Do Thin Combustible Ceilings Compromise Sprinklers?

Discussion created by dhefpe on Mar 13, 2017
Latest reply on Dec 22, 2018 by dhefpe


The Las Vegas Strip contains many of the largest and most unique buildings in the world.  To create the desired look and feel within a space, the owners and designers occasionally propose suspended ceilings out of thin combustible materials, such as fabrics or plastics.


At temperatures somewhat below 300 oF, most thermoplastics start to lose their tensile strength and liquefy (melt).  As such, plastics that exceed one-quarter of an inch thick may even fit into the category described in the following paragraphs.  Fabrics are also thermally thin materials and even if they are made out of natural products, they can be expected to degrade at moderate temperatures generated during a fire.


Whether these materials do or do not comply with building code requirements as a ceiling can be debated.  If one fully understands the intent of the codes, the required tests, and what constitutes a compliant application, it is easy to conclude that these products cannot be used.  Not only must the requirements for suspended ceilings be considered, but when these materials create a plenum, additional limitations are imposed.  Although these fire protection aspects must be considered, the primary concern addressed herein is that these materials may adversely affect sprinkler activation and effectiveness.


Depending on the depth of the void above a combustible ceiling, NFPA 13 requires sprinklers to be installed in the plane of the ceiling and at the deck above.  When a fire originates beneath a thin combustible ceiling, which set of sprinklers will activate first?


Several full-scale fire tests have been conducted on some of these materials.  Some of the tests indicate that these materials may be good performers.  Others indicate the opposite.  The physics applicable to proper sprinkler activation appears to be:


  1. When flames from a fire do not extend anywhere near the ceiling and the space is bounded (walls or draft curtains), temperatures at the membrane ceiling can be below the degradation temperature of the material.  The heat can then spread out on the ceiling, without breaching the membrane, and bank down.  In this situation, sprinklers in the plane of the ceiling activate as expected and can then keep the fire from spreading.


  1. When flames from a fire extend reasonably close to the ceiling and the temperature at the membrane exceeds the degradation temperature of the material, the membrane opens up where it is impacted by the heat.  The size of the hole created is approximately the diameter of the heat plume.  The heat then enters the space above the membrane.  As described in the following paragraph, proper sprinkler operation cannot be assured.  This scenario may also apply to a large space that is essentially unbounded (walls located a substantial distance from the fire).


Imagine a fire centrally located between four sprinklers with the heat having created a breach in the thin membrane approximately the diameter of the plume.  Once the heat enters the cavity above the membrane, we can expect sprinklers at the deck to activate before sprinklers in the plane of the membrane activate.  The sprinklers at the deck will then pre-wet the membrane below.  The intended sprinkler spray distribution will be obstructed by the pre-wetted membrane and not penetrate to the seat of the fire below the membrane.  Pre-wetting of adjacent combustibles on the floor below will also be restricted.  In addition, the piping/sprinklers that penetrate the membrane will be pre-wetted.  This scenario can allow the fire to spread below the membrane.  Even if lower level sprinklers activate, these thin membranes may drape down and restrict proper sprinkler spray distribution.


Of course sprinklers with activation temperatures below 300 oF can be used, but the ceiling temperature prior to sprinkler activation is the real question/concern.  Due to the thermal lag of the “fusible link”, the actual temperature at the ceiling can be expected to exceed the listed activation temperature of the sprinkler.  The actual ceiling temperature prior to sprinkler activation is further affected by the distance between the location where the heat plume hits the ceiling and the adjacent sprinklers.  Ceiling temperatures prior to sprinkler activation are also dependent on the growth rate of the fire.


Based on the preceding, it seems reasonable that appropriate fire tests must be developed to ensure that these products do not adversely affect proper sprinkler protection.  Until then, it may be necessary to consider combustible membrane ceilings only on a case-by-case basis with a documented engineered fire protection analysis.



Douglas H. Evans, P.E., FSFPE