Detector activation time

Detector activation time is the time from the start of a fire until an automatic fire detector, or a sprinkler acting as a detector, responds to it. In performance-based design it is the first part of the required safe egress time, because warning and escape can only start once the fire has been detected. A prediction needs the growth rate of the fire, the characteristics of the detector and the geometry of the building.

For a heat detector or a sprinkler, the calculation is a heat balance. Hot gas in the ceiling jet warms the sensing element until it reaches its operating temperature. How fast the element warms is expressed by its response time index. Smoke detectors are much harder to model, and engineers often treat them as very sensitive heat detectors that respond to a temperature rise of about 10 to 13 °C. Validation against experiments shows a wide scatter. For smoke detectors modelled with the empirical temperature rise method, the predicted activation times have a relative standard deviation of 0.58, against 0.34 to 0.39 for the fire models CFAST, MAGIC and FDS.

Role in fire safety design

In New Zealand, C/VM2 (second edition, 2025) writes the required safe egress time as a sum of four parts. These are the detection time, the time to notify the occupants, the time before they start to move, and the time spent moving to a place of safety. The detection time in this sum must come from deterministic modelling of automatic detection or sprinkler activation. In the Introduction to Fire Safety Engineering, the first warning for occupants likewise follows from detection and alarm, given either by installed systems or by people.

Fire detection and alarm systems can serve life safety, property protection, business continuity and environmental goals. For life safety, the warning must come early enough to allow complete evacuation of the danger zone before conditions become untenable. Schifiliti, Meacham and Custer therefore define a critical fire size: the heat release rate at which the fire must be detected to reach the design objective. The time between this critical fire and the design limit is the total system response time. It is the sum of fixed delays such as alarm verification, variable transport delays and the response lag of the actions that follow. The transport delays depend on the distance from the fire to the detector, the ceiling height and the convective heat release rate.

Prescriptive rules place detectors at a fixed spacing, for example one smoke detector per 84 m² or one every 9 m. Such rules do not state how quickly the detectors respond or how large the fire is at that moment. The listed spacing of a heat detector comes from a full-scale test with an alcohol pan fire under sprinklers rated at 71 °C on a 3 m grid. The fire is set so that a sprinkler operates after about 2 minutes, and the listed spacing is the largest at which the detector operates first. According to the same authors, this code approach does not allow a quantitative assessment of detector response.

Heat transfer to a heat detector or sprinkler

Heat reaches the sensing element of a heat detector or sprinkler by conduction, convection and radiation. In the early stage of a fire, radiation can be neglected. The elements of most heat detectors are thermally isolated from the rest of the unit, so convection from the hot gas dominates. The element is treated as a lumped mass, which means that its temperature is assumed to be the same throughout. Its heating is described by a time constant that grows with its mass and specific heat and falls with its surface area and heat transfer coefficient.

The heat transfer coefficient grows roughly with the square root of the gas velocity. The product of the time constant and the square root of the velocity is therefore nearly constant for a given device. This product is the response time index (RTI), with units of m^1/2s^1/2. It is measured in a plunge test, in which the device is suddenly lowered into a stream of hot gas with a constant temperature and velocity. Generic values range from 34 m^1/2s^1/2 for a quick-response sprinkler link to 235 m^1/2s^1/2 for a standard-response bulb. For listed heat detectors, NFPA 72 gives time constants derived from the listing tests, which convert to RTI values with a reference velocity of 1.5 m/s.

With the RTI known, the element temperature follows from a heat balance. The element warms at a rate set by the square root of the gas velocity and by the difference between gas and element temperature, divided by the RTI. A fixed-temperature device activates when the element reaches its rated temperature. When the gas temperature and velocity stay constant, the equation can be solved directly for the activation time: $$\frac{\mathrm{d}{T}_{\mathrm{d}}}{\mathrm{d}t}=\frac{\sqrt{u}(T_{\mathrm{g}}-T_{\mathrm{d}})}{\mathrm{R}\mathrm{T}\mathrm{I}}$$ $$t_{\mathrm{a}\mathrm{c}\mathrm{t}}=\frac{\mathrm{R}\mathrm{T}\mathrm{I}}{\sqrt{u}}\ln \left(\frac{T_{\mathrm{g}}-T_{\mathrm{a}}}{T_{\mathrm{g}}-T_{\mathrm{a}\mathrm{c}\mathrm{t}}}\right)$$

Symbol	Description	Unit
$T_{\mathrm{d}}$	temperature of the sensing element	°C
$T_{\mathrm{g}}$	gas temperature at the detector	°C
$T_{\mathrm{a}}$	ambient (initial) temperature	°C
$T_{\mathrm{a}\mathrm{c}\mathrm{t}}$	rated activation temperature	°C
$u$	gas velocity at the detector	m/s
$\mathrm{R}\mathrm{T}\mathrm{I}$	response time index	m1/2s1/2
$t$	time	s
$t_{\mathrm{a}\mathrm{c}\mathrm{t}}$	activation time	s

Not every heat detector behaves like a lumped mass. Rate-of-rise detectors respond when the temperature rises faster than a set rate, typically about 7 to 8 °C per minute. Most of them contain an air chamber with a small vent, so the RTI is only an approximation of their response. Rate-compensated detectors tend to respond when the surrounding gas reaches their rated temperature, as long as it rises by less than about 22 °C per minute. Heat detectors have the lowest false alarm rate of the automatic detectors, but they are generally the slowest.

Gas conditions at the detector

The heat balance needs the gas temperature and velocity at the device, which under a flat ceiling means the ceiling jet. For steady fires, the Alpert correlations give both values from the heat release rate, the ceiling height and the radial distance from the fire axis. They assume that the gas at the detector follows the instantaneous heat release rate, so they ignore the time the gas needs to travel from the fire. The height of the device below the ceiling matters too. In tests with small pan fires about 4.3 m below a ceiling, Marrion found the highest temperatures and velocities between 25 and 76 mm below the ceiling.

A growing fire can be split into short intervals with a constant heat release rate, which is called quasi-steady modelling. For each interval, the gas temperature and velocity are calculated, and the heat balance gives the detector temperature at the end of the interval. The calculation continues until the detector reaches its operating temperature. The computer program DETACT-QS by Evans and Stroup uses this method with Alpert's equations.

Many fires grow roughly with the square of time, and such a t-squared fire has its own ceiling jet correlations. Heskestad and Delichatsios developed them to include the moment at which the heat front reaches the detector. NFPA 72 describes these fires by a fire growth time, the time needed to reach 1,055 kW. Its slow, medium, fast and ultrafast fires have growth coefficients of 0.00293, 0.01172, 0.0469 and 0.1876 kW/s². These fires therefore reach 1,055 kW after about 600, 300, 150 and 75 s. Beyler later solved the detector heat balance for these correlations analytically, which allows the required spacing to be found by iteration.

Smoke detectors

The response of a smoke detector is much harder to predict than that of a heat detector. It depends on the smoke particles, their transport to the detector, the aerodynamics of the detector housing and the type of sensor. Ionisation and scattering detectors do not work by light attenuation, while research data mostly describe smoke by its optical density, which is a measure of attenuation. Detectors also use response algorithms that manufacturers generally do not publish. For smoke detectors there is therefore no measure that plays the role of the RTI.

The most common approximation treats a smoke detector as a sensitive heat detector with little or no thermal lag. NUREG-1805 uses a temperature rise of 10 °C, based on work by Heskestad and Delichatsios for typical fuels. The International Fire Engineering Guidelines (2005) use 13 °C above ambient with an RTI below 10 m^1/2s^1/2. The 2016 validation study for nuclear power plant fire models used an RTI of 5 m^1/2s^1/2 and a rise of 10 °C, as do the zone models CFAST and MAGIC. This approach assumes that temperature rise and smoke concentration go together.

A second approach compares the optical density of the smoke at the detector with the density needed for an alarm. C/VM2 uses an optical density at alarm of 0.097 m⁻¹ for spot and air sampling detectors, a value for normal sensitivity. The value at alarm depends strongly on the smoke. A scattering detector that alarms at 0.029 m⁻¹ in smoke from a smouldering cotton wick may need 0.15 m⁻¹ in the darker smoke of a kerosene fire.

Smoke also needs time to enter the sensing chamber. Heskestad described this entry lag as a characteristic length of the detector divided by the gas velocity, so the lag grows when the ceiling jet is slow. Residential detectors respond well to flaming fires that create gas velocities above 0.25 m/s, but may fail in smouldering fires with velocities below 0.15 m/s. Aspirating systems can be up to ten times as sensitive as spot detectors, but dilution and travel through the pipework usually delay them considerably. For gas sensing detectors, the guidelines know of no accepted method to determine activation times.

Worked examples

Schifiliti, Meacham and Custer give the example of a kerosene pool fire of 4 m² under a flat ceiling 6 m high, in a room at 20 °C. With a burning rate of 67 g/(m²·s) and a heat of combustion of 40.3 kJ/g, the heat release rate is about 10,800 kW. A heat detector with an RTI of 55 m^1/2s^1/2 and a rating of 57 °C hangs 6 m from the fire axis. The Alpert correlations give a gas temperature of 153 °C and a velocity of 2.4 m/s at the detector. The steady-state equation then gives the activation time: $$t_{\mathrm{a}\mathrm{c}\mathrm{t}}=\frac{55}{\sqrt{2.4}}\ln \left(\frac{153-20}{153-57}\right)=35.5\times 0.326\approx 12\mathrm{s}$$ A detector with the RTI of 30 m^1/2s^1/2 that C/VM2 prescribes for heat detectors would respond in about 6.3 s in the same fire. From the same values, the time constant of the detector element in this flow, its RTI divided by the square root of the gas velocity, is about 35.5 s. The calculated activation time is much shorter than this, because the gas at 153 °C is far hotter than the 57 °C rating, so the logarithmic term in the equation stays small. Such short times also arise because the fire is assumed to burn at its full size from the start.

The same authors show a quasi-steady calculation for a stack of wood pallets under the same ceiling, with the detector again 6 m from the fire. The heat release rate rises from 5 kW after 10 s to 675 kW after 120 s. By then the detector has warmed only from 20 °C to about 32 °C, so a 57 °C detector has not yet responded. Continuing the calculation with the tabulated heat release rates gives about 46 °C after 180 s, when the fire has reached 1,519 kW.

A third example compares a sprinkler and a heat detector under a 4 m ceiling. The sprinkler is rated at 74 °C with an RTI of 110 m^1/2s^1/2, the detector at 57 °C with an RTI of 42 m^1/2s^1/2. In a fast fire with a growth time of 50 s, the detector responds after 50 s and the sprinkler after 65 s. In a slow fire with a growth time of 600 s, they respond after 300 s and 370 s. The detector thus responds when the fast fire has reached about 1.06 MW, but the slow fire only about 0.27 MW (both values rounded up, to the safe side).

NUREG-1805 compares three methods for a smoke detector 3.05 m from a steady 1,000 kW pool fire in a compartment 3.96 m high. The method of Alpert gives 0.42 s, the method of Mowrer 0.74 s and the method of Milke 0.26 s. Mowrer's method, for example, adds the time for the smoke to rise to the ceiling and the time to travel along it to the detector. These results only describe smoke transport from a fire that is steady from the start. A medium t-squared fire, by comparison, needs about 300 s to grow to 1,055 kW.

Requirements and design methods

In New Zealand, C/VM2 (second edition, 2025) requires the detection time of an automatic system to be modelled with a ceiling jet correlation or with a CFD model. Detectors must be modelled with fixed values, regardless of the make and installation of the devices that will actually be fitted. A heat detector, for example, has an RTI of 30 m^1/2s^1/2, an activation temperature of 57 °C and a radial distance of 4.2 m. Smoke detectors are placed at a radial distance of 7.0 m, and all devices at least 25 mm below the ceiling.

C/VM2 allows a manual alarm only in spaces with a ceiling height of at least 5 m, with fewer than 50 occupants who are awake and familiar with the building. Occupants are then assumed to become aware of the fire once the ceiling jet has crossed the whole space. C/VM2 gives an equation for this time when zone modelling is used. For storage up to 5 m high, the time in seconds is 10 plus 2.4 times the length of the space, if the length is at most 1.4 times the width. A space 20 m long and 15 m wide would therefore give a detection time of 58 s.

Schifiliti, Meacham and Custer distinguish two kinds of problem. In a design problem, the required response time or fire size is given, and the engineer searches for the detector spacing. That spacing equals 1.41 times the largest radial distance that still gives activation in time. In an analysis problem, the spacing is known and the activation time or the fire size at activation is calculated. In one of their examples, a 50% larger sprinkler spacing, from 3.0 to 4.6 m, raised the fire size at activation by about 25%.

Limits and uncertainty

Many sources of error affect these calculations, including uncertainty in the operating temperature, the ambient temperature and the ceiling jet correlations. Because the size of these errors is largely unknown, the SFPE Handbook chapter states that a confidence interval for the result cannot be estimated. The RTI concept is itself a simplification. The heat transfer coefficient varies with the Reynolds number to a power between 0.330 and 0.805, and up to 10% of the heat gained may be lost by conduction to the detector body. The heat detector model in FDS does not account for thermal radiation.

For t-squared fires, errors in the predicted dimensionless gas temperature can reach 50%, which corresponds to 5 to 10 °C. The largest errors occur for fast fires and low ceilings. The ceiling jet methods also assume a flat ceiling without walls. In one handbook example, the leading edge of the ceiling jet travelled about 48 m in 120 s. The authors use such a distance to check whether the limits of the model have been exceeded. NUREG-1805 adds that detector response times increase as forced ventilation increases.

The 2016 supplement to NUREG-1824 compared predictions with experiments. For sprinkler activation time, the empirical correlation had a bias factor of 1.11 and a relative standard deviation of 0.41. For smoke detector activation with the temperature rise method, these values were 1.07 and 0.58, while the experimental uncertainty itself was 0.34. The evaluation only included sprinklers under smooth, flat and horizontal ceilings, and smoke detectors in the room of origin.

For smoke detectors, the SFPE Handbook chapter goes further. Because the accuracy of the estimation methods is unknown, it advises against comparing their results with egress time calculations or with the time to untenable conditions. It recommends using them instead to compare how the response of one detector changes with its spacing or location. C/VM2, by contrast, uses the detection time directly in the required safe egress time, with fixed detector values for every design.

See also

• Aspirating smoke detection
• Beam smoke detector
• Flame detector
• Multi-sensor detector
• Fire alarm system

References

• Alpert, R. L. (2002). Ceiling jet flows. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-18–2-31). National Fire Protection Association; Society of Fire Protection Engineers.
• Iqbal, N., & Salley, M. H. (2004). Fire Dynamics Tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program (NUREG-1805). U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation.
• McGrattan, K., Hostikka, S., Floyd, J., McDermott, R., Vanella, M., Mueller, E., & Paul, C. (2026). Fire Dynamics Simulator technical reference guide: Vol. 1. Mathematical model (6th ed., NIST Special Publication 1018-1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.1018
• Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
• National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
• Schifiliti, R. P., Meacham, B. J., & Custer, R. L. P. (2002). Design of detection systems. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 4-1–4-43). National Fire Protection Association; Society of Fire Protection Engineers.
• Stroup, D., Lindeman, A., McGrattan, K., Peacock, R., Overholt, K., Joglar, F., LeStrange, S., & Montanez, S. (2016). Verification and validation of selected fire models for nuclear power plant applications: Supplement 1 (NUREG-1824, Supplement 1; EPRI 3002002182). U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research; Electric Power Research Institute.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.