Water Vapour Resistance of Exterior Coatings—Influence on Moisture Conditions in Ventilated Wooden Claddings

Abstract

Increasing climate fluctuations and extremes due to climate change are particularly concerning for wooden building envelopes, especially in the Nordic region, which has harsh climatic conditions. The exterior coating’s barrier properties are crucial for maintaining building envelopes’ intended lifespans. Hence, it is unfortunate that the vapor resistance of exterior coatings is not openly accessed for commercial products. This study investigates the influence of the water vapour resistance of exterior coatings on the moisture conditions and mould growth risk of ventilated wooden claddings. The $s_d$-value (vapour diffusion-equivalent air layer thickness) is determined for nine free-standing coatings (alkyds, emulsions, and acrylics); in total, 100 specimens are tested with the wet cup method. Additionally, with WUFI Pro, one-dimensional hygrothermal simulations under Nordic climatic conditions investigate how the coatings’ vapour resistance might influence the moisture dynamics of wood. The mould risk is assessed by the add-on WUFI VTT Model. The determined $s_d$-values for the coatings range from 0.453 to 1.350 m (three layers) and from 0.690 to 2.250 m (six layers), showing a strong correlation with the dry film thickness. The vapour resistance of the coatings does not significantly influence the wood moisture content, but lower resistance may cause slightly faster drying. The importance of surface treatment is confirmed. The mould risk is notably higher in a Stavanger climate on a southwest-facing wall compared to Trondheim on a north-facing wall.

1. Introduction

The hygrothermal performance of buildings is affected by urban and local microclimates, considering the impacts on façade surfaces and surrounding air temperatures [1]. Due to global climate change, an increasing frequency and intensity of extreme weather and climate events is expected [2], as well as increased fluctuations in moisture levels, which is likely to implicate the degradation of the building envelopes that are especially vulnerable due to being directly exposed to the climate agents [1,3]. The Nordic region already experiences a particularly harsh climate, with strong seasonal variations, driving rain, freeze–thaw cycles, strong winds, etc. For this region, the anticipated climate change implies even more precipitation and rising temperature in the years to come [4], which will put a stronger demand on our building constructions.

Historically, wood has been an important construction material in Nordic countries due to its simple production and the local availability of raw materials [5]. Wood, as a building material, is of ever-rising global importance considering the necessary implementation of climate change adaptation strategies, being favoured especially for its versatile characteristics [5] and low carbon footprint [6]. However, increasing weather fluctuations are creating unfavourable circumstances for building with wood [7,8,9] as wood is a hygroscopic material highly affected by the surrounding moisture conditions, as well as vulnerable to attacking microorganisms. Mould growth criteria involve the temperature, RH, exposure time and wood moisture content [10,11,12]. The wood moisture content is typically stated as a percentage of the wood’s dry state mass in “weight-%” [13]. The air temperature, global radiation, wind velocity, and driving rain are the most significant contributors to the fluctuating moisture content in wood [14]. It is recognised that 75% of building defects are related to moisture, leading to problems such as mould growth, rot, swelling, and salt migration within the materials [15]. Hence, it is essential to produce designs and conduct good quality maintenance to withstand the future climate exposures and ensure the intended material lifespans.

The executional quality of the building envelope may be varied. The principle of “two-stage weatherproofing” or “dual-barrier weatherproofing” is common in Scandinavia and recommended for façades exposed to the harsh Nordic and coastal climates, as thoroughly described in several studies, i.e., [16,17,18]. In short, this is a ventilated façade system featuring an outer cladding separated from the wind barrier by an air cavity. It is primarily designed for avoiding driving rain penetration while still facilitating proper drainage and drying of embedded moisture that may have entered. The principle is a prerequisite for the use of external wooden cladding in Nordic climate as it allows drying out of the cladding from the rear side.

The most commonly used wood-type exterior cladding material in Norway is spruce (Picea abies) [19,20,21], among other reasons because it absorbs less moisture and adheres better to paint compared to other wood types (i.e., pine) [5,21,22]. Maintenance and treatment of the outdoor wood surfaces (e.g., by application of paint) are necessary to protect against harsh climate conditions that may cause biodeterioration [5,14,23]. Coatings on exterior wooden claddings not only provide an aesthetic appeal but also form a protective barrier of the wood [8,24,25,26] when executed and maintained properly. The protective properties of exterior coatings are addressed in several studies, e.g., [24,26,27,28].

The ever-reported issues of rot damage to wooden claddings in Southwest Norway may be attributed to multiple factors, i.e., climate change, changes in the quality of wooden cladding, alterations in building practices, better insulated walls, reduced maintenance, and variations in surface coating quality. Commercial exterior coatings are typically proclaimed by manufacturers to be “rainwater impermeable” with the ability to withstand the Nordic weather conditions. However, their water vapour resistance performance is not addressed in the technical declaration sheets. A question arises whether these exterior coatings might be so vapour impermeable that they potentially impede the drying of embedded moisture in wooden claddings, and thus, accelerate the biodeterioration processes. In this context, a common way to assess coatings’ resistance to water vapour diffusion is by the equivalent air layer thickness, $s_d$-value, which resembles the thickness in m of a motionless air layer with the same water vapour resistance as the material [29].

The existing literature recognizes factors influencing the moisture protection of coatings such as the binder type, dry film thickness, surfactants, weathering, and the age and condition of the paint film [8,22,30,31]. Hu et al. [32] tested highly permeable films with the wet cup method and found that the water vapour transmission rate depends on the film thickness. Brito et al. [33] found that the coating’s impact on substrate drying significantly varies with the substrate’s moisture content. A substrate is the surface to which a coating material is applied [34]. Gezici-Koc et al. [35] found that the interaction between wooden substrate and coating did not affect the water vapour permeability, yet contradictory conclusions were drawn by Yaseen and Raju [36], who found that the choice of substrate significantly affects the films’ water vapour permeability. The study by Hysek et al. [31] revealed that water vapour uptake was affected by the film thickness and coating composition. Topçuoǧlu et al. [27] found that the moisture barrier properties of acrylic coatings decrease with a decreasing binder content.

Some studies explicitly measuring or mentioning the $s_d$-value for coatings were identified. According to Table 5 of NS-EN ISO 10456:2007 [37], the $s_d$ is 0.1 m for emulsion paints and 3 m for glossy paints, with no mention of the layer thicknesses. Table B1.3 of Geving and Thue [38] reports an $s_d$-range of 0.049 to 0.49 m for alkyds with two layers. The wet cup measurements of combined coating and spruce board by Geving et al. [23] indicate an $s_d$-value of 2.5 m for oil-dilutable coatings and of 1.3 m for waterborne acrylic/alkyd, albeit with large uncertainties due to few specimens. Nore and Hundhausen [39] used in their study an alkyd coating with a measured $s_d$ of 2.8 m and an acrylic coating with an $s_d$ of 1.6 m; however, the measurement method remains unexplained. They concluded that surface treatments did not influence the hygrothermal performance. Grüll and Truskaller [30] tested both the liquid- and water vapour permeability of coatings on wood substrate and reported a strong correlation of these properties. Šadauskiene et al. [40] measured that $s_d$-value of coatings on exterior render façades should be less than 0.6 m to allow proper drying. The results also showed an increasing $s_d$ for an increased number of layers (i.e., film thickness), yet not linearly. Of these, acrylic coatings were the least vapour permeable of the different types. Tomren et al. [41] measured the $s_d$ of ceiling coatings on gypsum substrate, finding a range between 0.1 and 0.5 m depending on the number of paint layers. Conclusions were drawn that the $s_d$ did not linearly increase with the number of layers. Note that the substrate types of the abovementioned studies vary beyond solely wood, which is due to the limited number of identified studies focusing exclusively on coatings for wood.

The review of the existing literature confirms the anticipation of some knowledge gaps in determining the vapour resistance (in terms of the $s_d$-value) of exterior coatings on wood. The vapor resistance of exterior coatings is not openly accessed for commercial products, at least not in the Norwegian market. How these coatings’ diffusion properties may influence the moisture conditions of wooden panels in Nordic climates should be studied as well.

The current study aims to fill these knowledge gaps by determining the vapour resistance ($s_d$-value) for various commercial coatings for exterior wooden claddings and by investigating to what extent this vapour resistance affects the moisture-drying conditions of wooden claddings set in two Nordic climates. The following research questions have been formulated:

• What is the range of the water vapour resistance $s_d$ of exterior coatings?
• How does the water vapour resistance of coatings affect the moisture-drying conditions and mould growth risk of wooden claddings?

The novelty of this study is, thus, the mapping of the water vapour resistance of exterior commercial coatings for wooden cladding. However, the limitations of this study include reporting only the water vapour resistance in terms of the $s_d$-value by the “wet cup method” in NS-EN ISO 12572 [29], disregarding other characteristics, i.e., surfactants, contact angle, and liquid transport. The test did not involve cyclic weathering exposure. This study is confined to investigating the $s_d$ of free-standing films, excluding any possible influence of the interaction with the wood substrate. No other coating properties than the $s_d$-value and the film tickness were controlled. Secondly, the hygrothermal simulations are one-dimensional. The simulations are based on two case buildings in Norway, with some parameters chosen to reflect more realistic environmental conditions rather than those that typically would produce more extreme moisture conditions. The mould growth model used is highly sensitive to the end-user’s parameter choices.

2. Materials and Methods

2.1. Laboratory Measurements

Laboratory measurements for determining the water vapour resistance of commercial coating products intended for exterior wood have been carried out at the Laboratory of Building Physics at the Institute of Civil and Environmental Engineering at NTNU. The wet cup method, as specified in NS-EN ISO 12572:2016 [29] and in detail for paints and varnishes in NS-EN ISO 7783:2018 [42], has been applied for determining the properties describing the coatings’ abilities to resist water vapour diffusion. A total of 110 specimens were comprehended. Paint sampling and test conditioning were performed in October to December 2023, while test assembly and weighing procedure were conducted January to May 2024 for all coatings apart from the following:

• No. 53, which was fully tested in the fall of 2023 as a subject for method evaluation in the preliminary project.
• No. 56, which suffered substantial leakage and therefore was resampled and conditioned anew in the spring of 2024.

2.1.1. Materials

Nine commonly available commercial products were chosen for their relevance in the Norwegian market of exterior coatings for wooden claddings that are advertised as being “impermeable to rain” and “withstanding harsh Nordic weather”. The coatings were then purchased from two local consumer-oriented hardware stores, and they are detailed along with the technical specifications in

Table 1. Overview of the tested coatings. Each product consists of one series of three layers and one series of six layers, labelled -B and -B2, respectively.

Coating		Layers	Binder Type	Bulk Density [kg/m3]	Additional Film Support **
51	B	3	Alkyd resin	1260	Metal grid
	B2	6			Metal grid
52	B	3	Emulsion (alkyd/acrylic)	1260	Metal grid
	B2	6			Metal grid
53	B	3	Emulsion (alkyd/acrylic)	1230	-
	B2	6			-
53X *	B	3	Emulsion (alkyd/acrylic)	1230	-
	B2	6			-
54	B	3	Acrylic resin	1200	-
	B2	6			-
55	B	3	Acrylic resin	1200	Metal grid
	B2	6			Metal grid
56	B	3	Emulsion (alkyd/acrylic)	1200	Plastic cylinder
	B2	6			-
57	B	3	Acrylic resin	1250	-
	B2	6			-
58	B	3	Acrylic resin	1200	Plastic cylinder
	B2	6			-
59	B	3	Acrylic resin	1260	-
	B2	6			-

* The “X” signifies the second test of an already tested coating. ** Explained in Section 2.1.2.

. All the coating products are waterborne, semi-glossy, and white-hued. The binder type varies between alkyd resin, acrylic resin, and emulsion of acrylic/alkyd. Three and six layers were approximated to resemble, respectively, one and two rounds of surface treatment. The intent was to account for uncertainties in estimating the number of previous coating layers and the eroding effects of weathering.

2.1.2. Test Preparation and Assembly

The coating application process (e.g., application method by brush, uniform direction, consistent repaint intervals, etc.) was executed by a skilled laboratory technician and accords with NS-EN 927-10:2019 [43] and the manufacturers’ specifications. The re-paintable intervals between the layers were sufficient to allow proper drying. Each product had a separate brush and stirrer to prevent sample contamination. A Teflon film substrate was used to allow for easy detachment [42]. For each product, a total of twelve specimens were painted, six of which were of three layers and six of which were of six layers, then cured for seven days in freely circulating air, before being punched out to the circular specimens of 0.174 m diameter, see Figure 1. After which, they were stored in the controlled test enclosure for at least 28 days [42] until further cup assembly.

The Teflon substrate was removed after the curation time, and the dry film thickness was measured with a thickness gauge of 0.01 mm accuracy.

In total, ten specimens of each product (five with three layers, five with six layers) were assembled. Molten wax and a ring template were used for sealing, in accordance with Annex B of NS-EN ISO 7783:2018 [42]. The specimen was sealed to the rim of a metal cup containing a saturated aqueous solution of KNO₃. This liquid was poured until it measured 10 mm below the specimen surface. The KNO₃ creates a relative humidity of 94% RH inside the test cup, thus a lower partial vapour pressure $p_v$ than in the surrounding test enclosure [42]. Note that the diameter of the test surface after the sealing is changed to be 0.164 m.

Some coatings were too elastic to allow for self-support and needed a stabilizing support to provide enough stiffness to seal them on the cups’ rims, see Figure 2. This is not part of the standardised method and may yield uncertainty in the measurements. Table 1 displays the coatings to which this applies. For some specimens, the additional cup support created difficulties in sealing tightly. Though, upon any suspicion of leakage during the test, the specimen was sealed promptly.

2.1.3. Test Procedure

During testing, the cups were stored in the conditioned test enclosure of (23 ± 2) °C and (50 ± 5) % RH, as shown in Figure 3.

The difference in the partial vapour pressure across the specimen, $\mathsf{\Delta }{p}_v$, is the driving force of the vapour flow (from higher to lower partial pressure) [29,44]. As a result of this rate of vapour flow through the specimen, periodic weighings are conducted to determine the rate of mass change over time. From this, the water vapour transmission rate through the specimen, $V$, is found, and finally, the diffusion-equivalent air layer thickness, $s_d$, in meters is calculated using Equation (1):

$$s_d=\frac{{\delta }_a \times \mathsf{\Delta }{p}_v}{V}$$

(1)

where ${\delta }_a$ is the water vapour permeation coefficient of air at standard temperature and pressure [42].

It is worth mentioning that the $s_d$-value is only one of several ways to express the results of the wet cup method.

The first weighing occurred 24–48 h post-assembly to accommodate the initial stabilization of water vapour transmission. The time interval for the subsequent measurements was then set at 24 h, based on the expected range of the product’s $s_d$-value, which was between 0.5 and 2 m [29,42]. The weight measurements were carried out using a high-precision balance with an accuracy of 0.001 g, which was placed inside a plastic chamber to prevent disturbances from the surroundings, see Figure 4. The balance was calibrated with a control weight of 1000 g before each individual weighing.

The test concluded when the mass change across the last five consecutive measurements remained constant within ±5% of its average rate of change [29]. The temperature (T), relative humidity (RH), and partial vapour pressure $p_v$ of the test enclosure were monitored hourly by sensors and logged in the calculation spreadsheet alongside the weight measurements for the calculations.

The results are corrected for the surface heat resistance above the sample, the vapour transport through the overlapping zone, and the resistance of the air layer in the cup. The reader is referred to the two abovementioned standards for other specific details.

2.2. Hygrothermal Simulations and Mould Growth Risk Assessment

2.2.1. WUFI^® Pro

The simulated moisture conditions of the considered cladding construction have been produced with the one-dimensional software WUFI^® Pro 6.7 by the Fraunhofer-Institut für Bauphysik [45], which is validated in several studies.

The simulations are based on the climatic conditions of two case buildings, Fjogstad-Hus and ZEB-Laboratory, located in Stavanger and Trondheim, respectively, in Norway, previously analysed by Ingebretsen [46] and Ingebretsen et al. [47]. One sensor from each building was selected based on Ingebretsen’s results: SW1 for Fjogstad-Hus, due to the extreme moisture conditions (southwest orientation), and MN5 for ZEB-Laboratory, as a reference line. Note that MN5 is on the north façade, which was the only wooden one. A southwest orientation is more exposed to driving rain, while north usually reflects the poorest winter drying conditions due to lack of solar radiation [5].

Table 2. Input parameters dependent on the case building.

Parameters Dependent on Case Building	Fjogstad-Hus	ZEB-Laboratory
Outdoor climate	Stavanger	Trondheim
Sensor	SW1	MN5
Indoor climate “Sine Curve”	Curve fit of SW1 data	Curve fit of MN5 data
Wall orientation	Southwest	North
Building height	Short building (<10 m)	Tall building (10 to 20 m)
Driving rain coefficients	R1 = 0, R2 = 0.07	R1 = 0, R2 = 0.1

shows the parameters that depend on the climate and orientation of the case building.

As this study aims to investigate realistic climate boundary conditions, Typical Meteorological Year (TMY) weather files were used for the outdoor climate [48,49,50], and the cavity microclimate represents the indoor climate boundaries. To use WUFI’s “Sine Curve” option, least-square sinusoidal regression [51] was performed to estimate the best-fit sinusoids for the temperature (T) and relative humidity (RH) data of the cavities. To account for the most humid periods, the RH curve’s vertical offset D was eventually shifted by +5 offset. A thorough methodology and uncertainty assessment of the sinusoid-fitting are available in the referenced master thesis, see the Data Availability Statement. The resulting parameters of the four fitted sinusoids that are representing the cavity microclimate in the simulation are presented below in

Table 3. Sinusoid-fitted model parameters for WUFI’s indoor climate option “Sine Curve”.

Cavity Model	WUFI Parameter “Sine Curve”	Sensor SW1	Sensor MN5
Temperature T (best-fit)	Mean	9.89	7.6
	Amplitude	7.2	10.43
	Date of max	28th of July	20th of July
	R2	0.6751	0.7058
Relative humidity RH * (offset shifted +5)	Mean *	83.77	75.05
	Amplitude	9.2	12.54
	Date of max	31st of December	16th of December
	R2	0.1859	0.3907

* The shift of +5 offset only affects the “mean”, which resembles the parameter D of the sinusoid-fit.

.

Most parameters are set as constants throughout the simulations, while others are varied to understand how different parameters and combinations might influence the moisture and mould conditions. The parameters and variations are presented in

Table 4. Input parameters and variations for the WUFI Pro simulations.

Parameters	Setting	Variation(s)
Monitor position	Interior side of cladding
Cladding material	Scandinavian spruce transversal direction II
Wall inclination	90°
Initial wood moisture content	20 wt%	15 wt%
$\mathrm{Coating}’\mathrm{s}\mathrm{vapour}\mathrm{resistance}{s}_d$	2.25 m (maxB2 *)	0.453 m (minB *); 1.35 m (maxB *); 0.69 m (minB2 *); untreated
$\mathrm{Exterior}\mathrm{surface}\mathrm{heat}\mathrm{resistance}{R}_{se}$	0.0588 (m2K)/W
$\mathrm{Interior}\mathrm{surface}\mathrm{heat}\mathrm{resistance}{R}_{si}$	0.125 (m2K)/W	0.0588 (m2K)/W
Short wave radiation absorptivity	0.4 (white hue)
Ground short wave reflectivity	0.2
Adhering fraction of rain	0.7
Simulation start	October (beginning of wetting season)	April
Simulation duration	Five years

* B and B2 signify the number of layers.

.

The modelled component consists of the coating and the wooden cladding material. The cladding is 22 mm Scandinavian spruce with a density of 390 kg/m³. Initial attempts to insert the coating as merely an $s_d$-value did not successfully include the liquid transport properties. The coating was therefore modeled as a separate material layer, utilizing the liquid transport properties of a waterborne acrylic coating found in the built-in WUFI material database [45]. Personal measurements of the vapour resistance $s_d$ were converted to the vapour resistance factor $\mu$ and included, together with the respective bulk density (see Table 1), in the modelled component layer. Note that the coating needed to be inserted with a thickness of 1 mm due to numerical convergence reasons and that µ is adjusted accordingly.

The interior surface heat resistance $R_{si}$ on the inside of the cladding was varied between WUFI’s standard exterior value $R_{se}$ of 0.0588 (m²K)/W and the standard interior value of 0.125 (m²K)/W due to unknown cavity ventilation. However, a realistic value for the ventilated façade considered in this study is expected to lie within this range, though possibly closer to 0.125 (m²K)/W (which is also anticipated to yield more conservatively).

2.2.2. WUFI^® Mould Index VTT

The simulated T and RH conditions on the interior side of the cladding were evaluated for the mould growth risk using the additional plugin WUFI^® Mould Index VTT originally developed by Hukka and Viitanen [52]. The VTT Model is thoroughly described in several studies, e.g., Ojanen et al. [53] and Vereecken and Roels [54]. In short, it generates a mould growth index ranging from 0 to 6. The model considers factors such as the humidity and temperature, exposure time, decline during unfavourable conditions, and material qualities. Risk levels are indicated using a traffic light interpretation: green for acceptable risk, yellow for further evaluation needed, and red for unacceptable risk requiring action. However, the thresholds of the mould risk vary by structure and so-called “occupant exposure class”. This study evaluates spruce material stationed outdoors and is not expected to impact the occupants directly. Thus, the risk level thresholds in the current study are limited to solely green (0 to 3) and yellow (index 3 to 6), never showing any level of unacceptable risk [52].

The parameters chosen for simulating the planed wooden cladding are shown in

Table 5. Parameter settings in the WUFI Mould Index VTT model.

Parameter	Chosen Setting
Occupant exposition class	No impact on occupants expected
Material category	Wooden or natural materials
Material sub-category	Untreated pine or spruce (heartwood)
Sensitivity class	Sensitive
$\mathrm{Decline}\mathrm{coeff}.(C_{mat}$)	Almost no decline (0.1)
Type of surface	Planed
Type of wood	Softwood

. The sensitivity class and decline coefficient regarding the evaluated spruce material are based on the literature by the model developers [52,53,55], as well as on affirmative studies [46,54,56], for a realistic yet conservative model set-up. For the untreated wood cases, the sensitivity class “very sensitive” and mould decline coefficient “Significant decline (1.0)” were set [55].

In total, there are 64 different possible combinations of the parameter value choices in the VTT model: four values of class, two of surface, two of wood type and four values of $C_{mat}$ [56]. The end-user’s choices of input parameters can significantly impact the results [9,56]. A sensitivity analysis should be conducted to identify parameters affecting the mould growth risk, though this is outside the scope of the current study and should be explored in future research.

3. Results

3.1. Laboratory Measurements

3.1.1. Range of Water Vapour Resistance for Exterior Coatings

The measured dry film thicknesses and vapour resistances ($s_d$-values) of the wet cup experiments are presented in

Table 6. Measured $s_d$-values for the coatings and dry film thicknesses. Differentiated by the number of layers. Standard deviation of the mean in brackets.

		Three Layers (B)		Six Layers (B2)
Coating	Binder	Dry Film Thickness [mm]	${\mathit{s}}_{\mathit{d}}$-Value [m]	Dry Film Thickness [mm]	${\mathit{s}}_{\mathit{d}}$-Value [m]
51	Alkyd	0.34 (0.02)	1.350 (0.083)	0.56 (0.02)	2.250 (0.048)
52	Emulsion	0.28 (0.01)	0.836 (0.042)	0.48 (0.02)	1.280 (0.020)
53	Emulsion	0.26 (0.01)	0.669 (0.031)	0.52 (0.02)	1.090 (0.049)
53X	Emulsion	0.25 (0.01)	0.602 (0.026)	0.42 (0.02)	0.925 (0.047)
54	Acrylic	0.28 (0.01)	0.787 (0.037)	0.41 (0.02)	1.110 (0.058)
55	Acrylic	0.26 (0.01)	1.090 (0.027)	0.41 (0.02)	1.530 (0.065)
56	Emulsion	0.25 (0.01)	0.453 (0.024)	0.45 (0.01)	0.690 (0.022)
57	Acrylic	0.24 (0.01)	0.999 (0.051)	0.44 (0.03)	1.480 (0.063)
58	Acrylic	0.20 (0.01)	0.472 (0.024)	0.45 (0.02)	0.932 (0.019)
59	Acrylic	0.26 (0.01)	0.559 (0.006)	0.38 (0.01)	0.724 (0.023)

. The thickness reported in Table 6 is the mean of five random measurements for each specimen of a series. Official laboratory reports for each tested series can be found in the Data Availability Statement. The measured range of $s_d$-values for the B-series was from 0.453 (0.024) m for the most vapour permeable to 1.350 (0.083) m for the least vapour permeable. For the B2-series, the range was 0.690 (0.022) m for the most vapour permeable to 2.250 (0.048) m for the least vapour permeable. The standard deviations of both the $s_d$-values and dry film thicknesses are small, which suggests precision and consistency in terms of the measurement techniques and instruments used.

3.1.2. Correlation of Dry Film Thickness and Vapour Resistance

In Figure 5, the linear correlation between the measured dry film thickness and the determined $s_d$ for the nine coatings are shown. For each product, both series (B + B2) are displayed since they represent a greater span of thicknesses for the same coating. The coefficients of determination R² are provided in

Table 7. Coefficients of determination for the tested coatings.

Coating	51	52	53 + 53X	54	55	56	57	58	59
R2	0.9345	0.9254	0.9113	0.8088	0.9827	0.977	0.934	0.9734	0.8926

. The high R² values indicate that 81 to 98% of the variance in the $s_d$-value can be explained by the dry film thickness, suggesting a strong linear relationship for all the coatings. Higher dry film thickness generally results in a greater resistance against vapour diffusion.

3.1.3. Effect of Binder Type

The measured values in Table 6 suggest a relationship between the binder type and the vapour resistance of the coatings, as depicted in Figure 6 below. The alkyd coating no. 51 exhibited the higher vapour resistance compared to emulsions and acrylics; this is assessed further in the later discussion. However, the difference between the acrylics and the emulsions seems to be negligible in terms of the $s_d$-value.

3.2. Simulated Moisture Contents and Mould Growth Risk

3.2.1. Parametric Study and Mould Growth Indices

The results, see

Table 8. Parametric study and resulting yearly mould growth indices of the Fjogstad-Hus cases, where green colour indicates acceptable mould growth risk, yellow that further evaluation is needed.

Input Parameters		Case
		1	2	3	4	5	6	7	8	9	10	11	12	13	14
Outdoor Climate	Trondheim
	Stavanger	x	x	x	x	x	x	x	x	x	x	x	x	x	x
Cavity Climate Model	MN5 (T0, RH+5)
	SW1 (T0, RH+5)	x	x	x	x	x	x	x	x	x	x	x	x	x	x
Vapour Resistance $s_d$ [m]	2.250 (max-B2)	x					x					x		x	x
	0.690 (min-B)		x					x
	1.350 (max-B2)			x					x
	0.453 (min-B)				x					x			x
	Untreated					x					x
Initial Wood MC [wt%]	20	x	x	x	x	x	x	x	x	x	x			x	x
	15											x	x
Simulation Start	1st of October	x	x	x	x	x	x	x	x	x	x	x	x
	1st of April													x	x
$R_{si}$ [m2K/W]	0.125						x	x	x	x	x			x
	0.0588	x	x	x	x	x						x	x		x
Yearly Mould Growth Index [-]		2.64	2.63	2.64	2.63	4.54	3.30	3.28	3.30	3.28	5.20	2.63	2.63	3.30	2.64

and

Table 9. Parametric study and resulting yearly mould growth indices of the ZEB-Laboratory cases, where green colour indicates acceptable mould growth risk.

		Case
Input Parameters		15	16	17	18	19	20	21	22	23	24	25	26	27	28
Outdoor Climate	Trondheim	x	x	x	x	x	x	x	x	x	x	x	x	x	x
	Stavanger
Cavity Climate Model	MN5 (T0, RH+5)	x	x	x	x	x	x	x	x	x	x	x	x	x	x
	SW1 (T0, RH+5)
Vapour Resistance $s_d$ [m]	2.250 (max-B2)	x					x					x	x	x	x
	0.690 (min-B)		x					x
	1.350 (max-B2)			x					x
	0.453 (min-B)				x					x
	Untreated					x					x
Initial Wood MC [wt%]	20	x	x	x	x	x	x	x	x	x	x	x			x
	15												x	x
Simulation Start	1st of October	x	x	x	x	x	x	x	x	x	x		x	x
	1st of April											x			x
$R_{si}$ [m2K/W]	0.125						x	x	x	x	x			x	x
	0.0588	x	x	x	x	x						x	x
Yearly Mould Growth Index [-]		0.06	0.06	0.05	0.06	0.10	0.15	0.15	0.11	0.15	0.18	0.05	0.05	0.15	0.15

, present the combinations of moisture simulations in WUFI Pro, along with the corresponding mould indices for each considered case. Table 8 and Table 9 display the results for Fjogstad-Hus and ZEB-Laboratory, respectively. The interior heat resistance was initially varied due to uncertainties regarding the cavity ventilation conditions. However, after further assessment, the value of 0.125 m²K/W was deemed more realistic for this study and also confirmed to be more conservative by the simulated yearly mould indices. Consequently, cases with an $R_{si}$ of 0.0588 m²K/W have been set aside, as the ventilation of the claddings is not within the scope of this study.

In Table 8, all the cases with an $R_{si}$ of 0.125 (m²K)/W exhibit yellow risk levels ranging from 3.28 to 3.30 for coated claddings, and 5.20 for the untreated cladding. Table 9 shows that with the same interior heat resistance, the mould index ranges from 0.11 to 0.15 for coated surfaces and 0.18 for the untreated case. For both case buildings, the simulation starts as the wetting or drying season was insignificant, and the initial wood moisture content (MC) of 20 wt% versus 15 wt% had no significant effect on the mould index.

The development of the yearly mould growth indices when comparing the most and least vapour-permeable coated wood, as well as untreated wood, are shown in Figure 7. There are noticeable seasonal fluctuations in the yearly mould index, with peaks corresponding to the winter months when the moisture levels are higher. During the summer months, the index tends to decrease for all cases in both locations. Fjogstad-Hus shows an increasing annual risk of mould growth for the coated claddings.

3.2.2. Wood Moisture Conditions

The simulated wood MC for both locations over five-year and one-year periods are shown below. The effects of the maximum and minimum vapour resistance $s_d$ of the lab measurements are compared in Figure 8, while the other $s_d$-values, as suspected, will yield within this MC range. The other parameters that were varied have remained constant. Observe the fluctuating trends in the course of the MC for all three cases. The peaks are in the winter and the lowest MC during the summer. Despite the initial wood MC 20 wt%, the MCs for the coated claddings stabilize around the intervals 10 to 20 wt% and 10 to 19 wt% (for Fjogstad-Hus and ZEB-Laboratory, respectively).

4. Discussion

4.1. Laboratory Measurements

The range of the $s_d$-values was found to be 0.453 to 1.350 m (three layers) and 0.690 to 2.250 m (six layers). Regardless of the number of layers, coating no. 51 exhibited the highest vapour resistance, and no. 56 showed the lowest resistance among all the products. Some measured $s_d$-values are higher than identified in previously reported values for coatings for exterior wooden claddings [23,38,39]. However, due to unexplained method [38,39] and few specimen [23], the comparison of the $s_d$-values is not confident. The low standard deviations in both the $s_d$-values and dry film thicknesses across all the coatings for our measurements indicate consistency and precision in the measurements. This consistency enhances the reliability of the results and the importance of accurate application methods for achieving a predictable vapour resistance in a coating.

The high R² suggests that 81 to 98% of the variability in the $s_d$-value can be explained by the dry film thickness and that the relationship between these two variables is strong and consistent across different samples. The strongest linear relationship was observed for coatings no. 55 and 56 (R² 0.9827 and 0.9770, respectively), while no. 54 had the weakest, yet still significant, linear relationship (R² 0.8088), possibly due to variability in the application process or thickness measurements.

The additional film support of the cup assembly (metal grid or plastic cylinder) listed in Table 1 indicates a trend whereby coatings with a metal grid support (no. 51, 52 and 55) have higher $s_d$-values, suggesting a greater vapour resistance. Coatings with plastic cylinder support and those without support generally show lower $s_d$-values. This trend suggests that assembling with an additional support may influence the measured vapour resistance, and it introduces uncertainty into the results.

The most vapour-resistant coating was no. 51, an alkyd resin coating. Alkyds being more vapour resistant than acrylics and emulsions confirms the work of Geving et al. [23] and Nore and Hundhausen [39]. However, neither the acrylics nor emulsions showed any clear trends in terms of the binder type and $s_d$-value. The variability might be due to differences in either the formulation or film thickness, or simply due to other factors than the binder content. And with no. 51 being the only tested alkyd, this study cannot with certainty determine any such relationship of binder type. Its high vapour resistance may be related to other factors, i.e., the higher bulk density, solid content, or the additional metal grid support.

4.2. Moisture Content and Mould Growth Risk in Ventilated Wooden Claddings

Over the simulation period, one observes that the wood MC fluctuates with seasonable variations as anticipated, with the highest MC in the winter. During fall and winter, the MC increases, but it dissipates over the summer months, without signs of moisture accumulation over the simulated five-year period. This suggests that the vapour resistance of the coating does not significantly impede the drying of moisture.

During the drying period, however, Figure 8c,d indicate a slightly faster drying for the vapour-permeable coating, as indicated by the green line having a more rapid decline in the MC compared to the red line during this specific period. Untreated claddings (yellow) show both the fastest drying and wetting, as well as more variability and occasional spikes, indicating that they retain more moisture compared to treated wood cases. The amplitude of the MC fluctuations is notably greater for untreated cladding compared to coated cladding, which is in line with findings from several studies [14,23,39] regarding the positive effects of surface treatment of wood.

The variations in the initial wood MC of 20- and 15 wt% yield similar MC projections and mould indices for both Fjogstad-Hus and ZEB-Laboratory. This is probably due to the thinness of the cladding, which facilitates fast moisture drying regardless of factors, i.e., the vapour resistance of the coating. Similarly, a simulation start in April versus in October does not affect the results. This confirms the effect of the thin cladding’s ability to quickly dry toward the air cavity, with the cladding being uncoated on this interior side.

The seasonal fluctuations in the mould growth risk mirror the fluctuations in the wood MC, with a peak in the winter and a declining risk during the drying period (Figure 7). The two different vapour resistances of the coatings yield similar mould risk developments, which underscores that either the difference in the $s_d$ is too small or the value itself is insignificant. It is observed for Fjogstad-Hus that the coated claddings have an slight annual increase in the mould risk, indicating that the moisture exposure and conditions of this wall over time do not fully dry out, causing favourable conditions for mould growth. This confirms the analysis of Ingebretsetsen et al. [47].

The notably higher mould growth indices and MC levels of Fjogstad-Hus compared to ZEB-Laboratory support the previous findings of Ingebretsen [46,47] and were expected due to the drying conditions being worse (i.e., orientation and climate conditions in Stavanger climate). While north-facing orientations are usually considered the worst-case scenario for winter drying due to lack of direct solar radiation, the southwest façade of Fjogstad-Hus is more exposed to driving rain.

Both Fjogstad-Hus and ZEB-Laboratory show the higher risk of mould growth of untreated claddings compared to coated claddings. This is because the exterior surface coating, after all, protects against liquid moisture absorption in the cladding. These results confirm the importance of applying a coating to façade cladding to protect against biodeteoriation, thereby achieving longer lifespans.

5. Conclusions

The laboratory experiments revealed that the water vapour resistance (in terms of the $s_d$-value) for three layers ranged from 0.453 to 1.350 m, and from 0.690 to 2.250 m for six layers. There is a strong linear relationships between the $s_d$-value and the dry film thickness. The alkyd exhibited higher $s_d$-values than the emulsions and acrylics, though the sample size is not sufficient to determine this relationship.

Over the five-year simulation, the wood MC fluctuates seasonally. The vapour-resistant coating showed a slightly slower drying than the more vapour-permeable one. However, their different impacts on the MC are insignificant over the five-year period as no moisture is accumulating. The untreated claddings showcased both greater MC fluctuations with more variability and a higher mould risk than the two coated claddings, which affirms the importance of surface coating. The cases of Fjogstad-Hus exhibited greater wood MC levels and greater mould risks than the cases of ZEB-Laboratory due to the more extreme climatic conditions.

Future studies should aim to isolate and quantify the impact of additional film supports by conducting controlled experiments, where coatings are tested systematically with and without various supports, to help understand the extent of their influence on the water vapour permeability and resistance measurements. To establish a more definitive correlation between the binder type and the vapour resistance, more targeted experiments with a larger number of samples and various types of coatings are necessary, as the current findings only provide an indication of this relationship.