How Reliable Is Your Smoke Control System
Will my smoke control system work properly during a fire? That is the question all building owners/operators should ask themselves. Smoke control systems are unique in that they often sit dormant (sometimes for years) and problems that can affect their operation may not be identified until it is too late. To ensure the right answer when you need it, catch up on device monitoring, commissioning, and why the self-test concept has received a bad rap.
Smoke control systems have a reliability problem. Too often, one can find a building with an existing smoke control system that is no longer functional. Problems are often identified at the point where modifications/renovations are made to the building that trigger a re-investigation of the system, and it is found out that: a) the system “never operated properly” or “no one really understood how it functioned” and the system was disconnected at some point; b) it was found to be improperly working due to a loss of function of one or more critical components; or c) it was found to be operating, but operation of the system either fails to meet the intended design criteria or exceeds those criteria in such a way as to create an unsafe condition in the building (typically by creating unacceptably high door-opening forces).
The International Building Code (IBC) requires that smoke control systems be provided in buildings where occupant egress may take a significant amount of time or involves a greater than normal hazard, such as for high-rise or underground buildings. Smoke control systems are also required in buildings with large volume multi-story spaces such as atrium buildings, malls, or theaters where high ceilings may result in a lack of sprinkler operation to control or extinguish a fire. Smoke control systems are therefore required as a means to control the large quantity of smoke produced from these fires1. High-rise and underground buildings typically accomplish smoke control by containment, using pressurization to keep smoke confined to a pre-defined zone of a building. Large volume space typically accomplishes control by managing the smoke and exhausting smoke from the space of origin at a rate that maintains acceptable conditions within that space.
Smoke control systems are often complex in their design and involve components such as the building’s HVAC system, electrical equipment, fire alarm system, and architectural components (e.g. operable doors). There are many potential failure points that may cause the system to fail to operate properly, but these typically fall into one of the following categories:
• Device monitoring and control
• Commissioning procedures
• Automatic self-testing
• Periodic testing and maintenance
Device Monitoring and Control
In order to verify that smoke control system components are aligned properly to achieve their desired function, it is paramount that proper device monitoring and control be provided. There has been a great deal of confusion over the past 10-15 years as to how this should be done, which results in an almost jurisdiction-by-jurisdiction standard of care for how this is accomplished.
When smoke control system component monitoring language was first introduced into the codes (circa the 1997 Uniform Building Code and 1995/1996 editions of NFPA 92A/92B), the term “Supervision” was used for this intended monitoring function, but it was learned that this implied performance that could not be achieved electrically for devices (i.e., dedicated fans) that were not normally operating. Current editions of the IBC and NFPA 92 use the term “Verification” for the intended language and both documents include the following language:
Verification shall include positive confirmation of activation, testing, manual override, and the presence of operating poser downstream of all circuit disconnects.
As shown in Figure 1, a Firefighter’s Control Panel (FFCP) per IBC requirements has four status lights required for each device showing normal (white) and fault (amber) conditions and open/on (green) or closed/off (red) status. In order to properly annunciate the condition of each device on the FFCP it is critical that a proper end-of-line device or devices be provided to indicate status.
For fans, it is critical to monitor fan status using a differential pressure or current switch downstream of all disconnects. To ensure maximum reliability, the position of all disconnect switches should also be monitored so that a fault is indicated for the fan when a disconnect is in the off position. For fans equipped with VFDs, they are required to go to fixed setpoints to deliver the proper air flows in smoke control mode. Contacts should be provided to render a VFD’s “off” button inoperable in smoke control mode and to disable varying the fan speed while in this mode.
Dampers end switches that monitor the blade position are typically more reliable than a proximity sensor mounted to a damper’s operating shaft. When a shaft sensor is used it should be somehow permanently affixed to the shaft and not mounted by a set screw that may loosen over time. Large multi-section FSDs (Figure 2) often fail when the linkage loosens. When only one of the multiple sections of dampers fails (typically due to damper leakage loosening) but the one with the end switch still works, a non-functional damper is not always detected by the weekly self-test. To ensure reliability, all sections of the multi-section damper must have end switches.
For operable doors and windows, it is critical to be able to capture both the fully closed and fully open status of that component. Therefore, it is generally not sufficient to simply install a magnetic end switch in the door/window jamb to report closed status and then consider that device open if the end switch is not triggered, as the door/window could fail (or get stuck or blocked) in a partially open position. Door and window operators capable of reporting both fully closed and fully open are commercially available and should be used for this function.
Proper commissioning of a building smoke control system starts with hiring a qualified entity to perform this type of testing. The IBC requires a third-party special inspection of smoke control systems. The special inspector should ideally be one with expertise in smoke control system design and testing, so that failure modes can be anticipated and tested during commissioning. A registered fire protection engineer (or equivalent) should be considered to perform this type of testing.
Commissioning testing involves several steps, starting with inspection and testing of individual components, then proceeding with sequence of operations testing, and finally performance testing. While component testing can often be performed while construction is ongoing, it is critical that construction is essentially complete when performance testing is performed. Performance testing involves measuring system airflows and pressures across smoke barriers. Measuring pressures before critical components such as door hardware or door thresholds/carpeting have been installed, or when large unsealed openings still exist in the building’s curtain wall, could lead to differences in measured pressures versus the actual pressures that would be recorded had the construction been properly completed at the time of testing.
Other sources2 have provided a list of recommended steps to perform during commissioning of smoke control systems to verify that the system performs as designed and critically that door opening forces are maintained in an acceptable range. If a building is commissioned in the spring time in a cold weather city such as New York and Boston, and door opening forces are measuring 28-29 lbf, particularly at stair doors, it is safe to say that the maximum door opening force of 30 lbf may be exceeded under winter stack effect conditions. Therefore, the system should either be recommended to be retested or pressures should be backed off during commissioning to allow for an appropriate safety factor to anticipate the winter stack effect. This may require the commissioning agent to consult with the original system designer.
Underwriters Laboratories (UL) has included specific language pertaining to Smoke Control Systems in its standard UL 864 — Control Units for Fire Protective Signaling Systems since 1997, although listing of equipment under the UUKL product category date back to 1989. Per these requirements, UUKL listed panels are required to be capable of performing a weekly self-test of smoke control system components. The weekly self-test may be the most important development toward ensuring smoke control system reliability in the past 30 years. The self-test requires regular and automatic testing of system components that would otherwise, in the case of some systems, sit for years without being tested. This greatly reduces the likelihood that the system would work properly when needed in a fire event.
Over time, the weekly self-test was codified by both NFPA 92 (and previously NFPA 92A and 92B) and the IBC and, until recently, continued to be included as a prescriptive requirement for smoke control systems. Section 220.127.116.11 of the 2012 edition of NFPA 92 and Section 909.12 of the 2012 edition of the IBC specifically required verification of system function using the weekly self-test.
Unfortunately, the codes in their most recent edition eliminated or softened the requirement for the weekly self-test, which is a detrimental blow to the overall goal of smoke control system reliability. This has occurred largely due to the efforts of building owners and equipment vendors who find the test “onerous.” The 2015 edition of NFPA 92 eliminated reference to the weekly self-test altogether. The 2015 edition of the IBC retained the language pertaining to the self-test but added an exception that allows bypassing this test for components that could produce unwanted effects to normal building operations upon approval of the building official.
Opposition to the self-test stems from the fact that it is often wrongly thought of as a system test rather than a component test, and building owners balk at the thought of bringing in large volumes of untreated air into an atrium building or starting up a “noisy” system in an operating high-rise hotel. Contrary to this belief, the self-test is only intended to be a component test, meaning that individual components can be tested sequentially in such a way to cause minimal to no disruption to a building. Fans can be energized individually, and only briefly, to ensure that they are in an operating state. Once each fan is tested, dampers can be tested individually to verify that they cycled fully to their off-normal condition and then back to their normal condition. Finally, operable doors/windows can be briefly opened to verify their operation, bringing in no more outside air than a normal exterior door that opens and closes to allow occupants to enter and exit during the day.
It should be noted that components can be scheduled to be tested at different times of the day. Fans may be tested at night when a building is not occupied to lessen the impact of excessive noise, and exterior doors may be tested during the middle of the day to avoid security access issues. For years, a range of building types have had to comply with the self-test requirement and have successfully done so with minimal reported problems.
Periodic Testing and Maintenance
Of equal importance to the initial commissioning testing and automatic self-testing is a periodic manual testing and maintenance program for smoke control systems. NFPA 92 recommends a semi-annual frequency for dedicated systems and an annual frequency for non-dedicated systems.
Periodic testing includes inspection and performance testing that can identify problems not identified by the automatic component testing. Architectural modifications may have been made to the building that affect the system of pressure differentials created by the smoke control systems. This may require a modification to the system air flows or sequence of operations to bring the system’s function back into agreement with the original design intent.
There has been an increased tendency to design systems to modulate to achieve the design pressures where the pressure differential is measured using one or more pressure transducers. Pressure transducers will drift over time and must therefore be recalibrated on a frequency recommended by the manufacturer of that device. All too often, these devices are not recalibrated. This could lead to excessive pressures being created by the system as it modulates to hit the wrong target. These excessive pressures (and resulting door opening forces) would be identified during a full performance retest.
Ensuring proper operation of a building’s smoke control system should be considered essential to building owners and operators. It must always be remembered that the smoke control system was required to mitigate a hazard created by a feature of the building’s design, such as a multi-story communication between floors via an atrium or the creation of an underground space. The smoke control system is an essential life safety system and not a “ticket to punch” only necessary to obtain a building’s certificate of occupancy.
Maintaining a functional smoke control system can be achieved by following some basic guidelines:
1. Hire a third-party commissioning agent experienced with smoke control system design and testing. A licensed fire protection engineer (or equivalent) should be responsible for the commissioning who has an in-depth knowledge of the intended design of the system and can anticipate failure modes that can be addressed at the time of commissioning.
2. Properly monitor all smoke control system devices and control points. All devices that are critical to the function of the system should be monitored so that a malfunction is capable of being quickly identified when the system is tested. Manual disconnects should be monitored whenever possible.
3. Employ the UUKL weekly self-test. The weekly self-test is an essential component to ensure the overall reliability of the system. Perceived inconveniences caused by the test can almost always be worked around by a properly sequenced test.
4. Test the system periodically. Per NFPA 92, dedicated smoke control systems should be tested semi-annually and non-dedicated systems should be tested annually. Testing should include performance testing that includes measuring pressure differentials and door opening forces. Building modifications that can impact system function should be fully investigated and pressure transducers used to modulate system function should be properly calibrated. ES
1. Ferreira, Michael J. “Chapter 14: Smoke Control Systems,” NFPA Fire and Life Safety Inspection Manual (Ninth Edition), National Fire Protection Association, Quincy, MA, 2013.
2. Ferreira, Mike. “The Pitfalls of Pressurization Smoke Control Systems,” Engineered Systems, Vol. 32 No. 1, January 2015.