Moisture absorption of CLT elements during construction: insights from an on-site monitoring case study

Abstract

The recent growth of mass timber buildings (MTB) in number, height, and architectural complexity represents a major opportunity for the timber construction sector, but it also introduces significant challenges to overcome, particularly regarding moisture management and long-term performance in diverse climatic conditions. In this context, this study presents 240 days of hygrothermal monitoring data of a MTB during construction, focusing on the moisture behaviour of cross-laminated timber (CLT) elements and the influence of construction practices on moisture safety. Monitoring was carried out in high-risk zones, including wall–floor joints on north and south façades and interior moisture-prone rooms. Moisture content (MC) was continuously recorded alongside daily precipitation and key construction events. During the first 100 days, characterised by mild temperatures and limited rainfall, MC values remained mostly below 20%. However, short and intense rainfall events caused rapid increases above 30%, demonstrating the high sensitivity of CLT to water exposure. As precipitation increased and temperatures decreased, drying was limited, and MC remained above 20% for prolonged periods. The premature installation of gypsum boards and bituminous membranes before roof and façade waterproofing trapped moisture, resulting in visible mould growth. The construction schedule and lack of temporary protection significantly extended water exposure, while alternative strategies, such as early roof waterproofing, temporary shelters, or targeted end-grain protection were not adopted. These results demonstrate that the growing complexity of MTBs requires an integrated moisture management plan developed prior to construction and applied throughout the construction phase. Such a plan should consider local climate, construction sequencing, protective measures, and continuous monitoring to ensure moisture-safe building practices and support the long-term performance of timber structures.

Introduction

In recent years, the adoption of mass timber buildings (MTB) has grown considerably worldwide, driven by the need for sustainable alternatives to traditional construction materials. Advances in engineered wood products (EWP), particularly cross-laminated timber (CLT), have addressed many of the constraints associated with traditional sawn timber, facilitating the realisation of taller and more architecturally complex timber structures. The global production of CLT has increased rapidly in recent years. In 2018, estimative of annual production exceeded 1 million m³ and by 2020 it had nearly tripled to 2.8 million m³, being 43% oh this production in Europe.^1,2 Additionally, the market for EWPs, of which CLT represents a significant share, was valued at approximately USD 18.5 million in 2022 and is projected to grow at a compound annual growth rate of 9.4%, until 2028. ³

Despite these advantages, MTBs introduce novel challenges, particularly concerning the long-term performance of these buildings.^4,5 Several studies have reported early-stage issues in CLT structures across Europe. For example, in Norway, Austigard and Mattsson ⁶ documented cases of fungal development during the initial service period of CLT elements. Similar observations were made in Italy, where Gaspari et al. ⁷ reported CLT balconies failures in newly constructed timber buildings. In the United Kingdom, Clifford ⁸ highlighted instances of biological degradation in CLT assemblies. These challenges are particularly pronounced in regions with limited historical experience in timber construction and in climates that promote the growth of biological agents, such as fungi and insects, like Mediterranean countries.

Portugal has recently shown increased interest in MTB, with the establishment of domestic production facilities for timber-frame panels and CLT, and a growing portfolio of timber-based projects. ⁹ While the country had no CLT factories in the early 2020s, by 2025 at least three were operational. The most commonly used species for CLT production in Portugal is Norway Spruce (Picea abies). On the construction side, early examples such as the four-story glued-laminated timber (GLT) structure Redbridge School in Lisbon (2018) have been followed by more ambitious projects, including the eight-story CLT structure Antas Garden in Porto. These developments signal a strengthening confidence in mass timber as a structural solution in Portugal.

Moisture exposure remains a challenge affecting the performance of mass timber elements, both during construction and throughout their service life. Beyond the intrinsic durability of the wood species employed, the duration of exposure to conditions conducive to biological degradation, specifically high moisture content (MC) and temperatures favourable to fungal and insect activity, is critical.^10–12 Among the biological agents responsible for wood degradation, fungi and subterranean termites are generally considered the most critical. Moulds and sapstain fungi are able to colonise wood at relatively low MC levels, typically around 30%, whereas decay fungi require a wider moisture range, with optimum conditions usually between approximately 40% and 80% MC. ¹³ Subterranean termites, while capable of attacking wood at lower MC levels, show a clear preference for wood with MC above 20%. ¹²

Although moisture availability is considered to be the most critical factor influencing biological degradation agents growth in wood, no single MC threshold can reliably determine whether a mass timber element is safe from biological degradation agents attack, as different organisms exhibit distinct moisture requirements. Nonetheless, an MC of approximately 20% is commonly cited in the literature^14–17 and used in practice as a conservative limit above which the risk of biological activity increases. Importantly, MC should not be assessed solely as an instantaneous value, but rather in combination with the duration for which elevated moisture levels persist.^11,18 This duration-dependent risk is highly case-specific and influenced by multiple factors, including wood species and its natural (or acquired) durability, air temperature and relative humidity, exposure conditions, moisture transport pathways within the element, and design-related factors such as presence of covering layers, shading, and wind-driven drying potential. The presence of wood degradation agents is also essential, although its development is constrained by the aforementioned factors. Reflecting these complexities, recent approaches emphasise the importance of depth-dependent and time-resolved moisture measurements. For instance, Kalbe et al. ¹⁴ recommend introducing a secondary measurement point when MC values measured at 10 mm from the end grain exceed 19%, with additional monitoring at 30 mm depth to evaluate the persistence of elevated moisture and enable a more robust assessment of degradation risk.

The construction phase is particularly decisive in determining long-term performance, as improper handling or prolonged wetting can compromise structural integrity. Extending the service life of EWPs offers substantial advantages from a life cycle assessment (LCA) perspective. Most LCA studies of MTB assume a 50-year service life. Longer-lasting timber elements reduce the need for material replacement, thereby lowering energy and resource consumption.^19,20 Extending the service life and enabling reuse at the end of life prolongs carbon storage and delays greenhouse gas emissions, amplifying the climate benefits of timber construction. Ensuring that EWPs remain moisture-safe from the earliest construction stages is therefore fundamental for maximising their long-term environmental performance.

Nowadays, several techniques are available for measuring the wood MC, which can broadly be classified into direct and indirect methods. Direct methods, such as oven-drying, determine MC through destructive testing and are therefore unsuitable for in-situ monitoring of real buildings. On the other hand, indirect methods rely on physical properties of wood that are correlated with MC and are more appropriate for construction-phase and long-term monitoring. ²¹ Common approaches include the electrical resistance method and sorption isotherm method, both of which provide precise point-based MC measurements at the specific locations where sensors are installed. In addition to these point-based techniques, other monitoring systems are available that detect relative changes in moisture conditions over larger areas, such as conductive sensor tapes or glass-fibre mats, although these systems do not provide absolute MC values. ²² Each technique presents specific advantages and limitations in terms of accuracy, spatial resolution, invasiveness, and cost. Consequently, the selection of a monitoring method should be tailored to the specific project objectives (e.g., leak detection, construction-phase monitoring, or long-term performance assessment) as well as practical constraints such as the available budget.

This recognition has driven extensive research into moisture monitoring of mass timber components (e.g., Schmidt and Riggio ²³ ; Baas et al. ²⁴ ; Kalbe et al. ²⁵ ) and the development of targeted moisture management protocols to guide best practices during fabrication, transportation, and on-site assembly.^15,16,26 However, the implementation of watertight encapsulation of CLT at the manufacturing stage or use of full temporary shelter is often dismissed due to high costs.

Although MTBs are often associated with rapid assembly and reduced construction times, the growing scale and complexity of these structures, combined with limited experience among contractors, can result in prolonged exposure to environmental conditions during the construction phase. ²⁷ For example, Austigard and Mattson ⁶ analysed 12 MTB case studies and found that insufficient protection during construction or construction errors led to excessive water infiltration that resulted in mould growth in 75% and decay fungi in 50% of the case studies. Kalbe et al. ²⁵ monitored MC in six buildings in Estonia and observed that, although panel installation was relatively rapid, exposure durations were extended due to delays in the installation of façade protection systems. The study further showed that some temporary protective measures were often ineffective, permitting water ingress and restricting adequate drying. As a result, MC frequently exceeded critical thresholds (>20%) following individual rainfall events, demonstrating that rapid construction alone does not guarantee effective moisture control.

In contrast, Schmidt and Riggio ²³ reported generally lower MC in a CLT structure monitored during construction in Oregon (USA), with MC values remaining below 16% for most of the construction period. However, the authors emphasised that these results were strongly influenced by favourable conditions, including low precipitation and extended drying periods, which may not be representative of most construction scenarios. Even under these conditions, localised areas prone to water accumulation and limited drying, such as shaded connections between structural elements, exhibited elevated MC levels. In this context, Kalbe and Kalamees ²⁸ highlighted the importance of incorporating specific moisture safety drawings and strategies already during the building design phase after the evaluation of moisture safety strategies in two CLT buildings.

Effective moisture management must be considered at every stage, including transportation, on-site storage, and construction activities. This presents a significant challenge due to the differences and complexities that can arise in each of these phases, and, in some cases, the varying responsibilities involved. On the other hand, CLT displays heterogeneous wetting behaviour due to the intrinsic anisotropy of wood combined with the cross-laminated configuration of CLT panels. Moreover, these elements generally dry more slowly than conventional timber products, which can lead to localised moisture accumulation and the formation of moisture-trapping zones, thereby increasing the risk of biological deterioration as well as physical damage such as cracking and delamination. ²⁹ Such phenomena highlight that, although careful design detailing is essential,^27,30 it must be complemented by a comprehensive moisture management plan to ensure moisture-safe performance.

Lima et al. ⁹ emphasised the need to build a robust database of hygrothermal data to better understand biodeterioration processes in CLT exposed to weather, as well as the importance of protective measures, particularly at end-grain surfaces. In line with this, the objective of this study is to investigate the moisture behaviour of CLT elements during the construction phase of a MTB. Although the building includes other timber-based structural elements, the monitoring and analysis presented here focus exclusively on CLT components. The study aims to (i) monitor the moisture history in CLT elements located in high-risk zones during construction, (ii) evaluate the influence of construction planning, scheduling, and the adoption or absence of moisture-protection measures on moisture absorption and drying, and (iii) highlight the relevance of construction-phase moisture management in order to alert practitioners to potential risks and to motivate the development and adoption of good construction practices for moisture protection of CLT elements.

Materials and methods

Case study characterisation

Due to confidentiality requirements, no company names are presented in this work. Instead, the focus is placed on describing the construction characteristics relevant to the study objectives.

The case study object is a residential building located in Guimarães, northern Portugal. The building is implanted on a site with a relatively steep slope. The architectural layout consists of two elongated volumes extending from the terrain and converging into a third connecting volume and it is composed of three stories above ground and one basement level.

Its structural system combines concrete, timber, and steel. Reinforced concrete was used exclusively in the basement, foundations, and in the stair and elevator cores. The ground floor is supported by steel beams and columns. From the first floor upward, only timber elements were used for the structural system, while the elevator and stair cores remain in reinforced concrete. Floors, walls, and the roof were constructed from CLT panels, supported by GLT beams and columns where needed. All CLT and GLT components were produced from Norway Spruce, with no treatments or surface coatings. Figure 1 illustrates the three-dimensional representation of the structural system and a plan view of the building, including the indication of the north direction.

Figure 1.

Three-dimensional representation of the building structure and plan view illustrating the building configuration and north orientation.

The external wall system consists of a three-layer CLT panel (100 mm thick), combined with a ventilated façade. A breathable membrane was installed on the exterior side of the insulation layer. The façade assembly includes mineral wool insulation and an outer layer of thermally modified wood cladding. On the interior, walls were covered with two layers of gypsum board. The internal partition walls were also constructed from three-layer CLT panels (100 mm thick) lined with a double layer of gypsum board on each side.

The intermediate floor consists of a five-layer CLT panel (160 mm thick) covered with an acoustic membrane and a lightweight concrete topping. Below the CLT, two gypsum board layers provide the ceiling finish. An acoustic strip was also installed between the façade walls and floor panels to improve sound insulation. Figure 2 illustrates the layer configuration of both the façade walls and the intermediate floor.

Figure 2.

Construction layers of the façade walls and intermediate floor, including structural, insulating, and finishing elements.

Balconies were also integrated on the first and second floors. On the first floor, the five-layer CLT balcony slabs (160 mm) are supported by two steel beams and two GLT beams. On the second floor, the balconies are supported by three GLT beams and one three-layer CLT wall panel (100 mm). Above the balcony slabs, a drainage system for rainwater was installed, which was waterproofed with a bituminous membrane. On top, a timber deck made of thermally modified wood completes the finish.

The roof was designed as a flat garden roof, built from CLT panels supported by GLT beams positioned above the ceiling of the second floor. This configuration creates a ventilated cavity between the CLT roof panels and the CLT ceiling of the second floor. No ventilated layer was included above the CLT roof; instead, the top surface of the CLT was directly waterproofed to allow the installation of the garden roof system. Despite the known sensitivity of CLT to moisture, no temporary moisture protection measures were implemented during the construction of this case study. The building remained fully exposed throughout the assembly process, with no use of temporary shelters, protective weather membranes for CLT encapsulation, end-grain sealing, or mechanical drying of areas where standing water occurred.

In this study, the monitoring team gained access to the construction site only when the structural timber elements, namely CLT and GLT components, began to arrive. This occurred after a portion of the ground-floor structure had been completed, providing the necessary support for the upper levels. The first timber elements were delivered to the construction site on 24 May 2024, and the assembly of the timber structure started approximately one week later, during the final month of spring. The assembly phase was completed in the first half of December 2024, resulting in a total timber erection period of approximately 200 days. Due to budgetary constraints, construction activities were suspended in April 2025 and are currently expected to resume in 2026. However, at the time of writing, construction had not yet restarted. Consequently, the present study covers the 240-day monitoring period during which access to the site to data collection were possible.

Hygrothermal monitorization

Although the number of hygrothermal monitoring studies on MTBs has been growing, there is still no standardised methodology for this type of monitoring, which limits comparability across case studies. The present study adopted the methodological framework proposed by Baas et al. ³¹ for the monitoring of mass-timber structures during construction because, at the time the monitoring plan was developed, it was the only approach available in the scientific literature specifically designed for construction-phase monitoring of MTB. Its explicit focus made it the most suitable and systematically articulated framework for guiding this study.

The parameters monitored included the MC (%) of the CLT as well as environmental conditions, namely temperature (°C), relative humidity (%), and daily precipitation (mm). Environmental exposure was characterised using data from a public weather station located 3.5 km from the construction site, obtained through the weather underground database. ³²

MC measurements were obtained by the electrical resistance method (EN 13183-2 ³³ ), which requires the installation of insulated electrodes. This method was selected because it is a well-established technique for in situ monitoring of wood MC and provides good accuracy below the fibre saturation point (FSP). Although its accuracy decreases above the FSP, the measurements remain suitable as a reliable indicator of moisture presence and trends, which is particularly relevant for identifying critical wetting events during construction. Since this technique only allows measurements at electrode locations, the placement of electrodes required strategically defined measuring points. The selection of measuring points followed a risk-based approach aimed at identifying areas most susceptible to moisture accumulation. The analysis considered both the construction and the service phases of the building, given that the monitoring program is intended to continue during the early years of building operation.

The identification of critical zones was based on three main factors: (1) literature references highlighting areas of CLT elements prone to higher water absorption, (2) documented cases in which excessive humidification was detected in MTB, and (3) the research team's expertise in analysing the construction drawings of the case study.

Based on the risk analysis, four zones of the building were identified as particularly vulnerable to excessive moisture accumulation and therefore selected for monitoring. These are summarised as follows:

Façade wall–floor joints

The interface between the façade walls and the floor panels was considered a critical zone due to the presence of exposed end-grain surfaces. In the case of the floor panels, the end grain readily absorbs water but also dries rapidly when exposed to weather. This repeated wetting–drying cycle generates steep moisture gradients within the CLT lamellas, leading to the formation of cracks and delamination. Such defects can create preferential pathways for further water ingress, allowing moisture to remain trapped between lamellas and thereby significantly increasing the risk of biological deterioration. In contrast, the wall panels at this junction exhibit high water absorption but limited drying capacity, as their end grain is not ventilated. Consequently, in the event of excessive wetting, drying may require several months, further elevating the risk of long-term damage.^9,25,26 This zone is further classified as high risk because it typically accommodates structural connectors, such as angle brackets and hold-downs. In this context, the development of fungal deterioration and moisture-induced changes in the mechanical properties of wood are particularly critical, as they may directly compromise the structural performance of these connections.^34–36

Moisture-prone rooms (bathrooms, kitchens, laundry rooms, etc.)

Interior moisture-prone rooms were included in the monitoring strategy because of their potential vulnerability during the use phase of the building. Water leaks from pipes or appliances can remain undetected for extended periods, similar to failures in roof assemblies, and may lead to concealed water accumulation within the timber elements.^6,8 Although the focus of monitoring the moisture-prone rooms was to on a long-term perspective after the construction, these areas also provided additional data during the construction phase.

Flat garden roof and balconies

The flat roof and balconies represent high-risk elements. As a horizontal surface, it is directly exposed to precipitation and therefore receives the greatest water load. Although the CLT roof panels and balconies are protected by a bituminous waterproofing membrane, the risk of infiltration remains high because any membrane failure would result in hidden water ingress. Without continuous moisture monitoring, such failures would likely only become evident once substantial damage had occurred, requiring costly repairs.^7,26,37,38 In this study, the results for the flat garden roof and the balconies are not included. Both areas require a distinct analysis, as they are horizontal elements directly exposed to rainfall and rely on specific protective design systems to ensure their performance. These zones demand particular attention during both the construction and use phases. Therefore, their monitoring results will be presented and discussed in detail in a future publication.

Monitoring setup

The limiting factors regarding number and placement of electrodes were the project budget, the architectural design of the building, and the research team's capacity to manage and store data. The setup also aimed to balance redundancy and diversity in the dataset, ensuring both comparability across different locations and coverage of areas predicted to be more vulnerable to moisture ingress or entrapment, as identified in the risk analysis.^23,31 In total, more than 80 pairs of electrodes were installed, distributed across different layers of 34 CLT panels.

The electrode installation was designed to monitor the central region of each targeted lamella. For this purpose, insulated electrodes were fabricated with different lengths depending on the CLT layer to be monitored. Electrode insulation ensured that only the tip of each electrode was in contact with the wood material, with approximately 10 mm of the tip left uninsulated. Pre-drilling was carried out to allow the insertion of the electrodes without damaging the insulation, both the diameter and depth of the pre-drilled holes varied according to the monitored layer. For wall panels, electrodes were installed at a distance of approximately 30 mm above the floor panels, while for floor panels, electrodes were positioned as close as possible to the adjacent wall panels in order to capture moisture conditions at the wall–floor interface.

Figure 3 illustrates the electrode installation locations within the building floor plans, while Table 1 summarises the monitored zones, the specific CLT layers where electrodes were positioned, the number of panels monitored, and a schematic representation of the electrode position within each CLT element. In the schematic drawings included in Table 1, the circle indicates the location of the electrode tip within the panel cross-section. Electrodes were installed only in the two building volumes highlighted in Figure 3, due to restrictions imposed by the construction company.

Figure 3.

Location of electrode installations across the building: distribution in façade wall–floor joints and moisture-prone rooms.

Table 1.

Summary of electrode installations by monitored zone, CLT layer, and number of panels.

The electrodes installation was phased according to the timing of the CLT panel installation, the subsequent installation of adjacent panels (which could potentially damage the electrodes or measuring devices), and the detection of localised water accumulation.

Three different setups were used to measure the MC of the CLT panels: one automated system and two setups requiring manual measurements. The automated system, supplied by Tector^©, recorded daily measurements, and due to budget constraints, only a limited number of these devices were available. The first manual setup required weekly or biweekly measurements during site visits. In this configuration, electrodes were connected to wires routed to technical areas of the building, allowing MC readings to continue even after finishing layers were installed. The second manual setup involved electrodes without cable routing, which likewise required weekly or biweekly measurements using handheld wood moisture meters but could only be monitored until covering materials such as gypsum boards or impermeable membranes were installed. Manual measurements in both setups were performed using a PROTIMETER TimberMaster^® handheld wood moisture meter. The Tector^© monitoring devices and the PROTIMETER TimberMaster^® handheld moisture meter were set for Norway spruce, ensuring that the electrical resistance-based MC readings were calibrated to the specific properties of the CLT material used in the case study. Figure 4 illustrates examples of the three measurement setups: (a) automated system, (b) cable-routed system, and (c) standard manual system.

Figure 4.

Examples of the three different measurement setups: (a) automatic system, (b) cable-routed system, and (c) standard manual system.

The first MC measurement was conducted on 13 June 2024. Over the course of the monitoring period, a total of 1182 manual measurements were performed across the different CLT panels. The automated measurement system operated continuously for approximately 200 days, providing daily data and complementing the manual readings to ensure comprehensive coverage of the monitored zones.

Results and discussion

The climatic data are presented in Figure 5 and Table 2. For the analysis, the monitoring period was divided into two climatic phases: a warm and relatively dry period from 24 May to 23 September, corresponding to late spring and summer, and a cooler and humid period from 24 September to 7 February. The dataset is presented as average, maximum, and minimum air temperature (°C) and relative humidity (%), as well as the total accumulated precipitation (mm) and the relative number of days with precipitation exceeding 0.25 mm (%). Accumulated precipitation alone can lead to misinterpretation, as high totals may result from a small number of intense events, which differ substantially from frequent low-intensity rainfall in terms of wood wetting and drying behaviour. To better capture precipitation exposure, the number of precipitation days was therefore included. A threshold of 0.25 mm was adopted to classify a day as rainy, following the approach used in the Scheffer Climate Index. ³⁹ As highlighted by Scheffer, ³⁹ precipitation duration and frequency are often more influential on wood MC than total precipitation, since prolonged or repeated wetting events generally result in higher moisture uptake than short, intense rainfall delivering similar amounts of water.

Figure 5.

Climatic data recorded at the nearest public weather station: (a) average, maximum, and minimum air temperature (°C) and relative humidity (%), and (b) daily precipitation (mm) and accumulated precipitation (mm).

Table 2.

Climatic data recorded at the nearest public weather station divided in two seasonal periods.

Period [days]	Temperature [°C]			Relative humidity [%]			Precipitation
	MaxT	AvgT.	MinT	MaxRH	AvgRH.	MinRH	Acc. [mm]	Days>0.25 [%]
24/05–23/09	28.1	21.7	16.3	89.2	73.2	51.5	74.8	16.4
24/09–07/02	18.8	13.2	9.0	97.0	87.9	67.1	963.9	52.2
Total	23.2	17.2	12.4	93.4	81.0	59.8	1038.6	35.1

Note. Temperature (°C): average of maximum temperature (MaxT), average temperature (AvgT), and average of minimum temperature (MinT). Relative humidity (%): average of maximum relative humidity (MaxRH), average relative humidity (AvgRH), and average of minimum relative humidity (MinRH). Precipitation: total accumulated precipitation (Acc; mm) and percentage of days with precipitation above 0.25 mm (Days > 0.25; %).

From the analysis of Table 2 and Figure 5, it can be observed that the first period, spanning 122 days (24 May to 23 September), was characterised by mild temperatures and relatively dry conditions, with an average air temperature of 21.7 °C and a total accumulated precipitation of 74.8 mm. During this phase, only 16.4% of the days recorded precipitation exceeding 0.25 mm. In contrast, the second period, covering 136 days (24 September to 7 February), presented a colder and wetter conditions, with the average temperature decreasing to 13.2 °C, total accumulated precipitation increasing to 963 mm, 52.2% of days presenting precipitation above 0.25 mm, and increasing the average relative humidity from 73.2% to 87.9%.

These climatic conditions are consistent with the expected seasonal patterns for the region. According to the Climate Normal 1991–2020 provided by the Instituto Português do Mar e da Atmosfera (IPMA) for Braga, the average temperature and accumulated precipitation from June to September are approximately 20.5 °C and 170.7 mm, respectively, while the period from October to February presents an average temperature of 11.0 °C and accumulated precipitation of 848.8 mm.

Façade wall-floor joints

The MC results for the façade wall–floor joints are presented in Figure 6 for the south façade. Figure 6(a) displays the MC of the wall panels, while Figure 6(b) shows the MC of the floor panels. Continuous monitoring equipment was implemented in central part of the external lamella of one wall panel and in the central part of the upper lamella of one floor panel, while the remaining panels were monitored through manual measurements. A short period of missing data is observed in the continuous monitoring records, which resulted from an inadequate initial positioning of the gateway, preventing stable communication with the sensor. Once this issue was identified, the gateway location was adjusted, and continuous monitoring was successfully resumed. The graphs include the 20% and 30% MC thresholds, as well as daily precipitation to facilitate visualisation of the MC variation during precipitation events. The 20% MC threshold represents a commonly adopted reference value associated with the development of fungi (e.g., EN335:2013 ¹⁷ ). Although no single MC value can definitively define fungal activity ‒ since this depends on factors such as wood and fungi species ‒ it is generally accepted that development may occur once MC approaches approximately 18%, with optimal fungal growth conditions occurring at MC values well above 40%.^13,40,41 On the other hand, the 30% MC threshold corresponds approximately to the FSP of Norway spruce. ⁴² Above this level, additional moisture does not induce volumetric changes in the wood but significantly increases the risk of biological degradation due to the presence of free water.^12,43 This threshold also marks the range beyond which the electrical resistance method exhibits reduced measurement accuracy.

Figure 6.

Monitorization of the moisture content (MC; %) of the south façade wall–floor joints, measured in the (a) wall panels (exterior, middle and Interior layers) and in the (b) floor panels (upper, second and third layers), daily precipitation (Prec; mm), and MC thresholds (MC = 20% represented by the yellow line and MC = 30% represented by the red line). Highlighted construction phases affecting CLT exposure to precipitation include: (i) assembly of the CLT structure of the first and second floors (30 May 2024–10 July 2024); (ii) period with the structure assembled but without envelope protection, including the start of gypsum board installation on walls and ceilings (6a) and bituminous membrane installation over floors (6b), in the absence of roof waterproofing, ventilated façade, and windows (10 July 2024–13 November 2024); (iii) roof waterproofing phase (22 November 2024–10 July 2025); and (iv) installation of the ventilated façade system on the south façade (13 November 2024–20 December 2024).

Additionally, key construction periods that affected the exposure of the CLT panels to precipitation are highlighted in the graphs, referring exclusively to the two building volumes included in the monitoring program (Figure 3):

the assembly of the CLT structural system of the first and second floors (30 May 2024–10 July 2024);

a period during which the CLT structure was fully assembled but remained unprotected, characterised by the absence of roof waterproofing, ventilated façades, and windows, and including the start of gypsum board installation on CLT walls and ceilings and the installation of a bituminous membrane over the CLT floor (10 July 2024–13 November 2024);

the roof waterproofing phase (22 November 2024–10 January 2025); and

the installation of the ventilated façade system on the south façade (13 November 2024–20 December 2024).

In the first 122 days of monitoring, most MC measurements remained below 20% due to low precipitation and relatively high temperatures. However, individual rainfall events, like those observed toward the middle and end of period (i), resulted in rapid increments on the MC of exposed CLT elements, with both wall and floor panels reaching values above 30%. After the first 122 days, a rise in precipitation and relative humidity combined with gradually decreasing temperatures created conditions that enhanced water uptake and hindered the drying of the CLT panels. From this point forward, nearly all measurements consistently exceeded the 20% MC threshold.

Examining the key exposure events highlighted in Figure 6, it is evident that the installation of the second-floor ceiling panels (end of Period i), carried out before the installation of the ventilated façade system and the roof structure was fully waterproofed, did not effectively prevent water ingress into the building. In contrast, following the completion of the roof waterproofing (Period iii) and the installation of the ventilated façade system (Period iv), a clear downward trend in MC was observed in most wall and floor panels of the south façade. Exceptions were identified in the wall and floor panel equipped with the continuous monitoring system as well as in the Upper layer of one additional floor panel. These exceptions were associated with localised water ingress caused by leakages in the ventilated façade system, as documented in Figure 7, which shows moisture penetration observed from the interior (Figure 7(a)) and from the underside of the façade (Figure 7(b)).

Figure 7.

Evidence of water leakage associated with the ventilated façade system on the south façade: (a) moisture ingress observed from the interior side of the façade wall; (b) water runoff and accumulation observed at the bottom of the façade assembly.

Figure 8 presents the MC results for the north façade wall–floor joints, including the 20% and 30% MC thresholds, daily precipitation, and key construction periods. However, it should be noted that during the monitored period the ventilated façade system on the north façade had not yet been installed. The results generally follow the same trend observed for the south façade during the first 122 days, up until the installation of the ventilated façade system on the south side, which protected the wall panels from further direct exposure. From that point onward, it is evident that the north façade wall panels did not dry as effectively as those on the south façade. This difference is attributed to the earlier installation of the façade system on the south façade, combined with its more favourable sun exposure.

Figure 8.

Monitorization of the moisture content (MC; %) of the north façade wall–floor joints, measured in the (a) wall panels (exterior, middle and Interior layers) and in the (b) floor panels (upper, second and third layers), daily precipitation (Prec; mm), and MC thresholds (MC = 20% represented by the yellow line and MC = 30% represented by the red line). Highlighted construction phases affecting CLT exposure to precipitation include: (i) assembly of the CLT structure of the first and second floors (30 May 2024–10 July 2024); (ii) period with the structure assembled but without envelope protection, including the start of gypsum board installation on walls and ceilings (8a) and bituminous membrane installation over floors (8b), in the absence of roof waterproofing, ventilated façade, and windows (10 July 2024–13 November 2024); and (iii) roof waterproofing phase (22 November 2024–10 January 2025).

The results obtained are consistent with the findings of Lima et al., ⁹ where CLT specimens of wall–floor joints exposed to outdoor conditions exhibited rapid moisture uptake due to the direct exposure of end-grain surfaces to liquid water. In addition, water entrapment was observed in wall panels as a result of their contact with the floor, which restricted ventilation at the joint surface. Similarly, in floor panels, moisture accumulation within the inner layers was reported, highlighting the difficulty of drying once water had penetrated the system.

Regarding Period ii, it is possible to observe in Figures 6 and 8 that the installation of gypsum boards on the walls and ceiling and the application of a bituminous membrane over the CLT floors were carried out prior to the completion of roof waterproofing, window and the ventilated façade installation. The application of covering layers over wet CLT surfaces from the inside and before the building envelope was fully watertight limited the drying potential of the panels, and increases in MC were observed. As the precipitation and relative humidity increased (from day 122 onwards), moisture buildup and mould growth quickly became evident.

Figure 9 illustrates a wet façade wall–floor joint covered by a bituminous membrane, where the absence of window installation allowed further water ingress (Figure 9(a)), and a façade wall–floor joint covered by a bituminous membrane and gypsum board, showing visible signs of fungal growth on the CLT wall surface (Figure 9(b)). These observations are consistent with the findings of Kukk et al., ⁴⁴ who highlighted the increased risk of mould growth when façade walls with elevated MC due to weather exposure are covered prematurely.

Figure 9.

Examples of the effects of premature covering of the façade wall-floor joints: (a) joint with elevated MC covered by a bituminous membrane before window installation, allowing further water ingress, and (b) joint covered by a bituminous membrane and gypsum board, showing visible mould growth on the CLT wall surface.

Moisture-prone rooms

The corresponding results of the moisture-prone rooms are presented in Figure 10. The electrodes in this high-risk zone were installed at the wall–floor joints of partition walls and intermediate floors, adjacent to the cut-offs made in the CLT floor panels for pipe passages between floors. This configuration was selected because it represents a region where the end grain of the floor panels is exposed and where pipe joints commonly occur, as the pipes transition from vertical to horizontal. These conditions create a point of vulnerability both for potential pipe leakage and for water absorption by the CLT. The results shown in Figure 10 presents the MC results, including the 20% and 30% MC thresholds, daily precipitation, and key construction periods. Since the moisture-prone rooms are not part of the building façade, the installation of the ventilated façade system (Event iv) is not represented in Figure 10.

Figure 10.

Monitorization of the moisture content (MC; %) of the moisture-prone rooms wall–floor joints adjacent to cut-offs, measured in the (a) wall panels (outermost and middle layers) and in the (b) floor panels (upper and second layers), daily precipitation (Prec; mm), and MC thresholds (MC = 20% represented by the yellow line and MC = 30% represented by the red line). Highlighted construction phases affecting CLT exposure to precipitation include: (i) assembly of the CLT structure of the first and second floors (30 May 2024–10 July 2024); (ii) period with the structure assembled but without envelope protection, including the start of gypsum board installation on walls and ceilings (10a) and bituminous membrane installation over floors (10b), in the absence of roof waterproofing, ventilated façade, and windows (10 July 2024–13 November 2024); and (iii) roof waterproofing phase (22 November 2024–10 January 2025).

The results for the moisture-prone rooms’ wall panels (Figure 10(a)) indicate that although these interior zones were not directly exposed to rainfall, they did not exhibit lower MC values as might be expected. Instead, their limited ventilation and lack of sun exposure substantially reduced their drying potential, leading to persistently elevated moisture levels. This behaviour aligns with the findings of Schmidt and Riggio, ²³ who reported that shaded and poorly ventilated areas required the longest time to dry. The floor panels (Figure 10(b)), in contrast, exhibited more rapid fluctuations in MC, with sharp peaks followed by relatively fast drying. This behaviour is likely related to the exposed end grain, which, facilitated faster drying, as reported in Lima et al. ⁹ Nevertheless, the results show that even after roof waterproofing, MC levels in the CLT floors remained above the 20% threshold for more than 60 days, highlighting the moisture-trapping potential of these elements.

These high-risk zones also maintained MC levels above 20% for most of the period following the first 122 days, and visible fungal growth developed rapidly. As observed in the results presented in the previous section, the construction schedule played a major role in extending the duration of CLT exposure to moisture, primarily due to water ingress from above caused by the delayed roof waterproofing. Inside the rooms, water accumulation on the top of floor panels was frequently observed, and the CLT surfaces remained wet for prolonged periods. Consequently, mould growth on walls, floors, and ceilings occurred more quickly and extensively than in the façade areas, largely due to the combined effects of shading and poor ventilation. Figure 11 illustrates these conditions: (a) water infiltration from the roof running down the wall surface, (b) water accumulation on the top of floor panel of an interior room, and (c) visible fungal growth on a CLT floor panel.

Figure 11.

Examples of moisture-related issues observed in moisture-prone rooms: (a) water infiltration from the roof running down the wall surface, (b) water accumulation on the floor of an interior room, and (c) visible fungal growth on a CLT floor panel.

It was also possible to observe that the premature installation of gypsum boards on walls and ceilings, as well as the application of bituminous membranes, hindered the natural drying of the CLT panels. These interior covering layers were installed before the roof waterproofing and while the panels still exhibited elevated MC levels (>20%), which contributed to the retention of moisture and subsequent mould growth. Figure 12 presents two examples of premature gypsum board installation, showing visible mould growth developing on covered areas.

Figure 12.

Examples of the effects of premature installation of gypsum boards on CLT ceilings leading to visible mould growth on the gypsum surface.

Analysis of construction schedule and CLT moisture protection measures

The construction process can strongly influence the hygrothermal and long-term performance of CLT elements. On-site decisions ‒ such as task sequencing, storage conditions, handling practices, and protection measures ‒ can either mitigate or increase moisture-related risks. It is important to emphasise that, during the monitoring of this case study, no specific or target moisture management actions were adopted on-site. This section discusses the construction schedule and absence of protection measures during the construction phase that affected the moisture absorption of CLT panels and presents recommendations, based on both on-site observations and existing literature, for improved practices to minimise moisture exposure and enhance performance. Although design details also play a major role in long-term performance, they are not examined here.

The construction schedule of the case study building was organised by dividing the structure into three parts, hereafter referred to as volumes A, B, and C (Figure 13). The first and second floors of volumes A and B were erected simultaneously, while the first and second floors of volume C were assembled only after the completion of the corresponding levels in volumes A and B. The roof structure was installed once the second floor of all three volumes had been completed. This sequence was designed to allow two CLT assembly teams to work concurrently on volumes A and B, while a third team began the installation of interior gypsum boards and bituminous membranes in these completed sections. After volume C was finished, the roof assembly and waterproofing were carried out. Although this sequencing was intended to accelerate the construction process, it resulted in prolonged exposure of the CLT elements to precipitation, ultimately leading to mould development in early stages of autumn (period when precipitation and relative humidity started to rise). Figure 13 illustrates the described construction workflow and highlight the phase during which mould growth on the CLT surfaces intensified and became widespread. First signs of mould growth on CLT surfaces were noticed few days after the start of the second period of Table 2.

Figure 13.

Construction workflow adopted for the case study building, highlighting the phase which visible mould started to develop on the CLT surfaces.

In this context, Cappellazzi et al. ⁴⁵ emphasise that preventing contact between CLT and moisture is a more robust approach than relying on subsequent drying. This is particularly relevant given that CLT panels are typically manufactured at low MCs (generally MC ≤ 12%), and post-wetting drying requires careful control, as rapid drying may induce drying cracks and delamination. ⁴⁵

For this construction schedule to perform effectively, a comprehensive moisture management plan would have been necessary. Two approaches that could have been adopted are: (i) the use of a full temporary shelter to prevent direct exposure of CLT elements to precipitation throughout construction, or (ii) the adoption of localised protection measures at moisture-critical areas, such as the application of temporary membranes and tapes to protect exposed end-grain surfaces, horizontal CLT panels, openings, and wall–floor joints.^15,16,26 The first option, although more effective in protecting the structure and maintaining construction progress under dry conditions, is rarely implemented due to its high cost. The second option, while increase fewer material costs, could introduce indirect delays, since the installation of covering layers (gypsum boards and bituminous membranes) could only proceed after the façades were fully sealed and the CLT panels adequately dried. Without temporary protection, the natural drying process could take a considerable time, or alternatively, mechanical dehumidification would have been required, introducing additional costs.

When opting to construct without a full temporary shelter, two main approaches can be considered: (1) the application of a full protective membrane system or (2) the targeted protection of critical zones most susceptible to water ingress. The first option consists of covering all exposed CLT surfaces with breathable weather protection membranes specifically developed for mass timber construction. These membranes, available on the market, are designed to protect CLT panels from precipitation and moisture during the construction phase. Although this solution offers comprehensive protection and effectively prevents moisture accumulation, it involves additional costs and requires careful handling. Also, damage to the protective membrane during construction activities can significantly reduce its effectiveness.^15,16

For the second approach, Alsmarker ²⁶ provides a detailed guideline for safeguarding such critical areas during construction. In summary, all exposed end-grain surfaces should be protected, including façade floor edges, pipe cut-offs, window openings, wall-to-wall joints, floor-to-floor joints, and wall-to-floor joints. Protection should be applied using specialised sealing tapes designed for timber construction and installed immediately after panel placement. In the case of cut-offs and openings, temporary sealing with plastic film and tape should be maintained until the installation of windows or piping systems. For roof elements, the climatic conditions observed during the second half of the monitoring period indicated that approximately 50% of the days presented measurable precipitation (>0.25 mm). Under such circumstances, roof panels would particularly benefit from full-surface protection prior to water contact, as repeated rain exposure makes it difficult to achieve proper waterproofing under dry conditions and covering roof panels while their MC remains elevated can have severe long-term consequences for CLT.^37,38,46

Regardless of the chosen protection system ‒ full temporary shelter, full panel protection or localised end-grain protection ‒ a MC monitoring plan should be implemented to identify potential failures in the protective system. Additionally, any water accumulation observed on top of horizontal elements should be removed to prevent localised saturation.^15,16,26

Moisture protection measures and monitoring systems during the construction phase are often limited or entirely absent due to the additional costs associated with their implementation. However, Sharples ⁴⁷ emphasises that the financial consequences of moisture-related damage in mass timber construction can be substantial, as repairs or replacement of CLT elements are often complex, disruptive, and costly. These downstream costs frequently exceed the upfront investment required to implement an effective moisture management plan. Consequently, economic considerations should extend beyond initial construction expenses and account for the long-term financial risks associated with moisture damage when selecting and defining protective measures during construction. ⁴⁷

In this case study, the development of a moisture management plan prior to the start of construction could have contributed to a cost-effective application of moisture protection measures by aligning construction activities with the expected climatic conditions. As discussed in Results and discussion section, the first 122 days of construction were characterised by high temperatures and low precipitation, conditions that are typical for the late spring and summer months in the study region (late May to September). Under such conditions, the implementation of a full temporary shelter would likely have been excessive. Instead, sheltering measures could have been reserved for the second half of the analysed construction period. Furthermore, the construction schedule could have been adapted to local weather patterns so that Volumes A and B were assembled and fully waterproofed during the dry season, thereby limiting the extent of a full temporary shelter to Volume C only and reducing overall moisture protection costs.

To further contextualise the moisture exposure observed during construction, two complementary tables are introduced to relate the construction schedule with precipitation exposure of different CLT elements considering two different scenarios: the first is the actual construction timeline of the case study and no protection measure adopted (Real case) and the second a hypothetical construction scenario (Hyp. scenario) based on the same overall construction start date and phase durations but assuming earlier completion of the roof and ventilated façade for volumes A and B, the adoption of localised moisture protection measures in these volumes, and the use of a full temporary shelter for volume C due to increased precipitation during the later construction period. The localised protection measures considered were derived from relevant scientific literature, including case studies and practical guidance documents.^{15,16,25,26,28,29,38,48,49} However, their efficiency and the specific types of products used (e.g., tapes, membranes, liquid-applied coatings) are not assessed in this study. Instead, these areas are assumed to be adequately protected against moisture ingress. For consistency, construction durations were defined in calendar days rather than working days, with the only adjustment being that no new construction phase was allowed to start during weekends or major holidays (e.g., Christmas and New Year).

Localised measures included end-grain protection at all CLT joints (wall–floor, wall–wall, and floor–floor interfaces), sealing of all other exposed end-grain surfaces, temporary weatherproof sheeting at cut-offs and window openings, and the implementation of temporary drainage solutions on horizontal CLT surfaces to prevent water accumulation. Proper on-site storage of CLT elements was also assumed, consisting of storage in ventilated areas protected from direct exposure to precipitation and solar radiation, with storage durations kept as short as possible.

Table 3 presents the start and end dates (di and de) of each construction phase, and their duration (t; days). For the hypothetical scenario, roof assembly time was proportionally allocated based on comparable roof areas, and the duration of ventilated façade installation for volumes A and B was estimated from the installation rate observed in the real case.

Table 3.

Construction schedule and duration of the different construction phases for the real case study and the hypothetical construction scenario.

			A + B			C
			t [Days]	di [dd/mm/yy]	de [dd/mm/yy]	t [Days]	di [dd/mm/yy]	de [dd/mm/yy]
Real case	1st and 2nd floor		41	30/05/24	10/07/24	95	15/07/24	18/10/24
	Roof	Assembly	28	21/10/24	18/11/24
		Waterproofing	49	22/11/24	10/01/25
	Ventilated façade		86	13/11/24	07/02/25
Hyp. scenario	1st and 2nd floor		41	30/05/24	10/07/24	95	02/09/24	06/12/24
	Roof	Assembly	14	11/07/24	25/07/24	14	09/12/25	23/12/25
		Waterproofing	24	29/07/24	22/08/24	25	02/01/25	27/01/25
	Ventilated façade		44	19/07/24	01/09/24	53	16/12/24	07/02/25

Meanwhile, Table 4 presents the associated exposure to precipitation as accumulated precipitation (Acc.; mm) and days with precipitation above 0.25 mm (Days > 0.25). The data presented in Table 4 is divided by structural elements on both scenarios, including façade wall–floor joints, façade wall exterior surfaces, intermediate floor upper surfaces, interior wall–floor joints, and roof upper surfaces.

Table 4.

Precipitation exposure of different building elements during the construction phases defined in Table 3, for the real case study and the hypothetical construction scenario.

	A + B				C
	Real case		Hyp. Scenario		Real case		Hyp. Scenario
	Acc. [mm]	Days > 0.25	Acc. [mm]	Days > 0.25	Acc. [mm]	Days > 0.25	Acc. [mm]	Days > 0.25
Façade wall-floor joint	1037.5	89	‒	‒	978.8	76	‒	‒
Façade wall (exterior surface)	1037.5	89	70.6	16	978.8	76	‒	‒
Intermediate floor (upper surface)	719.2	76	70.4	16	660.5	63	‒	‒
Interior wall-floor joints	719.2	76	‒	‒	660.5	63	‒	‒
Roof (upper surface)	341.0	39	14.2	5	341.0	39	‒	‒

The comparison between the real and hypothetical construction scenarios presented in Tables 3 and 4 indicates that a cost-effective moisture management plan could have substantially reduced the exposure of CLT elements to direct precipitation. Under the hypothetical scenario, direct rain exposure would have been almost entirely eliminated for most CLT components. The only remaining exposure would have occurred at the exterior surface of the façade wall and the upper surface of the intermediate floor in Volumes A and B, both subjected to 70.4 mm of accumulated precipitation over 16 days, as well as at the roof upper surface, which would have been exposed to 14.2 mm over 5 days. Importantly, all joints and end-grain surfaces were assumed to be sealed, and temporary drainage systems were implemented on horizontal elements, thereby limiting moisture contact to the precipitation events duration. Considering that the longitudinal moisture transport pathways were effectively blocked, this level of exposure is not expected to pose a high risk to the long-term performance of the CLT elements.^9,25,50 When compared to the real construction scenario, these measures would have significantly improved moisture control during construction, reducing the likelihood of moisture accumulation, prolonged wetting, and associated risks to material performance. Overall, the construction schedule and the protective measures play a critical role in determining the moisture safety of MTB. There is no single universally optimal approach, as it must account for multiple project-specific factors, including budget, deadlines, workforce availability, and local climatic conditions.

Conclusions

The results obtained from this case study demonstrate that the increasing complexity of MTB demands a comprehensive, integrated, and dynamic moisture management plan developed prior to the beginning of construction. Such plan should include the analysis of the expected local climatic conditions, the definition of a construction schedule that minimises exposure of CLT elements to precipitation, the adoption of adequate protective measures, and the implementation of hygrothermal monitoring during the construction phase which would enable timely intervention in the event of moisture-related issues.

It is important to note that, under optimal design and execution, MTB can be safe from moisture-related problems. However, moisture infiltration remains one of the most common sources of building pathologies worldwide, and mass timber structures are not an exception. ²⁹ As water invariably finds potential entry points, the identification and monitoring of critical risk zones are essential for ensuring performance.

The MC data presented in Results and discussion section clearly showed that, once the rain period began, nearly all monitored points exceeded the 20% MC threshold. This behaviour, combined with the appearance of biological degradation agents, such as mould, during the construction phase, highlights the vulnerability of CLT to prolonged wetting and the importance of proactive moisture control strategies.

The analysis of a hypothetical, climate-adapted construction scenario demonstrated that a cost-effective moisture management plan ‒ combining construction scheduling aligned with seasonal weather patterns, localised protection measures, and selective use of temporary shelters ‒ could have substantially reduced direct precipitation exposure. In this scenario, direct wetting of CLT elements was nearly eliminated, with only limited exposure of selected exterior and horizontal surfaces during precipitation events.

It is important to reinforce that effective moisture management must be approached as a fundamental component of project planning rather than as a corrective response to problems encountered on site.^16,45 Ensuring the long-term performance of MTB requires early coordination between design, engineering, and construction teams to anticipate exposure risks and define preventive strategies tailored to the specific project and local climatic conditions.