Laboratory Durability Testing of Preservative-Treated Wood Products

Abstract

Recently, certain European standards have allowed for the classification of the biological durability of chemically modified wood and preservative-treated wood, including treated products, but necessary methods for representative sampling and testing are lacking. Instead of sampling from products that can contain areas of varying durability, this study aimed at testing full-size products. Sections of untreated and preservative-treated terrace decking and palisades were incubated with pure cultures of brown and white rot fungi. Instead of mass loss, the decayed cross-sectional area was determined. The spatial distribution of decay and wood moisture content was investigated. After 16 weeks of incubation, all untreated product specimens showed signs of decay independent of the test fungus. The treated specimens were less affected. The mean and the maximum decayed cross-sectional areas were well correlated, for both the total and the sapwood cross-sections. The wood moisture content after incubation was always favorable for fungal decay, but highest where the specimens were in direct contact with the malt agar. Different infestation pathways became evident: (1) from the sapwood mantle, (2) via radial checks, and (3) from the end-grain. The latter should be prevented in order to better mimic real outdoor exposure conditions.

1. Introduction

Durability classes (DCs) are commonly assigned to wood species [1,2,3,4,5] and sometimes also to treated timbers, such as thermally or otherwise modified timbers [6,7,8]. The basis for such a durability assessment is the result of different field and laboratory tests. The latter can include incubation with basidiomycete monocultures (e.g., EN 113-2 [9]) or exposure to unsterile soil (CEN/TS 15083-2 [10]).

Since its latest revision in 2016, the European standard EN 350 [5] also allows the durability classification of preservative-treated wood and wood products, which consequently include preservative-treated products, such as decking boards, posts, and poles. The basic idea of treating all wood-based materials and products equally appears fair and promising and could provide a transparent assessment scheme. However, the standards lack detailed guidance on the sampling, testing, and classification of wood products [11]. At second glance, there are numerous unanswered questions and pitfalls associated with the durability classification of preservative-treated wood products.

Material testing requires a homogenous matrix [5], while the presence of zones and gradients are typical characteristics of wood products. Therefore, such products need to be sampled in a representative manner, which can hardly be achieved without knowledge about the outdoor performance of the entire product. Furthermore, the latter is affected by the environmental conditions; the risk of weathering and leaching; and the type, retention, and distribution of the preservative itself [12,13,14,15,16,17]. The challenge of taking representative samples from wood products can be described using the example of preservative-treated utility poles made from pine wood (Pinus spp.). Pine roundwood can be divided into at least two zones, the outer non-durable sapwood and the inner heartwood (Figure 1A), which ranges from less durable to durable [18]. The distribution of biological durability within the pole is reversed after a pressure treatment with a wood preservative (Figure 1B). While the outer sapwood is supposed to be very durable, the durability of the heartwood is not changed by the impregnation due to its low permeability and preservative uptake. In principle, there are only two options for taking samples from a treated pole, i.e., (1) random sampling (Figure 1C) or (2) zone-wise sampling (Figure 1D). However, both options cannot reflect the actual exposure of the pole in contact with soil (Figure 1E), where the most critical hazards can be expected from the outer mantle of the pole, which is in direct contact with the soil and the micro-organisms in it. Furthermore, the pole is subject to various abiotic agents before and during exposure. Drying checks can occur before and after the treatment and can serve as entry port for moisture and decay organisms [19,20]. On the other hand, the outer, more durable zone of the pole forms a shell that is supposed to provide protection to the inner core as long as it is intact. The complexity of the latter makes any sampling guaranteed to fail. Keeping the basic configuration of a wood product during the test seems to allow a realistic assessment of its durability.

The aim of this study was therefore to investigate the feasibility and suitability of product testing for a durability classification of treated wood products. Sections of terrace decking and palisades should be incubated with pure cultures of basidiomycetes and ways of determining the durability of the tested products should be explored. The experimental set-up followed as closely as possible the standard procedure for testing the durability of wood against wood-destroying basidiomycetes, according to EN 113-2 [9]. Therefore, malt agar was used as a nutritious medium and all specimens were incubated for 16 weeks, although the volumes of the product specimens under test were larger than the standard specimens’ volume.

2. Materials and Methods

2.1. Wood Products and Preservative Treatment

Terrace decking boards (27 × 140 × 3000 mm³) and palisades (Ø 160 mm, 3000 mm long), both made from Scots pine (Pinus sylvestris), were investigated, as summarized in

Table 1. Treated wood products under test. Data for the entire sample set and data for the sample subset used in biological durability tests (in parentheses) are shown.

Product 1	Number of Replicates	Solution Uptake [kg/m3]	Retention [kg/m3]
Palisade PQ (poor treatment quality)	50	74 ± 28	3.8 ± 1.4
	(10)	(68 ± 22)	(3.5 ± 1.1)
Palisade HQ (high treatment quality)	50	347 ± 57	18.4 ± 3.0
	(10)	(353 ± 41)	(18.7 ± 2.2)
Palisade C (untreated control)	40	-	-
	(10)	-	-
Decking board HQ	60	268 ± 127	14.2 ± 6.7
	(5)	(208 ± 99)	(11.0 ± 5.2)
Decking board C	30	-	-
	(5)	-	-

¹ all products were made from Scots pine (Pinus sylvestris) and contained sapwood and heartwood portions.

. Therefore, the different products were partly impregnated in full size with a commercial water-borne copper-based wood preservative. Before impregnation, the decking boards and half of the palisades had a wood moisture content (MC) below cell wall saturation (approx. 30%, e.g., [14]) and were thus ‘ready to impregnate’. The other half of the palisades had been submerged in a basin for two weeks to increase the MC well above cell wall saturation in order to create a test material of lower treatment quality. The impregnation treatment was conducted in an industrial treatment plant at Fürstenberg-THP GmbH, Hüfingen, Germany. The process consisted of a pre-vacuum (≥150 min pre-vacuum at <25 mbar) and a pressure phase (min. 480 min at >9 bar). The solution uptake and the preservative retention were based on the preservative manufacturer’s recommendations and determined by weighing the palisades and decking boards to the nearest 100 g and 10 g, respectively, before and after impregnation. Data for both measures are summarized in Table 1.

2.2. Specimen Preparation and Incubation

Palisade and board sections of 200 mm length were cut from the products after those were stored for several weeks under roof. The end-grain of the palisade sections as well as those of 50% of the board sections were sealed with polyurethane (Sikaflex 221i, Sika GmbH, Tüffenwies, Switzerland). The palisade specimens were inoculated with the following fungi.

Brown rot:

• Coniophora puteana (Schumach.) P. Karst. (Eberswalde 15), C.p.
• Rhodonia placenta (Fr.) Niemelä, K.H. Larsson & Schigel (FPRL 280), R.p.
• Gloeophyllum trabeum (Pers.:Fr.) Murrill (Eberswalde 109), G.t.

White rot:

• Trametes versicolor (L.) Lloyd. F. (CTB 863a), T.v.

The specimens were incubated in plastic containers (170 × 270 × 235 mm³, H × W × D) filled with 850 mL malt agar (4%) and closed with a lid (Figure 2A–G). The incubation time was 16 weeks. Each container comprised two test specimens and two virulence control specimens (15 × 25 × 50 mm³) either made from untreated Scots pine sapwood or beech (Fagus sylvatica) for brown rot and white rot fungi, respectively (Figure 2F). Twelve inocula were placed on the malt agar, as shown in Figure 2D. During incubation, the containers’ lids were additionally sealed with a stretchable closure tape (Parafilm M, Pechiney Plastic Packaging, Chicago, IL, USA).

In analogy to the palisade specimens, board sections of 200 mm length were prepared and incubated with C. puteana (Figure 3). In total, n = 80 specimens were tested. The end grain of half of the specimens was sealed with polyurethane, and half of the specimens were oven-dried at 103 ± 2 °C after the treatment in order to examine to what extent drying checks affected the fungal decay of the specimens. All specimens were gamma sterilized at a dosage of 29.7 kGy at BBF Sterilisationsservice GmbH (Kernen, Germany) before incubation.

2.3. Decay Assessment

After 16 weeks of incubation, the specimens (Figure 2H) were removed from the containers, cleaned of the adhering mycelium, and cut into seven transverse sections each with a width of approx. 25 mm. The sections were consecutively numbered from one end to the other (#1–7), scanned in a wet state, air-dried, sanded, and scanned again using a flatbed scanner (Epson Expression 11000XL, 300 dpi, Suwa, Japan). The scan images of the dry specimens were saved in jpeg format and used to quantify the decayed cross-sectional area of the specimens with the help of GIMP (2022, The GIMP Team). The number of pixels in a marked area was displayed using a histogram. In order to determine the number of pixels corresponding to the cross-sectional area, the entire cross-section of randomly selected scans was repeatedly marked and the number of pixels corresponding to the area was noted in the histogram. The arithmetic mean was calculated from these and taken as the number of pixels corresponding to the cross-sectional area.

Both the percentage of decayed area related to the total cross-section (Equation (1)) and the absolute decayed area (Equations (2) and (3)) were calculated based on the number of pixels of the total cross-sectional area and the cross-sectional sapwood area. The proportion of heartwood in the cross-sectional area was also determined using this method.

$$P_{decay}=\left(1-\frac{{Pix}_{total}-{Pix}_{decay}}{{Pix}_{total}}\right)\times 100$$

(1)

$$A_{decay}=A_{total}\times {(P}_{decay}/100)$$

(2)

$$A_{decay,sap}=A_{sap}\times {(P}_{decay,sap}/100)$$

(3)

where:

• P_decay is the percentage decayed area (%);
• P_decay,sap is the percentage decayed sapwood area (%);
• Pix_total is the number of pixels, total cross-sectional area;
• Pix_decay is the number of pixels, decayed cross-sectional area;
• A_decay is the decayed cross-sectional area (cm²);
• A_decay,sap is the decayed cross-sectional sapwood area (cm²);
• A_total is the total cross-sectional area (cm²);
• A_total,sap is the sapwood cross-sectional area (cm²).

The mass loss (ML) of the virulence control specimens was determined by oven-drying the specimens at 103 °C until the mass was constant and weighing the specimens to the nearest 0.001 g before and after incubation.

2.4. Determining Wood Moisture Content and Its Spatial Distribution

The wood MC of the palisade sections after incubation as well as its distribution over the cross-section and the length of the specimen was determined on selected specimens. The cut slices #2, #4, and #6 of one specimen per combination of test fungus/material were cut and split into 36 segments, where the outer, central, and inner zones were distinguished, as illustrated in Figure 4. The wood MC of a total of 1.296 specimen segments was determined gravimetrically to the nearest 0.001 g before and after oven-drying at 103 ± 2 °C.

3. Results and Discussion

3.1. Fungal Growth

After two weeks of incubation, the agar plates were fully covered by mycelium of all four test fungi. Control specimens were also partly overgrown. C. puteana and partly G. trabeum grew on treated specimens as well. The PU sealed end-grain faces were also overgrown. T. versicolor and R. placenta did not grow on the preservative-treated specimens until the end of incubation.

After eight weeks of incubation, condensation took place in containers with untreated control specimens as well as with treated specimens inoculated with C. puteana. On the latter specimens, mycelium still spread after ten weeks of incubation. Differences in mycelial growth as well as the condensation of water beneath the containers’ lids are exemplarily shown in Figure 5.

3.2. Fungal Decay

The ML of the virulence control specimens exceeded the validity thresholds according to EN 113-2 [9], while C. puteana and T. versicolor caused mass loss (ML) even above 45 and 60%, respectively (

Table 2. Mass loss of virulence control specimens (validity threshold according to EN 113-2, 2021: minimum mass loss: 20%).

Test Fungus/Wood Species	Mass Loss (%)
	Palisade Test		Decking Board Test
	Mean	SD	Mean	SD
Trametes versicolor/Beech	67.3	11.1	-	-
Coniophora puteana/Scots pine sapwood	56.7	12.7	46.8	19.7
Gloeophyllum trabeum/Scots pine sapwood	25.2	5.3	-	-
Rhodonia placenta/Scots pine sapwood	27.0	7.3	-	-

).

The untreated control palisades were invariably infested by all four test fungi, which coincided with the low durability of Scots pine sapwood against both brown and white rot [5,18]. In

Table 3. Decayed cross-sectional areas (A_decay%) of palisade section specimens (n = 10) after 16 weeks of incubation.

Treatment	Test Fungus	Decayed Specimens [%]	Decayed Cross-Sectional Area Pdecay% [%]
			Total				Sapwood
			Mean	SD	Max	SD	Mean	SD	Max	SD
None	T.v.	100	29.4	7.5	34.4	8.2	46.8	9.9	53.2	11.0
	C.p.	100	51.2	7.9	54.6	4.7	81.6	4.7	84.3	3.6
	G.t.	100	16.9	6.3	21.2	5.9	33.8	13.7	42.0	13.7
	R.p.	100	12.4	5.4	15.5	6.0	18.8	7.8	23.4	8.4
Poor quality PQ	T.v.	10	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	C.p.	100	7.1	5.7	11.1	7.1	12.8	10.0	19.3	12.7
	G.t.	60	0.3	0.6	0.7	1.1	0.4	1.0	1.1	1.7
	R.p.	0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
High quality HQ	T.v.	0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	C.p.	60	0.1	0.4	0.5	1.1	0.1	0.5	0.6	1.1
	G.t.	0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	R.p.	0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

, the number of infested specimens, as well as the mean and maximum decayed cross-sectional areas (P_decay%), are summarized. The latter two measures differed by factor 1.1 and 1.3 for the untreated controls and between 1.6 and 2.3 for the treated sets showing decay (Figure 5). For the sake of better comparability, the mean maximum P_decay% might be the measure of choice, since it closely refers to the maximum depth of decay, which is the decisive measure in field experiments, such as graveyard and lap joint tests [21,22,23]. In addition, the P_decay% was calculated on the basis of the sapwood area (Table 3), which was higher compared to that based on the total cross section by factor 1.3 to 2.0. With regard to both practical applications and the failure criterion, preference should be given to the total cross-sectional area for the calculations. On the contrary, using the sapwood cross-sectional area can help to increase reproducibility and make the test results more independent regarding variations in sapwood percentage and preservative uptake of this more permeable stem zone.

As expected, the decayed cross-sectional area P_decay% varied significantly (i.e., coefficients of variation between 15 and >500%, Table 3). However, the variation of the mean P_decay% was generally more prominent compared to its maximum (by a factor of 1.1 to 2.4, see Table 3 and Figure 6), which again speaks in favor of using the latter. In addition, the mean and the maximum decayed cross-sectional areas were well correlated (Figure 6) for both the total and the sapwood cross-sections. To provide agreement with existing field test standards, such as EN 252 [21] and EN 12037 [22], the maximum value has been given preference.

The preservative-treated palisade sections showed no or very little decay except for the PQ palisades incubated with C. puteana; i.e., up to 11% of the cross-section was decayed. In two other sets, the A_decay% was well below 1%, but in both cases, six out of ten replicates showed slight signs of decay. Nevertheless, the basic result was that the outer preservative-impregnated shell protected the overall product and thus gave it better performance.

Apparently, the decayed cross-sectional area varied partly in terms of the specimens’ length, but such variations did not follow a clear rule, as seen in C. puteana and T. versicolor in Figure 7. For assessing both the mean and the maximum decayed cross-sectional areas, sampling should be carried out along the entire specimen length. In general, the number of studies using different decay assessment methods in comparison is scarce [24]. However, some comparative studies included a visual evaluation of cross sectional specimen areas [24,25,26] and concluded that weak points could be detected more easily compared to mass loss, mechanical loss, or schematized indentation measurements. Similar observations were made when examining decay in standing trees [27,28].

Similar to the palisade specimens, the majority of decking board specimens showed clear signs of decay after 16 weeks of incubation with C. puteana, but different infestation pathways did not become evident, since all specimens contained both sapwood and heartwood sections (Figure 8A,B,D). Only the treated, non-dried, and end-grain-sealed specimens were free from decay (

Table 4. Decayed cross-sectional areas (A_decay) of decking board section specimens (n = 10) after 16 weeks of incubation.

Treatment	Pre-Oven-Drying	End-Grain Sealant	Sapwood Area [%]	Decayed Specimens [%]	Decayed Area Adecay [%]
					Mean	SD	Max	SD
No	Yes	No	38.0	100	71.3	29.5	95.6	12.1
		Yes	35.8	100	50.3	19.5	60.3	17.3
	No	No	34.4	100	41.0	16.6	65.5	20.5
		Yes	35.4	100	25.6	19.7	35.8	21.7
Yes	Yes	No	53.6	100	31.4	20.9	55.3	28.8
		Yes	54.9	90	3.2	5.3	4.9	5.4
	No	No	51.6	60	10.9	16.0	25.8	26.8
		Yes	50.8	0	0.0	0.0	0.0	0.0

, Figure 8C). All untreated specimens were infested, and 90%–100% of the treated and subsequently oven-dried specimens showed signs of decay as well. It became evident that oven-drying had a positive effect on the progress of fungal decay, which might be the consequence of the occurrence of drying checks, but the latter could not be clearly identified at the end of the incubation experiments. In contrast, end-grain sealing had a negative effect on the progress of fungal decay, which is most likely the effect of blocking the axial pathways in wood, which are preferred not only by liquid water, but also by decay organisms such as fungi [29,30]). As previously shown for the palisades, the maximum A_decay was scattered less than the mean A_decay (Table 4).

3.3. Moisture Conditions

The wood MC within the palisade specimens varied throughout their cross-sections after 16 weeks of incubation, as illustrated exemplarily for C. puteana and both treatment qualities in Figure 9. The highest MC (i.e., 35%–68%) was observed at the bottom part of the specimens, which had been in direct contact with the malt agar. This part of the specimens was not decayed, not even within the untreated control specimens, which might be attributed to unfavorably high moisture content or the presence of malt agar. However, the MC range found in this specimen region did not exceed the physiological cardinal points of C. puteana [30,31,32]. In contrast, the adjacent area above the ‘malt–agar line’ showed the most severe decay (Figure 10). The remaining parts of the specimens had MCs of between approx. 22 and 45%, but all specimens contained sections where the moisture conditions were favorable for fungal decay [32,33].

3.4. Infestation Pathways and Decay Patterns

In summary, 63 of 120 specimens showed fungal decay. Three different infestation pathways became evident: (1) from the sapwood mantle (61% of the cases), (2) via radial checks (60%), and (3) from the end-grain (34%). The latter should have been prevented through the end-grain sealant but occurred in a few cases. However, a defect sealant led to infestation via end-grain in less than 7% of the specimens. Further potential reasons for ‘internal decay pockets’ are fungal growth from remote cracks moving into the specimen’s core or the deficient adhesion of the sealant, which might allow the fungus to penetrate the contact face and infest the specimen via the end grain. However, the latter could not be confirmed.

Many specimens were infested via more than one pathway (Figure 11). Similar infestation pathways had been reported from poles and other commodities in service [24,25,34,35,36]. Detailed knowledge about the spatial distribution of wood moisture and fungal decay in complex wood-based products could be gained from laboratory product tests (‘commodity testing’). This might be of interest for understanding infestation pathways and decay development in products containing glued or otherwise connected wooden elements [37,38].

4. Conclusions

After the standard incubation period of 16 weeks, all untreated product specimens showed signs of decay independent from the test fungus. The treated specimens were less affected. The mean and the maximum decayed cross-sectional areas were well correlated for both the total and the sapwood cross-sections. To provide agreement with existing field test standards, the maximum value was given preference. The palisade incubation experiments showed that the wood moisture content after incubation was always favorable for decay fungi, although both the absolute specimen volume and the wood–agar volume ratio were significantly higher compared to standard durability tests. Different infestation pathways became evident: (1) from the sapwood mantle, (2) via radial checks, and (3) from the end-grain. The latter should be prevented to better mimic real outdoor exposure conditions.

The decay testing of products provides information about their performance but cannot be used effectively for durability classification. In contrast to the durability testing of materials, it is essential to retain the composition of a product during the test, e.g., the treated outer protecting sapwood shell and the less durable inner heartwood core. The latter was largely achievable, but the insufficient adherence of sealants and the occurrence of cracks during drying lead to reduced reproducibility; however, this may mimic outdoor exposure scenarios of products in service. The main shortcoming of the approach is the lack of a non-durable reference product. While it might be feasible to produce terrace board sections containing 100% pine sapwood, full-size poles can hardly be made from such reference wood species. Generally, the time and effort for preparing reference products are significantly higher compared to small clear reference specimens. In addition, the effort for determining the decayed cross-sectional areas is significantly higher compared to traditional mass loss measurements, and it is associated with a higher degree of uncertainty, because the evaluation of fungal-induced discoloration of the cross section is subject to interpretation.

Although the results of this study are only preliminary, they suggest that the durability classification of wood should be limited to wood-based materials and should not include wood-based products. The latter could nevertheless be subjected to performance tests in order to gain valuable insights into their outdoor behavior.