Energy and Buildings 38 (2006) 1270–1282

www.elsevier.com/locate/enbuild

Moisture buffering capacity of hygroscopic building materials:

Experimental facilities and energy impact
Olalekan F. Osanyintola, Carey J. Simonson *
Department of Mechanical Engineering, University of Saskatchewan, 57 Campus Drive, Saskatoon, SK, Canada S7N 5A9
Received 8 September 2005; received in revised form 2 March 2006; accepted 20 March 2006

Abstract
Research into dynamic moisture storage in hygroscopic building materials has renewed interest in the moisture buffering capacity of building
materials and shown the potential for these materials to improve indoor humidity, thermal comfort and indoor air quality in buildings. This paper
complements previous research by estimating the effect of hygroscopic materials on energy consumptions in buildings. The results show that it may
be possible to reduce heating and cooling energy consumption by up to 5% and 30%, respectively, when applying hygroscopic materials with well-
controlled HVAC systems. The paper also describes two different experimental facilities that can be used to measure accurately the moisture
buffering capacity of hygroscopic building materials. These facilities provide different convective transfer coefficients between the hygroscopic
material and ambient air, ranging from natural convection in small, sealed jars to fully developed laminar and turbulent forced convection. The
paper presents a numerical model and property data for spruce plywood which will be used in a companion paper [O.F. Osanyintola, P. Talukdar,
C.J. Simonson, Effect of initial conditions, boundary conditions and thickness on the moisture buffering capacity of spruce plywood, Energy and
Buildings (2006), doi:10.1016/j.enbuild.2006.03.024.] to provide additional insight into the design of an experiment to measure the moisture
buffering capacity of hygroscopic materials.
# 2006 Elsevier B.V. All rights reserved.

Keywords: Moisture buffering capacity; Energy savings; Experimental facility; Uncertainty; Indoor air quality; Convective transfer coefficients; Spruce plywood

1. Introduction provides comfort and air quality near desired levels, while
consuming less energy than traditional methodologies.
In recent times, there has been increasing interest to reduce Furthermore, humidity is an important parameter in emergency
the energy consumption and green house gas emissions shelters [8] and supermarkets [9]. The RH is often too high for
associated with the use of mechanical (active) HVAC systems comfort in shelters that are passively heated by occupants and
in buildings. In view of this, researchers are investigating the solar gains even during cold weather, while the RH in
use of passive systems to assist or even eliminate some aspects supermarkets may vary significantly and is often closely linked
of these active systems or to control active systems more to the outdoor temperature. The indoor RH affects the
efficiently. One important aspect is moderating the indoor refrigeration load of freezer rooms and display cases and as
variations in relative humidity (RH) in buildings because indoor a result, indoor RH must be included when designing energy
humidity affects warm respiratory comfort [2], skin humidity recovery for supermarket refrigeration systems [9]. In addition,
[3] and perceived indoor air quality [4]. Also, moisture in conservation researchers have shown that a wide variety of
buildings has been shown to affect the sensible and latent artifacts displayed in museums require specific indoor
conduction loads [5] and may cause deteriorations in buildings conditions to minimize their deterioration. Yu et al. [10]
[6]. A recent study [7] included indoor RH as one of the control investigated the use of silica gel as an absorbent to control
parameters in a new HVAC control methodology which humidity in museums.
Due to the importance of indoor humidity, several
researchers [11–20] have studied the use of various hygroscopic
DOI of original article: 10.1016/j.enbuild.2006.03.024.
materials to moderate indoor humidity levels. These studies
* Corresponding author. Tel.: +1 306 966 5479; fax: +1 306 966 5427. have included laboratory, field and numerical studies and have
E-mail address: Carey.Simonson@usask.ca (C.J. Simonson). shown that hygroscopic materials are able to moderate the
0378-7788/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.enbuild.2006.03.026
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1271

indoor humidity levels and thus improve the thermal comfort

Nomenclature and perceived air quality in buildings, while still providing low
energy consumption [20,21]. The impact of hygroscopic
Cp specific heat capacity at constant pressure (J/
materials depends on many factors: the amount and type of
(kg K))
materials in a given room, the outdoor climate, the outdoor
Da binary diffusion coefficient for water vapor in air
ventilation rate and the moisture production rate, which also
(m2/s)
depends on the indoor temperature and RH [22]. During warm
Deff effective vapor diffusion coefficient (m2/s)
and humid outdoor conditions, hygroscopic materials may
Dh hydraulic diameter (m)
reduce the peak humidity in a bedroom by up to 35%, 30% and
ha convective heat transfer coefficient (W/(m2 K))
20% RH when the ventilation rate is 0.1, 0.5 and 1 ach,
hfg latent heat of vaporization/sorption (J/kg)
respectively [11,12]. At a ventilation rate of 0.5 ach, these
hm convective mass transfer coefficient (m/s)
reductions in peak indoor relative humidity result in a 10–20%
H enthalpy (kJ/kg)
reduction in the percent dissatisfied with warm respiratory
Hdesired desired indoor enthalpy (kJ/kg)
comfort and a 20–30% reduction in the percent dissatisfied with
Hindoor calculated indoor enthalpy (kJ/kg)
perceived air quality. The hygroscopic materials used in these
k thermal conductivity (W/(m K))
studies were wooden paneling, porous wood fiber board and
keff effective thermal conductivity (W/(m K))
cellulose insulation, but other studies [11–20] have used log
L effective thickness of specimen defined as the
panels, cellular concrete, furniture, fabric and other materials.
distance between the surface exposed to ambient
In addition, the porous materials in wall and flooring
air and the impermeable plane in the test speci-
constructions have been found to be capable of buffering
men (m)
indoor relative humidity and temperature [23]. This paper will
ṁ phase change rate per unit volume (kg/(m3 s)) or
extend this research by using published data [11,12] to
mass flow rate (kg/s) in Eq. (11)
determine the potential impacts that hygroscopic materials may
MBC moisture buffering capacity defined as a measure
have on the energy consumption in buildings.
of the mass of moisture that a material can absorb
The ability of materials to damp (or buffer) diurnal changes
and desorb during a specified humidity cycle per
in indoor humidity depends on their active thickness, vapor
unit exposed surface area (g/m2)
permeability and moisture storage capacity. The active
Nu Nusselt number
thickness (or depth to which moisture will penetrate for a
Q energy transfer (kWh)
given cycle) varies significantly for different materials. For
Re Reynolds number of air flow over the specimen
example, Olutimayin and Simonson [24] measured the
RH relative humidity (%)
development of the vapor boundary layer in a bed of cellulose
t time (s)
insulation following a step change in ambient humidity and
T temperature (8C)
found that the moisture penetration depth was 300 mm, 10 h
TMT transient moisture transfer
after a step change in the ambient humidity. They also
u mass of moisture adsorbed per kg of dry spruce
introduced a moisture property (moisture diffusivity) that is
plywood (kg/kg)
analogous to thermal diffusivity for heat transfer, which takes
x distance from the top of plywood specimen (m)
into account moisture storage. This property can be use to
Greek symbols calculate the active thickness of a hygroscopic material.
d water vapor permeability (kg/(m s Pa)) Neglecting moisture storage over predicts the active thickness
D difference by a factor of 10 for cellulose insulation.
e volume fraction As noted previously, proper RH control is an important
r density (kg/m3) environmental factor not only for humans but also for artifact
r0 dry density of the plywood specimen (kg/m3) preservation and passive systems (mostly hygroscopic materials)
f relative humidity in fraction have some potential to assist in controlling indoor relative
humidity. Research indicates that there are many materials that
Subscripts have the capacity to store moisture and buffer indoor humidity
a dry air changes, but a standard test is needed to compare the ability of
eff effective porous media property various materials to buffer indoor humidity changes [25,26].
g gas phase (air and water vapor) Furthermore, two test methods developed to measure the moisture
i initial buffering capacity of building materials propose different
‘ adsorbed phase boundary conditions and material thicknesses [27,28]. The
s solid experimental facilities and numerical model presented in this
v vapor paper support such standard development. A companion paper [1]
1 ambient or free stream property will apply the facilities and model to investigate the moisture
buffering capacity of spruce plywood and document the effect of
boundary conditions (convective transfer coefficient and humid-
ity cycle) and specimen thickness on moisture buffering capacity.
1272 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

aluminium-foil tape to eliminate lateral moisture transfer

through the edges. This allows the results to be compared on a
per area basis and also represents the application of plywood in
practice where moisture transfer through the edges is minimal.
In both facilities, the hygroscopic material can be exposed to
different humidity cycles, but in this paper, the plywood is
exposed to the same 24-h humidity cycle in both facilities—
75% RH for 8 h which is followed by a step change to 33% RH
Fig. 1. Scanning electron microscope picture of spruce heartwood showing the for 16 h. This cycle is intended to represent diurnal changes in
cell walls and the cell lumens as well as the main direction of moisture transfer
indoor humidity and is repeated several times. The change in
for plywood [28].
mass is measured gravimetrically with time in each facility.

2. Plywood material 3.1. Natural convection in 1 L glass jars

Plywood is a common building material that gains or This facility tests a small sample of spruce plywood
releases moisture and heat as the outdoor and indoor conditions (60 mm
60 mm
9 mm) using glass jars containing still air
change. It is manufactured from different species of peeled and saturated salt solutions (as shown in Fig. 2). In this facility,
wood veneers, such as spruce, oak and pine. These veneers are both faces of the plywood are exposed to the air in the jar, which
glued together, layer by layer, to form a panel. Plywood creates an impermeable layer at half the thickness of the
products are produced to be able to withstand extreme weather plywood. The numerical model, presented in Section 4, uses a
conditions by using phenolic formaldehyde resin in the gluing convection boundary condition at x = 0 and an impermeable
process. boundary condition at x = L. Therefore, an effective thickness
All wood products contain moisture, from saturated fresh cut (L) defined as the distance between the surface experiencing
logs to the fairly dry wooden indoor structures and furniture. convective moisture transfer and the impermeable plane in the
Moisture in wood is stored as either bound water or free water. test specimen is introduced and will be used throughout this
Bound water is held within cell walls by bonding forces paper. It will be useful when comparing the results of the two
between water and cellulose molecules. Free water is contained facilities in the companion of this paper [1]. In the glass jar
in the cell lumens/cavities and is held by surface tension. A facility, the effective thickness (L) for moisture penetration is
microscopic view of wood is shown in Fig. 1. Since plywood 4.5 mm from each exposed side.
veneers are made by rotating the log and peeling a thin veneer Prior to testing, the plywood needs to be conditioned to a
from the log, moisture transfer in the veneer in the direction of uniform moisture content. In this paper, the plywood samples
the thin dimension is equivalent to moisture transfer in the are conditioned for a long time (2 months) in the laboratory and
radial direction of the log. When these veneers are assembled the initial moisture content of the wood is 0.028 kg/kg, which
and used in buildings and furniture, the moisture transfer corresponds to a relative humidity of about 55%. The plywood
through the exposed surface and into the plywood is equivalent sample is then placed in a jar containing a saturated solution
to moisture transfer in the radial direction of the original log. and the jars are kept in an environmental chamber that is
Therefore, the direction of moisture transfer considered in this maintained at 23.3  0.3 8C during the test. The plywood is
paper is across the cell walls and lumens (in the radial direction subjected to a step change in relative humidity by moving it to a
of the original log) as shown in Fig. 1. Because of the rotary jar with a different salt solution. NaCl is used to create the high
peeled veneers, plywood will have more uniform moisture humidity condition, which creates a humidity of 75.3  0.1%
transfer characteristics than raw timber for example, which will RH at 23 8C [48] and MgCl2 is used to create the low humidity
have moisture transfer in directions that are both radial and condition, which creates a humidity of 33.1  0.2% RH at
tangential to the wood grains. Therefore, plywood is a good
material for the experimental and numerical investigation of
moisture buffering capacity.

3. Experimental facilities

In this paper, two facilities are presented that can measure

the moisture buffering capacity (MBC) of hygroscopic
materials. Each of these facilities creates different convective
transfer coefficients between the humid air and the plywood
and thus different boundary conditions ranging from natural
convention in sealed small jars to fully developed, forced
convection air flow (laminar and turbulent) in a small wind
tunnel. The size of the sample in the test facilities is also Fig. 2. Picture showing the spruce plywood in a sealed glass jar containing still
different, but the edges of all plywood pieces are sealed with air and a saturated salt solution.
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1273

the plywood is first placed in the jar, a temperature/humidity

transmitter with uncertainties of 0.1 8C and 1% RH is
placed (along with the plywood) in the jar to monitor the
temperature and humidity in the jar during a typical test. The
transmitter is moved along with the plywood from one jar to
another jar during the test and the measured temperature and
humidity of the air in the jar are presented in Fig. 3. As can be
seen in Fig. 3, the plywood is initially exposed to 75% RH air
for 8 h, followed by 33% RH air for 16 h. This cycle is repeated
three times. The data in Fig. 3 is recorded every 1 min and
therefore it can be seen that it takes about 4 min to realize the
change in humidity between 75 and 33% RH. As a comparative
test, the temperature/humidity transmitter alone (without the
plywood) is moved from one jar to another to check the
transient response of the humidity sensor alone. It was found
Fig. 3. Measured relative humidity and temperature of the air in the glass jar that it takes the sensor less than 1 min to equilibrate with the
facility during a typical test, showing the transitions from 75.2  0.8 to ambient air in the jar. These tests show that when plywood is
33.3  0.8% RH at a constant temperature of 23.3  0.3 8C. first placed in a jar with a salt solution, the plywood has a minor
effect on the ambient RH maintained by the salt solution.
23 8C [48]. To determine the actual fluctuation of humidity in
the jar, the humidity in the jar was measured during a typical 3.2. Fully developed air flow in a transient moisture
test and the results are presented in Fig. 3. The standard transfer (TMT) facility
deviation of the humidity is 0.8% RH at both the high and low
humidity conditions. The air inside the jar is not mixed and thus In the TMT facility, the experimental apparatus in Fig. 4(a)
a natural convection boundary layer exists between the vertical is used to create fully developed air flow over a bed of
surface of the plywood and the air in the jar. The change in mass hygroscopic material. The convective boundary is controlled by
of the plywood is monitored by periodic weighing using an passing air at different flow rates above the material to be tested.
electronic mass balance with a bias uncertainty of 3 mg and a In this paper, five pieces of plywood (each with dimensions of
precision uncertainty of 0.1 mg. The plywood is not removed 600 mm
280 mm
9 mm) are placed in a container made of
from the jar during weighing and the change in mass of the Lexan plastic. The five pieces are held together with nylon
plywood during the humidity cycle can be used to calculate the screws to reduce air gaps between the pieces (Fig. 4(b)). Other
moisture buffering capacity of spruce plywood. The glass jar materials could be investigated using different containers. As in
facility measures moisture accumulation in spruce plywood, the glass jar facility, the four lateral edges of each plywood
which will be used to calculate the moisture buffering capacity, sample are sealed to minimize lateral moisture transfer. In the
with a bias uncertainty of 0.4 g/m2. TMT facility, there is only one side of the test specimen
In order to determine if the surface area of the salt solution is exposed to moisture transfer. Since the Lexan container is
adequate to maintain a constant RH in the jar especially when impermeable to moisture transfer, the impermeable plane is at

Fig. 4. (a) Schematic of the TMT facility showing spruce plywood and the sensors and (b) picture showing the spruce plywood held together by nylon screws inside
the Lexan container.
1274 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

Table 1
Summary of the convective heat and mass transfer coefficients at different
Reynolds numbers in the TMT facility
Reynolds number, Re
1000 2000 4000
Convective heat transfer 2.5  0.2 3.5  0.2 8.1  0.4
coefficient, ha (W/(m2 K))
Convective mass transfer 2.1  0.2 2.9  0.2 6.7  0.3
coefficient, hm (
103 m/s)

5176-1 [29]. The mass flow rate is varied to create different

Reynolds numbers (Re) in the air flow channel above the
plywood, which results in different convective transfer
coefficients between the flowing air and the spruce plywood.
The facility is capable of creating air flow Re numbers of up to
nearly 10,000. In this paper, Re of 1000, 2000 and 4000, which
Fig. 5. Plot of relative humidity and temperature of the air entering the TMT corresponds to average air velocities of 0.4, 0.8 and 1.6 m/s in
facility during a typical test, showing the transition from 75 to 33% RH the channel above the plywood, will be investigated. The
controlled within 2% RH and temperature of 22.6 8C controlled within
uncertainty in the Re numbers calculated from the measured air
0.2 8C.
flow is 8% for this facility. A separate test is conducted to
determine the convective mass transfer coefficients for these Re
the bottom of the spruce plywood bed and the effective numbers. In this test, a tray of water is placed in the TMT
thickness (L) of the sample is 45 mm. Prior to testing, the facility and air is passed over the free surface of water. As the
plywood samples are conditioned for a long time (4 months) in air with a controlled Re number passes over the test section,
the laboratory and the initial moisture content of the wood was water evaporates into the air and the mass of water in the tray
0.025 kg/kg, which corresponds to a relative humidity of about decreases, which is recorded by the load sensors. The
48%. temperature and relative humidity of the air entering and
The air flow is provided by a variable speed vacuum pump, leaving the test section are measured to determine the vapor
which provides a fully developed air flow over the top of the density of the air flowing above the water. The temperature of the
specimen. The air is drawn from an environmental chamber water is also measured to determine the water vapor density at the
maintained at a constant temperature (22.6  0.2 8C in this surface of the water. From the mass readings and vapor densities,
paper) and a specified humidity (75 and 33% RH in this paper) the convective mass transfer coefficient is determined. The
that can be controlled within 2% RH. Fig. 5 presents the measured convective mass transfer coefficient is then used to
temperature and humidity measured by the sensor at the inlet of determine the Sherwood number. With an assumption that the
the test section during a typical test. As can be seen in Fig. 5, the Sherwood number equals the Nusselt number, the convective
plywood is initially exposed to 75% RH air for 8 h, followed by heat transfer coefficient is determined. The convective heat and
33% RH air for 16 h. This cycle is repeated three times. The mass transfer coefficients determined are shown in Table 1 for the
data in Fig. 5 is recorded every 5 min and therefore it can be three Reynolds numbers investigated.
seen that it takes about 20 min to realize an increase in RH and
about 30 min to realize a decrease in RH. It should be noted that 4. Numerical model and material properties
this step change is not as rapid nor is the RH control as good as
in the glass jar facility presented previously (Fig. 3). Spruce plywood, like any wood material, is a porous
The moisture accumulation/loss in the material under material that is made up of solid cell walls and lumens (Fig. 1).
isothermal conditions is measured by four load sensors on The cell walls are irregularly shaped, which makes it extremely
which the Lexan container (housing the plywood) is resting difficult to analytically define the boundary between each cell
(Fig. 4(a)). The load sensors are calibrated with calibration wall and the surrounding fluid. Therefore, local volume
masses and have a bias uncertainty of 2 g. Therefore, any averaged equations and properties are used in the model
change in mass during the test will be continually measured and [30]. The assumptions that reflect the experimental conditions
will be the moisture accumulated or lost by the spruce plywood. and allow the problem to be simplified are as follows: (1) heat
These calibrated load sensors allow the TMT facility to and moisture transfer through the spruce plywood is one-
measure moisture accumulation in spruce plywood, which will dimensional; (2) the transport process within the spruce
be used to calculate the moisture buffering capacity, with a bias plywood is pure diffusion of heat and water vapor; (3) air and
uncertainty of 1.1 g/m2. water vapor behave as ideal gases; (4) the only heat source in
A tapered orifice plate embedded in the supply line the medium is the heat of phase change resulting from the
downstream of the test section measures the mass flow rate adsorption and desorption of water vapor within spruce
of the air with an accuracy of 6% according to ISO standard plywood (hfg = 2.5
106 J/kg); (5) the solid and fluid states
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1275

are in thermal equilibrium; (6) Knudsen and Fickian diffusion

can be combined and thermal diffusion (i.e., Soret effect) can be
neglected [31]; and (7) the volume changes (swelling and
shrinkage) with changes in humidity are negligible [32].
The resulting conservation equations for mass and energy
are listed below [24,32]. The symbols are defined in the
nomenclature:
@e‘
r‘ þ ṁ ¼ 0 (1)
@t


@ðrv eg Þ @ @rv
 ṁ ¼ Deff (2)
@t @x @x


@T @ @T
rC p eff þ ṁhfg ¼ keff (3)
@t @x @x
@u Fig. 7. Effective thermal conductivity curve for spruce plywood showing the
ṁ ¼ r0 (4) measured data with the 95% uncertainty bars and the curve fit.
@t
The boundary conditions for the one-dimensional heat and
moisture transfer problem are convective heat and moisture 0.0001 kg/kg, which corresponds to an uncertainty in
transfer between the spruce plywood and the air above it moisture content of 1% at 11% RH and 0.1% at 97%
(Table 1), and an impermeable and adiabatic boundary below RH. The experimental data are curved fitted with a continuous
the spruce plywood. The initial conditions are constant polynomial relationship between moisture content (u) and
temperature and relative humidity throughout the spruce relative humidity (f) in fraction. The polynomial equation for
plywood and these are determined from the experiment. The the curve fit is given as
relative humidity is calculated based on the initial moisture
content and the sorption curve. a þ cf þ ef2
u¼ (5)
1 þ bf þ df2 þ f f3
4.1. Properties
where a = 1.0147E04, b = 0.2339, c = 0.06754, d = 2.3603,
4.1.1. Sorption isotherm e = 0.06574, f = 1.1329. Eq. (5) fits the measured data with
The sorption isotherm data are measured using the glass jars r2 = 0.999.
(Fig. 2) according to the method of Wadso et al. [33] and the
data are presented in Fig. 6. The dry mass is obtained by drying 4.1.2. Effective thermal conductivity
in a vented oven at 50 8C until the change in mass between two The effective thermal conductivity data are measured using a
successive measurements, with a time interval of at least 24 h, heat flow meter apparatus that measures according to ASTM
is lower than 0.1%. The uncertainty in the mass measurement is Standard C518 [34]. The samples are conditioned to the
3 mg and the uncertainty in the moisture content is different RH values using saturated salt solutions [48] in order
to quantify the change in thermal conductivity with equilibrium
RH (Fig. 7). The uncertainty in the measured effective thermal
conductivity is 1%. It should be noted that the thermal
conductivity measurements took about 30 min to complete
while it took about 14 days to condition the samples to
equilibrium, therefore moisture movement during the thermal
conductivity test is expected to be minimal. The curve fitted
relationship is represented by a polynomial given below:

keff ¼ a þ bf þ cf2 þ df3 (6)

where a = 0.08185, b = 0.02212, c = 0.02313, d = 0.01291.

Eq. (6) fits the measured data very well as shown in Fig. 7
(r2 = 0.999).

4.1.3. Effective vapor diffusion coefficient

The vapor diffusion coefficient (Deff) can be determined
using
Fig. 6. Sorption isotherm for spruce plywood showing the measured data and
the curve fit. Deff ¼ dRv T (7)
1276 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

moisture buffering capacity by less than 0.1%, but increases the

solution time by over five times.

5. Energy impacts of moisture buffering

Hygroscopic materials have the potential to improve indoor

climate, comfort and air quality [12,13,15,18,25,36], but the
effect of hygroscopic materials on energy consumption has not
been studied to the same degree. Therefore, this section
attempts to address the question of whether the application of
hygroscopic materials can reduce the energy needed to heat,
cool and ventilate buildings. The main intent is to identify the
magnitude of possible savings and suggest the most promising
areas of future work. To accomplish this, results from Simonson
Fig. 8. Water vapor permeability curve for spruce plywood showing the et al. [11], which are for a bedroom in a wooden apartment
measured data with the 95% uncertainty bars and the curve fit.
building, will be extrapolated to estimate the potential
magnitude of energy savings. This study [11] used hourly
where the water vapor permeability (d) is determined using the weather data from four different cities (Helsinki, Finland; Saint
cup method [35]. In the dry cup test, CaCl2 (0% RH) is used in Hubert, Belgium; Holzkirchen, Germany and Trapani, Italy) to
the cup and Mg(NO3)2 (53% RH) is used to condition the calculate the indoor temperature and humidity in a 32.4 m3
surrounding air. The wet cup test uses KNO3 (94% RH) in the bedroom (floor area of 12 m2 and internal surface area of
cup and Mg(NO3)2 (53% RH) in the surrounding air. An 60 m2) occupied by two adults for 9 h each night (22:00–7:00).
additional cup measurement is made at high humidities using The total moisture production was 60 g/h during occupation
KNO3 (94% RH) in the cup and KCl (84% RH) in the and the ventilation rate was constant at 0.5 ach. The main
surrounding air. The uncertainty in the measured value of d hygroscopic materials in the external and internal walls are
is 13%. Eq. (8) gives the curve fitted relationship: porous wood fiberboard (11 mm), building paper (0.3 mm) and

0:5 cellulose insulation (150 mm). All of the internal surfaces are
bf permeable (5
109 kg/(s m2 Pa)) except for the floor in the
d¼ aþ (8)
‘nf hygroscopic case, while all internal surfaces are impermeable
(5
1012 kg/(s m2 Pa)) in the non-hygroscopic case. These
where a = 2.3573E25, b = 8.1601E24 (Fig. 8).
two extreme cases will be used to investigate the potential for
hygroscopic materials to affect the energy consumption in
4.1.4. Density and specific heat capacity
buildings. More details of the input data can be found in the
The equations that quantify the changes in density and
literature [11,12]. Even though the simulations of Simonson
specific heat capacity due to moisture adsorption result from the
et al. [11] can be viewed as representing periodic occupation in
local volume averaging of the governing equations and are
any building, the extrapolations should be used with caution
r ¼ es rs þ eg rg þ e‘ r‘ (9) because the loads, ventilation rate and other factors vary
significantly in different buildings.
and In this paper, potential savings are divided into ‘‘direct’’ and
es rs C p s þ e‘ r‘ Cr‘ þ eg fðrC p Þa þ ðrC p Þv g ‘‘indirect’’ energy savings. Direct savings are savings in the
C p eff ¼ (10) required heating and cooling of a building, while indirect
r
savings are the possible savings that could result due to a lower
ventilation rate, a lower indoor temperature in the winter or a
4.2. Numerical solution higher indoor temperature in the summer.

The coupled partial differential equations are discretized 5.1. Direct energy savings
using the finite difference method with second order accuracy for
the spatial nodes and the implicit scheme for the time derivative. 5.1.1. Heating energy
For the spatial nodes at the boundary, the backward or forward In the heating season, direct energy savings are possible
scheme is used for the discretization, while the central scheme is because moisture accumulation in hygroscopic materials
used for the central nodes. To provide a stable solution, an under releases 2.5 kJ/kg of moisture, which will decrease the required
relaxed, Gauss–Seidel iteration method is used and the solution is heating energy. Since humans are an important source of
considered to have converged, when for any time step, the change moisture in buildings, this moisture accumulation will occur
in any dependent variable (T, rv ) is less than 106. A sensitivity during occupation. The energy required to heat the 12 m2
study showed that a uniform grid size of 0.1 mm and a time step bedroom [11] with one west-facing external wall (150 mm
of 30 s provide a numerically accurate solution. Decreasing the insulation) during the occupied hours (22:00–7:00) is presented
grid size to 0.05 mm and the time step to 10 s changes the in Fig. 9(a) for the hygroscopic and non-hygroscopic cases. In the
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1277

Fig. 9. The heat generated during moisture accumulation in hygroscopic building materials (a) decreases the heating energy consumption during occupation (22:00–
7:00), but (b) has a small affect on the total energy consumption during the heating season.

simulations, the heating power is adjusted to keep the bedroom Hindoor is the indoor enthalpy and Hdesired is the desired indoor
between 20 and 21 8C during the heating season, which was set to enthalpy. Cooling the room is expected to increase the humidity
be from 1.9 to 31.5 in Finland and from 1.10 to 30.4 in Belgium of the indoor air and building materials and will likely increase
and Germany. The average indoor temperature was 20.7 8C in the moisture transfer from that calculated in [11] because the
Finland and 20.5 8C in Belgium and Germany. Fig. 9(a) shows slope of the sorption curve typically increases with increasing
that the energy consumption during occupation is about 10% humidity. These effects are neglected here.
lower in the hygroscopic case than in the non-hygroscopic case, The calculated cooling energy and demand are presented in
which means that moisture accumulation in the building Fig. 11 when the desired indoor enthalpy is 47 kJ/kg, which
materials during occupation can decrease the needed heating would result in a percent dissatisfied of 32% if the air was
energy. On the other hand, energy is needed to dry this moisture unpolluted [4]. This is comparable to the recommended
from these materials during unoccupied periods and the net result perceived air quality of 2.5 dp (30% dissatisfied) [38].
is that the total energy consumption during the heating season is Fig. 11 shows that the required cooling energy is quite low
nearly equal for both cases (Fig. 9(b)). The slightly higher total for the bedroom because the only internal heat loads are
heating energy consumption in the hygroscopic case is likely due 100 W of lighting for 1 h and two people for 9 h. Never-
to a slightly higher thermal conductivity due to higher material theless, the required cooling energy during occupation is
moisture contents in the hygroscopic case. The results in Fig. 9 lower (from 10% in Italy to 35% in Finland) with hygroscopic
show that it may be possible to save heating energy with materials than with non-hygroscopic materials as shown in
hygroscopic materials, but a control strategy is required to realize Fig. 11(a). The peak cooling demand is also lower (from 10%
these savings. Such control strategies could be temperature and in Italy to 30% in Finland, Fig. 11(c)) with hygroscopic
ventilation set back during unoccupied periods. materials than with non-hygroscopic materials. Similarly as
was discussed with the heating energy savings, a control
5.1.2. Cooling energy strategy is needed to realize these savings because they
During the cooling season, hygroscopic materials are able to represent the energy consumption and demand during
reduce the indoor humidity and consequently reduce the indoor occupied hours.
enthalpy [11]. Decreasing the enthalpy of indoor air decreases Fig. 11(b) shows that the savings in cooling energy
the energy required to cool the building and also improves the consumption for all hours during the year are lower than
indoor air quality [4,36,37]. The potential for hygroscopic during occupation which is similar to the findings of Fairey and
materials to reduce cooling energy consumption can be Kerestecioglu [39]. Simulation results [39] show that if a
estimated from the calculated indoor enthalpy. The bedroom building is continuously conditioned regardless of occupation,
studied by Simonson et al. [11] had no cooling, but the energy the cooling energy savings due to hygroscopic mass are in the
required to cool the room to a desired enthalpy of 47 kJ/kg order of 5% (for the month of July in Atlanta, GA), but if
(24 8C and 50% RH) can be estimated by the multiplying the ventilation and cooling are controlled according to occupation
area under the curve in Fig. 10 with the ventilation rate greater savings can be realized. It is expected that a control
(0.5 ach = 4.5 L/s) according to strategy to optimize the benefits of hygroscopic mass would be
Z similar to that recommended to optimize the benefits of thermal
Q ¼ ṁventilation DH dt (11) mass for the cooling of buildings (e.g. [40–42]). Peak cooling
loads can be reduced by as much as 50% by precooling the
where building mass during unoccupied periods [41], but such savings
can be overestimated if the moisture adsorbed in the building
DH ¼ Hindoor  Hdesired (12) structure and furnishings during unoccupied periods is not
1278 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

Fig. 10. Frequency distribution of the difference between indoor enthalpy and an enthalpy of 47 kJ/kg during occupied hours. The shaded regions are proportional to
the energy required to cool the room to 24 8C and 50% RH (47 kJ/kg).

included [39]. Nevertheless, it is not unreasonable to expect occupation. The relative heating/cooling energy savings are
peak cooling load reductions of 10–30% when hygroscopic relative to the total heating/cooling energy consumption (i.e.,
materials are applied, as shown in Fig. 11(c). This could have a including both occupied and unoccupied hours) and are
large impact on the size, cost and efficiency of cooling therefore lower than the relative savings presented previously.
equipment in buildings.
5.2. Indirect energy savings
5.1.3. Summary of direct savings
Fig. 12 summarizes the magnitude of the potential savings of The main purpose of conditioning buildings it to provide an
heating and cooling energy considering the occupied hours and indoor environment that is comfortable and an indoor air
all hours. In the case including only occupied hours, it is quality that is acceptable, where temperature, humidity and
assumed that the HVAC control system is optimized to take ventilation (among other factors) affect comfort and air quality
advantage of the lower heating and cooling loads during [2–4,43]. Therefore, since hygroscopic materials can improve

Fig. 11. Cooling energy required to cool the bedroom to an enthalpy of 47 kJ/kg (24 8C and 50% RH) during (a) occupation and (b) all hours, and (c) cooling demand
during occupation.
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1279

Fig. 12. Potential direct (a) heating and (b) and cooling energy savings when applying hygroscopic materials. The percent savings are relative to the total heating or
cooling energies.

indoor humidity conditions, it may be possible to alter the still provide a comparable IAQ. Simonson et al. [12] found that
temperature and outdoor ventilation rate of buildings that use even larger ventilation reductions were possible, ranging from
hygroscopic materials and still provide a similar comfort and 20% to 90% depending on the criteria selected. Even though
indoor air quality as in buildings without hygroscopic materials more research is needed before ventilation reductions can be
[12]. safely brought into practice, Fig. 13 presents the savings that
could result with a modest estimation of a 15% reduction in
5.2.1. Reducing outdoor ventilation ventilation rate. Here it is important to note that decreasing the
Research has quantified the effect of humidity on perceived ventilation rate is expected to have a greater effect on IAQ as
air quality and warm respiratory comfort using laboratory the ventilation rate decreases [44]. Meanwhile, decreasing the
experiments [2,4]. In addition, these findings have been ventilation rate is expected to have a smaller effect on energy
confirmed in a field study where it was found that the consumption as the ventilation rate decreases.
perceived indoor air quality was moderately better at a Reducing the ventilation rate by 15% would have a
ventilation rate of 3.5 L/s per person and an indoor enthalpy of significant impact on the energy consumption in Finland and
35 kJ/kg (20 8C/40% RH) than at a ventilation rate of 10 L/s per could save an estimated 3 TWh/a of heating energy and
person and an indoor enthalpy of 45 kJ/kg (23 8C/50% RH) 0.6 TWh/a of electricity (Fig. 13(a)). These savings are based
[37]. Therefore, the perceived indoor air quality will be similar on the estimated heating and electricity consumption due to
if the ventilation is reduced by 75% and the indoor enthalpy ventilation (21 and 4 TWh/a, respectively) presented by
decreased by 10 kJ/kg. Since the average indoor enthalpy is Seppänen [45].
about 2 kJ/kg (1.9 kJ/kg in Finland and Belgium, 1.7 kJ/kg in In a case study in Austin, Texas, it was found that reducing
Germany and 1.5 kJ/kg in Italy) lower during occupation in the the outdoor ventilation rate increased comfort and reduced the
hygroscopic case than in the non-hygroscopic case [11], the measured energy consumption in a 9200 m2 office building
ventilation rate in buildings with hygroscopic materials could [46]. To improve the indoor air quality in the building, the
possibly be reduced by 15% (i.e., 2 kJ/kg (75%/10 kJ/kg)) and outdoor ventilation rate was reduced by 86% (from 74 to 10 L/s

Fig. 13. Possible energy savings in (a) Finnish buildings and (b) an office building in Austin, Texas due to reducing the ventilation rate by 15%.
1280 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

per person). Decreasing the ventilation rate, decreased the peak not an important comfort parameter at low temperatures, but
indoor relative humidity from 70% to 55% and resulted in a has a strong effect on the risk of condensation and mould
significant improvement in comfort conditions during the growth during the heating season [44]. The average indoor
summer. In addition to improving comfort, the reduced temperature in the hygroscopic and non-hygroscopic cases of
ventilation rate decreased heating energy use by 48%, Simonson et al. [11] are nearly identical (less than 0.2 8C
electricity use by 21% and total energy use by 27%. Assuming difference). On the other hand, the indoor temperature in the
that the decrease in energy use is proportional to the decrease in hygroscopic case can be reduced by an average of 1.6 8C in
ventilation rate, a 15% reduction in ventilation rate would Finland and Belgium and 1.5 8C in Germany, while maintaining
reduce the heating energy consumption by 8% and the total the same indoor relative humidity during the occupied hours of
energy consumption by 5% (Fig. 13(b)). It is interesting to note the heating season. The potential energy savings due to a
that the percent savings in heating energy are similar in Finland reduced indoor temperature are estimated by multiplying the
and in Texas. ratio of the temperature reduction (1.5 or 1.6 8C) to the average
Woloszyn et al. [47] present the numerical analysis of a two- temperature difference between indoors and outdoors by the
bedroom apartment (140 m3, 60 m2) for one evening (19:00– heating energy used during occupation. Since the average
0:00) in a mild humid climate using the mean January climate of temperature difference between indoors and outdoors during
Macon, France. Macon is in east-central France and the average the occupied hours of the heating season are 21.6, 18.9 and
outdoor conditions during the simulation where 3 8C and 93% 21.3 8C in Finland, Belgium and Germany, the estimated
RH (4.4 g/kg). Woloszyn et al. [47] compared the indoor heating energy savings due to reducing the indoor temperature
humidity and energy consumption of the apartment for two cases. are 2% of the total heating energy (7–9% of the heating during
One case is where the moisture buffering capacity of the structure occupation) as shown in Fig. 14(a).
and furnishings is included in the simulation and the other case is
where the moisture buffering capacity is neglected. The 5.2.3. Increasing indoor temperature in the summer
apartment had a mechanical ventilation system and a ventilation In the summer, the perceived air quality (PAQ) and warm
rate that varies between 0.12 and 0.7 ach depending on the indoor respiratory comfort during occupation may be significantly
relative humidity. The simulation results show that the indoor poorer in a building with non-hygroscopic materials than in a
humidity in the kitchen and living room is over 15% RH greater building with hygroscopic materials [36]. As a result, it is
in the non-hygroscopic case than in the hygroscopic case, even possible to allow the indoor temperature in a building with
though the average ventilation rate is about 10% higher in the hygroscopic materials to be higher than in a building with non-
non-hygroscopic case than in the hygroscopic case. The total hygroscopic materials and still have comparable indoor
energy consumption during the 6 h period is 45% higher in the comfort and air quality. Increasing the indoor temperatures
non-hygroscopic case than in the hygroscopic case. New will reduce the energy needed for cooling the building during
ventilation units that control the ventilation rate based on warm weather. The results of Simonson et al. [11] show that the
measured humidity and CO2 could result in additional savings for indoor temperature can be increased by about 1 8C in a
buildings with hygroscopic materials. hygroscopic building and yet provide similar conditions of
warm respiratory comfort. Similarly, a hygroscopic building
5.2.2. Reducing indoor temperature in the winter can have up to 2 8C higher indoor temperature than a non-
Since the indoor humidity during occupation is reduced hygroscopic building and still have similar perceived indoor air
when applying hygroscopic materials, the indoor temperature quality. The energy savings that could result from such a
can be reduced and yet result in the same indoor relative temperature change are estimated by changing Hdesired in
humidity. Here it is important to note that relative humidity is Eq. (12) and the results are in Fig. 14(b).

Fig. 14. Possible energy savings when the indoor temperature in the hygroscopic case is (a) decreased while maintaining the same indoor relative humidity as in the
non-hygroscopic case, and (b) increased while maintaining the same comfort and air quality conditions as in the non-hygroscopic case.
 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282 1281

Table 2 facility and are known within 10%. The bias uncertainty in
Potential reductions in the total consumption (%) of heating and cooling energy
the measurement of moisture accumulation is 0.4 using the
when applying hygroscopic materials in buildings
glass jar facility 1.1 g/m2 using the TMT facility.
Heating Cooling The potential for hygroscopic materials to reduce energy
Direct energy savings consumption in buildings is also presented in this paper. The most
Optimized control of 2–3 5–30 promising energy savings are for buildings with mechanical
HVAC system cooling equipment located in hot and humid climates, but there
No control of HVAC system 0 0–20
are potential savings in all climates if the HVAC system can be
Reduction in energy demand 0 10–30
optimally controlled to regulate the indoor climate, comfort and
Indirect energy savings air quality. The results show that moisture transfer has the
Reducing ventilation rate 5 5
Changing indoor temperature 2 2–10% (comfort),
potential to reduce the energy consumption of buildings
5–20% (PAQ) ‘‘directly’’ and ‘‘indirectly’’. Direct savings are defined as
savings in the heating and cooling of a building that result when
applying hygroscopic materials. Indirect savings are defined as
5.3. Summary of potential energy savings savings that result from adjusting the ventilation rate and indoor
temperature while maintaining adequate indoor air quality and
The approximate potential energy savings calculated by comfort with hygroscopic materials. The potential direct energy
different methods is summarized in Table 2 as a percentage of savings are small for heating (2–3% of the total heating energy),
the total heating or cooling energies. It is important to note that but significant for cooling (5–30% of the total cooling energy).
these values are estimates based on numerical results of These savings require the integration of hygroscopic materials
Simonson et al. [11,12] and a few, mainly numerical, studies and a well-controlled HVAC system. The potential indirect
reported in the literature [37,39,45–47] and must be used with savings for heating are in the order of 5%, while they range from 5
caution. to 20% for cooling.

6. Conclusions
Acknowledgements
Based on the literature reviewed in this paper, the moisture
The experimental test facilities presented in this paper have
storage capacity of hygroscopic materials during transient been developed with funding from the Canada Foundation for
changes in ambient air relative humidity (moisture buffering
Innovation (CFI) and testing was funded by the Natural
capacity) is an important parameter that requires further
Sciences and Engineering Research Council of Canada
research into standard test methods and facilities that can
(NSERC) Discovery Grant program and Special Research
quantify it accurately and repeatably. To help with this standard
Opportunities program. The energy impact study was funded by
development, two different test facilities are developed and
Wood Focus Oy. The financial assistance of CFI, NSERC and
presented in this paper together with a numerical model, which
Wood Focus are greatly appreciated.
can be used to compare the results from these different
facilities. The model can also be used to investigate other test
conditions and materials and help formulate a testing standard. References
Since the experimental facilities and numerical model will be
applied in a companion paper [1] to investigate the moisture [1] O.F. Osanyintola, P. Talukdar, C.J. Simonson, Effect of initial conditions,
boundary conditions and thickness on the moisture buffering capacity
buffering capacity of spruce plywood, the thermal and moisture of spruce plywood, Energy and Buildings 38 (10) (2006) 1283–1292.
transfer properties of spruce plywood are included in this paper. [2] J. Toftum, A.S. Jorgensen, P.O. Fanger, Upper limits of air humidity for
Both of these facilities described in this paper are able to preventing warm respiratory discomfort, Energy and Buildings 28 (1998)
provide well-controlled temperature and humidity boundary 15–23.
[3] J. Toftum, A.S. Jorgensen, P.O. Fanger, Upper limits for indoor air
conditions for the hygroscopic material being investigated and
humidity to avoid uncomfortably humid skin, Energy and Buildings 28
permit these boundary conditions to be changed rapidly. One (1998) 1–13.
facility (termed the glass jar facility) uses a hanging mass [4] L. Fang, G. Clausen, P.O. Fanger, Impact of temperature and humidity on
balance and small, sealed jars containing saturated salt the perception of indoor air quality, Indoor Air 8 (1998) 80–90.
solutions located in a controlled environmental chamber. The [5] N. Mendes, F.C. Winkelmann, R. Lamberts, P.C. Philippi, Moisture
other facility (termed the transient moisture transfer (TMT) effects on conduction loads, Energy and Buildings 35 (2003) 631–
644.
facility) is more complex and provides fully developed air flow [6] F. Lucas, L. Adelard, F. Garde, H. Boyer, Study of moisture in buildings
above the hygroscopic material using a small scale wind tunnel for hot humid climates, Energy and Buildings 34 (2002) 345–355.
connected to an environmental chamber. The glass jar facility [7] S. Atthajariyakul, T. Leephakpreeda, Real-time determination of optimal
provides more constant humidity conditions and is able to indoor-air condition for thermal comfort, air quality and efficient energy
change these conditions more rapidly than the TMT facility, but usage, upper limits for indoor air humidity to avoid uncomfortably humid
skin, Energy and Buildings 36 (7) (2004) 720–733.
is limited to investigate natural convection moisture transfer [8] C. Crawford, P. Manfield, A. McRobie, Assessing the thermal perfor-
and small samples. On the other hand, the air flow and mance of an emergency shelter system, Energy and Buildings 37 (5)
convective transfer coefficients are well controlled in the TMT (2005) 471–483.
1282 O.F. Osanyintola, C.J. Simonson / Energy and Buildings 38 (2006) 1270–1282

[9] J. Arias, P. Lundqvist, Heat recovery and floating condensing in super- [27] JIS A 1470-1. Test method of adsorption/desorption efficiency for build-
markets, Energy and Buildings 38 (2) (2006) 73–81. ing materials to regulate an indoor humidity-Part 1: Response method of
[10] D. Yu, S.A. Klein, D.T. Reindl, An evaluation of silica gel for humidity humidity, Japanese Standards Association, Tokyo, 2002.
control in display cases, WAAC Newsletter 23 (2.) (2001). [28] D.J. Gardner, The relevance of surface properties & wood finishes to the
[11] C.J. Simonson, M. Salonvaara, T. Ojanen, Improving indoor climate and wood science & technology research community, in: Proceedings of the
comfort with wooden structures, VTT Building Technology, Espoo: VTT Second SWST Annual Fundamental Disciplines Session, 2004 http://
Publications 431, http://virtual.vtt.fi/inf/pdf/publications/2001/P431.pdf, www.swst.org/meetings/AM04/Gardner.pdf.
VTT Building and Transport, 2001. [29] ISO, Measurement of fluid flow by means of pressure differential devices,
[12] C.J. Simonson, M. Salonvaara, T. Ojanen, Moderating indoor conditions ISO 5176-1, Switzerland, 1991.
with hygroscopic building materials and outdoor ventilation, ASHRAE [30] M. Kaviany, Principles of Heat Transfer in Porous Media, Springer-
Transactions 110 (2) (2004) 804–819. Verlag, New York, 1991.
[13] C.J. Simonson, S.O. Olutimayin, M. Salonvaara, T. Ojanen, J. O’Connor, [31] H.M. Kuenzel, Simultaneous Heat and Moisture Transport in Building
Potential for hygroscopic building materials to improve indoor comfort Components—One-and Two-dimensional Calculation using Simple
and air quality in the Canadian climate, in: Proceedings (CD) of the Parameters, IRB Verlag, Fraunhofer-Informationszentrum Raum and
Performance of the Exterior Envelopes of Whole Buildings IX Interna- Bau, Stuttgart, Germany, 1995.
tional Conference, Clearwater Beach, FL, December 5–10, 2004, 15 pp. [32] O.F. Osanyintola, Transient moisture characteristics of spruce ply-
[14] C.J. Simonson, M. Salonvaara, T. Ojanen, Heat and mass transfer between wood, M.Sc. Thesis, Department of Mechanical Engineering, Univer-
indoor air and a permeable and hygroscopic building envelope. Part I. sity of Saskatchewan, 2005, http://library.usask.ca/theses/available/
Field measurements, Journal of Thermal Envelope and Building Science etd-12222005-082100/.
28 (1) (2004) 63–101. [33] L. Wadso, K. Svennberg, A. Dueck, An experimentally simple method for
[15] A.H. Holm, H.M. Kunzel, K. Sedlbauer, Predicting indoor temperature measuring sorption isotherms, Drying Technology 22 (10) (2004) 2427–
and humidity conditions including hygrothermal interactions with the 2440.
building envelope, ASHRAE Transactions 110 (2) (2004) 820–826. [34] ASTM, Standard test method for steady-state heat flux measurement and
[16] C. Rode, T. Mitamura, J. Shultz, T. Padfield, Test cell measurements of thermal transmission properties by means of the heat flow meter apparatus,
moisture buffer effects, in: Proceedings of the Sixth Nordic Building ASTM C 518, Philadelphia, 2003.
Physics Symposium, Trondheim, Norway, (2002), pp. 619–626. [35] ASTM, Standard test method for water vapor transmission of materials,
[17] R. Peukhuri, C. Rode, K.K. Hansen, Moisture buffering capacity of ASTM E96/E96M, Philadelphia, 2005.
different insulation materials, in: Proceedings (CD) of the Performance [36] C.J. Simonson, M. Salonvaara, T. Ojanen, The effect of structures on
of Exterior Envelopes of Whole Buildings IX International Conference, indoor humidity—possibility to improve comfort and perceived air qual-
Clearwater Beach, FL, USA, 2004, 14 pp. ity, Indoor Air 12 (2002) 243–251.
[18] M. Salonvaara, T. Ojanen, A. Holm, H.M. Kunzel, A.N. Karagiozis, [37] L. Fang, P. Wargocki, T. Witterseh, G. Clausen, P.O. Fanger, Field study on
Moisture buffering effects on indoor air quality—experimental and simu- the impact of temperature, humidity and ventilation on perceived air
lation results, in: Proceedings (CD) of the Performance of Exterior quality, in: Proceedings of the Indoor Air’99, vol. 2, Edinburgh, (1999),
Envelopes of Whole Buildings IX International Conference, Clearwater pp. 107–112.
Beach, FL, 2004, 11 pp. [38] CEN, Ventilation for buildings—design criteria for the indoor environ-
[19] K. Svennberg, L. Hedegaard, C. Rode, Moisture buffer performance of a ment, European Committee for Standardization, Report CR 1752, 1988.
fully furnished room, in: Proceedings (CD) of the Performance of Exterior [39] P.W. Fairey, A.A. Kerestecioglu, Dynamic modeling of combined thermal
Envelopes of Whole Buildings IX International Conference, Clearwater and moisture transport in buildings: effect on cooling loads and space
Beach, FL, 2004, 11 pp. conditions, ASHRAE Transactions 91 (2A) (1985) 461–472.
[20] S. Hameury, Moisture buffering capacity of heavy timber structures [40] K. Keeney, J.E. Braun, A simplified method for determining optimal
directly exposed to an indoor climate: a numerical study, Buildings cooling control strategies for thermal storage in building mass, Interna-
and Environment 40 (2005) 1400–1412. tional Journal of HVAC&R Research 2 (1) (1996) 59–78.
[21] C.J. Simonson, Energy consumption and ventilation performance of a [41] I. Andresen, M.J. Brandemuehl, Heat storage in building thermal mass—a
naturally ventilated ecological house in a cold climate, Energy and parametric study, ASHRAE Transactions 98 (1) (1992) 910–918.
Buildings 37 (1) (2005) 23–35. [42] J.E. Braun, Reducing energy costs and peak electrical demand through
[22] J. Liu, Y. Aizawa, H. Yoshino, Experimental and numerical study on optimal control of building thermal storage, ASHRAE Transactions 96 (2)
indoor temperature and humidity with free water surface, Energy and (1990) 876–888.
Buildings 37 (4) (2005) 383–388. [43] ASHRAE, Thermal environmental conditions for human occupancy,
[23] S. Hameury, T. Lundstrom, Contribution of indoor exposed massive wood ANSI/ASHRAE Standard 55, Atlanta, 2004.
to a good indoor climate: in situ measurement campaign, Energy and [44] J. Sundell, What we know, and what don’t know about sick building
Buildings 36 (2004) 281–292. syndrome, ASHRAE Journal 38 (6) (1996) 51–57.
[24] S. Olutimayin, C.J. Simonson, Measuring and modeling vapor boundary [45] O. Seppänen, Estimated cost of indoor climate in Finnish buildings, in:
layer growth during transient diffusion heat and moisture transfer in Proceedings of the Indoor Air’99, vol. 4, Edinburgh, (1999), pp. 13–18.
cellulose insulation, International Journal of Heat and Mass Transfer [46] M. Liu, Y. Zhu, B.Y. Park, D.E. Claridge, D.K. Feary, Airflow reduction to
48 (2005) 3319–3330. improve building comfort and reduce energy consumption—a case study,
[25] C. Rode, A. Holm, T. Padfield, A review of humidity buffering in the ASHRAE Transactions 105 (1) (1999) 384–390.
interior spaces, Journal of Thermal Envelope and Building Science 27 (3) [47] M. Woloszyn, G. Rusaouën, G. Fraisse, J.-L. Hubert, Predicting indoor
(2004) 221–226. climate using electric analogy estimations of condensation potential, in:
[26] C. Rode, R. Peuhkuri, K.K. Hansen, B. Time, K. Svennberg, J. Arfvidsson, Proceedings of the Healthy Buildings 2000, vol. 3, Helsinki, (2000), pp.
T. Ojanen, Moisture buffer value of materials in buildings, in: Proceedings 165–170.
of the Nordic Building Physics Conference vol. 1, Reykjavı́k, Iceland, [48] ASTM, Maintaining constant relative humidity by means of aqueous
June 13–15, 2005, pp. 108–115. solutions, ASTM E 104, Philadelphia, 1985.