In a seminar presentation at the ASHRAE Winter meeting in January 1985, commissioning was defined as "the process of assessing system performance."1 Commissioning has evolved since those early days, and different commissioning agents now define commissioning in various ways. Many focus on documentation. The preceding definition forms the underlying philosophy in this article.

As with any building system, commissioning smoke control systems contemplates that the systems have been started, tested and adjusted so they deliver the design performance at full load. That means motors have been checked for proper rotation and airflows have been measured and adjusted to the design quantities. Commissioning is not a substitute for thorough equipment startup. Commissioning goes beyond balancing, so problems discovered during balancing need to be corrected before commissioning can begin. On the fire alarm side, before commissioning can begin, smoke detectors must be shown to activate when exposed to the appropriate level of smoke obscuration, and audiovisual units must sound and flash when the fire alarm system is in alarm. Commissioning brings two more measures of performance evaluation to the table:

  1. Commissioning verifies that the systems work properly together, so a smoke detector in alarm or a sprinkler flow switch causes the smoke control systems to activate and the appropriate dampers to position properly in response to that alarm. These are the "process" and " system" parts of the earlier definition.
  2. Commissioning checks how well actual system performance delivers the design concept. These are the "assessing" and "performance" parts of "the process of assessing system performance."

Commissioning and Design

Commissioning begins with design.-A clear expression of the design criteria or the design intent is a key tool for successful commissioning. It tells everyone what performance they can expect the smoke control system to deliver during commissioning and, therefore, during a fire.

The Massachusetts State Building Code2 requires a fire protection system narrative as part of the project construction documents. Whether required in other jurisdictions or not, this document is useful. Preparing it helps designers organize their thoughts about what the fire protection system is supposed to do. It also provides written documentation of what commissioning agents and inspection authorities should expect the system to do. To help with commissioning, the design narrative needs to express performance in terms of the measurements and observations that will be performed during commissioning. The measurements might be air quantities or pressure differences. The observations might be airflow patterns or general direction of airflow. For example, a smoke control system might be expected to confine smoke to the area of origin and prevent its spread to egress stairs and other adjacent spaces. A different type of smoke control system might be expected to allow smoke to collect in the upper strata of the space and prevent it from descending to a level lower than a defined elevation above the floor.

System Type

Developing meaningful commissioning procedures requires understanding smoke control system design concept and performance goals. Smoke control systems can be divided into two classes: differential pressure (NFPA 92A) systems and buoyancy or volume (NFPA 92B and NFPA 204) systems.

Differential Pressure Systems

NFPA 92A3 is the basic design guide for pressure differential systems. These systems are typically applied as stair pressurization systems and exit stair vestibule ventilation systems. As stated in the section of the International Building Code:4

909.6 Pressurization method. The primary mechanical means of controlling smoke shall be by pressure differences across smoke barriers. Maintenance of a tenable environment is not required in the smoke control zone of fire origin. (emphasis added)

The goal of these systems is to create and maintain a pressure difference that discourages smoke from migrating outside the fire zone.

Chapter 8 of NFPA 92A provides recommended test procedures for commissioning differential pressure-type smoke control systems. These tests involve:

1. Using a manometer to measure the pressure difference between the fire floor and the stairwell. The design pressure difference is the primary performance measure for this type of smoke control system, so commissioning should include measuring it.

For a stairwell pressurization system, the design criteria-should state the number of doors expected to be open during a fire. Open doors create huge leaks. The stair pressurization system should not necessarily be expected to achieve the design pressure difference with more than the design number of open doors.

If the system does not achieve the desired stair pressurization, the first step is to verify fan performance. If the fans check out, the commissioning procedure should direct looking for unintended air leaks and sealing any that are found.

2. A spring-type scale can be used to measure stairwell door-opening force. The building code usually dictates the maximum allowable door-opening force. For example, the 2003 International Building Code4 allows a maximum of 15 pounds force (67 N) to unlatch the door plus another 15 pounds (67 N) to set the door in motion (30 pounds (133 N) total force to open the door).

Section 909.6.2 provides the following formula for calculating the door-opening force:


A = Door area, square feet (m2).
d = Distance from door handle to latch edge of door, feet (m).
F = Total door opening force, pounds (N).
Fdc= Force required to overcome closing device, pounds (N).
K = Coefficient 5.2 (1.0). When using English units, K converts pressure from inches of water to pounds per square foot.
W = Door width, feet (m).
ΔP = Design pressure difference, inches of water (Pa).

The pressure exerted by the stair pressurization system is considered uniform over the surface of the door. It is analyzed as a single point force applied at the center of the door. The stair pressurization system force applied to the door, K(AΔP), times its moment arm, (W/2), can be no greater than the allowable opening force, (F), times its moment arm, (W-d).

The pressures that could create door-opening forces that are greater than the maximum permitted door-opening force are typically greater than the pressures usually required for effective smoke control performance. Therefore, a door that requires excessive opening force during commissioning might be opening against a misadjusted closer or with more pressure than needed to control smoke. The closer force can be tested by measuring the door-opening force with the smoke control system off.

The door-opening force need only be measured for one door with all other stairwell doors closed. As soon as a door opens, the pressure in the stairwell decreases, and so will the force required to open the other doors into the stairwell.

In its simplest form, commissioning involves measuring the door-opening force. However, commissioning could include computing the door-opening force to predict what to expect when performing the field test. If the force is not as expected, commissioning includes figuring out why.

Strictly speaking, the purpose of commissioning is to verify that the system performs in accordance with the design. A design that requires more door-opening force than the building code allows could still be commissioned successfully. However, the calculation part of commissioning is more helpful when performed during the design phase of the project. That way, planning for commissioning can contribute to designing a system that will meet the building code requirements.

The following have occasionally been sources of confusion when smoke control systems have been commissioned:

  1. Section 909.20 of the International Building Code4 states specific prescriptive requirements in terms of air changes per hour for the mechanical ventilation alternative for the smokeproof enclosures that section 403.13 requires for egress stairs in high-rise buildings. Therefore, delivered airflow is the only performance measure that commissioning these systems can address. (Whether the prescribed airflow is sufficient to exclude smoke from the vestibule, or clear smoke that might enter the vestibule as the door opens, is not part of the design criteria, and therefore is irrelevant to commissioning the system.)
  2. Although NFPA 92A pressure differential systems utilize smoke exhaust fans, there is no design intent that these exhaust fans should keep the fire zone clear of smoke during the fire. Smoke bombs can be useful visual indicators of smoke leakage or flow patterns. Viewing the discharge from the smoke exhaust fan can provide a meaningful visual measure of whether the smoke exhaust fan actually removes smoke from the fire zone.
  3. There is also no intent that these fans should clear the fire zone of smoke within a certain period of time after the fire is out, even if they are sized on the basis of air changes per hour.
  4. The design goal for these systems is to maintain a pressure differential between the fire zone and surrounding spaces. Therefore, a manometer to quantify pressure differenceand smoke pencils to show the direction of airflow are appropriate tools for commissioning these systems.

Buoyancy or Volume Systems

NFPA 92B5 is the basic design standard for buoyancy systems. NFPA 2046 also provides some guidance on buoyancy systems. The fundamental difference between the two standards is that NFPA 92B systems often utilize fans while NFPA 204 systems traditionally rely exclusively on buoyancy. Like the title of NFPA 92B, the International Building Code4 reserves these systems for large, open spaces:

909.8 Exhaust method. When approved by the building official, mechanical smoke control for large enclosed volumes, such as in atriums or malls, shall be permitted to utilize the exhaust method. The design exhaust volumes shall be in accordance with this section.

The primary objective of buoyancytype smoke control systems is to keep the bottom of the smoke layer above some stated level for a stated period of time. The system works by providing an upper-level reservoir where hot smoke from the fire can collect. If needed, a mechanical smoke control system exhausts smoke-laden air from that reservoir at the calculated rate of smoke production from fire. By exhausting smoky air (usually from the top of the reservoir) at the same rate as the fire adds smoke to the bottom of the reservoir, the interface between the smoke layer and the clear air below does not descend below the design elevation.

It is difficult to simulate the amount of smoke the design fire in a mall or atrium is expected to create. Section 909.9 of the IBC4 specifies 5,000 BTU/sec (5 MW) as the design fire size for buoyancy-type smoke control systems. The taller the space, the more smoke is produced because the plume entrains more air. The additional ambient air cools the smoke plume, but the increase in air induced into the smoke plume more than offsets the reduction in density from the cooling effect, so the smoke extraction rate increases.

Since it is typically not practical to simulate the design fire in these large spaces, the smoke control system air quantities (supply and exhaust) are the key performance measures for commissioning these systems. Commissioning should also attempt to verify that two common problems, plugholing and destratification, do not compromise system performance.

Plugholing occurs when the air velocity entering an exhaust fan is high enough to entrain clear air from below the smoke layer into the exhaust inlet airstream. The comparatively high-velocity clear air jet "drills" through the smoke layer and creates a "plughole." The problem with plugholes is that they are paths for the exhaust fan to suck clear air from below instead of smoke from the upper-level reservoir. As a result, the bottom of the smoke layer can descend below the design interface level. Plugholing is slightly self-correcting. For a given exhaust inlet velocity, the deeper the smoke layer, the lower the potential for plugholing. Therefore, it is possible that a system prone to plugholing at the design smoke level interface will successfully keep the bottom of the smoke layer above some lower interface level.

Destratification occurs when a jet of make-up air enters the space with enough velocity to entrain smoke from the bottom of the smoke layer and suck smoke down to the clear air zone. An airstream that enters a space as a free jet entrains the adjacent-air mass and expands its volume. The higher the jet velocity entering the space, the longer the throw to a negligible terminal velocity and the larger the jet becomes.

Both plugholing and destratification are products of airflow patterns. Various types of "puffers" or smoke pencils that release visual indicators in different locations throughout the space might be useful tools to check for plugholing and destratification. One difficulty is that both plugholing and destratification are flaws in the buoyant smoke layer in the upper strata of the smoke control zone. That means testing for these problems could require getting up on a lift to measure airflow patterns and velocities in the upper-level smoke zone of the space.

Chemical Smoke

Cold smoke from smoke candles has been a favorite of some fire officials for gauging smoke control system performance. NFPA 92A3 cautions against using smoke bombs to test smoke control systems. Chemical smoke can provide some indication of airflow patterns and leakage paths, and might help identify plugholing and destratification. However, chemical smoke does not duplicate the buoyancy and thermal expansion of real smoke from a real fire. The lack of buoyancy of chemical smoke can make a satisfactory buoyancy-or volume-type system appear to fail. The low temperature of chemical smoke compared to real smoke can reduce the pressure increase due to thermal expansion of air in the fire zone and misrepresent the performance of a differential pressure-type system.

That admonition is cold comfort for an engineer or commissioning agent in the midst of knocking heads with a fire official who has not read NFPA 92A or does not care what it says. One compromise might be to use chemical smoke in a way that gives it some of the temperature and buoyancy of real smoke from a real fire. The smoke test rig illustrated in Figure 1 has been used successfully to test a buoyancy-type smoke control system. The insulation on the container that holds the smoke candles helps retain heat to keep the smoke warm and make it more buoyant. The insulated duct gives the smoke some loft and contributes to chimney effect to get the smoke to rise up into the upper strata of the space.

Kenneth M. Elovitz is with Energy Economics, Inc.


  1. Elovitz, K., "Commissioning Building Mechanical Systems," ASHRAE Trans. 1992, vol.98, part 2, paper number BA-92-8-1, pp. 543-552.
  2. The Massachusetts State Building Code, 6th Edition, Commonwealth of Massachusetts, Boston, MA, 1997.
  3. NFPA 92A, Recommended Practice for Smoke-Control Systems, National Fire Protection Association, Quincy, MA, 2006.
  4. International Building Code, International Code Council, Falls Church, VA, 2003.
  5. NFPA 92B, Standard for Smoke Management Systems in Malls, Atria, and Large Areas, National Fire Protection Association, Quincy, MA, 2005.
  6. NFPA 204, Standard for Smoke and Heat Venting, National Fire Protection Association, Quincy, MA,2002.