Available online at www.sciencedirect.


Procedia Engineering 11 (2011) 233–241

The 5th Conference on Performance-based Fire and Fire Protection Engineering

Estimation of Fire Detection Time
LI Qianga,b,*

Department of Fire Protection Engineering, Chinese People’s Armed Police Forced Academy, Langfang 065000, China b State Key Laboratory of Fire Science, University of Science and Technology of China, Hefei 230027, China

Abstract Estimating the fire detection time is one of the important contents of performance-based fire protection design, which can be used to compute required safety egress time (RSET) and to determine the time of aerosol extractors start-up. In spite of that, the computing methods of fire detection time do not exist in current codes of China. Engineers often set a fixed fire detection time depend on their experience. In this thesis, a review is present demonstrating the detection time estimation methods of fire detectors, include fixed temperature heat detectors, rate of rise temperature heat detectors, light-scattering type smoke detectors, ionization-type smoke detectors, beam unit smoke detectors and radiant energy fire detectors.

© 2011 Published by Elsevier Ltd.
Keywords: performance-based fire protection design; fire detector; fire detection time;

Nomenclature A c cp d Dm H Hc g surface area of the detector’s element, m2 detector element’s specific heat, kJ/(kgć) specific heat of air, kJ/(kgć) distance between the fire and the detector, m mass optical density, m2/g ceiling height or height above fire, m convective heat transfer coefficient for the detector, kW/(m2ć) gravitational constant, 9.18m/s

* Corresponding author. Tel.:+86-316-206-8513; fax: +86-316-206-8513. E-mail address: sterops@mail.ustc.edu.cn .
1877–7058 © 2011 Published by Elsevier Ltd. doi:10.1016/j.proeng.2011.04.652

kg/(kg s) ambient temperature. m/s * Ta Td Tg OD  Q  ′′ Q  Q conv  Q total r RTI S t u up Į ǻH ǻT ǻT p ȗ ț ȡ IJ * dimensionless fire gases velocity power law fire growth coefficient. m1/2s1/2 radiant power reaching the detector. m the response time index. kW heat release rate per unit area. s . ć dimensionless smoke temperature extinction coefficient of air at detector operating wavelengths proportionality constant for the detector density of ambient air. kW total heat transfer to a detector. kg/m3 detector time constant. of the detector. g smoke fraction per unit time.234 LI Qiang / Procedia Engineering 11 (2011) 233–241 m s m mass. kJ/g increase above ambient in temperature of gas surrounding a detector (=T g -T a ). kW/sn heat of combustion. kW radial distance from fire plume axis. m-1 heat release rate. ć the optical density per meter. ć temperature of fire gases at the detector. kW/m2 convective heat transfer. s velocity of fire gases. or set point. ć temperature rating. W time.

tuned to detect the extremely small particles of combustion.1. A typical alarm may sound when the rate of temperature rise exceeds 6. and to alert the building's occupants and fire emergency organizations. Recent technological developments have enabled the perfection of detectors that activate at a temperature of 47ć. thereby increasing the time available for evacuation and minimizing fire damage. Smoke-sensing fire detectors Smoke detectors are designed to identify a fire while in its smoldering or early flame stages. such as a kitchen. which are placed along ceilings or high on walls in a manner similar to spot thermal units [4].3. 2. Estimating the fire detection time is one of the important contents for performance-based fire protection design. Consequently. they are useful in places where smoke detectors have a high rate of nuisance alarms. As smoke migrates between the two components. Rate-of-rise (ROR) heat-sensing detectors react to the sudden change or rise in ambient temperature from a normal baseline condition [3]. Any sudden temperature increase that matches the predetermined alarm criteria will cause an alarm. This is interpreted as a smoke condition. such as ionization smoke detector and scattering smoke detector. For large open spaces such as galleries and atria. 2. the compute methods of fire detection time do not exist in current codes of China. 2. In spite of that. radiant energy-sensing fire detectors.LI Qiang / Procedia Engineering 11 (2011) 233–241 235 1. and the alarm activation signal is transmitted to the fire alarm panel. Fire detectors A key aspect of fire protection is to identify a developing fire emergency in a timely manner. The most common fixed temperature point is 58ć. Engineers often set a fixed fire detection time depend on their experience.7ć to 8. with each type having advantages in different applications. Fixed temperature heat detectors react when the ambient temperature reaches a fixed point [3]. such as heat-sensing fire detectors. They operate on either an ionization or photoelectric principle. Radiant energy-sensing fire detectors . Most air-sampling smoke detection systems are capable of a higher sensitivity than spot type smoke detectors. it is necessary to give the way to estimate the fire detection time.2. However. The most common smoke detectors are spot type units. This is the role of fire detectors. Introduction On the basis of fire safety engineering. which are often used in buildings. The fixed detection time is lack necessary theoretical support and also reduces the credibility of performance-based fire safety design. 2. performance-based fire protection design is a new way for building fire safety design [1]. a frequently used smoke detector is a projected beam unit [4]. that are mounted at some distance apart. Heat-sensing fire detectors Heat-sensing detectors are designed to respond when the ambient temperature or rate-of-rise of temperature exceeds a predetermined value [2]. providing increased time to escape. An air-sampling smoke detector is capable of detecting microscopic particles of smoke [5]. Air samples are drawn through a network of small-bore pipes and past a sensitive optical device.3ć per minute. often a solid-state laser. a light transmitter and a receiver. and so on. which can be used to compute required safety egress time (RSET) and to determine the time of aerosol extractors start-up. This type of heat detector can react to a lower threshold condition than would be possible if the threshold were fixed. This provides earlier notification of a developing fire allowing manual intervention or activation of automatic suppression systems before a fire has developed beyond the smoldering stage. smoke-sensing fire detectors. They are also useful in areas of the home where smoke detectors cannot function effectively because it is too hot or cold. This detector consists of two components. This paper addresses the methods to calculate the fire detection time. Heat-sensing detectors are not effective early-warning devices because they must be very close to a fire to be set off. the transmitted light beam becomes obstructed and the receiver is no longer able to see the full beam intensity.

Method for estimating the detection time of heat fire detectors Detection typically occurs during the initial stages of a fire. especially when compared to the convective heat transfer rate. so we can neglect the radiant heat transfer. As radiant energy in the approximate 4. which are recommended by NFPA 72 [6]. their sensing equipment recognizes the fire signature and sends a signal to the fire alarm panel. and so forth. and is thus approximately proportional to the square root of the velocity of the gases passing the detector. This time constant is a function of the mass. convective heat transfer coefficient. Because the majority of the heat transfer to the detection element is via convection. refineries and fuel loading platforms. specific heat. the change in temperature over time can be expressed as follows: dTd H c A (Tg − Td ) = dt mc (2) The use of a time constant (IJ) was proposed by Heskestad and Smith [7] in order to define the convective heat transfer to a specific detector’s heat-sensing element. As the mass. as well as from the ceiling.236 LI Qiang / Procedia Engineering 11 (2011) 233–241 Radiant energy-sensing fire detectors are line of sight devices that operate on either an infrared. the heat-sensing elements of most of the heat detectors are thermally isolated from the rest of the detection unit. The advantage of flame detection is that it is extremely reliable in a hostile environment. Common uses include locomotive and aircraft maintenance facilities. the following relationship can be expressed as the response time index (RTI) for an individual detector: RTI = τ u1/ 2 (4) . we address 6 calculation methods of fire detection time.1. They are usually used in high value energy and transportation applications where other detectors would be subject to spurious activation.000 to 7. 3. and area of the element and can be expressed as follows: τ = mc ( H c A ) (3) Literature [8] shown that the convective heat transfer coefficient for heat detection elements is similar to that of spheres. as indicative of a flaming condition. 3. so we can assume that the conductive portion of the heat release rate can be considered negligible. thermal capacity and area of the detection element remain constant. ultraviolet or combination principle [4]. the total heat transfer on the detectors can be written as:  =Q  Q total conv = H c A ( Tg − Td ) (1) For a mass (m) of the detection element. and mines.700 angstroms range occurs. cylinders. Method of fire detection time In this section. On the other hand.

These fires are therefore referred to as t-squared fires.(2).LI Qiang / Procedia Engineering 11 (2011) 233–241 237 Substituted Eqs.115 ) =® −0.59 ( r / H ) (12) * t2 = A1/5α 1/5 H −4/5t * t2 f = 0.87 / ( 9.146 + 0.63 r / H > 0. the following correlations were presented for fires that had heat release rates that grew according to the power law equation. Heskestad and Delichatsios [11]. ¸ © H¹ (7) ΔTp* = A 2/ ( 3 + p ) (Ta / g ) α t −1/ (3 + p ) ΔT 2/ ( 3 + p ) H − ( 5− p ) / ( 3+ p ) (8) where t* p = A −1/ (3 + p ) α H 4/ (3+ p ) (9) A = g ( c pTa ρ ) (10) Using the preceding correlations. with p = 2. and with later updates from another paper by Heskestad [12].242r / H .3 °3. the following equation can therefore be used to calculate the heat transfer to the detection element and thus determine its temperature from its local fire-induced environment: dTd u = dt 1/ 2 (T g − Td ) RTI (5) On the assumption that fires grow according to the following power law relationship:  = αtn Q (6) Relationships have been developed by Heskestad and Delichatsios [9] for temperature and velocity of fire gases in a ceiling jet. ¸ © H¹ r · § = g ¨ t* p.861 (1 + r / H ) (13) (14) (15) D = 0. ­ 0 ° ° * * ΔT2* = ®§ t2 − t2 f °¨ ¨ D ° ¯© * u2 * * t2 ≤ t2 f · ¸ ¸ ¹ 4/3 * * > t2 t2 f (11) ( ΔT ) where * 1/ 2 2 1/ 2 ­ r / H ≤ 0.3 ° ¯0.(3) and (4) into Eq. These have been expressed as follows [10]: u A 1/ ( 3 + p ) 1/ ( 3 + p ) u* p = u H ( p −1) / ( 3 + p ) r · § = f ¨ t* p.

Table 1. Similarly. Beyler[13] given the solution for the preceding temperature of thermosensitive element: 1/ 4 § 4 · § ΔT · ΔT2* ) (1 − e −Y ) ¨ ¸¨ * ¸( © 3 ¹ © ΔT2 ¹ dTd ( t ) dt = (t / t ) D * 2 (16) −Y § ΔT · * ª1 − (1 − e ) º « » Td ( t ) = Td ( 0 ) + ¨ T Δ 2 * ¸ Y « » © ΔT2 ¹ ¬ ¼ (17) where 3§ u · Y= ¨ *¸ 4 © u2 ¹ 1/ 2 ª º * « u2 » « ( ΔT * )1/ 2 » 2 ¬ ¼ 1/ 2 § ΔT2* · § t · ¨ ¸¨ * ¸ D © RTI ¹ © t2 ¹ (18) Consequently. the ratio of temperature rise in a fire driven flow to smoke concentration can be expressed as follows:  ′′ s Q OD ΔT = 3330 ρ c p m (19) So. we can estimate the detection time of rate-of-rise heat fire detectors by Eq. 3. rearrange and integrate Eq.1 Optical density versus temperature rise method Work by Evans and Stroup [14] on a large. the detection time estimation method of fixed temperature heat fire detectors is presented. Temperature rise for spot type smoke detector response [6] . flat ceiling compartment.238 LI Qiang / Procedia Engineering 11 (2011) 233–241 Substituting Eqs. the temperatures rise to response of a smoke detector for many materials are list in Table 1.(11)-(15) into Eq. Methods for estimating the detection time of smoke fire detectors 3.(5) and integrated.2.(17).(16).2.

However. the flow velocity at a certain distance between the fire plume centerline and the detector expressed by Eq.(7) change as * 1/5 1/5 u = u2 A α H 1/5 (24) .(21) and integrated. we can write the relation: m = DAπ r 2 h / Dm Since Q(t) = mǻH c and Q(t) = (Įt3)/3.7 27.2 Mass optical density method This method can be used to estimate the fire detection time of beam type smoke detectors. formula is obtained for estimating the fire detection time.LI Qiang / Procedia Engineering 11 (2011) 233–241 239 Material Wood Cotton Polyurethane PVC Ionization-Temperature Rise (ć) 13.152 m/s.2 7. not t-square fires.7 7. So. In the UL smoke box test. we can write the relation: m = α t 3 ( 3ΔH ) (21) (22) Substituted Eq.2 7.1 Average Identifying the fires grow processing. a t-square fire can be modeled as a succession of steady-state fires for slow and medium growth rate fires.3 Critical velocity method Research by Brozovski [16] shows that a minimum critical velocity is necessary before smoke can enter the sensing chamber of the smoke detector.8 7. 3. by the optical density in the room produced by the burning material can be calculated from the following equation [15]: OD = Dm m (π r 2 h ) (20) In the case of the smoke detectors respond at an optical density of D A .(8).8 Scattering-Temperature Rise (ć) 41. This method assumes that. then the fire detection time can be estimated using Eq. 2 t=ª ¬3DAπ r hΔH c / (α Dm ) º ¼ 1/3 (23) 3. Modeling the smoke layer as a cylindrical volume centered about the fire plume having a depth (h) equivalent to the ceiling jet thickness or some multiple of it. if the critical velocity has been attained.2 21.15 m/s. Therefore.(22) into Eq. sufficient smoke concentration is in the ceiling jet gas flow to produce an alarm signal.2 7. Ceiling jet velocity correlations exist for steady-state fires. Brozovski [16] suggests that the critical flow velocity at the detector is approximately equal 0.2. the velocity is 0.9 1.2.

Where flammable or combustible liquids are the fuel load and are unconfined.115 )3/ 4 uc ( 0. k=1.146 H + 0.861( H + r ) º ° −0.87 / ( 9. model the fuel as a circular pool. m.(24) yields 2/3 ­ ª 1/5 1/5 H 1/5t − 0.5h f w f (29) where c a is power per unit area proportionality constant.3 1/ 2 1/5 ° 0.861( H + r ) º 1/ 2 1/5 1/5 1/5 A α °3.3 °0.240 LI Qiang / Procedia Engineering 11 (2011) 233–241 For t 2 *•t 2f *. Substituted Eqs.59 ( r / H ) » 0.87 ) ( Aα H ) ( Aα H ) t=® 3/ 2 0.945 ° uc (r / H ) (0.(11)(15) into Eq. k=4). the fire detection time is then estimated as: ª 0.242r ° « » ° ¬ ¼ u=® 2/3 ª 1/5 1/5 H 1/5t − 0.242r ) 0.146 H + 0. u c . The radiant power output of the fire to the detector (P) can be approximated as being proportional to the radiating area of the flame: P = ca Ar Ar = 0. 3.242r ) 0. h f is flame height ( = 0.364Sd 2 eζ d º » t=« 2/5 « ¬ κ cW f ( kα ) » ¼ 5/ 4 (30) . fuel package is in a corner.182(kQ ) ). m.593/ 2 ( Aα H ) ( Aα H ) ¯ (26) for a certain critical flow velocity.861( H + r ) + r / H ≤ 0.146 H + 0.242r « » ° ¬ ¼ ¯ (25) and thus. and began spread under the ceiling. the fire detection time can be expressed as: 3/ 2 ­ ( 9.3.3 ° 3/ 2 1/ 2 1/5 ° ( 3. fuel package is near a wall. For t-square fire.861( H + r ) + r / H ≤ 0. which means the smoke reached ceiling. Compute the radiating area (A r ) using the following equation: (28) 2/5  where w f is flame width.63 1/5 1/5 1/5 A α A α H « r / H > 0.115 ) A α H « r / H ≤ 0.3 » 0. Method for estimating the detection time of radiant energy -sensing fire detectors Work by Robert [10]. the response of flame detectors as shown in the following equation: S = κ Pe−ζ d d 2 (27) Determine the minimum anticipated flame area width.146 H + 0. k is wall effect factor (no nearby walls. k=2.

Fire Safety Journal 1989.nedcc. php [5] http://en. In: Murthy SNB.LI Qiang / Procedia Engineering 11 (2011) 233–241 241 4. Evaluation of smoke detector response estimation methods. Section Four. Glassman I. 2003. DETACT-QS. Prediction of specific optical density for smoke obscuration in an NBS smoke density chamber. and so on. Methods to calculate the response time of heat and smoke detectors installed below large unobstructed ceilings. Delichatsios M. Fire detector methods today have been comparatively mature through long-term research and practice.org/resources/leaflets/3Emergency_Management/02IntroToFireDetection. Heat Transfer.15:471–475. [15] Chien WP. College Park. Several special application computer models.22:54–65. .org/wiki/Heat_detector [4] http://www. Delichatsios MA. “Environments of Fire Detectors . National Fire Protection Association. Design of Detection Systems. VA 22151. Investigation of a New Sprinkler Sensitivity Approval Test: The Plunge Test. In: DiNenno PJ. 17th international symposium on combustion. University of Maryland. Stroup DW. II (NBS-GCR-77-95). The Initial Convective Flow in Fire. [16] Brozovski E.hometips. National Fire Protection Association.com/buying-guides/home-heat-detectors. Hefei: Anhui Technology and Science Press. PA: In The Combustion Institute. editors. Chapter 1:4-1~4-43. SFPE Handbook of Fire Protection Engineering. Norwood. I (NBS-GCR-77-86).11:206–218. 2010. but also the basis of aerosol extractors start-up time determine for numerical simulation. National Technical Information Service (NTIS). p. [7] Heskestad G. JET. Master of Science Thesis. References [1] Huo R. [12] Heskestad G. Report 22485. [17] Geiman JA. such as Argos. Master of Science Thesis. NFPA72. MA. In: National fire alarm and signaling code. [8] Hollman JP. [9] Heskestad G.Phase 1: Effect of Fire Size. Third edition 2002. Worcester. Analysis vol. Factory Mutual Research Corporation. Fire Technology 1986. Update: The Initial Convective Flow in Fire. are available to assist in the design.html [3] http://en.org/wiki/Smoke_detector [6] Engineering guide for automatic fire detector spacing. 1113–1123.. Patton J R. Quincy. [11] Heskestad G.wikipedia. A Preliminary Approach to Siting Smoke Detectors Based on Design Fire Size and Detector Aerosol Entry Lag Time. Delichatsios MA. [10] Robert PS.11:53–75. Quincy. [13] Beyler CL. 1991. DETACT-T2. In view of fire position environmental characteristics and fire detector characteristics. New York: McGraw-Hill. FMRC Tech. [14] Evans DD. Fire Technology 1975.wikipedia. Springfield. Buliding performance-based fire protection analysis and design. Seader JD. it is able to utilize fire dynamics models to estimate fire detection time. Measurements vol. [2] http://www. Massachusetts. Fire Safety Journal 1986. Ceiling Height and Material”. 2003. LAVENT. Fire plumes and ceiling jets. Pittsburgh. 1151 Providence Turnpike. MA. MA 02062. Yuan HY. Worcester Polytechnic Institute. MD. 1979. 1976. Summary Estimate a reasonable fire detection time is not only the condition of performance-based safety evacuation design.

Sign up to vote on this title
UsefulNot useful