Author(s): James Lathrop. Published on July 3, 2014.

Old Test, New Test
Why NFPA 286 testing is preferred for plastics that melt and drip

All the major model codes in the U.S. now require that interior finishes using polypropylene (PP) or high-density polyethylene (HDPE) be tested in accordance with NFPA 286, Standard Methods of Fire Tests for Evaluating Contribution of Wall and Ceiling Interior Finish to Room Fire Growth. This full-scale test is better at determining the hazard of an interior finish—especially with certain plastics that can drip or melt and fall to the floor—than its predecessor, ASTM E84, Standard Test Method for Surface Burning Characteristics of Building Materials, also known as the “tunnel test.”

Although regulated as far back as 1924 in NFPA’s Building Exits Code, the provisions for interior finishes were refined in NFPA 101 in the 1950s, when all interior finishes were evaluated using the tunnel test, developed in the 1940s at Underwriters Laboratories and so called because of the tunnel-like design of the test apparatus. In this test method, all interior finishes, regardless of their intended use in the real world—walls, ceilings, or floors—were tested on the “ceiling” of the apparatus. (ASTM is currently developing a test for floor-mounted specimens as well.) A gas burner was located at one end; a draft was induced down the tunnel, and the test operator followed the flame front using observation panels in the side of the tunnel. The test generated a flame spread index using a relative scale, where fiber-cement board has a value of 0. That index was translated by the model codes into an “A,” “B,” or “C” rating, with “C” indicating the highest level of flame spread permitted. A flame spread index of 0–25, as determined by the ASTM E84 test, is Class A; flame spread of 26–75 is Class B, and 76–200 is Class C. A smoke developed index is also determined to describe how dense, or how black, the smoke is; toxicity and other important data are not measured, recorded, or reported. Designers, specifiers, and other stakeholders should be wary of manufacturers who offer only a given flame spread index but do not provide information on the smoke developed index.

In the 1960s, it was determined that the tunnel test was inappropriate for materials that melt and drip away from the underside of a horizontal ceiling surface, such as exposed cellular or foamed plastics, and that full-scale testing was a more accurate away of assessing the performance of these materials in fire events. Over time, NFPA 286 was developed as a full-scale room-corner test, where the test room is 8 feet wide, 12 feet deep, and 8 feet high. The test has numerous advantages over the tunnel test, starting with the fact that the product being tested is installed as it would be in the real world: wall materials on the wall, ceiling materials on the ceiling. Testing wall materials in a vertical orientation is preferable, especially for plastics, as they drip or flow onto the floor as they would in actual fires.

The NFPA 286 test also features a realistic burn, using a 40 kilowatt gas diffusion fire for the first five minutes, similar to a trash can fire, followed a 160 kilowatt gas diffusion fire for the next 10 minutes, similar to a large trash bag filled with paper towels. The test also yields real-world data versus an arbitrary scale. Data such as heat-release rates, total heat released, oxygen depletion, carbon monoxide, carbon dioxide, and total smoke released are also measured and provided as part of the product information. Information on other gases, such as hydrogen cyanide and hydrogen chloride, can also be determined. Full-scale testing yields better data with some materials, but it’s also more expensive; a tunnel test can be run for less than $1,000, while a room-corner test costs around $5,000, plus the cost of additional test material.

Chapter 10 of the 2012 edition of NFPA 101 states that PP and HDPE materials are not permitted as interior wall or ceiling finishes unless they are tested in accordance with NFPA 286 and that the tests must be performed on a finished assembly with the maximum thickness intended for use.