Farm Forestry New Zealand

Durability performance and structural timber

Wednesday, November 15, 2017, Dean Satchell's blog

Durability performance and H1.2

What is the level of durability performance that should be expected from framing/structural timber that becomes wet and stays wet? This question could be redefined as "how long should we expect the timber element to last without significant decay in this situation, before it is found and remedied". The regulator (now MBIE) appears to have not provided an answer to this question since introduction of the H1.2 hazard class, this despite New Zealand's performance-based building code that specifies number of years durability required for building elements. For timber framing "a 50 year durability is required", something not achievable for permanently wet framing.

H1.2 boron treatment was introduced to offer "opportunity time" for detecting and repairing leaks. Framing tests undertaken by Scion between 2001 and 2009 resulted in preservative formulations designed to conservatively prevent decay in framing subjected to intermittent wetting (Hedley et al. 2009). Although the concentration of borates required to inhibit brown rot fungi were found to only be "in the order of 0.15-0.25% boric acid equivalent" (Hedley et al 2009), it was decided that because boron potentially leaches in service situations where timber becomes wet, a 0.4% boric acid equivalent level of treatment would be "the cross-sectional retention required to inhibit fungal growth on framing" in the H1.2 specification (Singh et al. 2014a).

This is a simple, clean solution because it provides consumers with security that there is plenty of time to remedy the leak. Perhaps, being satisfied that this conservatively high level of treatment was the only available option, the regulator has never felt it necessary to specify how long framing should last once it gets wet?

To summarise, all we know is that:

• Perishable wood was not good enough;
• If more than enough boron is in the timber, then treated timber is unquestionably good enough.

The Reference material

In New Zealand the required level of durability for materials is set as minimum years in service. Timber framing is required to last for at least 50 years. However, if it gets wet and stays wet, timber framing will not last 50 years, so assessing performance would require some method for verifying durability, a test that measures rate of decay over time and where the test timber either passes or fails.

The building code is clear that tests need to be relevant to field and service conditions. However, as described by Hedley et al. (2010) "While relative rates of decay in laboratory tests are reasonably easy to determine, absolute times under real-life conditions are considerably more difficult". Therefore, decay rates for any test method should simulate real-life conditions, which will inevitably vary. 

So why not just compare relative decay rates for other materials against H1.2 boron treated pine as the reference material? After all, H1.2 radiata seems to have become the gold standard, the de-facto benchmark with which to compare all others. The reason is that there is no performance benchmark defining duration of wetness. The test method would need to meet the requirements as defined in the building code, but there aren't any requirements on how long the wet framing element should last before finding the leak. Perhaps more importantly, because service conditions conducive to decay vary so greatly, in order to meet the requirements in the building code a range of tests would be necessary to establish a set of lab protocols that adequately represents the real-life conditions of a leaky building.

Test methods should simulate real-life conditions, using a range of tests. Initial accelerated decay framing tests were undertaken by Scion between 2001 and 2009 using simulated wall units, exposed to an environment that enhanced decay i.e. 25° C, 95% relative humidity and "periodically wetted" (Hedley et al. 2009). Although this method exposed wood "to a warm and wet environment containing active decay fungi" (Hedley et al. 2009) this does not appear to simulate a worst-case leaky-house scenario at all. Under these test conditions boron penetrated "into the core of most components in a sufficient amount to control decay" and was found to diffuse "through the whole of the cross section" (Hedley et al. 2009). This experimental design aided boron diffusion through the wood, without leaching it. The test method did not actually simulate the scenario where leaching would be sufficient to reduce protection, but held the boron in the wood at high concentrations for long periods of time. I am not aware of any subsequent experiments by Scion that sought to test a range of leak and leach scenarios as required by the building code, although supposedly development of protocols for assessing preservative systems "remains an ongoing activity at Scion" (Singh et al. 2014b).

The Protocols

Despite clear deficiencies in experimental design, the simulated wall unit test methods developed by Scion have been introduced into the Australasian Wood Preservation Committee (AWPC) Protocols for Assessment of Wood Preservatives (2015) as the "Wall frame cavity test" (also called the "simulated wall unit test" by Scion) and the "I Frame sample test". For both tests, the protocols state that samples "shall be periodically sprayed with water to maintain the wood moisture content at a level suitable for decay to progress". No further methods are prescribed for how this should be undertaken or how much water to use, or over what time period, but apparently Scion, who undertake such testing for clients, spray their wall unit tests "with water for a short period each week to simulate occasional rain wetting" (Singh et al. 2014b). Another method used by Scion, the "enclosed tank method" involves the tanks being "regularly opened (usually weekly) so that the samples could be sprayed with water" (Singh et al. 2014b). Given that risk of boron leaching was the clear rationale for the introduction of 0.4% BAE retention required for the H1.2 hazard class, and keeping in mind that the building code specifies that tests need to be relevant to field and service conditions, did the research that led to development of these methods answer the following questions:

• Does "spraying with water" sufficiently address the risk of leaching, or actually only aid diffusion by keeping the wood moist?
• How would different methods of periodically wetting boron treated wood affect the retention of boron?
• What is understood about boron leaching in service conditions and what importance would this have to designing accelerated decay test methods? 

What we do know is that under the test exposure conditions of a "warm and wet environment containing active decay fungi" (Hedley et al. 2009) Scion reported that "periodic wetting helped diffusion into the timber without causing serious loss of boron" (Singh et al. 2014a). The "periodic wetting" in this case involved samples being "intermittently wetted to maintain high moisture content (HMC), greater than 30% moisture content." by being "sprayed with water". Results showed that where timber was surface treated with only 0.22% BAE it was in the same condition after 3 years as timber treated to a retention of 0.42% and 0.65% BAE. Index of condition did not change over the whole 160 week period for all three treatments, apparently because very little leaching occurred from the intermittent wetting. Only with low initial retention levels of 0.1% BAE did the wood progressively decay over time under these test conditions. Importantly, what this experiment uncovered is that concentration of boron determines life of treated timber rather than how wet or how warm the exposure conditions are. Therefore the two key variables that influence rate of decay for boron treated timber appear to be:

1. initial concentration of boron; and
2. rate of leaching.

Of course the next step in scientific enquiry would be to verify or reject this hypothesis with further experiments. Then, if this were indeed the case, methods for assessing durability performance of H1.2 boron treated timber would need to refocus on leaching potential. Unfortunately, it appears that methods currently used by Scion to periodically wet the timber only purport to represent wetting cycles as found in leaky buildings. The reality is that there is much work to do before suitable verification methods become available for clients that offer compliance with the building code for H1.2 situations.

Meanwhile a decade has passed without anybody questioning the validity of Scions test methods, with these methods available to clients wishing to apply for inclusion as an acceptable solution in the building code via NZS 3640 and NZS 3602. What appears to be a biased test method has gone unnoticed allowing H1.2 boron treatment to cement its monopolistic place in timber construction, a position currently reflected in both the building code and complete market dominance of that product. This method also opens the door for envelope treatment because leaching is minimised and diffusion is maximised.

Indeed, weather data across New Zealand suggests that 1 hour of water dripping in and through framing every 3 days is a pretty close approximation of what occurs in service in an enclosed (wet) frame subject to a rainwater leak. Until further experiments establish an appropriate rate of leaching that can be incorporated into design of methods for laboratory tests, the playing field cannot be levelled, with the door wide open for inferior products that also might pass the test regime on offer.

Leaching and boron

So what levels of boron leaching can be expected in a leaking building with wet framing? Is it possible to replicate this in lab conditions and prescribe an appropriate level of leaching? Water flow will be different for every real-life leak situation, as will be resulting levels of decay for boron treated timber. At one extreme is the Scion framing tests and at the other extreme is graveyard testing where the boron quickly diffuses out into the soil and the wood decays. What about the situation where a building leaks, then is eventually repaired, then leaks again in the same place? What about framing in an enclosed situation, saturated with water, and with through-flow of water every time it rains? The only study I am aware of that looked at this used water dosage of 100ml per lineal metre of framing every four days, with the conclusion that on average boron reduced by approximately 30% over the first 6-8 months of the testing (Drysdale et al. 2011). This suggests that further work is necessary on replicating real-life service conditions and understanding the leach scenario.

Questions in my mind relevant to field and service conditions are:

1. Does dripping water induce leaching in wet boron-treated framing?
2. If so, how does rate of dripping and period of dripping affect boron levels in wet framing?
3. How rapidly can boron deplete if leaching does take place?

Assuming that different leaching scenarios will affect the durability performance of boron-treated timber, leaching rate would need to be introduced into the equation before H1.2 boron treated radiata could be the reference material (experimental control) for any comparative lab testing procedure. It should be kept in mind that the rationale for introducing such a high BAE content (0.4%) was as an insurance policy against leaching, a redundancy, not for the specific level of durability performance attained by 0.4% BAE in wood that is "periodically wetted" just to keep it moist. 

It also needs to be kept in mind that alternatives to boron, such as fixed preservatives or naturally durable timber, would have little variability related to water flow, with moisture content and temperature being the important variables influencing decay. 

However, even if a fixed preservative were to be used as the reference material, we should be careful not to put the cart before the horse. The preservative formulation for the control in any test procedure would need to be set at the exact level of treatment that meets predetermined durability performance criteria. In addition to temperature and moisture, this would require:

• a defined duration of wetness (which is the time it should reasonably take to find and remedy a leak); along with
• an index of condition expected after that time.

H1.2 boron treated timber had an index of condition above 8 even after ten years of exposure in Scions accelerated decay test units (Page et al. 2011), so what does this explain? Is the level of treatment excessively high or is the boron just not being leached? Is this material and test method really appropriate as a reference for testing other products against? Other materials that are seeking to conform with the performance-based requirements of the building code?

What level of decay is acceptable to the regulator for a defined duration of wetness, under accelerated decay test conditions? Is this just a subjective call made by Scion staff for client applicants to subsequently present to Standards committees (or Codemark) for approval of their products, or should performance test criteria be set by the regulator? Singh et al. (2014) found that only when the decay condition reached moderately severe (ratings 6 or lower in the AWPA scale, standard E7-93) did timber stiffness suffer.

Natural durability and the Reference material

The AWPC protocols specify test methods for assessing performance of preservative treatments. Leaching is not important when comparing equally diffusible preservatives. There is no issue comparing different species with a radiata H1.2 reference material, each treated to H1.2 boron retention levels. For example H1.2 poplar compared with the H1.2 radiata pine reference material.

There is also no doubt that the AWPC test methods are highly effective in determining relative durability of untreated timber or fixed preservatives, because conditions are highly conducive to decay (i.e., warm & wet). Fixed preservatives might even be compared with naturally durable timbers, but what would be a suitable reference material? H1.2 boron treated radiata pine is unsuitable because leaching potential confounds the results – you can't compare apples with peaches. Untreated radiata developed severe decay in only six months and components were beginning to fail after only 12 months (Hedley et al. 2010), so radiata pine is clearly not a very suitable reference material. Douglas fir heartwood lasted for more than 3 years under these accelerated decay conditions. Could this be the reference material that would satisfy the regulator as being sufficiently durable for framing? Hedley et al. (2010) found that Douglas fir heartwood lasted between 3 and 6 years before decay was unacceptable in the accelerated decay model frame units. Only until a suitable reference material were available that represented the level of durability that the regulator requires, would a reference test be available for naturally durable species and fixed preservative-treated timber. It should also be understood that starting with a fixed preservative retention level that is arbitrary (but more than adequate) and then expecting other products and species to match or exceed that arbitrary performance level is the cart before the horse. 

Although Mick Hedley's original intention was to set up a method for testing materials under accelerated conditions, researchers who subsequently picked up his pioneering methods have never looked beyond the assumption that the leak scenario was adequately represented by "intermittent wetting" as prescribed in the original experimental design. Studies have shown that extensive loss of borates occurs only when timber remains wet throughout its cross-section (Obanda et al., 2008, as cited in Singh et al. 2014a) and where there is a sink for boron migration (Lloyd, 1995) i.e. leaching takes place. Where there is little leaching as per the Scion frame test, 0.4% BAE is sufficient boron preservative to achieve how many years decay prevention? So far more than ten years under accelerated conditions and counting. So has duration of wetness become the horse, with H1.2 treatment being the cart?

The question remains: How many months or years should wet framing remain free of decay before the situation is discovered and remedied? The NZ building code is, after all, performance based and durability is defined as years in service, not a specific level of boron treatment.

References

AWPC Protocols, September 2015
Drysdale, J. Marston, N. Hedley, M. (2011) A Method for Studying Boron Redistribution and Leaching in Timber Framing. IRG/WP 11-20476
Hedley, M. (2005) The decay resistance of Douglas fir, macrocarpa, Lawson cypress and European larch framing. Summary of results after 52 weeks exposure. Report prepared for the Building Research Association of New Zealand. Scion, Rotorua.
Hedley, M. Page, D. van der Waals, J. (2009) Application of Boracol 200rh (Framesaver) to control decay on pre-decayed model frame units. Wood Protection Issue No. 43.
Hedley, M. Page, D. van der Waals, J. (2010) Summary of Tests on Untreated Douglas-fir, Treated and Untreated Radiata pine for Use as Framing in Domestic Construction in New Zealand. Report No. FFR- DS029
Lloyd, J.D. (1995) Leaching of boron wood preservatives - A reappraisal. B.W.P.D.A. Annual Convention 1995.
Page, D. van der Waals, J. Singh, T. (2011) Decay Resistance of Radiata Pine Framing The Condition of Test units after Ten Years Exposure. Scion.
Singh, T. Page, D. Bennet, A. (2014a) Effectiveness of on-site remediation treatments for framing timber. International Biodeterioration & Biodegradation 86, 136-141.
Singh, T. Page, D. van der Walls, J. (2014b) The development of accelerated test methods to evaluate the durability of framing timber. International Biodeterioration & Biodegradation 94, 63-68.

---

Disclaimer: Personal views expressed in this blog are those of the writers and do not necessarily represent those of the NZ Farm Forestry Association.