Society of Fire Protection Engineers

New Zealand Chapter


About Members


Events Links



Chris Mak, Company Engineer, Wormald New Zealand Limited


  • As both merchandisers and installers of sprinklers, we are commonly seeing sprinkler heads being specified as requiring to meet certain RTI criteria. For example, sprinklers are often specified as having an RTI £ 35 m½ s½. We question whether design to such precise levels is sound engineering practice, and as such, provide the following information for discussion.

  • Sprinkler RTI’s are measured by plunging a sprinkler assembly into a heated gas stream, of certain velocity, and measuring the average operating time. The concept of RTI was developed by Factory Mutual, to be independent of both gas temperature and gas velocity, as the concept of sprinkler sensitivity. The RTI is a physical attribute of the sprinkler, much as is the operating temperature of the head. Another physical attribute of the heads is its "conductivity", which is a measure of the heat transfer characteristics of the sprinkler head bulb and frame assembly. This attribute also has the same units as RTI, namely, m½ s½.

    An international working group on sprinkler and water spray product performance standards agreed tentatively on specific ranges of RTI’s and conductivity of sprinklers, to categorise them as standard, special or fast response sprinklers. Generally, fast response heads have an RTI of less than 50 m½ s½, special response heads have an RTI between 50 and 80 m½ s½, and standard response heads will have an RTI of between 80 m½ s½ and 350 m½ s½. Note that "abnormal" conductivities would affect these definitions.


  • The four major sprinkler manufacturers represented in New Zealand (Grinnell, Viking, Reliable and Central) do not publish sprinkler RTI’s in their data sheets. The information is informally available through the manufacturers.
  • The sprinkler bulb manufacturers do publish this information in the form of a Technical Specification.


  • In use of RTI’s in fire engineering design, we question whether any tolerance is applied to the figures in design. For example, in testing bulbs, 50 bulbs are tested at any one time, and for 57° C and 68° C, the mean operating temperature is allowed to exceed the nominal temperature by upto 6° C. The UL requirement for fast response sprinklers allows a temperature tolerance of ± 3.5%.
  • How does this effect fire engineering calculations?

  • Sprinkler RTI figures are being used in engineering calculations, using algorithms such as "DETACT", which is available in FPETool and FireCalc. However, use of manufacturer’s figures such as the previously quoted RTI £ 35 m½ s½ ignores other aspects which may contribute to sprinkler detection and fire control or suppression, including:
  • i) Height of fusible element with respect to ceiling. What happens if the head is used in a warehouse, below beam and purlin construction? What happens if the roof slopes?
  • ii) Cleanliness/loading of head. Private correspondence with Grinnell’s Research and Development personnel indicates that the apparent RTI of heads has doubled after the heads have been loaded with dust, in a matter of weeks. If heads are installed in freezers, and the heads are lightly coated in ice, what impact does this have?
  • iii) New Zealand’s common use of a drop in pressure to fire alarm signalling necessitates trapped air to act as a thermal buffer, and thus avoid false alarms. Is there the potential of time delays prior to water actually being available to control the fire?

    As fire engineers, are concerned about the use of sprinkler RTI’s, when designing to the limit. There are a number of parameters that may be significant in delaying sprinkler detection and operation times. We also question whether it is necessary to have a maintenance contractor know what sprinkler head RTI is required, if the heads need replacing, especially as the information is not published by the sprinkler head manufacturer. How do we control long term that activated sprinklers are replaced with an equivalent RTI?

    Thus, we would suggest that some conservatism is used in designing and specifying sprinkler head RTI’s. We would suggest the following figures, based on the ISO recommendations:

    i) Fast Response (3mm bulb) 50 m½ s½

    ii) Intermediate Response (4mm bulb) 80 m½ s½

    iii) Standard Response (5mm bulb) 135 m½ s½, (8mm bulb) 250 m½ s½

    This will allow any head to be replaced with any head with equivalent hydraulic characteristics, and meet the full requirements of the fire engineering design for the building.

    We note that these figures are offered for discussion purposes only, and we undertake no warranty as to their suitability for use in specific fire engineering calculations.


    We have taken the opportunity to discuss other issues with respect to sprinkler heads :

    i) Residential Heads

    We have seen residential heads specified to meet not only RTI criteria, but also a temperature rating of 57° C. The National Fire Sprinkler Association 1995 listing of residential heads lists two heads rated below 68° C, namely, GEM’s FR1 - Fusible Link Horizontal Sidewall, and Reliable F4/Res Concealed head. Both these are specialised heads, and not suitable for most installations. Residential heads have been designed to operate quickly, and control a fire prior to any life threatening situation.

    Thus, we question whether it is really necessary to specify sprinkler design parameters? If so, we would ask that the industry norms are specified, viz, either 68° C (bulb) or 71° C (fusible element).

    ii) Quick Response vrs Fast Response

    We note that there is some confusion with respect to the definition of quick response and fast response. We offer the following interpretation.

    A fast response head is just that. A head with a fast response element. A quick response head also considers the coverage of the head, which is fire tested. Thus, a quick response head is tested to operate within a set time, in a set room geometry. To illustrate this, an extended coverage ordinary hazard head uses fast response elements, with an extended coverage deflector, which has been demonstrated to equal a standard sprinkler head, via fire tests. The head is not a quick response, as it is listed for a maximum spacing of 6.1m. The fast response element with extended spacing is considered equal to a standard response element, with conventional spacing.

    (c) 2002
    Maintained by Mike Spearpoint