Heat Detector RTI – New Developments

Fire Suppression and Detection Research and Applications – A Technical Working Conference
(SUPDET 2011), March 22–25, 2011, Orlando, Florida

Daniel T. Gottuk, Ph.D. and Andrew T. Pomeroy, Hughes Associates, Inc.,

3610 Commerce Drive, Suite 817, Baltimore, MD 21227, Phone: 410-737-8677,
Fax: 410-737-8688, dgottuk@haifire.com

NFPA 72, the National Fire Alarm Code, specifies that the response time index (RTI) of heat
detectors must be listed along with the operating temperature. The RTI is a measure of how
quickly a detector’s thermal element will respond when exposed to a gas temperature at or above
its alarm threshold. To be able to adequately predict the response of a heat detector, the RTI of
the device must be known. Originally derived as a parameter for sprinkler activation [1–3], there
has been limited work performed to validate the use of the RTI concept for heat detectors [4–6].
The utility of the RTI in predicting response times of detectors for many applications is the
supposition that the RTI is constant for a range of temperatures and velocities. The current
standard being used by manufacturers for approval determines RTI values by plunging a device
into a hot wind tunnel at 197°C and 1.6 m/s [7]. However, these test conditions are
unrealistically severe for the range of expected ceiling jet profiles that heat detectors will
experience when alarming. This conclusion was demonstrated in this work by analysis of
bounding conditions using Alpert’s correlations over a range of expected fire sizes, ceiling
heights, and radial distances between detectors and the fire [8,9]. These findings raise the
question that if RTI values are not constant over applicable temperatures and velocities, then
reported RTI values (i.e., tested at high levels) may not be applicable to most realistic conditions.

While the RTI correlation is purported to be independent of temperature and velocity, data
from this and previous studies indicate otherwise. This study examined the effects of low
temperature and low velocity plunge test conditions on the constancy of the RTI for several
common heat detectors. The RTI correlation was found not to be constant across the ranges of
temperature and velocity applicable to realistic conditions. Consequently, the current RTI test
conditions being utilized for listing/approval purposes are deemed unsatisfactory for achieving
the principle purpose of listing RTI values (i.e., to provide a means for predicting heat detector
response).

The current plunge test apparatus (as used for sprinkler activation) was determined to be
unsuitable for performing low temperature, low velocity plunge tests due to the effects of
thermal stratification, as revealed by the Richardson number. Therefore, all tests were performed
using a modified plunge test apparatus. The test apparatus featured a vertically oriented, down-
flow test section designed to avoid thermal gradients at low temperatures and velocities.
Detailed measurements of the test section showed very uniform velocity and temperature profiles
where the detector is plunged into the apparatus.

Heat detectors were tested in the modified plunge test apparatus using temperatures of 75°C
and 200°C (167°F and 392°F) and velocities of 0.5, 1.0, 1.5, 2.5 m/s (1.6 ft/s, 3.2 ft/s, 4.9 ft/s and
8.2 ft/s). Four detectors were tested representing the three most common methods of heat

1
detection on the market: rate compensated thermal expansion, fusible link, and thermistor heat
detectors. Multiple alarm operating temperatures were evaluated. All of the heat detectors
showed relatively constant RTI values across test velocities at 75°C (167°F). However, the RTI
values varied significantly with velocity when the heat detectors were exposed to test
temperatures of 200°C (392°F). At the high test temperature, the RTI correlation appears to have
a positive relationship with velocity, leading to large standard deviations from the average value.
RTI varied by a factor of 2 to 4 over the range of varying velocity and temperature.

The findings indicate that the current measurement standard for RTI values of heat detectors
does not provide adequate results that can be used to accurately predict the response time of heat
detectors. It is apparent that a single RTI value for all conditions is not appropriate. However,
additional work will likely show that RTI values can be adequately specified for specific ranges
of velocity and temperature.

[1] Heskestad, G. and Smith, H.F. (1976), “Investigation of A New Sprinkler Sensitivity
Approval Test: The Plunge Test.,” Factory Mutual Research Corporation.

[2] Heskestad, G. and Bill Jr., R.G. (1988), “Quantification of Thermal Responsiveness of
Automatic Sprinklers Including Conduction Effects,” Fire Safety Journal, 14 (1–2), July
1988, pp. 113–125.

[3] Heskestad, G. and Bill Jr., R.G. (1989), Comments on “Thermal Response of Sprinkers,
Part II. Characteristics and Test Methods” and “Thermal Response of Sprinklers – A
Theoretical Approach,” Fire Safety Journal, 14 (3), January 1989, pp. 189–190.

[4] Bissel, W.G. (1988), “An Investigation Into The Use of the Factory Mutual Plunge
Tunnel and the Resulting RTI for Fixed Temperature Fire Detectors,” Unpublished
Thesis, Worchester Polytechnic Institute.

[5] Nam, S. (2006), “Predicting Response Times of Fixed-Temperature, Rate of Rise, and
Rate-Compensated Heat Detectors by Utilizing Thermal Response Time Index of
Detectors,” Factory Mutual, Norwood, MA.

[6] Nam, S., Donovan, L.P., and Kim, J.G. (2004), “Establishing heat detectors' thermal
sensitivity through bench-scale tests,” Fire Safety Journal, 39 (3), April 2004,
pp. 191–215.

[7] Factory Mutual (2008), FM 3209 American National Standard for Evaluating the
Response Time Index for Fixed, Rate of Rise, and Rate Compensated Heat Detectors.

[8] Alpert, R.L. (1972), “Calculation of Response Time of Ceiling-Mounted Fire Detectors,”
Fire Technology, 8 (3), pp. 181–195.

2
[9] Pomeroy, A.T., “Analysis of the Effects of Temperature and Velocity on the Response
Time Index of Heat Detectors,” Masters of Science, Department of Fire Protection
Engineering, University of Maryland, 2010.