Construction Technology Update No.

72

Duct Smoke Detectors:
The Impact of Various Factors
on Their Effectiveness
By G.D. Lougheed
This Update presents the results of research on smoke movement through
HVAC systems and the effectiveness of duct smoke detectors. It provides
guidance to practitioners and regulatory authorities in the context of North
American code requirements for such detectors.

Requirements on the use of smoke detectors inlets, filters and fans (see Figure 1). In some
in HVAC ducts are included in most build- jurisdictions, detectors must be located in
ing codes, including the National Building the return air duct as well. Installation
Code of Canada (NBC).[1] The intent of requirements are provided in standards,
the requirements, based on a recommenda- including CAN/ULC-S524.[3]
tion made by the National Board of Fire Because there had been little or no research
Underwriters in 1939 [2], is that the HVAC or data analysis to support this requirement,
system be shut down during a fire to mini- the Fire Detection Institute1 undertook
mize the circulation of smoke through the research on duct smoke detection to provide
building by the HVAC fan system. In most technical data to codes and standards com-
cases, the detectors must be located in the mittees and system designers. The research,
supply air duct downstream of the fresh air conducted jointly by NRC-IRC and the
University of Maryland (UMD), examined the
use of duct smoke detectors both as a means
of detecting fires or smoke within the HVAC
system and as part of a building’s smoke
management system. UMD conducted
small-scale experiments and modelling
studies while NRC-IRC conducted full-scale
experiments using its ten-storey test facility.
The research project addressed issues that
had been raised regarding the need for duct
smoke detectors and how they are installed
and used. The key question was: Does duct
smoke detection work and is it worth the
added cost, considering the potential for false
and nuisance alarms? The specific issues
investigated were the comparative driving
forces of the HVAC fans relative to those pro-
Figure 1. Typical constant-volume, single-zone HVAC system duced by the fire; how detection is affected
1 The Fire Detection Institute has recently merged with the Fire Protection Research Foundation.

However. little pling and multi-sensor. photoelectric. where the levels at which the become larger with aging).[5] 2 .06 50 17 – 25 advantage of shutting down the system 0. depend- ing on the type of building. Detector response (analog output) relative ing the fire itself. output from the detectors. As indicated in ative to that resulting from the fire itself Table 1. It does not elimi- nate smoke movement through building shafts As shown in Table 1. decreased by a factor of 10. those stemming from other factors. shape and colour. NRC-IRC researchers size. and the familiarity of occu- The shutdown of the HVAC system is intended pants with the means of egress. * Visibility range for front-illuminated object. smoke in the HVAC system once it reached It was found that the HVAC-related pres. detectors—ionization. concentrations that are comparable to those sure differences were generally larger than used as criteria for safe building evacuation. The measured optical densities can be used to estimate the visibility (i. This suggests alarm signal would sound are selected. it cools and changes they are generally expected to detect. The analog output at by the fire and ambient conditions (stack and which the detectors would produce a trou- wind effects). Each detector used in the research had a different range of outputs. stairs and service) and ductwork detectors was proportional to the optical as a result of pressure differences produced density of the smoke. typically in the range of 5–25 m. smoke aging. type of duct. sured at the same location in the return air Researchers investigated smoke aging. as required by codes and standards. the location in Comparative Driving Forces the building.15 100 7 – 10 depends on the specific characteristics of the building in question. small particles at the duct inlet compared the results would be the same for detectors to the number at a point 3 m downstream located in the supply duct. The research examined the ble signal or an alarm is set at the control issue of whether or not smoke movement unit and is typically less than the maximum created by the HVAC fans is significant rel. in ductwork it to a percentage scale based on the maximum output occurs predominantly within the first few for each detector. gases. Table 1. The extent of the 0. to minimize smoke transport through the building by the HVAC fans. sam. In addition. all the detectors would respond to and other effects. These greater pressure to smoke optical density and visibility differences also led to higher flows and the distribution of smoke to floors where there Smoke Detector was no fire. includ. As smoke moves detectors to respond reliably to the fires away from a fire source..04 – 0. They compared the was known about how this so-called aging analog output2 (signal) from each detector of smoke affects the response of commercial with the optical density of the smoke mea. and stratified flow in the ity criteria used in performance-based HVAC ducts. metres. smoke detectors. the dis- tance one can see) in the smoke in the by smoke dilution. The measure. duct in the HVAC system. To characteristics.e. Until this research was conducted. tified using computational methods. and agglomerates of these in the HVAC ducts might be too low for three classes of matter. For purposes of comparison. and the efficacy of sampling evaluations of fire-protection systems are tubes used for duct detection. soot conducted experiments using four different deposition on surrounding surfaces reduces types of commercially available duct smoke the amount of smoke in the air stream. This can be quan. the HVAC fans should indeed be shut 0 0 down upon detection of a fire. while the num- ber of large particles increased by a factor 2 The analog signal from the detector is sent to the of two (smoke particles agglomerate and alarm control unit. visibil- HVAC filter. liquid concerned that the concentration of smoke droplets.1 – 0. The results confirm that unless Optical Response Visibility* an active smoke management strategy is in Density (OD/m) (%) (m) place. including smoke particle address this concern.[4] Dilution Effects Smoke Aging Engineers and code officials have long been Smoke consists of solid particles. the outputs were converted than might be expected. measuring the number and size of smoke ments related to dilution and aging effects particles. the output of the (elevators. that the aging of smoke occurs more rapidly For comparison purposes. They found that the number of were made in a return air duct.

researchers were able HVAC Filters to demonstrate that at low velocities. near the fire source. Thus the the two locations showed that the type of smoke had ample opportunity to cool. the full. 30 – 35 55 40 ters would have an impact on the detection pleated wet. and efficiency of HVAC filters must ing duct smoke detectors to be located at be considered when evaluating the perfor. the filter. which was located the filter. the mid-length of a straight run. But as the velocity with the amount removed varying according increases. the conditioning area. it had detector response (the concern being that cooled to near ambient temperature and stratification would concentrate the smoke mixed with air and smoke from eight storeys at the top of the duct and could go unde- in the facility. Practitioners have cautioned that the return air duct in the mechanical room. all the duct smoke detectors was primarily inlets and outlets in the duct to allow for dependent on the optical density of the uniform mixing of the smoke in the air smoke. The greater effect tion. 3 . Detectors located they are less sensitive to the small-diameter near bends and outlets had the lowest particles that pass through the filters. The results indicated that tected). the duct smoke detector is mended. gest that there is no justification for requir- type. Effect of HVAC filter type on detector output Decrease in detector output (%) Dust-spot Filter efficiency Photoelectric Ionization type (%) detector detector and before any filters. the filters Group 1 would not have an effect on detector (Glass fibres response. response (analog output).Table 2. and HVAC filter affected duct smoke detector there was no observable thermal stratifica- performance (Table 2). It is recom- return air ducts. that they be located at typically placed after all return air inlets. The measurements were taken in stream. The project results Previous research suggested that photo. upper part of the duct. On the supply side. inlets and outlets. laid cellulose) as they can reduce the smoke to a level at which it is not a concern. or fans (see Figure 1).[5] using the dust-spot efficiency measure to Further experiments were conducted on indicate the effectiveness of the filter in a series of duct smoke detectors with both removing particulates from the airflow. fresh air inlets. formance for all of the detectors. the fil- (Extended area. A comparison of the results from remotely from the fire source. large distances (3–10 duct diameters) from mance of duct smoke detectors. regardless tric smoke detectors is due to the fact that of location or orientation. If used in bends. The results sug- The results suggest that the position. By the time the smoke the duct and thus negatively affect duct reached the measurement location. standards have recommended scale experiments conducted in the NRC-IRC that duct smoke detectors be located some facility indicated that the signal output from distance (3–10 duct diameters) from bends. uninterrupted straight runs of duct located a considerable distance from the may cause stratification of smoke within fire compartment. In this case. the duct 10 – 15 35 20 in a cardboard smoke detector is placed after the fresh air frame) inlet. The performance uniform. buoy- HVAC filters remove some of the smoke as ancy causes the smoke to concentrate in the the contaminated air passes through them. the distribution becomes more to the type of filter used. vertically and horizontally oriented sam- Measurements were conducted at the pling tubes on an extended length of duct NRC-IRC facility in both the return air duct in the NRC-IRC test facility’s HVAC and in the supply air duct downstream of mechanical room. more dependent on particle diameter than on Stratified Flow particle concentration. However. with the degree of stratification of the filters is quantified through testing in depending on the temperature of the smoke accordance with ASHRAE standards [6]. relative to that of the ambient air. indicate the duct detectors will respond if electric detectors may be activated more the smoke in the HVAC system reaches quickly as a result of smoke aging than other concentrations that are typically used as types of detectors because their response is criteria for safe building evacuation. Therefore. and Group 2 the fan (see Figure 1). smoke aging did not affect smoke detection By taking measurements in a duct system in the case of duct detectors. however. Historically. of fires (other than a fire in the filter itself). The studies showed very similar per- (decrease in detector output) on photoelec. long.

Duct detectors demonstrate the efficacy of sampling tubes with sampling tubes may also be installed for the range of velocities typically found in with a vertical orientation. G. tubes being used. Duct Smoke Detection. In this case. HVAC systems. 2001. 8. Mower. surveys of 65 commercial Milke. GA. J. energy to the surrounding environment. Summary The results of this research project indicate that Dr. Leprince. F. The Efficacy of Duct Smoke sampling tubes were shown to be effective Detection. the results provide guidance on detectors in the upper part of horizontal the location of duct smoke detectors and ducts is the best approach.A. Internet: http://irc.. J. see References 7 and 8. Standard for the Installation attached to the exterior of the duct.gc.0 m/s to 19 m/s. 4.06 m/s to 40.A.. For more information.D. • smoke dilution and smoke aging do not have an impact on the effectiveness of © 2008 duct smoke detectors.. References Efficacy of Sampling Tubes 1.. p. National Research Council. Mowrer F..L. Fire Protection Engineering. 2001. contact Institute for Research in Construction. and Torero. evacuation is still possible.52 m/s to Fire Protection Engineering. The • detectors in either the return or supply distance at which stratification becomes a system will typically respond to smoke at factor is dependent on the temperature of concentrations at which safe occupant the smoke relative to that of the ambient air. showing the greatest effect on be expected until the smoke travels a sufficient photoelectric detectors. p. If the smoke temperature is above the ambi. American system air velocities of 4. 2005. National Board of Fire Underwriters. 363-395. National Research Council of Canada September 2008 ISSN 1206-1220 “Construction Technology Updates” is a series of technical articles containing practical information distilled from recent construction research. Spring 2006. Ryder.W. lation of dust in the detection system.64 m/s. Refrigerating.[5] Spot Procedures for Testing Air-Cleaning Other experiments measured the response Devices Used in General Ventilation for of duct smoke detectors as a function of Removing Particulate Matter. and Torero. with only two of the 6. in the ducts. Fire Technology.L. detectors should be mounted at the top of For further information on this research the duct to minimize the potential accumu.1. in Construction. 1939. Society of Heating. 237-264. Because concerns had been raised 2004. Gravimetric and Dust- HVAC systems exceeding the maximum air velocity stipulated in the standard test. J. Sampling of Fire Alarm Systems. Facsimile: (613) 952-7673.D.32 m/s. 20.L. regarding the performance of the sampling 5. which are less sen- distance to mix with the ambient air and loses sitive to smaller diameter smoke particles. Volume 14. Underwriters’ tubes enter the duct to collect a representative Laboratories of Canada. S. Journal of at five velocities in the range of 1. installed on the ceiling. sample of the air flowing through the duct.ca . Standard tests for duct smoke detectors A Comparison of Driving Forces for include an evaluation of the sampling system Smoke Movement in Buildings. Ottawa. • the type of HVAC filter affects detector ent temperature. buildings in the Baltimore/Washington area Measurements of Smoke Characteristics were conducted. The 7. National Building Code of Canada. and Air- Over this range.. The other uses a NFPA Quarterly 33. The surveys determined in HVAC Ducts. Ottawa K1A 0R6 Telephone: (613) 993-2607. stratification in the duct can response. ANSI/ASHRAE 52.W. This finding suggests that mounting smoke In addition. F. There are two types of duct smoke detection. no significant variation in Conditioning Engineers. similar detector contained in a housing 3. p. Ottawa. the HVAC systems. 1992. which are the same as those Smoke hazards of air-conditioning systems. Atlanta. N. that the airflow velocities ranged from Volume 37. National Research Council of Canada. Lougheed is a principal research officer • the shutdown of the HVAC system will in the Fire Research program at the National reduce smoke movement through the Research Council Canada Institute for Research HVAC system. CAN/ULC S524.nrc-cnrc. Fire Protection Engineering. 2. Wolin. over the range of velocities typically found in Winter 2006. Milke. J. NEMA. One uses spot-type smoke detectors installed 2. NEMA. 113-122.. detector performance was observed.