methods must be formally documented prior to testing so that the preparations can occur prior to the testing day. A smoke control system designed in accordance with NFPA 92 and/or the IBC criteria is expected to perform admirably during an actual fire event.

Erik Anderson is a manager at Koffel Associates. Shaun Wrightson and Nick Sealover are registered fire protection engineers at Koffel Associates.