Case Studies in Thermal Engineering 28 (2021) 101703

Contents lists available at ScienceDirect

Case Studies in Thermal Engineering

journal homepage: www.elsevier.com/locate/csite

Assessment of effect of climate change on hygrothermal

performance of cross-laminated timber building envelope with
modular construction
Seong Jin Chang a, Yujin Kang b, Beom Yeol Yun b, Sungwoong Yang b, Sumin Kim b, *
a
Department of Interior Materials Engineering, College of Construction and Environmental Engineering, Gyeongsang National University, Jinju,
52725, Republic of Korea
b
Department of Architecture and Architectural Engineering, Yonsei University, Seoul, 03722, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Cross-laminated timber (CLT) modular construction possesses the advantages of wood, such as
Future climate excellent carbon storage and thermal insulation, and of modular construction, such as consid­
Modular construction erably reduced construction period and cost as well as high productivity. This study evaluates the
Hygrothermal performance hygrothermal performance of CLT walls considering modular construction in future climatic
Cross-laminated timber conditions. Firstly, CLT walls with plywood applied to a core layer were manufactured. A mock-
Mock-up experiments
up of a CLT building was produced and its construction process was analyzed. Hygrothermal
behavior of the CLT walls was simulated using WUFI simulation program, and the predicted
results were verified against measurements obtained from the mock-up experiment. Finally, the
hygrothermal performance of the CLT wall was evaluated for four types of insulation and future
climate in eight cities of USA. The coefficient of variation—root mean square error (CV(RMSE))—
of the temperature and relative humidity inside the ply-lam CLT wall from mock-up experiments
and simulation evaluation were 6.43% and 7.02%, respectively, which met the validation criteria.
Based on the hygrothermal performance, the ply-lam CLT wall with extruded polystyrene insu­
lation was evaluated as safe from moisture problems in all the eight cities considered in this study.
However, the risk of mold growth in all regions and insulation types increased under climate
change with a rise of average annual temperature.

1. Introduction
In view of these advantages, modular construction has been widely adopted for low-rise or medium-rise buildings, especially in
high-density cities, such as Singapore, where shortages of labor and urban land supply are critical issues. Modular construction in­
volves off-site prefabrication and on-site assembly of modular units. In a modular construction interior elements, such as walls, floors,
ceilings, and furniture, are pre-fabricated in a factory prior to the field assembly [1]. About 85% of the modular construction work is
performed off-site, and only the remaining 15% is completed on-site at the final project site. Therefore, modular construction saves
time and cost, and simultaneously offers better quality control and productivity improvement compared to traditional on-site con­
struction [2,3]. In addition, it helps increase site safety, and reduce manpower requirements and resource wastage [4–6].
Although considerable amount of research has been conducted on modular construction composed of steel-frame structures [1,6,

* Corresponding author.
E-mail address: kimsumin@yonsei.ac.kr (S. Kim).

https://doi.org/10.1016/j.csite.2021.101703
Received 21 June 2021; Received in revised form 1 December 2021; Accepted 6 December 2021
Available online 7 December 2021
2214-157X/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

7], interest in eco-friendly building materials has increased significantly in recent times owing to global warming and climate change.
Wood has excellent thermal insulation properties, which can reduce building energy consumption, and contribute to reducing fuel
consumption and carbon emissions. Wood is also a building material that can reduce the total carbon emissions during its production,
construction, and disposal [8–11]. Cross-laminated timber (CLT) is a bulk wood product made from relatively small pieces of wood
using good quality wood. CLT has been widely used in medium-to high-rise buildings because of its excellent structural stability,
insulation performance, fire resistance, and dimensional stability to other building construction systems [12–15]. Over the past few
decades, CLT production has increased dramatically and CLT is competing with mineral-based construction materials [16]. A CLT
modular structure is similar to that made of steel frames in many respects because it has similar structural properties as the latter, and
provides high durability [17].
Movement of moisture and water vapor through porous building materials such as timber and wood-based materials can cause
moisture and durability issues in building envelopes. To address this drawback, it is necessary to predict or evaluate the severity of
moisture-related problems and design building envelope accordingly. WUFI (Wärme Und Feuchte Instationär) [18] is a dynamic
analysis program for predicting hygrothermal behavior of building envelopes [19]. Chang et al. analyzed the hygrothermal behavior of
a CLT wall, using a miniature model experiment and also evaluated the moisture risk of the CLT wall considering the climatic con­
ditions prevailing in 20 major cities in Korea through WUFI simulation analysis [20]. Fu et al. predicted relative temperature, relative
humidity (RH) and water content of two timber-framed buildings based on relevant material parameters applicable in Nanjing using
WUFI simulations. It was demonstrated that the thermal performance of the entire wall of the building with an insulating cork-board
was better than that with an anti-corrosion pine board [21]. McClung et al. monitored moisture content and temperature in a building
envelope for one year at a test facility in Waterloo, Ontario. The moisture content as predicted by them through a WUFI simulation
generally matched well with the field data. In particular, the hygrothermal behavior of the CLT panel was influenced more by the
construction of the wall assembly than by the wood species [22].
Recently, based on the climate change scenarios envisaged by the Intergovernmental Panel on Climate Change (IPCC), it is pre­
dicted that the average atmospheric temperature will increase by 1.5–6 ◦ C by the year 2100. This rise in temperature will have several
implications for buildings, such as an increase in the building energy consumption and changes in the hygrothermal behavior of the
building envelope. Although the hygrothermal performance of building walls is determined by the choice of materials, further research
is needed on hygrothermal performance of buildings in future climates [23]. There have been some investigations on the effect of
future climate on building performance in recent years; however, a majority of these studies were focused only on the building’s future
heating and cooling needs [24–27].
Therefore, in this study, the hygrothermal performance of CLT walls in future climatic conditions was evaluated considering
modular construction. The research framework of this study consists of the following accomplishments: 1) CLT walls with plywood
applied to a core layer was manufactured, 2) a mock-up composed of CLT walls was produced and the construction process was
analyzed, 3) Using WUFI simulation software, temperature and RH inside the walls of the structure were predicted and these results
were validated by comparing them against measurements obtained through mock-up experiments, and 4) Finally, the hygrothermal
performance of the CLT wall was evaluated for different types of insulation and future climates in eight US cities.

2. Methods and materials

2.1. Setup of mock-up experiment with ply-lam CLT
The plywood-laminated cross-laminated timber (ply-lam CLT) is an engineered wood, in which plywood is applied to a core layer of
an existing CLT. It has the advantage of replacing the low yields of glued timber with high yields of plywood. Fig. 1 presents the top
view and cross-section of the mock-up structure assembled with ply-lam CLT. The size of the mock-up is 2600 mm × 2450 mm. The
roof has a slope of 3◦ , and the height at the center of the structure is 2250 mm. A door and a window are placed on the north and south
sides, respectively, and the roof protrudes 200 mm on the north and south sides.

Fig. 1. Top view and cross-section of mock-up with ply-lam CLT.

2
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

The wall consists of a 9-mm plywood, 150-mm expanded polystyrene (EPS) insulation, and 90-mm ply-lam CLT. The red dots in
Fig. 1 represent measurement points, at which temperatures and RH were measured in the mock-up experiments inside, outside of the
structure, as well as within the wall. The temperature and RH inside the wall were measured from the temperature at the interface of
the ply-lam CLT and the insulation.
The construction of the mock-up using the ply-lam CLT includes following four steps: 1) ply-lam CLT manufacturing, 2) mock-up
frame construction, 3) insulation and exterior finishing construction, and 4) on-site construction of mock-up. Fig. 2 shows various
stages in building the mock-up with ply-lam CLT applied. The commonly used CLT consists of 3, 5, or 7 layers, and the choice of
number of layers depends on the scale and purpose of the application. In this study, a 3-layer ply-lam CLT and 30-mm plywood applied
to the core layer was prepared. Firstly, a 30-mm laminar plate, horizontally connected to a lamina, was placed on the bottom layer.
Next, a thin layer of Phenolic-resorcinol adhesive was applied to the top surface. Then, a 30-mm plywood coated with the adhesive on
both sides was placed on the lamina and finally, a 30-mm laminar plate was placed on the top layer. Fig. 2 (a-2) shows the compression
step of the ply-lam CLT, which was compressed at ambient temperature for 8 h under a pressure of 980.67 kPa. After the compression,
the adhesive that flowed out to the side was treated and the ply-lam CLT board was cut to size taking into consideration the locations
for the openings in the mock-up.

Fig. 2. Various stages in building the ply-lam CLT mock-up; a) Manufacturing process of ply-lam CLT, b) Construction process of frame for mock-up with ply-lam CLT,
c) Completion process of mock-up with ply-lam CLT, d) On-site construction process of mock-up with ply-lam CLT.

3
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Fig. 2 (b) shows the mock-up frame construction phase. The ply-lam CLT was erected and assembled on a floor made of 25-mm thick
antiseptic wood. Temperature sensors (thermocouple K-type) and RH sensors (B530) were installed on the ply-lam CLT surfaces of the
north and east walls of the mock-up. Once the frame was constructed, the walls and roof were completed, as shown in Fig. 2 (c). The
roof had ventilated cavities improve the building envelope performance with respect to moisture safety. After the assembly was
completed, the modular mock-up was moved to the site and mounted, as shown in Fig. 2 (d). The site was located near an environ­
mental experiment factory in Chungnam National University, Daejeon, Korea.
One key feature in the process of building the mock-up is that all stages other than the on-site construction were completed in one
factory, located in Sunchang, Jeollanam-do, Korea. The construction and completion work were carried out by four workers in just 10
h. In the same factory, the ply-lam CLT was also manufactured, thus saving the transportation cost and shortening the construction
time in the field. In particular, it was a safe and stable construction process, free from obstacles often faced in various natural envi­
ronments, such as rain, snow, wind, extreme cold, and heat. This demonstrates the significant advantage of modular buildings con­
structed out of CLT.

2.2. Validation of WUFI simulation

In this study, the hygrothermal behavior of the ply-lam CLT wall was evaluated through a WUFI software simulation and these
results were validated by comparing the same with those measured in the mock-up using thermocouples and RH sensors. The WUFI
simulation program is a dynamic analysis program for hygrothermal behavior of building envelopes, and can evaluate heat and
moisture [19]. It was developed through various studies and experiments under natural climatic and artificial conditions [28–30].
Fig. 3 shows the image of the wall of the mock-up with the ply-lam CLT simulated in the WUFI Pro 5.6 simulation program. On the
exterior plywood, EPS, and ply-lam CLT were applied. The outdoor and indoor climates were set up by converting the actually
measured temperature and RH data into climate data. In the WUFI simulation, it is possible to monitor the moisture content of the
entire wall or of individual materials, as well as the temperature and RH at a specific point. In this step, the same point where the sensor
was located was set as a monitoring point to compare the change in temperature and RH inside the mock-up wall with the change in
temperature and RH through simulation.
Among various physical properties of the material, the WUFI simulation analyzes the heat and moisture based on the hygrothermal
properties, which include the bulk density, porosity, specific heat capacity, thermal conductivity, and water vapor resistance factor of
the constituent materials. Table 1 shows the hygrothermal properties of the applied materials, which were adopted from a previous
study [20]. Indoor and outdoor boundary conditions were input into the simulation program based on the measured temperature and
RH.

2.3. Evaluation of moisture problems in ply-lam CLT walls

Table 2 shows the types of insulation that can be applied to CLT walls, based on the climate zones of eight major cities in the USA, as
suggested in the CLT handbook [20,31]. These U.S. cities were classified as 1–7 according to their thermal environment, and were
further categorized into moist (A), dry (B), and marine (C) regions, based on their hygrothermal environment [32]. The CLT handbook

Fig. 3. Section of the ply-lam CLT wall; a) for simulation verification of mock-up walls, b) for case study according to wall composition and climatic conditions.

4
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Table 1
Hygrothermal properties of the applied materials in mock-up made of ply-lam CLT.

Material Bulk density [kg/ Porosity [m3/ Specific heat capacity [J/ Thermal conductivity [W/ Water vapor resistance factor
m3] m3] (kg⋅K)] (m⋅K)] [− ]

Plywood board 500.0 0.5 1400.0 0.1 700.0

Expanded 15.0 0.95 1500.0 0.034 30.0
polystyrene
Ply-lam CLT 600.0 0.486 960.0 0.104 75.44

Table 2
Applicable insulation in CLT walls based on the climate zone of major cities in the USA.

City Climate Zone Exterior insulation type

Extruded polystyrene Expanded polystyrene Rigid mineral wool

Miami 1A Yes No No
Houston 2A Yes No No
Atlanta 3A Yes Yes No
San Francisco 3C Yes Yes Yes
Seattle 4C Yes Yes Yes
Boston 5A Yes Yes No
Minneapolis 6A Yes Yes No
Anchorage 7 Yes Yes Yes

contains an evaluation of the moisture problem in CLT walls by applying extruded polystyrene (XPS), EPS, and rigid mineral wool
insulation in the eight cities included in Table 2 [31]. Here, in each row, under the column “Exterior insulation type”, a “Yes” implies
that it can be applied as there would be no moisture problem in that particular city, and a “No” implies that it is inadvisable to apply
that insulation in that city. For example, Atlanta and San Francisco are in the same climate zone, temperature-wise; both have 3 as the
temperature designator, but differ in terms of moisture content; Atlanta is designated as a “moist” place (“A”), while San Francisco is
designated as a “marine” location (“C”). As a result, the applicability of rigid mineral wool insulation is different for the two cities.
Marine climates are typically characterized by narrower annual temperature ranges than other regions of the same latitude. Generally,
they experience neither extremely dry nor wet subtropical summers. Hence, it is possible to apply rigid mineral wool in marine climate
because it is sufficiently dry in summer; however, the application of the same rigid mineral wool in a moist climate causes moisture
problems because the walls are not sufficiently dried as the annual RH is such places is high.
The data summarized in Table 2 have more implications than simply serving as a guide for choosing the insulation type. The climate
can constantly change, and various problems have arisen owing to rapid climate changes in recent years. For a stable CLT structure
from long-term perspective, an evaluation considering the future climate conditions is necessary.
The CLT handbook provides examples of CLT walls, including rain screens [31]. The biggest difference between a wooden frame
structure and the CLT wall is that the latter has no studs. As the insulating material is not interrupted by the studs, there is an advantage
in that the heat loss that would occur through the studs does not occur in a CLT wall, and the insulating material could be applied in the
form of a seamless plate [33]. Furthermore, to solve the moisture problem, the moisture permeability is controlled by applying a
weather-resistive barrier (WRB) between the CLT and the insulation, and the role of rain-screen is fulfilled by a ventilated and drained
cavity outside the insulation.
Fig. 3 (b) shows the composition and monitoring position of the ply-lam CLT wall evaluated in this study. The ply-lam CLT wall
consisted of a 10-mm wood cladding, 20-mm air layer, 80-mm insulation material, 1-mm vapor permeable weather resistive barrier
(WRB), and 120-mm ply-lam CLT from the outside. A vertical wood strapping was applied to the air layer for the construction of wood
cladding and the rain-screen. Karacabeyli and Douglas classified vapor permeance categories according to the water vapor perme­
ability of the material [31]. In this study, as suggested in the CLT handbook, vapor permeable WRB having a moisture permeability of
10 US perms or more was constructed between the insulation and the ply-lam CLT. The hygrothermal properties of vapor permeable
WRB and four other insulation materials are summarized in Table 3. It can be noticed that the water vapor resistance factor for WRB is
340, and the sd - value considering the applied thickness is 0.43 m.

Table 3
Hygrothermal properties of four insulation materials and vapor permeable WRB.

Material Bulk density [kg/ Porosity [m3/ Specific heat capacity [J/ Thermal conductivity [W/ Water vapor resistance
m3] m3] (kg⋅K)] (m⋅K)] factor [− ]

Mineral wool 60.0 0.95 850.0 0.034 1.3

Expanded polystyrene 15.0 0.95 1500.0 0.034 30.0
Extruded polystyrene 40.0 0.95 1500.0 0.034 100.0
Wood–fiber board 155.0 0.981 1400.0 0.034 3.0
Weather resistive 130.0 0.001 2300.0 2.3 340.0
barrier

5
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

The main variables in the predictive simulation of the ply-lam CLT walls were the outdoor climatic conditions and type of insu­
lation. The insulations considered were XPS, EPS, mineral wool (MW), wood-fiber board. The thermal conductivity of all the insulation
materials was fixed at 0.034 W/(m⋅K) when evaluating the moisture problems on the wall based on the difference in the water vapor
resistance performance. The hygrothermal properties of XPS, EPS, MW, and WF are summarized in Table 3. It can be noted that the
water vapor resistance factors of these materials are 1.3, 30.0, 100.0, and 3.0, respectively.
To compare the results of this study with those presented in the CLT handbook, outdoor climatic conditions in the eight cities
evaluated in the CLT handbook were selected. These were: Miami (1A), Houston (2A), Atlanta (3A), San Francisco (3C), Seattle (4C),
Boston (5A), Minneapolis (6A), and Anchorage (7). The outdoor climate data of each city was calculated using the Meteonorm 7.2
program, which is a climate data calculation program. Climate data based on the measured past climate conditions and predicted
climate data for 2050 and 2080 of each city, considering a rapid climate change, were calculated and set as variables. Meteonorm
program can calculate the future climate data from the Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC
AR4) B1, A1B, and A2 change scenarios. In this study, the IPCC AR4 A1B scenario was selected, which denotes high and medium
emission levels [34].

3. Results and discussion

3.1. Validation of WUFI simulation against the mock-up experiment
The mock-up experiment was performed for 3 weeks from July 1 to July 22, 2018. Fig. 4 (a) shows the hourly temperature and RH
both indoor and outdoor during the mock-up experiment period. The outdoor temperature and RH were affected by the external
environment, and the indoor temperature was controlled through an air conditioner. The indoor temperature varied between 22.8 ◦ C
and 28.3 ◦ C, while the indoor RH varied between 49.6% and 64.5%. The temperature and RH variations with time were recorded as
indoor and outdoor climate data, and used as input data for the WUFI simulation.
Fig. 4 (b) shows a comparison of the variations in temperature and RH between the insulation and the ply-lam CLT in the north wall
between the mock-up experiment and WUFI simulation. In Fig. 4 (b), the temperature in the mock-up experiment exhibited a more
rapid change, while that from the WUFI simulation showed a less abrupt change; however, the temperatures of the two results showed
very similar trends. The RH also changed in a similar manner but showed a slightly larger difference compared to the temperature.
In this study, the mean bias error (MBE), root mean square error (RMSE), and the coefficient of variation of the root mean square
error (CV(RMSE)) presented by ASHRAE were calculated to compare the simulation and experiment results. On the basis of ASHRAE,
the models built with monthly data are deemed suitable for predictive simulations if the values of MBE and CV(RMSE) are within ±5%
and 15%, respectively. Similarly, the models built with hourly data are deemed suitable for simulation prediction evaluations if the
value of MBE and CV(RMSE) are within ±10% and 30%, respectively [35]. In the mock-up experiment of this study, the temperature
and humidity at each hour were measured for a 3-week period, and as they were compared with the simulation results, the criteria of
MBE within ±10% and CV(RMSE) within 30% were followed. The formulae for MBE, RMSE, and CV(RMSE) are as follows.

Fig. 4. a) Indoor and outdoor hourly temperatures and RH, b) Comparison of temperature and RH in the north ply-lam CLT walls between the mock-up experiment
and WUFI simulation.

6
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Table 4
Validation of temperature and RH results from simulation with measurements from mock-up experiments.

Validation method Temperature RH

MBE [%] − 0.95 − 0.79

RMSE 0.22 0.49
CV(RMSE) [%] 6.43 7.02

/
∑
n ∑
n
MBE = y i − yi )
(̂ yi × 100 (1)
i=1 i=1

√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
/
√ n
√∑
RMSE = √ (̂y i − yi )2 n (2)
i=1

CVRMSE = RMSE/y × 100 (3)

Here, n is the size the population (− ), yi is either the temperature ( C) or RH (%) from the mock-up experiment, ̂
◦
y i is either the
temperature (◦ C) or RH (%) from the simulation, and y is the mean of either the temperature (◦ C) or RH (%) from the mock-up
experiment. Table 4 summarizes the validation results for temperature and RH from the mock-up experiment and simulations. MBE
of temperature and RH were − 0.95% and − 0.79%, respectively, which are less than 10%; therefore, the ASHRAE standards were
satisfied, and the temperature data can be considered reliable. CV(RMSE) of temperature and RH were 6.43% and 7.02%, respectively,
which shows that these results also met the ASHRAE standards. Therefore, these results proved that WUFI simulation software is
suitable and sufficiently accurate for the prediction of moisture problems.

3.2. Evaluation of moisture problems in ply-lam CLT walls considering future climate
3.2.1. Total water contents (TWC) in ply-lam CLT walls
In this study, the moisture problem was evaluated by analyzing the TWC, moisture content of the insulation, and mold growth risk
of the ply-lam CLT walls based on the outdoor climatic conditions and insulation type from WUFI simulation. The TWC was evaluated
for the drying potential of the wall based on the simulation results over three years. The evaluation of TWC indicates whether a
moisture increase or decrease takes place in the wall component during the investigated period. The increase in TWC during the
investigated period indicates that the wall components cannot dry themselves in the long term under climatic conditions, which means
that moisture problems may occur not only now but also in the future. In addition, when the initial moisture content is too low, the
moisture content increases over several years, in which case the TWC should be re-evaluated with a longer investigated period.
Fig. 5 shows the results of the TWC. The gray line shows the results of all the variable conditions corresponding to the three climatic
conditions (past, year 2050, and year 2080) of the eight cities, and the red line shows the average of the total water content for each
insulation type. As seen from Fig. 5, the wall composed of ply-lam CLT was considered to be free from long-term moisture problems
under all conditions because the TWC was consistent and did not increase. Although the difference in absolute amount appeared to be
depending on the application of each insulation material, the rate of change was the key in the analysis of the TWC.

3.2.2. Water content of insulation materials

WUFI also calculates the moisture content at different layers. In this study, the moisture content of the insulation materials was
analyzed separately. The reason for analyzing this parameter is that the thermal performance of the material decreases as the moisture
content of the material increases. Owing to the moisture in the pores, the thermal conductivity of the material increases, which leads to
a decrease in building energy performance [37,38]. In the case of wood, numerous problems occur when the moisture content is more
than 20% [36,39]. Fig. 6 shows the change in the moisture content of the insulation based on the climate change. The moisture content
of the insulation in decreasing order was from Miami (1A), which had the highest annual average temperature, to Anchorage (7),
which had the lowest annual average temperature.
At the same RH, high temperature air has a higher partial pressure of water vapor than low temperature air. Because the average
annual RH values of the eight cities were similar, cities with a high annual average temperature had a relatively high partial pressure of
water vapor, which helped increase the moisture content of the insulation applied therein. For the same reason, the moisture content of
the insulation increased in almost all the variable conditions, when future climate was applied. At this time, the thickness of the
insulation material is the same. In particular, Minneapolis (6A) showed the largest increase because its annual average temperature in
2080 increased by 3.0 ◦ C and the RH also increased by 2% compared to the past. Therefore, these results indicate that the increase in
temperature not only affected the increase in energy use of the building but also affected the moisture problem of the building
envelope.

7
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Fig. 5. Total water content with different insulation materials: a) XPS, b) EPS, c) mineral wool, and d) wood–fiber board.

Fig. 6. Moisture content of the insulation as a function of climate change for different insulation materials: a) XPS, b) EPS, c) mineral wool, and d) wood–fiber board.

8
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

3.2.3. Assessment of mold growth risk

Factors affecting mold germination and growth are temperature, RH, substrate, pH value, light, oxygen content, and biotic in­
fluences [20,40]. Among mold germination factors, it is most affected by temperature and RH, and the WUFI simulation program
evaluates mold growth risk through critical relative humidity curves based on temperature and RH. This curve is the result of analyzing
the conditions for most mold germination that can occur in building materials based on temperature and RH, and it is a curve
expressing the minimum RH at which mold can germinate at each temperature. In this study, the risk of mold growth was evaluated
based on temperature and RH. Since mold growth risk was assessed only with temperature and RH, if the critical line is exceeded, this
does not necessarily mean that mold will grow, but that mold growth cannot be excluded. The initiation conditions for mold growth
can be expressed by a mathematical formula with critical RH as a function of temperature, as given by Equation (4) [41,42].
no growth is assumed, when T ≤ 0 or 50 < T
RHrisk = { 80, when 20 < T ≤ 50 (4)
− 0.0026 × T 3 + 0.16 × T 2 − 3.13 × T + 100.0, when 0 < T ≤ 20
Here, RHrisk is the minimum RH for risk of mold growth (%) and T is the temperature (◦ C). Table 5 shows the results from an
evaluation of the mold growth risk when climate data calculated based on past climate conditions were applied. In Table 5, “None”
means that there was no risk of mold growth, and “Risk” means that there was a possibility of mold growth risk. The application of XPS
insulation with the high water vapor resistance factor of 100 did not cause moisture problems in any climatic condition, while the
application of EPS insulation with a water vapor resistance factor of 30 exposed the structure to the risk of mold growth in Miami (1A)
and Houston (2A). Due to the relatively high annual temperatures and RHs of these two cities, the high annual water vapor partial
pressure of the outdoor air appeared to have a significant impact on the inside of the wall. Similarly, the application of mineral wool
and wood-fiber board also exposed the structure to the risk of mold growth in Miami (1A) and Houston (2A) owing to their high
temperature and RH. The application of XPS insulation did not cause a problem because the moisture permeating to the inside could be
properly controlled by the high resistance of the material to moisture permeation. Another finding was that there were no moisture
problems with the application of any insulation material in San Francisco (3C) and Seattle (4C). The two cities have a marine climate
among the evaluated cities, and the influence of water vapor pressure on the outdoor air was low owing to a low RH in summer;
therefore, the effect on the inside of the wall was insignificant. The results of mold growth of ply-lam CLT walls with XPS, EPS, and
mineral wool insulation were the same as those presented in the CLT handbook. This means that the results of this study are highly
reliable. The wood-fiber board had a water vapor resistance factor of 3.0, which was similar to that of mineral wool, and the results of
the ply-lam CLT wall with these two insulation materials were also similar.
Fig. 7 shows the results of mold growth risk assessment under future climatic conditions, showing the hourly temperature and RH
for a total of three years of simulation. The black dots represent the results of applying past climate data, the gray dots represent the
results of applying future climate data, and the red line represents the critical RH of Equation (4). An interesting finding is that the risk
of mold growth increased when future climate data were used compared to when past climate data were used.
This was because, as explained earlier, the temperature of the outdoor air had increased due to the climate change. As with the
results in Table 5, the XPS application was evaluated to have no moisture problem in 2080 climatic conditions in all cities, and showed
very stable results for the CLT wall structure. In contrast, the ply-lam CLT wall with EPS and wood-fiber board insulation had increased
the risk of moisture problems in future climatic conditions, which had not occurred in the past.
Table 6 summarizes the results of applying the future climate of the cities for insulation and climatic conditions assessed as posing a
risk of mold growth in past climatic conditions. The number represents the time in hours exceeding the mold growth risk condition for
a total of three years (26,280 h) of simulation. The application of mineral wool insulation caused moisture problems in more cities than
the other insulations, and the application of future climate data increased the time to exceed the mold growth risk conditions. Mold is
affected not only by temperature and humidity, but also by how long it is exposed to poor environments; therefore, moisture problems

Table 5
Results from the evaluation of mold growth risk in past climate conditions.

City Climate Zone Exterior insulation type

Extruded polystyrene Expanded polystyrene Mineral wool Wood–fiber board

Miami 1A None Risk Risk Risk

Houston 2A None Risk Risk Risk
Atlanta 3A None None Risk None
San Francisco 3C None None None None
Seattle 4C None None None None
Boston 5A None None Risk None
Minneapolis 6A None None Risk None
Anchorage 7 None None None None

9
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Fig. 7. Mold growth risk assessment under future climatic conditions.

10
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

Table 6
Results of evaluation of mold growth risk in future climatic conditions.

Insulation type Climate Time that exceeds the risk criterion for mold growth for 8760 h [hours]

Past 2050 2080

Mineral wool Miami (1A) 3467 6013 5965

Houston (2A) 2504 3085 3769
Atlanta (3A) 101 718 1184
Boston (5A) 14 66 130
Minneapolis (6A) 2 115 247
Expanded polystyrene Miami (1A) 756 3732 3808
Houston (2A) 375 1476 2255
Wood–fiber board Miami (1A) 2851 5903 5590
Houston (2A) 1759 2736 3627

Fig. 8. Mold growth risk assessment of ply-lam CLT walls with EPS insulation applied in a) Atlanta and b) Minneapolis.

in the walls can be caused by climate change. Similarly, if EPS and wood-fiber boards had problems in past climatic conditions,
problems also developed in future climates, and the risk was even greater.
The increased risk for moisture problems occurred even when there were no problems in the past climatic conditions. Fig. 8 shows
the results of the mold growth risk assessment of ply-lam CLT walls with EPS insulation applied to Atlanta (3A) and Minneapolis (3C).
When climatic conditions of the past were applied, the ply-lam CLT walls with EPS insulation had problems only in Miami (1A) and
Houston (2A); there were no problems of exceeding the standard in Atlanta (3A) and Minneapolis (3C). However, as shown in Fig. 8
(a), when the 2050 and 2080 climate data of Atlanta (3A) were applied, the RH gradually increased, and for the 2080 climate data, the
mold growth risk criterion was exceeded for a total of 180 h. This result offers a straightforward explanation that moisture problems
may occur in the future with rapid climate changes, even for the walls considered to be free of moisture problems, as per the CLT
handbook. In the case of Minneapolis (3C), although RHrisk did not exceed the standard in 2080 climatic conditions, it gradually
increased according to the climate change. This result also means that the CLT wall design should be carried out considering the future
climate. To revitalize the modular construction using CLT, it is necessary to design a wall that secures moisture stability through an
analysis of the future climate of the building site.

4. Conclusion
In this study, WUFI simulation was validated against mock-up experiments with ply-lam CLT, and the hygrothermal performance of
ply-lam CLT walls according to the insulation type and future climate data was evaluated. Starting with an analysis of the mock-up
construction process, the ply-lam CLT-applied mock-up was assembled and completed in a factory, where the ply-lam CLT was also
manufactured. On-site construction of the mock-up involved moving and placing the completed modular mock-up on-site. This implies
that the modular construction with CLT is free from the effects of various natural environmental conditions. From the results of the
WUFI simulation validation, the CV(RMSE) of the temperature and RH were 6.43% and 7.02%, respectively, which indicate that the
prediction of hygrothermal performance using WUFI was reliable. In this study, the moisture problem was evaluated by analyzing the
total water content, moisture content of the insulation, and mold growth risk of ply-lam CLT walls according to outdoor climatic
conditions and insulation type using WUFI simulation. As a result of the analysis of total water content, it was evaluated that no long-
term moisture problem occurred under any variable conditions. The future climate would increase the moisture content of the
insulation. This would not only cause moisture problems in the insulating materials but also degrade the thermal insulation perfor­
mance of the wall; this would increase the building energy consumption. Finally, the risk of mold growth was evaluated based on the
critical relative humidity as a function of temperature. The results of mold growth of ply-lam CLT walls with XPS, EPS, and mineral
wool insulation agreed well with those presented in the CLT handbook. Only the ply-lam CLT walls with XPS insulation were not
exposed to the criteria for mold growth risk in any of the cities in 2080 climatic conditions; however, the risk of mold growth increased

11
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

with climate change for all insulation conditions. Therefore, for the CLT building with modular construction to be free from moisture
problems, it is necessary to predict and evaluate the hygrothermal performance considering both the wall composition and future
climate conditions of the building site.

Author statement
Seong Jin Chang: Writing - Original Draft, Conceptualization, Yujin Kang: Writing - Review & Editing, Beom Yeol Yun: Concep­
tualization, Sungwoong Yang: Methodology, Sumin Kim: Supervision, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgment
This work was supported by a Gyeongsang National University grant in 2020–2021.

References
[1] Y.S. Chua, J.Y.R. Liew, S.D. Pang, Modelling of connections and lateral behavior of high-rise modular steel buildings, J. Constr. Steel Res. 166 (2020), 105901,
https://doi.org/10.1016/j.jcsr.2019.105901.
[2] M. Kamali, K. Hewage, R. Sadiq, Conventional versus modular construction methods: a comparative cradle-to-gate LCA for residential buildings, Energy Build.
204 (2019), https://doi.org/10.1016/j.enbuild.2019.109479.
[3] S.V. Sendanayake, D.P. Thambiratnam, N. Perera, T. Chan, S. Aghdamy, Seismic mitigation of steel modular building structures through innovative inter-
modular connections, Heliyon 5 (2019), e02751, https://doi.org/10.1016/j.heliyon.2019.e02751.
[4] Z. Wang, W. Pan, A hybrid coupled wall system with replaceable steel coupling beams for high-rise modular buildings, J. Build. Eng. 31 (2020), 101355, https://
doi.org/10.1016/j.jobe.2020.101355.
[5] A.W. Lacey, W. Chen, H. Hao, K. Bi, Structural response of modular buildings – an overview, J. Build. Eng. 16 (2018) 45–56, https://doi.org/10.1016/j.
jobe.2017.12.008.
[6] A.W. Lacey, W. Chen, H. Hao, K. Bi, Review of bolted inter-module connections in modular steel buildings, J. Build. Eng. 23 (2019) 207–219, https://doi.org/
10.1016/j.jobe.2019.01.035.
[7] F.J. Luo, Y. Bai, J. Hou, Y. Huang, Progressive collapse analysis and structural robustness of steel-framed modular buildings, Eng. Fail. Anal. 104 (2019)
643–656, https://doi.org/10.1016/j.engfailanal.2019.06.044.
[8] M. Head, A. Levasseur, R. Beauregard, M. Margni, Dynamic greenhouse gas life cycle inventory and impact profiles of wood used in Canadian buildings, Build.
Environ. 173 (2020), 106751, https://doi.org/10.1016/j.buildenv.2020.106751.
[9] H. Pangh, H.Z. Hosseinabadi, N. Kotlarewski, P. Moradpour, M. Lee, G. Nolan, Flexural performance of cross-laminated timber constructed from fibre-managed
plantation eucalyptus, Construct. Build. Mater. 208 (2019) 535–542, https://doi.org/10.1016/j.conbuildmat.2019.03.010.
[10] A. Hassanieh, H.R. Valipour, M.A. Bradford, Experimental and numerical investigation of short-term behaviour of CLT-steel composite beams, Eng. Struct. 144
(2017) 43–57, https://doi.org/10.1016/j.engstruct.2017.04.052.
[11] H.M. Cho, S. Wi, S.J. Chang, S. Kim, Hygrothermal properties analysis of cross-laminated timber wall with internal and external insulation systems, J. Clean.
Prod. 231 (2019) 1353–1363, https://doi.org/10.1016/j.jclepro.2019.05.197.
[12] Z. Wang, M. Gong, Y.-H. Chui, Mechanical properties of laminated strand lumber and hybrid cross-laminated timber, Construct. Build. Mater. 101 (2015)
622–627, https://doi.org/10.1016/j.conbuildmat.2015.10.035.
[13] F. Stazi, M. Serpilli, G. Maracchini, A. Pavone, An experimental and numerical study on CLT panels used as infill shear walls for RC buildings retrofit, Construct.
Build. Mater. 211 (2019) 605–616, https://doi.org/10.1016/j.conbuildmat.2019.03.196.
[14] M. Izzi, D. Casagrande, S. Bezzi, D. Pasca, M. Follesa, R. Tomasi, Seismic behaviour of Cross-Laminated Timber structures: a state-of-the-art review, Eng. Struct.
170 (2018) 42–52, https://doi.org/10.1016/j.engstruct.2018.05.060.
[15] L. Setter, E. Smoorenburg, S. Wijesuriya, P.C. Tabares-Velasco, Energy and hygrothermal performance of cross laminated timber single-family homes subjected
to constant and variable electric rates, J. Build. Eng. 25 (2019), 100784, https://doi.org/10.1016/j.jobe.2019.100784.
[16] L. Franzoni, A. Lebée, F. Lyon, G. Forêt, Closed-form solutions for predicting the thick elastic plate behavior of CLT and timber panels with gaps, Eng. Struct. 164
(2018) 290–304, https://doi.org/10.1016/j.engstruct.2018.02.073.
[17] Temporhousing, CLT modular homes. http://www.tempohousing.com/products/clt-modular-homes/, 2019.
[18] Wufi homepage, (n.d.). https://wufi.de/en/.
[19] H.M. Künzel, Simultaneous Heat and Moisture Transport in Building Components: One- and Two-Dimensional Calculation Using Simple Parameters, Fraunhofer
IRB Verlag Stuttgart, 1995. ISBN v.3-8167-4103-7.
[20] S.J. Chang, S. Wi, S.G. Kang, S. Kim, Moisture risk assessment of cross-laminated timber walls: perspectives on climate conditions and water vapor resistance
performance of building materials, Build. Environ. 168 (2020), 106502, https://doi.org/10.1016/j.buildenv.2019.106502.
[21] H. Fu, Y. Ding, M. Li, H. Li, X. Huang, Z. Wang, Research on thermal performance and hygrothermal behavior of timber-framed walls with different external
insulation layer: insulation Cork Board and anti-corrosion pine plate, J. Build. Eng. 28 (2020), 101069, https://doi.org/10.1016/j.jobe.2019.101069.
[22] R. McClung, H. Ge, J. Straube, J. Wang, Hygrothermal performance of cross-laminated timber wall assemblies with built-in moisture: field measurements and
simulations, Build. Environ. 71 (2014) 95–110, https://doi.org/10.1016/j.buildenv.2013.09.008.
[23] A. Fang, Y. Chen, L. Wu, Transient simulation of coupled heat and moisture transfer through multi-layer walls exposed to future climate in the hot and humid
southern China area, Sustain. Cities Soc. 52 (2020), 101812, https://doi.org/10.1016/j.scs.2019.101812.
[24] A. Sehizadeh, H. Ge, Impact of future climates on the durability of typical residential wall assemblies retrofitted to the PassiveHaus for the Eastern Canada
region, Build. Environ. 97 (2016) 111–125, https://doi.org/10.1016/j.buildenv.2015.11.032.
[25] M.R. Gaterell, M.E. McEvoy, The impact of climate change uncertainties on the performance of energy efficiency measures applied to dwellings, Energy Build.
37 (2005) 982–995, https://doi.org/10.1016/j.enbuild.2004.12.015.
[26] K.J. Lomas, R. Giridharan, Thermal comfort standards, measured internal temperatures and thermal resilience to climate change of free-running buildings: a
case-study of hospital wards, Build. Environ. 55 (2012) 57–72, https://doi.org/10.1016/j.buildenv.2011.12.006.
[27] T. Berger, C. Amann, H. Formayer, A. Korjenic, B. Pospischal, C. Neururer, R. Smutny, Impacts of climate change upon cooling and heating energy demand of
office buildings in Vienna, Austria, Energy Build. 80 (2014) 517–530, https://doi.org/10.1016/j.enbuild.2014.03.084.
[28] S. Charisi, T.K. Thiis, P. Stefansson, I. Burud, Prediction model of microclimatic surface conditions on building façades, Build. Environ. 128 (2018) 46–54,
https://doi.org/10.1016/j.buildenv.2017.11.017.
[29] B. Moujalled, Y. Aït Ouméziane, S. Moissette, M. Bart, C. Lanos, D. Samri, Experimental and numerical evaluation of the hygrothermal performance of a hemp
lime concrete building: a long term case study, Build. Environ. 136 (2018) 11–27, https://doi.org/10.1016/j.buildenv.2018.03.025.

12
S.J. Chang et al. Case Studies in Thermal Engineering 28 (2021) 101703

[30] M. Ibrahim, H. Sayegh, L. Bianco, E. Wurtz, Hygrothermal performance of novel internal and external super-insulating systems: in-situ experimental study and
1D/2D numerical modeling, Appl. Therm. Eng. 150 (2019) 1306–1327, https://doi.org/10.1016/j.applthermaleng.2019.01.054.
[31] E. Karacabeyli, B. Douglas, CLT Handbook, – US edition, 2013.
[32] International Energy Conservation Code, U.S. Department of Energy, 2012.
[33] S.J. Chang, S. Wi, S. Kim, Thermal bridging analysis of connections in cross-laminated timber buildings based on ISO 10211, Construct. Build. Mater. 213 (2019)
709–722, https://doi.org/10.1016/j.conbuildmat.2019.04.009.
[34] Y.L. Song, D.L. Chen, Y.J. Liu, Y. Xu, The influence of climate change on winter wheat during 2012-2100 under A2 and A1B scenarios in China, Adv. Clim.
Change Res. 3 (2012) 138–146, https://doi.org/10.3724/SP.J.1248.2012.00138.
[35] ASHRAE Guideline 14-2002 – Measurement of Energy and Demand Savings, 2002.
[36] D. Zirkelbach, T. Schmidt, M. Kehrer, H.M. Künzel, WUFI Pro Manual, Fraunhofer Institute for Building Physics (IBP), 2007.
[37] W. Yang, J. Liu, Y. Wang, S. Gao, Experimental study on the thermal conductivity of aerogel-enhanced insulating materials under various hygrothermal
environments, Energy Build. 206 (2020), 109583, https://doi.org/10.1016/j.enbuild.2019.109583.
[38] R.H. Nosrati, U. Berardi, Hygrothermal characteristics of aerogel-enhanced insulating materials under different humidity and temperature conditions, Energy
Build. 158 (2018) 698–711, https://doi.org/10.1016/j.enbuild.2017.09.079.
[39] J.H. Park, Y. Kang, J. Lee, S.J. Chang, S. Wi, S. Kim, Development of wood-lime boards as building materials improving thermal and moisture performance based
on hygrothermal behavior evaluation, Construct. Build. Mater. 204 (2019) 576–585, https://doi.org/10.1016/j.conbuildmat.2019.01.139.
[40] K. Sedlbauer, Prediction of Mould Fungus Formation on the Surface of and inside Building Components, Stuttgart University, 2001.
[41] F. Fedorik, M. Malaska, R. Hannila, A. Haapala, Improving the thermal performance of concrete-sandwich envelopes in relation to the moisture behaviour of
building structures in boreal conditions, Energy Build. 107 (2015) 226–233, https://doi.org/10.1016/j.enbuild.2015.08.020.
[42] H. Viitanen, J. Vinha, K. Salminen, T. Ojanen, R. Peuhkuri, L. Paajanen, K. Lähdesmäki, Moisture and bio-deterioration risk of building materials and structures,
J. Build. Phys. 33 (2010) 201–224, https://doi.org/10.1177/1744259109343511.

13