1.      Why is fire performance testing necessary for decks?

Though there are currently no fire performance codes or ratings for decks, concerns have arisen that decks and other appendages might present a special hazard to structures during wildfires. Particular questions have been raised about the structural integrity and potential combustibility, during fire exposure, of the many new plastic-containing lumber products that are being introduced as deckboard materials.

2.      How did you arrive at the test protocols?

In preliminary under-deck firetests, we exposed redwood, ChoiceDek, and Trex decks to a continuous 80 kW flame until the deckboards collapsed. With ChoiceDek this occurred after about 12 min of exposure, while redwood and Trex collapsed after about 24 min.

However, this kind of test did not permit us to observe possible self-sustained combustion or dripping or dropping of combustibles.  Such an extended exposure was also considered unrealistic in most UWI fires.

Therefore, the final protocol was designed with a shorter, 3-min 80 kW flame exposure [References].

Tests were terminated when any of four conditions were met:

• flaming combustion stopped;
• a deckboard collapsed;
• flaming combustion accelerated rapidly;
• 40 minutes elapsed.

A burning-brand fire test was also developed, since in a real UWI fire, a deck could be ignited not only by burning flammable materials growing or stored under the deck, but also by burning materials falling onto the surface.  In preliminary testing with burning brands placed on a deck, but without wind, there was very poor reproducibility in the rate of charring. However, a protocol based on the ASTM E108 method for roof tests, combining brands on the deck surface with controlled air velocity, appeared to give meaningful results and was adopted for our tests.

An ASTM standard for testing various physical properties of plastic lumber decking is currently being prepared. The fire test portion being considered for this document incorporates the small "C" brands used in ASTM E108. The intent of the "C" brand is to simulate charcoal briquettes that could land on a deck if a grill tipped over. Our reasoning for selecting an "A" brand for our tests was simply that if an "A" brand exposure is deemed necessary for testing roof coverings on homes in the UWI, the same exposure should apply to decks.

3.      How were the 15 deckboard materials selected?

The overall project requires us to test not just deckboards but roofs, windows, and walls as well. This broad scope, with limited time and budget, made it impossible to acquire and test every brand of material for each of these building components.  We therefore selected materials that represented the range of types available on the market.  We especially wanted to test a wide variety of plastic-containing deckboards, but since many decks in California are still constructed of 8/4 (38 mm) Construction Heart redwood, it was decided to also include this material in the tests.

As other comercially available deckboard materials are tested, results will be added to the web pages.

4.      What materials properties most influence deckboard fire performance?

We considered several physical and chemical properties of samples of the deckboard materials (see Table 1) with the objective of gaining insight into fire performance. Though more study is needed, the following tentative conclusions are suggested:

Form:  All of the deckboard materials with a “channeled” form on the underside (Eon, ChoiceDek and TimberTech) had early degradation effects in the under-deck fire tests, presumably from the increased surface area. In the burning brand tests, this early degradation was not seen.

On the other hand, all of the "hollow" construction products, which generally performed well in the under-deck tests, exhibited board collapse within the 40-minute test period in the burning brand tests. This was no doubt due to the reduced thickness of the upper surface, since burn-through occurred in the hollow core areas, where flaming was sustained until the board collapsed.

Thermoplastic material:  Although the polyvinyl chloride sample, EverNew, collapsed very quickly in the under-deck fire test, it did not exhibit sustained combustion, as did most of the polyethylene-based products. Eon, which appeared to be an ABS-like material, underwent very rapid and intense runaway combustion, as did the all-polyethylene Maxituf.

5. Where was combustion concentrated in the under-deck fire tests?

In deck construction, gaps between deckboards are necessary to prevent standing water and to allow a deck to properly ventilate.  Our test decks were constructed with a typical between-board spacing of 5 mm (³/₁₆ in.)

During fire exposure, virtually every product initially began burning where burner flames penetrated through the deckboard gaps. In the cases where flaming combustion was self-sustained after the burner was turned off, it was generally concentrated at these gaps.  This is due to a bootstrapping process where the facing edges are mutually heated to sustain combustion.  Some materials [WeatherBest and Bedford (reinforced)] did not show this effect due to rapid blocking of the gaps with ash or char.

Most materials showed more sustained combustion at the open front edge of the deck assembly than elsewhere, though this had little effect on the degradation results.  In the case of Trex, however, the small amount of falling burning chunks that was observed in the flame impingement test was shed solely from the exposed front edge.

Ends of the deckboards were generally shielded from the burner flame by the joists, but on occasion, fire spread under or around the joists, igniting the exposed ends. In such cases, we discounted negative degradation effects that were solely due to the burning ends and repeated the test. On the other hand, if deckboard ends are left exposed in an installation, ignition of the core material in some products could make them more vulnerable to degradation. 

6.      Why are there no performance recommendations?

Our tests were conducted to determine the behavior of the materials under fire exposure, not to judge their suitability for a particular end use. How and where these materials are to be used is dependent on applicable codes and regulations. Also, since the composition of the deckboards is proprietary and subject to change by the manufacturers, there is no certainty that a material obtained currently will have the same fire performance as the sample we tested. 

7.      Are pictures of the actual tests available?

All tests were videotaped and most had still photos taken. These tapes were used to verify our direct observations. Any person interested in obtaining a copy of tapes of particular tests may do so with payment to cover the cost of copying, handling, and postage. E-mail UCFPL for ordering information.