COMBINED HEAT AND MOISTURE TRANSPORT

MODELING FOR RESIDENTIAL BUILDINGS

Sponsored by
U.S. National Institute of Standards and Technology
HL 2008-3

Project Award No: 60NANB4D1091

Submitted by:

Zhipeng Zhong, Graduate Research Assistant

James E. Braun, Principal Investigator

Approved by:

Patricia Davies, Director

Ray W. Herrick Laboratories

AUGUST 2008

iv

TABLE OF CONTENTS

Page
LIST OF TABLES..........................................................................................................................vi
LIST OF FIGURES .......................................................................................................................vii
NOMENCLATURE ........................................................................................................................ x
ABSTRACT..................................................................................................................................xiv
CHAPTER 1. INTRODUCTION .................................................................................................... 1
1.1. Building Structure Problems Caused by Moisture ................................................................ 1
1.2. Comfort and Health Problems Related to Moisture .............................................................. 5
1.3. Introduction of Moisture Transfer......................................................................................... 8
1.4. Objectives of this Research ................................................................................................. 12
1.5. Organization of this Thesis ................................................................................................. 14
CHAPTER 2. MODELING OF HEAT AND MOISTURE TRANSFER IN EXTERIOR
WALLS.............................................................................................................................. 17
2.1. Modeling of The Heat and Moisture Transfer in Building Envelope.................................. 17
2.1.1. Major Assumptions of the Heat and Moisture Transfer Model.................................... 17
2.1.2. Mathematical Description of the Heat and Moisture Transfer Model.......................... 18
2.1.3. Comparison of Moisture Transfer Modeling Tools...................................................... 24
2.1.4. Properties of Building Material .................................................................................... 27
2.1.5. Boundary Conditions .................................................................................................... 29
2.2. Numerical Scheme and Flow Chart of the Program............................................................ 31
2.3. Example of Moisture Build up Caused by Solar Heating after Rain................................... 38
2.4. Summary ............................................................................................................................. 42
CHAPTER 3. EXPERIMENTAL VALIDATION OF EXTERIOR WALL MODEL ................. 43
3.1. Methodology for the Experimental Validation ................................................................... 44
3.2. Equipment and Calibrations ................................................................................................ 45
3.3. Experimental Procedure ...................................................................................................... 56
3.4. Uncertainty Analysis ........................................................................................................... 62
3.5. Model Validation................................................................................................................. 65
3.6. Summary ............................................................................................................................. 73
CHAPTER 4. A SIMPLIFIED MODEL FOR GROUND-COUPLED HEAT TRANSFER IN
FLOOR SLABS ................................................................................................................... 75
4.1. Introduction ......................................................................................................................... 75
4.2. Methodology ....................................................................................................................... 78
4.2.1. Case Study Descriptions ............................................................................................... 78
4.2.2. Finite-Element Model................................................................................................... 81
4.2.3. Simplified Ground-Coupled Floor Modeling ............................................................... 82
4.2.4. Process for Correlating Perimeter Heat Loss Factor and Soil Depth............................ 84

4.3. Results and Discussion........................................................................................................ 86

4.3.1. Floor Configuration A: 4heavy-weight concrete on soil............................................ 86
4.3.2. Floor Configuration B: 10 cm (4 Inches) HWC Slab with 4 inches of Gravel............. 90
4.3.3. Numerical Validation ................................................................................................... 91
4.4. Summary ............................................................................................................................. 92
CHAPTER 5. WHOLE BUILDING HEAT AND MOISTURE MODELING............................. 93
5.1. Description of the Whole Building Heat-Moisture Balance Model .................................... 93
5.1.1. Infiltration, Inter-zonal, and Ventilation Air Flows...................................................... 96
5.1.2. Internal Moisture and Heat Generation ........................................................................ 96
5.1.3. Wall, Ceiling, and Floor Heat and Moisture Fluxes ..................................................... 99
5.1.4. Windows..................................................................................................................... 100
5.1.5. Moisture Buffer .......................................................................................................... 101
5.1.6. HVAC Equipment and Thermostat ............................................................................ 102
5.1.7. Weather Processing and Boundary Conditions .......................................................... 103
5.1.8. Model Implementation and Numerical Solution ........................................................ 104
5.1.9. Comparisons with Other Models ................................................................................ 108
5.2. Case Study Results ............................................................................................................ 109
5.2.1. Case Study Description .............................................................................................. 109
5.2.2. Results for Indoor/Attic Conditions and Equipment .................................................. 111
5.2.3. Results for Envelope Moisture ................................................................................... 116
5.2.4. Importance of Whole-building Analysis for Wall Moisture Performance ................. 119
5.3. Summary ........................................................................................................................... 122
CHAPTER 6. CASE STUDIES FOR BUILDING HEAT AND MOISTURE
PERFORMANCE .............................................................................................................. 124
6.1. Heating Climate-Minneapolis, MN................................................................................... 124
6.2. Mixed Climate-Nashville, TN........................................................................................... 131
6.3. Hot and Dry Cooling Climate- Phoenix, AZ..................................................................... 137
6.4. Hot and Humid Cooling Climate- Houston, TX ............................................................... 141
6.5. Summary ........................................................................................................................... 146
CHAPTER 7. CONCLUSIONS AND FUTURE WORK ........................................................... 147
7.1. Conclusions ....................................................................................................................... 147
7.2. Challenges for Future Studies ........................................................................................... 150
LIST OF REFERENCES............................................................................................................. 154
APPENDIX . INTRODUCTION OF CONTAMW .................................................................... 164
A.1. Background on the Multi-zone Model CONTAM ........................................................ 165
A.2. Building Representation in CONTAM............................................................................. 166
A.3. Solution Methods ............................................................................................................. 168
A.4. Boundary Conditions........................................................................................................ 172
A.5. Simplified Model Applied in Whole Building Model...................................................... 173
VITA............................................................................................................................................ 175

vi

LIST OF TABLES

Table
Page
1.1 Survey of moisture problems in residential buildings. .............................................................. 3
3.1 Comparison of the measurement technologies for moisture content. ...................................... 47
3.2 Material properties and geometry of the test specimen: Sugar-pine........................................ 48
3.3 The position of the moisture pin-pairs for the 1-layer specimen. ............................................ 50
3.4 Radiation Measurements at Different Locations on the Wood Specimen (W/m2). ................. 52
3.5 Calibration for the relative humidity sensor and transmitter. .................................................. 54
3.6 Calibration for the thermal couples.......................................................................................... 54
3.7 Air velocity, Temperature and Re number for computing hr/hc............................................... 58
3.8 Uncertainties of directly measured values/material properties. ............................................... 62
3.9 Uncertainty analysis for hmix_cold............................................................................................... 64
3.10 Uncertainty analysis for convective heat transfer coefficient. ............................................... 64
3.11 Uncertainty analysis for wood specimen surface absorptance............................................... 64
3.12 Uncertainty analysis for specimen surface vapor flux. .......................................................... 65
3.13 Experimental condition summary. ......................................................................................... 65
3.14 Validation model condition summary.................................................................................... 67
4.1 Base case thermal properties of floor and ground materials. ................................................... 80
5.1 U-factor and SHGC for windows. ......................................................................................... 101
5.2 Adsorption isotherm constants for curtain and bedding. ....................................................... 102
5.3 Parameters for building structure........................................................................................... 110
5.4 Sizing results for the HVAC equipment. ............................................................................... 111
6.1 Locations for the parametric studies. ..................................................................................... 124
6.2 Parameters for building structure in Minneapolis.................................................................. 127
6.3 Sizing result for the HVAC equipment in Minneapolis......................................................... 127
6.4 Parameters for building structure for Nashville_sandwich type. ........................................... 134
6.5 Sizing results for the HVAC equipment in Nashville............................................................ 134
6.6 Parameters for external wall structure for Nashville_conventional wall structure ................ 136
6.7 Parameters for building structure in Phoenix......................................................................... 138
6.8 Sizing result for the HVAC equipment in Phoenix................................................................ 138
6.9 Sizing result for the HVAC equipment in Houston. .............................................................. 142

vii

LIST OF FIGURES

Figure
Page
1.1 Examples of problems caused by moisture................................................................................ 2
1.2 Microscope and macro scope view of mold............................................................................... 4
2.1 The schematic for the moisture balance................................................................................... 19
2.2 The schematic for the energy balance...................................................................................... 21
2.3 The schematic for sorption isotherm........................................................................................ 29
2.4 Depiction of wall nodes for numerical solution....................................................................... 31
2.5 Depiction of wall nodes for the exterior boundary. ................................................................. 33
2.6 Flow diagram for envelope heat and moisture transfer model................................................. 36
2.7 Comparison for the node temperature with result from FEHT. ............................................... 37
2.8 Comparison for the inner surface temperature with result from Laplace Transform............... 38
2.9 Comparison of wall temperatures after rain shower with/without solar heating. .................... 41
2.10 Comparison of wall moisture contents after rain shower with/without solar heating............ 42
3.1 Experimental set up.................................................................................................................. 45
3.2 Delmhorst moisture content pin pairs and Kil-Mo-Trol Plus system. ................................... 48
3.3 Plan for the location of moisture pin pairs............................................................................... 49
3.4 Illustration of the effect of shrinkage and grain orientation on shape...................................... 50
3.5 Heating lamps, reflective shields and pyranometer. ................................................................ 52
3.6 Blower (driven by Variable Frequency Drive) and honey-comb-core outlet. ......................... 53
3.7 Relative humidity sensor and the ambient air temperature thermocouples. ............................ 54
3.8 Calibration for the moisture content pin pairs. ........................................................................ 56
3.9 Schematic of setup and measurements for estimating hc, hr and ........................................... 57
3.10 Specify the overall and the radiant heat transfer coefficient for the unheated side. .............. 58
3.11 Comparison of hc from tests and hc from a textbook correlation.......................................... 59
3.12 Preconditioning to soak the specimen. ............................................................................... 61
3.13 Room air dry bulb temperature during the experiment.......................................................... 66
3.14 Room air relative humidity during the experiment. ............................................................... 66
3.15 Measured and prediction surface temperatures for the wood specimen. ............................... 68
3.16 Moisture content at inch beneath surface B of the specimen............................................. 69
3.17 Moisture content at the center line of the specimen............................................................... 70
3.18 Moisture content at inch beneath surface A of the specimen............................................. 71
3.19 The specimen weight change for the drying process. ............................................................ 72
3.20 The specimen weight change for the drying process (with edge vapor leakage included in the
model)..................................................................................................................................... 73
4.1 Schematic of the slab-on-ground geometry. ............................................................................ 79
4.2 Example variations in ambient sol-air and zone temperatures................................................. 80

viii

Figure
Page
4.3 Schematic of the two-dimensional finite-element modeling mesh. ......................................... 82
4.4 Simplified 3-node models for a one-dimensional ground-coupled floor. ................................ 83
4.5 Example optimization of soil depth for centerline heat flux.................................................... 86
4.6 Comparison of centerline heat flux for finite-element and simplified models (Configuration A,
base case)................................................................................................................................ 86
4.7 Curve fit for optimized soil depth for centerline heat flux (Configuration A)......................... 87
4.8 Comparison of average floor surface heat flow flux for finite-element and simplified models
(Configuration A, base case). ................................................................................................. 88
4.9 Effect of soil conductivity and half-floor length on optimized perimeter heat loss factor
(Configuration A, no edge insulation). ................................................................................... 88
4.10 Curve-fit results for edge heat loss factor (Configuration A). ............................................... 89
4.11 Curve fit for optimized soil depth for centerline heat flux (Configuration B). ...................... 90
4.12 Curve-fit results for edge heat loss factor (Configuration B). ............................................... 91
4.13 Example simplified model accuracy for New York in summer (Configuration A)............... 92
4.14 Example simplified model accuracy for Chicago in winter (Configuration B). .................... 92
5.1 Schematic of moisture room paths and driving forces............................................................. 95
5.2 Example for typical daily indoor heat and moisture generation for a family of four............... 98
5.3 Schematic for attic and attic ventilation................................................................................. 100
5.4 The modular elements that form the whole building simulation. .......................................... 105
5.5 Flow diagram for the explicit solution scheme. ..................................................................... 107
5.6 Comparison of zone air temperatures with results from TRNSYS........................................ 109
5.7 Ambient air temperature and relative humidity for Indianapolis........................................... 111
5.8 The yearly attic air temperature and relative humidity. ......................................................... 112
5.9 Whole building simulation result for winter case. ................................................................. 114
5.10 Whole building simulation result for summer case.............................................................. 115
5.11 Cooling load condition for the equipment. .......................................................................... 117
5.12 Whole building simulation result for summer case.............................................................. 118
5.13 Isothermal absorption curve for selected building materials. .............................................. 119
5.14 Comparison of interior/exterior layer equilibrium RH for whole-building and stand-alone
wall modeling (vapor barrier on inside of insulation). ......................................................... 121
5.15 Comparison of interior/exterior layer equilibrium RH for whole-building and stand-alone
wall modeling (vapor barrier on outside of insulation). ....................................................... 122
6.1 Ambient air temperature and relative humidity for Minneapolis........................................... 126
6.2 Roof and ceiling equilibrium RH in Minneapolis.................................................................. 128
6.3 Wall structures internal equilibrium RH and temperature in Minneapolis (north)............... 129
6.4 Solar driven rain moisture build up for building structure..................................................... 130
6.5 Effect of vapor retarder location for cold climate condition.................................................. 131
6.6 Ambient air temperature and relative humidity for Nashville. .............................................. 132
6.7 Internal equilibrium RH and temperature for cold climate wall structure in Nashville......... 133
6.8 Sandwich wall structure internal equilibrium RH and temperature for Nashville. ................ 135
6.9 Conventional wall structures equilibrium RH for Nashville (no vapor retarder). ................ 136
6.10 Ambient air temperature and relative humidity in Phoenix. ................................................ 137
6.11 Roof and ceiling equilibrium RH for Phoenix. .................................................................... 139

ix

Figure
Page
6.12 Wall structure internal equilibrium RH and temperature for Phoenix. ................................ 140
6.13 Wall structure equilibrium RH and temperature for Phoenix (no vapor retarder). .............. 141
6.14 Ambient air temperature and relative humidity in Houston................................................. 142
6.15 Roof and ceiling equilibrium RH for Houston..................................................................... 143
6.16 Moisture performance for the original wall structure (facing north) in Houston................. 144
6.17 Modified building structure moisture performance for Houston. ........................................ 145
Appendix Figure
A.1 Illustration of building idealization in CONTAM. ............................................................... 166
A.2 Default wind pressure profile in MOIST whole building model (S&C_2_Long). ............... 174

NOMENCLATURE

absorptance for solar radiation, dimensionless

building envelope orifice opening area, m2

A1,A2,A3

coefficient of sorption isothermal equilibrium, A1,A2 and A3 dimensionless

Af

total floor surface area, m2

As

amplitude of the ground surface temperature wave, C

B1,B2,B3

coefficient of water vapor permeability function, B1 and B2 in kg/m-s-Pa or

Perminch, B3 dimensionless

cp

specific thermal capacity, J/kg-C or BTU/lbm-F

node thermal capacitance, J/C

C1,C2

coefficient for liquid diffusivity function, C1 in m2/s or ft2/hr, C2

dimensionless

Cd

discharge coefficient, dimensionless

Cf

mass flow coefficient, dimensionless

Cq

volume flow coefficient, dimensionless

CT, C

temperature and moister modification factor for k, W/m-C2 or BTU/hr-ft-F2

DAB

diffusion coefficient, m2/s or ft2/hr

liquid diffusivity, m2/s or ft2/hr

D*

ratio of the soil depth to the soil depth associated with the baseline (45 cm)

transfer function coefficient for previous outputs, W/m2

E0,E1,E2,E3 amplitude for temperature wave at 0,1,2,and 3 order for sin function, C
EH

node enthalpy flow term, J/m3-C-s or BTU/ft3-F-hr

xi

air mass flow rate, kg/s

Fp

perimeter heat loss factor, W/m-K

enthalpy, J/kg or BTU/lbm

hfg

heat of vaporization, J/kg or BTU/lbm

hc

convective heat transfer coefficient, W/m2-C or BTU/hr-ft2-F

hm

mass transfer coefficient, kg/m2-Pa-s or lbm/hr-ft2-inHg

hr

linearized radiation coefficient, W/m2-C or BTU/hr-ft2-F

height, m or ft;

Heqp

evenly distributed equipment long-wave radiation, W/m2 or BTU/ft2-hr

Hin

solar radiation intensity on the envelope inner surface, W/m2 or BTU/ft2-hr

Hsol

solar radiation intensity on envelope exterior surface, W/m2 or BTU/ft2-hr

specified moisture generation source, kg/m3-s or lb/ft3-h

acceleration of gravity (9.81 m/s2)

g wall

moisture flux from building envelope, kg/m2-s or lbm/ft2-hr

G& gen

indoor moisture generation rate, kg/s or lbm/hr

thermal conductivity, W/m-C or BTU/hr-ft-F

k*

ratio of the thermal conductivity to the baseline k of 1.73 W/m-K

kv

vegetation shade factor

hydraulic conductivity for liquid, kg/m-s-Pa or lbm/ft-hr-inHg

length of the internal zone, m

Le

Lewis number (Le = 0.927 for air)

mass, kg or lbm

vapor flux, kg/m2-s or lb/ft2-hr

Me

effective mass transfer coefficient, kg/m2-s-Pa or lbm/ft2-hr-inHg

xii

Mf

air film moisture transfer conductance, kg/m2-s-Pa or lbm/ft2-hr-inHg

Mp

surface moisture transfer conductance, kg/m2-s-Pa or lbm/ft2-hr-inHg

mass flux, mol/ m2-s or mol/ft2-hr

pressure, Pa or inHg

Pj, Pk

total pressure at zones j and k, Pa

Pm

length of the internal zone, m

Ps

pressure difference due to density and elevation differences, Pa

Pw

pressure difference due to wind, Pa

heat flux, W/m2

qf

heat flux from floor, W/m2

volumetric flow rate, m3/s;

Q&

total heat transfer rate or heat generation rate, W

node thermal resistance, K/W

transfer function coefficient for previous temperatures, W/m2-K

mass generation source, kg/m3-s or lb/ft3-hr

SHGC

Solar heat gain coefficient

time, s or hr

temperature, C or F

Tsol-air

sol-air temperature, C

Tg

far depth soil temperature, C

Tm

annual ambient air average temperature, C

concentration in dimensions of amount of substance, mol/m3

U-factor for windows heat conduction, W/m2-C

air velocities, m/s

xiii

VH

approach wind speed at the upwind wall height, m/s;

length, m or ft

entry and exit elevations, m

xiv

ABSTRACT

Zhong, Zhipeng. Ph. D., Purdue University, August, 2008. Combined Heat and Moisture
Transport Modeling for Residential Buildings. Major Professor: Dr. James Braun, School of
Mechanical Engineering.

Residential buildings are meant to provide a safe, healthy, and comfortable indoor
environment for occupants. However, many residential buildings suffer from a variety of
moisture problems. Unfavorable indoor relative humidity can make occupants uncomfortable.
Whats more, high humidity within building envelopes can lead to deterioration of material, and
cause some serious health problems due to the growth of mold and mildew.
This research addresses modeling of building envelope transient heat and moisture transfer
for structures used in residential buildings. The wall model development is based on a previously
developed one-dimensional model called MOIST 3.0, which incorporates water vapor diffusion
and water liquid capillary transfer. An important aspect of the current effort has been the
development of a wetted surface model that allows consideration of the moisture transfer caused
by wind-driven rain, where capillary liquid water transfer is an important mechanism.
In order to validate the MOIST model for heat/vapor/capillary transfer, an experiment was
devised and carried for drying of a sugar-pine panel. The panel was initially soaked until
saturation and then exposed to typically ambient conditions and surface radiant heating with
controlled air flow. The experiment was designed to simulate the effects of solar-driven moisture
transfer that would follow rain. An automated weighing system was used to trace the overall
wood moisture content and moisture pin-pairs measured the local moisture content within the test
specimen. During the transient drying test, radiant heating was projected on one surface then
switched to the other so as to develop a better understanding of the nature and significance of
solar-driven inward vapor diffusion. There was relatively good agreement between moisture

xv

content predictions and measurements from the moisture pin-pair sensors and very good
agreement of the overall moisture content from the weighing system.
Another contribution of this research is a simplified model for predicting transient heat
transfer in ground-coupled floor slabs. The model simplifies the traditional two-dimensional
approach by employing two parallel one-dimensional transient heat transfer paths. The model
incorporates correlations for the effects of ground soil properties and edge insulation that were
developed for two common floor configurations. Hourly heat flux predictions compared very
well with predictions from a two-dimensional finite-element program for these two geometries.
The method can be integrated within hourly simulation programs so that fast estimation of
transient heat transfer for the indoor air to the slab-on-ground can be realized.
A major contribution of the research is the coupling of the detailed envelope model with a
whole building model that allows investigation of the impacts of envelope design on occupant
comfort, energy use, and wall material conditions that can lead to mold growth, insulation
degradation, etc.

In addition to the external wall model, the whole building model incorporates

several individual models that were developed, including: 1) weather data treatment including
wind driven rain and solar radiation, 2) air infiltration and inter-zonal air flow, 3) indoor heat and
moisture generation, 4) heat transfer through slab-on-ground floors, 5) indoor moisture storage
within furnishings and other soft materials, and 6) HVAC equipment. The model allows a
detailed analysis of indoor/attic air conditions and moisture information for building materials.
The original MOIST model only considered individual walls and assumed constant indoor
air conditions. In order to evaluate the importance of coupling the wall analysis to indoor
conditions, results of the whole building analysis were compared with a single wall analysis
performed with constant indoor conditions. The two approaches gave significantly different
predictions for moisture levels within materials located near the indoor space but gave essentially
the same results for layers located away from the interior space. The stand-alone wall analysis
resulted in relatively stable moisture for the interior gypsum layer because of the constant indoor
boundary condition, whereas the interior surface for the whole building analyses varied over a
relatively large range. Although the moisture levels would not cause mold or material damage
problems for the case study considered, the stand-alone wall analysis would not identify potential
problems that could occur due to more significant indoor moisture gains that could potentially

xvi

occur, such as moisture gains within a bathroom not having an exhaust fan. The whole building
model has the potential for evaluating moisture problems caused by indoor conditions and also
can consider the impacts of design choices on indoor air moisture levels. On the other hand, the
standalone model enjoys much lower computational cost and may be adequate for evaluating
many moisture problems caused by ambient effects.
The whole building analysis tool was used to perform some case studies.

Moisture

performance of building envelopes was analyzed in some featured climates (heating climate,
mixed climate, and dry or humid cooling climate) with representative building constructions. It
was shown that for heating climates, a vapor retarder should be placed close to the room side so
as to prevent vapor excursion from the relatively warm and moist indoor air. For humid cooling
dominated climates, a vapor retarder should be positioned close to the ambient side to stop vapor
incursions. On the other hand, for hot and dry climates it is not necessary to use a vapor retarder.
For mixed climates, a conventional wall structure without a vapor retarder can work adequately.
However, a sandwiched structure with vapor retarders on both sides of a low-permeability
insulation is another option because it can handle either vapor incursion from the ambient during
summer or vapor excursion from the indoor air during winter.

CHAPTER 1. INTRODUCTION

Building envelopes separate indoor from outdoor environments and play the most important
role in sustaining necessary indoor comfort conditions for occupants. According to Rengin et al.
(2004), an optimum building envelope should provide visual, thermal and acoustical comfort in
accordance with the function of the room. In addition, Rousseau (2003) notes that successful
building design requires an understanding of moisture movement in building materials and
techniques for managing moisture that come from controlling heat, air and moisture transport
through the careful choice of materials properties.

1.1. Building Structure Problems Caused by Moisture

HVAC designers must consider and deal with moisture in almost all of their work. Moisture,
from whatever sources, is commonly regarded as the single greatest threat to the durability and
long-term performance of the housing stock (Newport Partners Report, 2004). Figure 1.1 shows
some examples of damage to building materials caused by moisture. Failure to properly manage
the transport of heat, air and moisture across the wall assembly can cause the following problems:
1.

Electrochemical corrosion of metal components such as HVAC equipment, ducts,

structural framing, reinforcement bars, masonry anchors, etc.;

2.

The chemical deterioration and dissolution of materials such as gypsum sheathing,

ceiling tiles, especially wood products on the exterior walls (Roussau, 2004);

3.

Discoloration of building finishes;

4.

Volume changes (swelling, warping and shrinkage) that can cause degradation of
appearance, structural failure, cracking, etc.;

5.

Freeze-thaw deterioration of concrete, stone, and masonry, especially for buildings in

cold areas if the building materials contain moisture (e.g., if the concrete holds more
than 44% moisture by pore volume, freeze-thaw damage to the concrete block may
happen if the temperature drops below the freezing point (Fagerlund, 1977);

6.

The increase of material thermal conductivity due to the moisture within the material;

7.

The growth of biological forms, including molds, mildews, mites, etc.

a.) corrosion of metal

b.) deterioration of ceiling

c.) discoloration of finish

d.) structure cracking

e.) freeze-thaw deterioration of concrete

Figure 1.1 Examples of problems caused by moisture.

A survey and overview of the findings from cases studies of moisture problems in residential
crawl spaces, basements, roofs, and inside surfaces of exterior walls was presented by Tsongas
(2000). Most of the emphasis was on relevant U.S. and Canadian studies. It was found that most
residential buildings suffer from a variety of moisture problems. For instance, based on a survey
of 334 Iowa households, 98% of the residents reported at least one type of moisture problem.
The most common types of moisture problems were: condensation on windows (62%), exterior
paint peeling (41%), staining of interior windows frames and sills (31%), mildew on
walls/ceilings or closets (23%), decay/rotting of interior window frames/sills (20%),
moisture/mildew problems in summer (18%), frost/condensation on walls/ceilings (13%), and
interior paint peeling (10%).
Results from a survey made in the early 1980s in Canada (Jacques Rousseau, 1982) for case
studies related to the occurrence and significance of bulk moisture problems is summarized in
Table 1.1.

Table 1.1 Survey of moisture problems in residential buildings.

Location

Sampling
Units

Newfoundland

10,400

*Serious
MoistureRelated
problem
27%

Maritimes

32,800

Quebec

Mold
and
Mildew

Window
Condensation

Attic
Condensation

Siding
Damage

3.0%

1.2%

2.2%

24%

1.4%

0.6%

0.3%

0.2%

0.8%

164,00

0.7%

0.3%

0.2%

0.6%

0%

Ontario

276,000

0.7%

0.4%

0.2%

0.4%

0.004%

Prairies

135,000

1.4%

0.1%

1.3%

0.7%

0.04%

B. Columbia

71,300

3.0%

1.3%

0.2%

2.7%

0%

*Serious moisture-related problems were regarded as:

Serious Condensation in Attic wood moisture content exceeding 22% or mold and mildew growth
covers more than 50% of the attic sheathing and roof joists.
Serious Wall Cavity Moisture wood moisture content exceeding 22%
Serious Exterior Siding Moisture buckling or warping of more than 50% of the wall area, or paint
damage affecting more than 0.3 square meters

Some other surveys can be found for a variety of climates from different countries such as
Finland (Lappalainen et al., 2001), Mauritius (Bholah and Subratty, 2003) and Portugal
(Loureno et al., 2006). There are many common problems identified in these studies, such as

growth of mold and mildew. Molds are forms of fungi, which are distinctly different from plants
and animals. Over 1,000 types of molds have been found in houses in North America. Mildew is
a term that is often used interchangeably with mold, generally referring to mold growing on
fabrics and bathroom tiles. Mold can appear cottony, velvety, grainy, or leathery, and it can be
any color, including white, pink, yellow, green, brown, gray, or black. Since the application of
biocides to kill fungi is always accompanied by additional health risks especially when used
indoors, the biocides are used selectively based on knowledge of the mechanism of mold growth.
Molds can produce and send out spores, which can be seen through a microscope, as shown in
Figure 1.2a, and which act like seeds. The following conditions are necessary and sufficient for
mold growth to occur on surfaces (Lstiburek etc. 1991):
1.

Mold spores must be present;

2.

A nutrient base must be available (most surfaces contain nutrients);

3.

Temperature range between 4~38C (40~100F);

4.

Relative humidity near the surface is above 70%.

Of these conditions, relative humidity near surfaces is the most practical to control. Figure
1.2b gives an example for the mold on books. Molds and mildews are one of the major causes of
the deterioration and decay of building materials. In addition, moisture-related biological growth
has taken on new significance due to the fact that the growth can have a major effect on indoor air
quality (IAQ) and occupant health, which will be discussed later.

a.) Micro view of molds

b.) Macro view of molds

Figure 1.2 Microscope and macro scope view of mold.

A 70% relative humidity criteria for mold growth has been adopted by the Moisture Control
Handbook (Lstiburek and Carmody, 1991). In Europe, the lowest boundary line for possible
fungus activity is called LIM (Lowest Isopleth for Mold). DIN (German Institute for
Standardization) 4108 and DIN EN ISO 13788 mention 80% surface relative humidity as the
critical growth limit for mold, and DIN 68800-2 gives a material humidity criteria (e.g. 20 MC%)
for building products made of wood or wood materials. In addition, HUD (the Department of
Housing and Urban Development of the U.S) uses 80% average relative humidity on a surface
over a certain period of time during summer as the critical condition for mold growth, and
requires manufactured houses to have interior vapor retarders to avoid moisture problems (U.S.
Department of Housing and Urban Development, 1994).
Detailed discussion of the germination of molds is presented by Sedlbauer (2000). A
biohygrothermal method was developed so as to predict mold fungus formation based on all
mentioned biological growth conditions (temperature, humidity and substrate) for mold fungi at
transient boundary conditions.
Mold can grow on lots of materials found in homes, such as stucco (Kuenzel, 2001), organic
coatings (Van der Wel, 1999), wood (TenWolde, 1994), insulation materials (Andreas etc., 1999),
and everyday dust and dirt. Mold can grow on both internal and external surfaces of building
envelopes. Building components, such as windows, closets, attics, crawl spaces and basements,
are all places that mold is frequently found.

1.2. Comfort and Health Problems Related to Moisture

It is well known that temperature and humidity of indoor air are key factors that influence
directly the thermal sensation of the human body (Fanger, 1972). Furthermore, it has been found
that skin humidity is a major reason for discomfort at high air humidity (Toftum et al., 19981).
For humans, the respiratory track acts as an air-conditioning system that regulates the humidity
content and temperature of inhaled air on its way to the lungs. It has been verified that insufficient
respiratory cooling is a cause of local thermal discomfort, and the respiratory system has more
stringent requirements for air humidity than the skin (Toftum et al., 19982).

Whats more, Fang et al. (2000) conducted human body exposure experiments and
discovered that:
1.

Air temperature and humidity have a significant impact on both the immediate and the
adapted perception of IAQ;

2.

Decreasing the air temperature and humidity may ameliorate the perceived IAQ
significantly, and the acceptability of air can be increased linearly with decreasing
enthalpy of air;

3.

Ventilation required for comfort may be significantly reduced when decreasing indoor
air enthalpy. The ventilation rate can be decreased from 10 L/s/person (21.2 cfm/person,
prescribed in existing ASHRAE standard 62-1999 for office buildings) to 3.5
L/s/person (7.4 cfm/person) when the indoor air enthalpy decreases from 45 kJ/kg (19
Btu/lb) at 23C/50% RH to 35 kJ/kg (15 Btu/lb) at 20C/40% RH without sacrificing
the perceived air quality.

Although high humidity levels are not good for occupant comfort, very low humidities can
lead to increased infections for the respiration system. In addition, a low humidity level is also
responsible for electro-static shocks of clothing, carpets, etc.
In addition to thermal comfort, it is a generally-known fact that fungi (e.g., mold) caused by
moisture can harm occupant health, and this type of IAQ problem has received increasing
attention recently.
Fungi can cause health hazards to human beings through inhalation or contact for allergic
reactions, eye and respiratory irritation, infection and toxicity. For people that are sensitive to
molds, symptoms such as nasal and sinus irritation or congestion, dry hacking cough, wheezing,
skin rashes or burning, watery or reddened eyes may occur. People with severe allergies to molds
may have more serious reactions, such as hay-fever-like symptoms or shortness of breath. People
with chronic illnesses or people with immune system problems may be more likely to get
infections from certain molds, viruses and bacteria. Molds can also trigger asthma attacks in
persons with asthma. Headaches, memory problems, mood swings, nosebleeds and body aches
and pains are sometimes reported in mold complaints (www.cdc.gov).

The health problems that come from exposure to molds can be summarized as:
1.

Respiratory problems, such as wheezing, and difficulty in breathing;

2.

Nasal and sinus congestion;

3.

Burning and watering eyes;

4.

Dry, hacking cough;

5.

Sore throat;

6.

Nose and throat irritation;

7.

Shortness of breath;

8.

Skin irritation;

9.

Mood problems.

Sick Building Syndrome (SBS) is mentioned more and more frequently in the context of
mold. Occupants inside sick buildings can suffer several nonspecific symptoms (e.g. mucosal
irritations, smarting eyes, repeated common colds, fatigue and weakness of concentration)
without there being a definitive cause. Occurrence of SBC has been attributed to a variety of
causes, including viruses, pollen, mites, pesticides, tobacco smoke, carbon dioxide, carbon
monoxide, nitrogen oxides, ozone, radon, volatile organic compounds (VOC) emissions from
building and facility materials, and, of course, fungus spores. A recent paper addressed the
general hypothesis that Sick Building Syndrome (SBS) is associated with exposure to waterdamaged buildings (Shoemaker et al., 2004).
In addition to fungi, some other organisms that spawn in moisture damaged materials, such
as amoebae, are found to favor co-occurrence with bacteria and fungi and may also cause IAQ
problems (Terhi et al., 2004). Based on a careful investigation in Finland within 124 building
material samples from moisture-damaged buildings, it was found that amoebae were detected in
22% of the samples.
It is reasonable to assume that phenomena leading to harmful exposures and health effects
may include chemical deterioration of moistened building materials, and microbial growth on
moistened materials, with chemical and particle emissions of biological origin. Therefore, a
moisture damage index model was built by Haverinen et al. (2003). Using this model, it was seen
that the predicted risk for respiratory symptoms increases with increasing severity of damage.
Also, a linear relationship between residential moisture damage and occupant reported health

symptoms, which was developed as a compromise between the knowledge-based and statistical
models, is preferred because of its simplicity.

1.3. Introduction of Moisture Transfer

Moisture moves under different mechanisms in each of it phases. The primary transport
processes, beginning with the least powerful to the most, are:
1.

Vapor diffusion within some porous materials;

2.

Vapor convection (i.e., air movement);

3.

Liquid water capillarity (i.e., wicking) through porous materials;

4.

Liquid gravity flow (including hydrostatic pressure) through cracks, openings;

Vapor diffusion moves water vapor from regions of high vapor pressure to low pressure.
Water vapor pressure is a function of both temperature and relative humidity. High temperature
and relative humidity will result in high water vapor pressure the vapor diffusion driving force.
Diffusion acts to move vapor through air, or through the air within the porous materials. Water
vapor does not diffuse through non-porous materials like steel, glass, some plastics, etc.
Water vapor diffusion plays an important role in transporting water vapor into porous
building enclosures where it can sometimes condense when moisture builds up. To control this
flow, vapor barriers are often specified. However, vapor diffusion alone is not usually the cause
of moisture damage in walls. Vapor diffusion can be important in roofs and walls with absorptive
claddings. Rainwater is absorbed into the cladding and subsequently heated by the sun. Even in
cold climates, very high vapor pressure gradients can form in this situation and move damaging
quantities of moisture inward.
Capillary suction moves liquid moisture slowly and steadily through porous materials from
regions of high liquid concentration to regions of low concentration. The smaller the pores, the
more powerful the capillary suction but the slower the flow. Although the rate of moisture
transport by this mechanism is relatively slow, it can act for years. Capillary transfer is important
in constructions contacting soil, and rain-wetted surfaces. Capillary flow can be controlled or

eliminated by installing a barrier to capillary flow. A small air gap or a capillary inactive material
is often sufficient.
Gravity flow can be the most powerful means of moisture transport. Very large quantities of
liquid water, often measured in liter per second, can flow downward through openings, cracks,
pipes, or air spaces when driven by gravity. Gravity flow requires relatively large openings,
which require the dimension to be larger than 1 mm or larger, since capillary suction forces tend
to overwhelm gravity forces in small pores. Hence, water will not flow out of a saturated brick,
but can flow through a screw hole in a plastic windowsill.
Small flows of air can move much greater quantities of water vapor than diffusion can.
Convection through openings in the building enclosure is a major cause of interstitial
condensation, sometimes ten to hundreds of times more significant than diffusion. Therefore,
durable, stiff and strong air barrier systems must be provided in all building enclosures to control
or eliminate convective moisture transport.
Transport processes rarely act alone to move moisture within and through buildings. In
reality, a number of transport processes act in parallel and series. For example, liquid water from
the groundwater may wick upward to below the surface of a crawlspace floor, where it evaporates
and moves by diffusion through the soil into the crawlspace. Small air pressures then act to
transport the water vapor into the main space of the building, raising the interior humidity and
resulting in condensation on a cold water pipe within a suspended ceiling. The condensation
accumulates at this point until it begins dripping onto the drywall ceiling below. Here mold
accumulates and the ceiling is damaged.
Moisture gets into homes from both the outside and the inside. Typical sources of dampness
from the outside are:
1.

Floods, ground water;

2.

Roof leaks;

3.

Inadequate or poor flashing;

4.

Missing downspouts and gutters;

5.

Window or door leaks;

6.

Damaged building materials, etc.

10

Typical sources of dampness from the inside are:

1.

Moisture in floors;

2.

Unventilated bathrooms and kitchens;

3.

Leaky plumbing;

4.

Condensation in walls and windows;

5.

Unventilated dryers;

6.

People, pets, plants, etc.

Ground soil can be a significant source of moisture near basements, crawl spaces
foundations and the first floors of buildings. Soil is a large source of moisture in both liquid and
vapor forms. Liquid water draining directly from the surface or from the water table tends to
penetrate through cracks, holes and other unintentional openings. The liquid water stored within
the soil matrix will wick through the soil and porous building materials like concrete, stone,
wood, etc. Stored liquid water deep below the surface and bound to the soil also provides a
practically inexhaustible supply of water vapor. Since diffusion is a less powerful mechanism,
soil water vapor is less significant than liquid water, but it is still a large moisture source. Water
vapor from soil enters buildings primarily by diffusion, although convection (air leakage) may
occur through soil in some cases. Soil in a wet basement or an uncovered crawlspace has been
found to evaporate at a rate of 100 to 500 g/m2 per day (Straube, 2002). Gravels or crushed stones
act as a capillary break between moist soil and the building enclosure, while the air gap between
stones allows vapor diffusion to act relatively unhindered. Hence, sheet or paint-applied vapor
barriers are often widely used near the exterior of below grad assemblies.
Pipe leaks and rain penetration are sources of water that must be avoided. Rain deposition on
roofs is usually of the order of several hundred to one thousand kg/m2 (200 lb/ft2) in most
climates. Walls typically receive about 25% to 50% of this load (Straube et al., 2000). Even little
leakage can cause serious problems, since liquid water from these sources can quickly reach
catastrophic proportions. Rain leaks or plumbing pipe failures can result in hundreds of gallons of
water being discharged into a building.
Moisture built-in to building materials can be important, but is specific to the type of
building construction and only plays a role for the first few years of a buildings life. Wood
framing typically loses close to 10% of its weight in moisture. A normal concrete mix contains

11

about 200 kg (440 lb) of water per cubic meter, of which about half is later released as vapor.
Hence, a typical house basement system containing 20 to 30 m3 (700 to 1050 ft3) of concrete will
release thousand liters of water over the first year or two. Similarly, a 200 mm (8 in.) thick
reinforced concrete floor slab in an office building can be expected to release 20 liters of water
per m2 during the first two years (Straube, 2002).
Water vapor can be almost as problematic as direct liquid water sources, although the
magnitude of the moisture involved is typically much less. Water vapor condensation may happen
on cold surfaces like water pipes, walls and window surfaces. Water vapor from the exterior
enters the building both through intentional ventilation and unintentional air infiltration through
the building enclosure and ducts.
In many types of buildings, a significant amount of moisture can be released or generated by
occupants, their activities and process. Also pets and plants raised inside the building can also
contribute to moisture generation.
For moisture-damage problems to occur, it is necessary for at least the following conditions
to be satisfied:
1.

A moisture source must be available;

2.

There must be a route or means for this moisture to travel;

3.

There must be some driving force to cause moisture movement, and

4.

The materials involved must be susceptible to moisture damage.

It is possible to avoid a moisture problem by eliminating any one of the four conditions. In
reality, it is practically impossible to remove all moisture sources, to create buildings with no
imperfections, or to remove all forces driving the moisture movement. It is also uneconomical to
use only materials that are not susceptible to moisture damage. Hence, in practice, it is often
advantageous to address two or more of these prerequisites so as to reduce the probability of
having a problem.
In practice, for new constructions, building envelope assemblies with a high tolerance for
moisture are, of course, highly recommended. While, for existing buildings with moisture
problems, a change in building operation is often the practical option to control moisture. This

12

usually involves manipulating indoor temperature and humidity. However, moisture

accumulation in the building envelope also can be minimized by controlling the dominant
direction of airflow. Whats more, ventilation can be an effective way for removing moisture in
many cold winter climates: ventilation is intended to not only provide acceptable indoor air
quality, but also control humidity (TenWolde et al., 1994).

1.4. Objectives of this Research

This project was motivated by a desire to improve an existing model developed by NIST
called MOIST 3.0 (Burch et al., 1997). MOIST 3.0 is a public software that models transient
one-dimensional hygrothermal performance of external building envelope structures subject to
limited boundary conditions. The primary deficiencies in the MOIST model include:
1.

Wetting of exterior surfaces by rain is not considered;

2.

Very simple model for coupling of external walls and ambient to internal space:
a)

does not couple the determination of the indoor air humidity to the moisture
transport through exterior structures,

b)

fully-mixed internal zone state with lumped storage for energy and moisture within
internal structures and furnishings (transport of moisture by air movement is
neglected),

c)

neglects the effects of solar transmitted through windows and internal radiative
gains,

d)

doesnt include the effects of wind and ambient temperature on infiltration,

e)

doesnt consider ground coupling;

3.

Doesnt consider water pooling around foundation due to groundwater or snow melt;

4.

Snow insulating effects for horizontal surfaces are neglected;

5.

Transport of heat by liquid movement in walls is neglected;

6.

Only considers one-dimensional heat and mass transfer;

The last three assumptions are relatively common and reasonable for most applications.
This thesis will address the first two deficiencies. Although the third deficiency can be very
important, it is difficult to address and beyond the scope of the current effort.

13

The first objective of this thesis is to develop a wetted surface model for exterior walls that
considers the effects of wind-driven-rain where water capillary transfer dominates. A number of
different codes have been developed for moisture transport in walls that consider capillary
transfer model and include a wetted surface condition (see Chapter 2). However, very little work
has been performed for validation of the water capillary transfer. Therefore, a key contribution of
the current work is a detailed validation of the water capillary wall model.
The second objective is to develop a detailed whole-building model for residences that more
appropriately models coupling of the envelope and ambient to the internal space, while
considering both energy and moisture transport. This will allow evaluations of the impacts of
envelope design on occupant comfort, energy use, and wall material conditions that can lead to
mold growth, insulation degradation, etc. In addition to the transient heat and mass transfer
model for exterior walls, the whole-building model incorporates models for air/moisture
infiltration and air/moisture movement within the space, moisture and heat generation, windows,
transient heat transfer within a floor slab and ground, energy and moisture storage within internal
structures and furnishings, and equipment.
One of primary considerations in the model development is computational speed. The
transport mechanisms for moisture within structural materials and heat within the ground are
relatively slow and require long simulation periods (e.g., multiple years) to erase the effects of
initial conditions and to identify worst case conditions. In addition, the simulation time step
needs to be significantly less than an hour in order to accommodate faster dynamics associated
with the zone air and internal furnishings that respond to driving conditions (internal gains, solar
radiation, etc.) that may vary significantly on an hourly basis.
The primary reason for incorporating internal air movement is to accommodate large
differences in moisture conditions that can occur within a house on different floors and within
confined spaces having high moisture generation (e.g., bathrooms).

For instance, moisture

migration caused by air flow from a lower level to a higher level because of the stack effect can
cause severe condensation on the top floors for multi-floor buildings (Clarkin and Brennan,
1998). It is not necessary to determine detailed spatial variations in air conditions in order to
consider these effects. As a result, a simplified airflow network modeling approach incorporated
within CONTAMW 2.0 (Dols and Walton, 2002) will be used to solve air flow and vapor

14

migration for multiple zones. A simplified version of this model has been developed to provide
computationally efficient estimates of air flow rates from ambient to the building and from one
zone to another zone inside a multi-zone building.

The coupling of the air flow and vapor

migration model to the hygrothermal wall model and whole building model will be a unique
contribution of this thesis.
Existing modeling approaches are also being used for other elements of the whole building
model. However, a key contribution of the research will be to identify modeling simplifications
and coupling approaches that lead to computational improvements. For instance, a simplified
model has been developed for transient ground heat transfer with slab-on-grade floors.

In

addition, different approaches are considered for solving the system of equations that characterize
energy and moisture transport in buildings.
The third objective of this thesis is to utilize the whole-building moisture model to
investigate the impact of design choices and external driving factors on moisture buildup in
materials and internal moisture levels. The results of this analysis will be used to identify
guidelines for appropriate design practices so as to prevent and correct specific moisture
problems.

1.5. Organization of this Thesis

This thesis presents an enhanced heat and moisture transfer model for building materials,
which includes heat diffusion, water vapor diffusion and water liquid capillary transfer. The wall
model is driven by different boundary conditions, including dry bulb temperature, water vapor
pressure, solar radiation, and wind-driven-rain (which may bring the building material to a wet
regime where capillary transfer dominates). In order to validate the liquid capillary transfer, an
experiment which employed moisture pin-pairs and automated weighing system, was performed
for the drying process of the soaked wood specimen. Then, the validated single wall model was
incorporated into a whole building heat and moisture balance model that includes building
structure heat and moisture transfer, heat loss from a slab-on-ground floor, indoor heat and
moisture generation, inter-zonal and infiltration air flow, soft furnishing (bedding and curtain,
etc) moisture buffering, and HVAC equipment heat and moisture removal. This whole building

15

simulation toolkit was applied to evaluate building envelope heat and moisture performance for a
number of case studies and some general guidelines were developed.
There are seven chapters in this thesis as follows:
Chapter 1,

Introduction. This chapter provided basic background on problems caused by

moisture, such as building material damage, thermal comfort problems, and
occupant health problems due to mold and mildew growth. The purpose was to
provide a motivation for the objectives, which were also presented in this chapter.

Chapter 2, Modeling of heat and moisture transfer in building exterior walls. The onedimensional mathematical heat and moisture transfer model that is incorporated in
MOIST 3.0 is developed. The assumptions and limitations of this model are
discussed and compared with other existing models from the literature. The model
from MOIST 3.0 is then enhanced to consider wetted surfaces caused by winddriven rain and is used to demonstrate the effects of solar driven moisture build up
within building materials.

Chapter 3,

Experimental validation of exterior wall model.

A laboratory test was

performed for the drying process of a sugar-pine panel from a soaked initial
condition.

The surfaces were exposed to radiant heating and controlled air flow

boundary conditions and the data was used to validate the MOIST model for
heat/vapor/capillary transfer. An automated weighing system was used to trace the
overall wood moisture content and moisture pin-pairs measured the local moisture
content within the test specimen. During the transient drying test, radiant heating
was projected on one surface then switched to the other so as to develop a better
understanding of the nature and significance of solar-driven inward vapor
diffusion.

Chapter 4, Ground-coupled slab floor heat transfer model. A simplified model is developed
for predicting transient heat transfer in ground-coupled floor slabs that can be
integrated within hourly simulation programs. Since residential buildings usually
have only one or two floors, heat transfer to or from the slab can be a significant

16

component in the overall energy analysis. Results from the simplified model are
compared with results from a finite-element model.

Chapter 5, Whole building heat and moisture balance modeling. The whole-building model
couples the exterior wall and floor models to other models necessary to determine
moisture and mass balances within the space. These models include:
1. Weather data treatment: wind driven rain, sky temperature and solar radiation,
2. Indoor heat and moisture generation models,
3. Simplified multi-zone air flow model for air flow from ambient to the zone and
between zones,
4. Ground floor heat transfer model,
5. Indoor moisture buffering model associated with soft furnishing,
6. HVAC equipment (moisture removal by coils and energy consumption by
compressor).
With these detailed modules, the whole building simulation toolkit can realize heat
and moisture analysis for room/attic air temperature and relative humidity, and
moisture performance of building structures. In addition to presenting the whole
building analysis model, this chapter presents comparisons of results from the
whole building analysis with results for a single wall analysis performed with
constant indoor conditions.

Chapter 6, Case studies. Case studies for building heat and moisture performance are presented
to investigate the impact of design choices and external driving factors on moisture
buildup in materials and internal moisture levels. The results of this analysis are
used to identify guidelines for appropriate design practices so as to prevent and
correct specific moisture problems.

Chapter 7, Conclusions and future work. This chapter presents the overall conclusions,
remaining challenges and recommendations for future work.

17

CHAPTER 2. MODELING OF HEAT AND MOISTURE TRANSFER IN EXTERIOR

WALLS

Building envelope hygrothermal analysis is complex, because heat and moisture transfer are
transient, and highly coupled with each other. Even though experimental studies are the most
direct approach for studying the moisture performance of building envelopes, testing is very
costly and time consuming. In particular, very long experimental durations would be necessary
due to very low speed of moisture migration. Therefore, it would be very difficult to
experimentally investigate varieties of building assemblies under different kinds of indoor and
outdoor conditions.

On the other hand, numerical simulation techniques can provide relatively

fast estimates of the heat and moisture performance of the building materials with acceptable
accuracy.
This chapter presents the basic assumptions and modeling approach that will be applied for
heat and moisture within external wall assemblies. This approach is compared with other existing
models and improvements are developed to handle wetted surface effects. Some demonstration
results are then given for the effects of solar driven vapor diffusion after wind-driven rain.

2.1. Modeling of The Heat and Moisture Transfer in Building Envelope

2.1.1. Major Assumptions of the Heat and Moisture Transfer Model

In this thesis, the building envelope transient heat and moisture transfer model is developed
based on MOIST 3.0 (Burch and Chi, 1997) in order to analyze the moisture performance of
building materials. This model incorporates the following assumptions:
1.

Heat and moisture transfer are one-dimensional;

2.

The construction is airtight, and the transport of moisture by air movement is neglected;

18
3.

Snow accumulation on horizontal surfaces, and its effect on the solar absorptance and
thermal resistance is neglected;

Other specific assumptions will be discussed later.

2.1.2. Mathematical Description of the Heat and Moisture Transfer Model

The migration of moisture is primarily the result of vapor diffusion and capillary transfer of
liquid. It turns out that the laws describing the migration of heat, vapor, and liquid look very
similar. Temperature is used as the potential or driving force for heat flow, water vapor pressure
is the driving force for the vapor transfer, and capillary suction pressure is the driving force for
liquid flow.

For a wall assembly, the following equations can be used to represent these three

mechanisms for one-dimensional transport:

1. Law of Fourier (for heat conduction)

q " = k

T
y

(2.1)

where q is the heat flux, k is the thermal conductivity of the material, T is temperature, and y is
the length.

2. Ficks Law (for vapor diffusion)

n " A = DAB

u A
y

(2.2a)

where nA is the amount of flux for substance A, is the density, DAB is diffusion coefficient of
substance A in substance B, and uA is the concentration of substance A. The diffusion flux is
usually expressed as a mass flux, so that Eqn. (2.2a) can be transformed to:

m "v =

Pv
y

(2.2b)

where mv is the mass flux for vapor, is the water vapor permeability, and Pv is water vapor
pressure.

19
3. Darcys Law (for capillary transfer):

m "w = K

Pl
y

(2.3)

where mw is the mass flux for liquid water, K is the hydraulic conductivity for liquid, and Pl is
capillary pressure.

It should be noted that the signs of the equations for Ficks law and Darcys law are
different. This is because water vapor flows in the opposite direction of the gradient in water
vapor pressure, whereas capillary water flows in the same direction of the gradient in capillary
pressure (typically called capillary suction pressure).

Mass balance:
Figure 2.1 depicts the mass transfers within a differential element of a wall having a source
generation. For a given phase k (vapor, liquid, or solid), a transient mass balance on the water
yields:

k
m
= k + Sk
t
y

(2.4a)

where d is the building material density for dry conditions (subscript d means dry condition), is
the moisture content in kg/kg or lbm/lbm, t is time, and S is the source term. In subsequent
developments, the subscript k will be replaced with v for vapor, w for water and ice for ice.
Sk

mk

mk +

mk
y

dy

Figure 2.1 The schematic for the moisture balance.

20
For the moisture transfer problem within a building envelope, there is no source term, but
there is phase change, so it is reasonable to change Eqn. (2.4a) to:

k
m
= k + Ik
t
y

(2.4b)

where Ik represents a transfer of moisture from the phase k due to phase change.
1. For vapor, the application of Eqn. (2.4b) gives:

v
m
= v + Iv
t
y

(2.5a)

2. For liquid water:

w
m
= w + Iw
t
y

(2.5b)

3. For ice, there is no ice flow so that:

ice
= I ice
t

(2.5c)

The summation of Eqn. (2.5a, b, and c) gives:

( v + w + ice )
( mv + mw )
=
+ Ik
t
y
ice , w , v

(2.6)

However, the individual transfers associated with the phase changes must balance internally, so
that:

Ik = 0

(2.7)

ice , w, v

Thus, Eqn. (2.6) can be simplified to:

( mv + mw )

=
t
y

(2.8)

21
After plugging Eqn. (2.2b) and (2.3) into Eqn. (2.8), the overall moisture balance is:

Pv Pl

K
= d
y y y y
t

(2.9)

Energy balance:
The energy flow for a differential element within a wall is depicted in Figure 2.2 There is no
energy generation within the envelope, and in the absence of freezing or thawing, the primary
energy flows are heat conduction and enthalpy flow caused by liquid water transfer and water
vapor transfer. An overall energy balance on the wall material and moisture within the differential
element can be written as:
dy

m v" h v +

mv"hv

mw"hw

q"

T,
= w+v

m v"
h
hv dy + v mv"dy
y
y

m w" h w +

q" +

m w"
h
hw dy + w mw"dy
y
y

q"
dy
y

Figure 2.2 The schematic for the energy balance.

22

m v"
h
hv dy + v mv"dy ) +
y
y
m w"
h
hw dy + w mw"dy ) +
m w" h w - (m w" h w +
y
y

m v" h v - (m v" h v +

q"- (q" +

(2.10a)

d ( c p ,d + w c p , w + v c p ,v ) T + d v h fg
q"
dy ) =
dy
y
t

where h is the enthalpy, and hfg is the heat of vaporization. Since v is much smaller than w, it is
reasonable to assume that =w. Furthermore, because cp,v is much smaller than cp,w, the sensible
heat of the water vapor within the building material is negligible and Eqn. (2.10a) can be
simplified as:
m v"
h
hv dy + v mv"dy ) +
y
y
m w"
h
hw dy + w mw"dy ) +
m w" h w - (m w" h w +
y
y

m v" h v - (m v" h v +

q"- (q" +

(2.10b)

d ( c p ,d + c p , w ) T + d v h fg
q"
dy ) =
dy
y
t

Eqn. (2.10b) can be simplified by eliminating common terms, so that:

-

m v"
h
m w"
h
q" d ( c p ,d + c p , w ) T + d v h fg
hv v mv"
hw w mw"
=
y
y
y
y
y
t

(2.11)

In addition, the enthalpy for vapor at its given temperature can be expressed in term of heat of
fusion and liquid enthalpy for the same temperature:

h v =h fg +h w

(2.12)

Eqn. (2.12) can be substituted into Eqn. (2.11) to yield:

-

m v"
m w"
h
h
q" d ( c p ,d + c p , w ) T + d v h fg
( hfg + hw )
hw v mv" w mw"
=
y
y
y
y
y
t

(2.13a)
and then rewritten as:
-

m w" m v"
m v"
h v
h
q" d ( c p ,d + c p , w ) T + d v h fg
hfg
+
mv" w mw"
=
hw
y
y
y
y
y
t
y

(2.13b)
Substituting the mass balance of Eqn. (2.8):

23

m v"
h
h

q" d ( c p , d + c p , w ) T + d v h fg
hfg + d
hw v mv" w mw "
=
y
t
y
y
y
t

(2.13c)

The mechanistic equations for heat conduction and mass transfer can then be inserted to give:
T
k
y y

d ( c p ,d + c p , w ) T + d v h fg

h v
h w
Pv

"
"
+
h

m
=
d
w
v
w
fg

y y
t
y
y
t

(2.14a)
Rearranging:
T
k
y y

h
h
Pv d ( c p ,d + c p , w ) T + d v h fg

hw + v mv" + w mw"
d
+ h fg
=
y
y
t
t
y

(2.14b)
and expanding the energy change rate term on the right-hand side for T and assuming constant
specific heats for the wall material and liquid water:
T
k
y y

Pv
T
( c p,d + c p,w ) + d tv h fg + d t c p,wT d t cp,wT + hyv mv" + hyw mw"
+ h fg
= d

y
y
t

(2.15)
which results in:

h
h
T
Pv T
d ( c p ,d + c p , w ) + d v h fg + v mv" + w mw"
k
+ h fg
=
y y
y y t
t
y
y

(2.16)

An order of magnitude analysis can be used to show that the last three terms on the righthand side of this equation are negligible compared to the other terms. Even though the heat of
vaporization is large, the second term on the right-hand side is small because water vapor transfer
occurs over a long period of time leading to a very small rate of change in vapor content. The last
two terms are small because the liquid water and vapor fluxes, water vapor and liquid specific
heats, and temperature gradients within a building envelope are all small. As a result, Eqn. (2.16)
can be simplified as:
T
k
y y

Pv T
d ( c p ,d + c p ,w )
+ h fg
=
y y t

(2.17)

Eqn. (2.9) and (2.17) are the resulting governing equations that need to be solved in order to
evaluate moisture and energy transport in a wall. However, in order to facilitate the numerical
solution, the vapor diffusion and capillary transfer equations are decoupled to give the following:

24

Pv

y y

mv

= d v
=
y
t

Pl mw
= d w
K
=
y y y
t
= v + w
T
k
y y

Pv
T
+ h fg
= d ( c p ,d + c p ,w )
y y
t

(2.18a)
(2.18b)
(2.18c)
(2.18d)

These equations are non-linear (due to variable properties for , K, and k) and strongly coupled
and have time-varying boundary conditions. The boundary conditions and numerical solution are
described in later sections.

2.1.3. Comparison of Moisture Transfer Modeling Tools

There are several models that have been developed for modeling moisture transport in walls
that are reviewed briefly below.
1)

MATCH - Moisture and Temperature Calculations for Constructions of Hygroscopic

Materials, was developed at Technical University of Denmark (Peterson, 1990). This
is a one-dimensional transient heat and moisture transfer model that was the basis for
the model incorporated in MOIST 3.0;

2)

MOIST 3.0 was developed at NIST in the mid 1990s, and there has been no further
development since 1997. It utilizes a 1-dimensional model for transient heat and
moisture transfer that was briefly described in the previous section. A comprehensive
laboratory experiment was conducted by Zarr et al. (1995) to verify the accuracy of
the MOIST 3.0 in the hygroscopic regime. Good agreement between predictions and
measured results for the moisture content and surface heat flux was found. However,
a separate validation by Sipes et al. (2000) did not show as good results, possibly due
to two-dimensional transfer phenomena and the assumption of negligible sky
radiation in the model. MOIST was applied to assess the moisture performance of
building envelops under constant indoor conditions (Burch et al. 1995), variable
indoor relative humidity (Tsongas et al. 1995) and roofs (Burch et al. 1996 and

25
Tsongas et al. 1996) in terms of building material surface relative humidities and
internal moisture contents.
3)

HygIRC-1D is a one-dimensional hygrothermal computer model developed at the

Institute for Research in Construction in Canada for the aim of heat and moisture
analysis of building materials in Canadian climates. Compared to MOIST, moisture
migration due to air flow within the porous wall materials is included. A series of
research papers was published that covers modeling (Maref et al., 2004), parametric
studies for the moisture management (Kumaran et al., 2002 and 2003), and mid-scale
and large-scale measurements (Maref et al., 2002 and 2004). In addition to the onedimensional analysis, hygIRC was extended to a two-dimensional model, which is
called hygIRC-2D. HygIRC-2D can treat both air and water leakage for the building
materials, and considers gravity effects for problems related to the capillary transfer.
This model has been used for moisture performance analysis of building components
(Mukhopadhyaya et al., 2004).

4)

MEWS - Moisture management for Exterior Wall Systems, was developed at

National Research Council Canada in 1998 (NEWS Project reports Task 2~8, 2002).
Its simulation model is based on hygIRC ;

5)

HAM, for Heat, Air and Moisture transport, is a building simulation program that
provides one-dimensional calculations of heat, air and moisture transport processes in
a building enclosure. The program is based on the finite difference technique with
explicit forward differences in time. Analytical solutions for the coupling between
the computational cells for a given air flow through the construction are used.
Moisture is transferred by diffusion and convection in vapor phase. No liquid water
transport is considered. HAM was initiated in the European Union and sponsored by
IEA-Annex 24. It is has developed as part of HAM-Tools, an integrated simulation
tool for heat, air and moisture analyses for whole buildings (Sasic Kalagasidis, 2003).
The HAM-Tools wall block was compared with other models within the HAMSTAD
project (Heat Air and Moisture Standardization) and had reasonable agreement.

6)

WUFI (Wrme und Feuchte instationr German for Transient Heat and Moisture)
was developed at IBP (Institute for Building Physics, Germany) for calculation of the
transient hygrothermal behavior of multi-layer building components exposed to
natural climate conditions. WUFI considers the influence of wind-driven rain, solar

26
radiation, and night sky radiation on the hygrothermal performance of wall systems.
It has both one-dimensional and two-dimensional options for analysis. WUFI was
validated through experiments and excellent agreement in total water content
between experiments and predictions was found (Knzel, 1995). WUFI has been
used to calculate overall moisture transfer for the drying of single layer brick walls
after impregnation (the impregnation could be assumed to be the effect of long term
rain) and good agreement was found between the predicted values and the results of
field test (Knzel et al. 1996).
WUFI has an international cooperation with the Oak Ridge National Laboratory
(USA) to develop a hygrothermal design tool named WUFI-ORNL/IBP. A complete
data set (including temperature, relative humidity, wind speed and orientation,
driving rain, and solar radiation) for more then 50 North American locations is
included in the model. This hygrothermal design model can assess the response of
building envelope systems in terms of heat and moisture loads and can also provide a
very useful and fair method for evaluating and optimizing building envelope designs
(Karagiozis et al., 2001). This software has hundreds of users in North America, and
John Straube, from the University of Waterloo, stated that: WUFI-ORNL/IBP could
well put an end to some of the controversies that have persisted in the building design
community.
7)

UMIDUS is a tool to model coupled heat and moisture transfer within porous media
in Brazil. In order to analyze hygrothermal performance of building elements when
subjected to any kind of climate conditions, both diffusion and capillary regimes are
taken into account. UMIDUS has been built in an OOP language to be a fast and
precise easy-to-use software (Mendes et al., 1999).

8)

CHAMPS, Coupled Heat, Air, Moisture, and Pollutant Simulations, is being

designed at Syracuse University, USA (Zhang, 2005) to allow one-, two- and threedimensional modeling of coupled heat and moisture transport. CHAMPS will include
the ability to predict contaminant transfer through building materials in addition to
heat and moisture transfer,

There are also some other codes that have been developed for heat and moisture transfer
calculations. For detailed comparison of these codes, please refer to:

27
http://www.cmhc-schl.gc.ca/publications/en/rh-pr/tech/03-128-e.htm
After careful study of the publications related to the leading and active software in this field
(e.g., MOIST, HygIRC, HAM, WUFI and UMIDUS), it can be concluded that the basic
governing equations for moisture and energy transport in the building envelope are the same
under the assumptions of one-dimensional transfer and no internal air flow.

However, they do

differ in terms of the surface boundary conditions, material property correlations, and numerical
solution schemes.
In hygIRC, WUFI, UMIDUS and CHAMPS, capillary transfer for the liquid water is
included, so that they have the ability to handle a wetted surface condition caused by wind driven
rain. In addition, some validation of wetted surface moisture transfer in single layer materials has
been performed.

However, no one has performed validation with wetted surface boundary

conditions in combination with solar heating effects or considered multiple layers. In addition,
previous validation studies involving capillary transfer have not provided transient moisture
distribution comparisons. Although some research groups have tried to realize whole building
heat and moisture analysis, e.g., HAM-Tools and UMIDUS, they utilized some simplifying
assumptions such as neglecting liquid water capillary transfer,, not accurately considering
infiltration, or neglecting differences in moisture concentrations in different parts of the building.

2.1.4. Properties of Building Material

One of the most important aspects of moisture transport modeling is material properties. In
this project, material property correlations were inherited from MOIST 3.0. MOIST incorporates
property correlations for a number of common building materials. Walls are normally subjected
to both thermal and moisture gradients so that an accurate heat transfer determination requires a
simultaneous calculation of both sensible and latent effects. The transfer of moisture through
common building materials, such as wood, concrete and brick, depends on the complex
morphotopological characteristics of the pores in these materials. The thermal conductivity (k) is
a function of temperature and moisture content and is calculated as:

k = kd + CT (T Tref ) + C

(2.19)

28
where kd is the thermal conductivity of the building material under dry conditions and CT and C
are temperature and moisture modification coefficients, respectively. The effect of moisture on
porous building material properties is quite large (Mendes et al. 2003) and neglecting it can lead
to large errors in conduction heat transfer.
The water vapor permeability () is a function of relative humidity ( ) within the cavity of
the porous materials:

= B1 + B2 exp( B3 )

(2.20)

where B1, B2 and B3 are coefficients of the water vapor permeability function based on building
material property testing curve fits.

The hydraulic conductivity (K) is related to liquid diffusivity (D) by the relation:

K =

d D
pl /

(2.21)

with the liquid diffusivity D specified as:

D = C1 exp(C2 )

(2.22)

where C1 and C2 are coefficients for the liquid diffusivity function.

A sorption isotherm correlation provides a link between the equilibrium moisture content ()
and the relative humidity of the air ( ) within the cavities of the porous materials. Figure 2.3
provides an example adsorption isotherm, which is correlated using the following functional
relation:

A1
(1 + A2 )(1 A3 )

(2.23)

where A1, A2 and A3 are coefficients for the sorption isothermal equilibrium function (Tsongas
etc., 1996).

29

Figure 2.3 The schematic for sorption isotherm.

It should be noted that the three transfer coefficients (k, and K) are functions of the state (T
and ).

Therefore, the moisture and energy transport equations are non-linear and tightly

coupled.

2.1.5. Boundary Conditions

Heat transfer boundary conditions

The moisture transport is a strong function of the temperature distribution, which is coupled
to heat transfer boundary conditions at the indoor and outdoor surfaces. The boundary condition
for the indoor surface results from an energy balance under some simplifying assumptions:

(h

r ,i

+ hc ,i ) (Ti T ) + asol H in + along H eqp = k

T
y

(2.24)

where hr,i is a linearize indoor radiation coefficient, hc,i is indoor convective heat transfer
coefficient, Ti is indoor temperature, sol is the absorptance of solar radiation for the wall inner
surface, Hin is the solar radiation through windows (treated as a source that is evenly distributed
to inner surfaces) long is the absorptance for long-wave radiation, and Heqp is indoor long-wave
radiation caused by equipment, etc. (assumed to be evenly distributed to all interior surfaces).
For the exterior surface, the boundary condition is:

30

hc ,o (To T ) + hr ,o (Tsky T ) + asol H sol = k

T
y

(2.25)

where hr,o is a linearized outdoor radiation coefficient, hc,o is indoor convective heat transfer
coefficient, To is ambient air temperature, Tsky is the effective sky temperature, sol is the
absorptance for solar radiation, and Hsol is the incident solar radiation.

Water vapor transfer boundary conditions

For the indoor and outdoor boundary conditions, the moisture transfer through an air film
and finish layer (e.g., paint) is equated to the diffusion transfer into the solid material surface
according to:

Pv
(2.26)
y
The effective mass transfer coefficient Me is the combination of the surface conductance (Mp) and
M e ( Pv ,a Pv ) =

conductance of air film (Mf) associated with convective mass transfer coefficient defined by:

Me =

1
1
1
+
Mf Mp

(2.27)

1
1
+
hm M p

where hm, the convective mass transfer coefficient for air in contact with vertical services, is
determined using the Lewis relation (Burch, 19951):

Wetted surface condition on the vertical wall / capillary transfer B.C.

It is assumed that there is no liquid flux at the indoor boundary surface. This is also the case
for the exterior boundary, when there is no wind-driven rain. However, rain can wet the surface
and create a saturated condition. Two conditions are possible: 1) when the incident rain is
insufficient to saturate the surface and all of the available liquid at the surface is transferred into
the wall and 2) the incident rain exceeds the capillary transfer rate limit at the surface condition
and the excess is drained from the surface. These two cases are handled with the following logic:

Pl
y

when

m "w = m "rain

when

m "w = K

Pl

mrain
y
P

K l mrain
y
K

(2.28)

31

2.2. Numerical Scheme and Flow Chart of the Program

In order to illustrate the numerical scheme employed, the energy equation will be discussed
first.

Consider the three adjacent elements within a wall depicted in Figure 2.4.

This

nomenclature is used in applying central differencing in space and backward differencing in time
to the energy Eqn. (2.18). For time step i and node N, the finite difference equation is:

N-1

yN-1

yN

Latent
heat

N+1

yN+1

Figure 2.4 Depiction of wall nodes for numerical solution.

TNi 1 TNi
TNi +1 TNi
T i TNi 1
+
+ h fg mv , N = N
d ( Cd + Ni 1Cw ) yN
yN 1 yN
yN yN +1

t
+
+
2k Ni 11 2k Ni 1 2k Ni 1 2k Ni +11

(2.29)

where N stands for the node number, i stands for the time step number, and

mv, N =

Pv iN11 Pv iN1
P i 1 P i 1
+ v N +1 v N
yN 1 yN
yN yN +1
+ i 1
+
i 1
2 N 1 2 N
2 Ni 1 2 Ni +11

(2.30)

is the summation of the vapor flux for the previous time step (i-1) with Pv for the water vapor
pressure.
The following symbols are defined to simplify the equations for solution at each time step i:

EH N = d ( Cd + Ni 1Cw ) yN / t

(2.31a)

32

CTN =

1
yN yN +1
+
2k Ni 1 2k Ni +11

and

ATN =

1
yN 1 yN
+
2k Ni 11 2k Ni 1

= CTN 1
(2.31b)

This allows Eqn. (2.30) to be written as:

ATN (TNi 1 TNi ) + CTN (TNi +1 TNi ) + h fg mv , N = EH N (TNi TNi 1 )

and:

ATN TNi 1 ( ATN + CTN + EH N ) TNi + CTN TNi +1 = EH N TNi 1 h fg mNi 1

(2.32a)
(2.32b)

Defining:

BTN = ( ATN + CTN + EH N )

and : DTN = EH N TNi 1 h fg mNi 1

(2.33a)
(2.33b)

then Eqn. (2.32a) can be expressed as:

ATN TNi 1 + BTN TNi + CTN TNi +1 = DTN

(2.34)

Similar equations result for the nodes that are at the exterior and interior surfaces of the wall.
Only the exterior wall surface boundary condition will be presented as an example. A schematic
showing the exterior boundary condition is given in Figure 2.5. The exterior surface is in contact
with the ambient environment where the boundary is affected by the ambient temperature (To),
water vapor pressure (PPo), and solar radiation (Hsol). Long-wave radiation caused by the sky is
also included in the simulation tool. For the exterior node, the finite difference equation for
energy transfer is:

33

N-1

Hsol, Tsky
Latent
heat

To
PPo
yN-1

yN

Figure 2.5 Depiction of wall nodes for the exterior boundary.

T i TNi
TNi 1 TNi
T i TNi
T i TNi 1
+ o
+ sky
+ h fg mv , N + asol H sol = N
y N d ( Cd + Ni 1Cw )
1
1
y N 1 y N
y N
y N

t
+
+
+
2k Ni 11 2k Ni 1 2k Ni 1 hc ,o 2k Ni 1 hr ,o

(2.35a)
where

mv ,N =

Pv iN11 Pv iN1
PP i 1 Pv iN1
+ o
1
y N 1 y N
y N
+ i 1
+
i 1
i 1
2 N 1 2 N
2N
hm ,o

(2.35b)

Defining:

BND =

yN
1
+
i 1
hc ,o
2k N

and

ATN =

1
= CTN 1
yN 1 yN
+
2k Ni 11 2k Ni 1

(2.36)

This allows Eqn. (2.35a) to be written as:

ATN (TNi 1 TNi ) + BND(Toi TNi ) + h fg mv , N + asol H sol = EH N (TNi TNi 1 )

and:

(2.37a)

ATN TNi 1 ( ATN + BND + EH N ) TNi = BND Toi EH N TNi 1 h fg mNi 1 sol H sol
(2.37b)

Defining:

BTN = ( ATN + CTN + EH N )

and : DTN = BND Toi EH N TNi 1 h fg mNi 1 sol H sol

(2.38a)
(2.38b)

34
then, if CTN=0, Eqn. (2.37a) can be expressed as:

ATN TNi 1 + BTN TNi + CTN TNi +1 = DTN

(2.39)

Application of this formulation approach to both surface and the interior nodes leads to a
system of equations that are linear in the node temperatures for time step i. The resulting
equations can be written in matrix form, with a coefficient matrix that is tri-diagonal with the AT,
BT, CT, and DT coefficients updated for each node at each time step. The tri-diagonal matrix can
be inverted efficiently to determine each nodes temperature.

The use of an explicit time

integration method allows the solution of the energy equation at each time step to be decoupled
from the solution of the moisture transfer equations.
Moisture transfer equations for water vapor diffusion and liquid water transfer are treated
separately and solved in a manner that is analogous to the energy equation. For vapor diffusion
within the wall, the finite difference equation is:

Pv iN11 Pv iN1
P i 1 P i 1
i v iN1
+ v N +1 v N = d v N
yN
yN 1 yN
yN yN +1
t
+
+
2 Ni 11 2 Ni 1 2 Ni 1 2 Ni +11

(2.40)

At the exterior boundary node

Pv iN11 Pv iN1
PP i 1 Pv iN1
i v iN1
+ o
= d v N
yN
yN 1 yN
yN
1

t
+
+
2 Ni 11 2 Ni 1 2 Ni 1 hm,o

(2.41)

A similar boundary condition is realized at the interior surface. The resulting set of finitedifferenced equations for vapor diffusion is linear in terms of vapor pressure at time step i, and
can be expressed as:

AM N pv iN 1 + BM N Pv iN + CM N Pv iN +1 = DM N

(2.42)

Capillary transfer is only enabled if the local moisture content is close to a saturated
condition. The criteria used to enable capillary transfer is the equilibrium wall moisture content
associated with an air relative humidity of 97% ( =0.97) as determined using the sorption
equilibrium isotherm. If the moisture content is greater than this value, then it is assumed that a
continuous liquid water bridge has been generated within the cavities of porous building materials

35
and capillary transfer will be considered. When capillary transfer is enabled, then the finite
difference form of the transport equation for a node within the wall is:

PwiN11 PwNi 1
P i 1 P i 1
i wiN1
+ wN +1 wN = d w N
yN
yN 1 yN
yN
yN +1
t
+
+
2 K Ni 11 2 K Ni 1 2 K Ni 1 2 K Ni +11

(2.43)

For the nodes that do not meet the capillary criteria, a very small (1E-20) value was assigned to
the hydraulic conductivity K, so that a whole matrix system can be built. If rain is occurring at the
exterior surface, then the boundary condition of Eqn. (2.28) is applied in finite-difference form.
The boundary condition at the interior surface or at the exterior surface when no rain occurs is
capillary adiabatic, which means no direct liquid water transfer to the ambient air.

The resulting set of finite-differenced equations for capillary transfer is linear in terms of
capillary pressure at time step i, and can be expressed as:

ACN pl iN 1 + BCN Pl iN + CCN Pl iN +1 = DCN

(2.44)

The finite-difference moisture transfer equations are decoupled from the energy transfer
equation by employing explicit time integration.

Figure 2.6 shows a flow diagram for

implementation of the numerical scheme. In this flow diagram, TOLD is the backup for the node
temperature for the previous time step, IT is the time step number, and NN is the node number.
The energy, vapor, and liquid transport equations are solved sequentially at each time step using
properties and some states from the previous time step. Then, subroutine NEWM determines the
overall moisture flux (both vapor and liquid water) and overall node moisture content. The
original MOIST 3.0 code was written in FORTRAN.

The model has been enhanced and

rewritten in C++ language for the current work.

Limiting Case Evaluations

In order to check proper implementation of the computer code, some limiting case results
were compared with results from other models for heat transfer only.

The heat transfer

evaluations were performed by switching off the water vapor diffusion and capillary transfer and
utilizing a constant thermal conductivity (independent of temperature and moisture content).

36

Back up each nodes temperature in TOLD

Call COEFT(IT) Tri-diagonal matrix AT,BT,CT,DT
Call TRIDAG(1,NN,AT,BT,CT,DT,T) T
Call COEFM(IT,TOLD) Tri-diagonal matrix AM,BM,CM,DM
Call TRIDAG(1,NN,AM,BM,CM,DM,Pv) Pv -vapor pressure
Consider Capillary?

Y
Call COEFCAP Tri-diagonal matrix AC,BC,CC,DC
Call TRIDAG(1,NN,AC,BC,CC,DC,Pl) Pl -capillary pressure
Call NEWM(IT,NN) compute vapor and cap. flux update moisture content
Next time step: IT = IT+1

Figure 2.6 Flow diagram for envelope heat and moisture transfer model.

I: ambient temperature step change

A step change in ambient temperature was considered for the following parameters:
Material: 8 inches brick, single layer, 17 nodes
Boundary condition:

ho =2.008 BTU/ft2-F, hi=0.4579 BTU/ft2-F

Ti = 50F constant
To = 100F @ t=0

Initial condition:

T = 50F

Time step:

5 min

The results from the modified heat and moisture transfer model are compared in Figure 2.7
with results from FEHT, a commercially available finite-element program for heat transfer (FChart Software, 2006). The transient node temperatures for the two approaches are nearly
identical for all times and positions. In addition, heat fluxes are the same for both models when
allowed to reach steady state after a long period of time.

37

Temperature (F)

90

0:00 MOIST

85

0:00 FEHT

80

1:00 MOIST
1:00 FEHT

Time increase

75

1:10 MOIST
1:10 FEHT

70

2:00 MOIST

65

2:00 FEHT
4:00 MOIST

60

4:00 FEHT

55

10:00 MOIST

50

10:00 FEHT

Node #

45
1

13

17

Figure 2.7 Comparison for the node temperature with result from FEHT.

II: Ambient air temperature sinusoidal change

The predictions from the model were also compared with a Laplace Transform solution for a
sinusoidal ambient temperature change at the exterior surface for the following parameters:
Material: 8 inches brick, single layer, 17 nodes
Boundary condition:

ho =2.008 BTU/ft2-F, hi=0.4579 BTU/ft2-F

Ti = 70F constant
To = 70 +32 sin(t) F. is 24hr based

Initial condition:

T = 70F

Time step:

15 min run infinite long

A theoretical solution for steady-periodic behavior of the inner surface node temperature is
available in terms of time lag and decrement factor from a Laplace Transform.

The numerical

model was run for a long time and results for the interior surface node (Node #1) are very close to
results from the Laplace transform solution, as shown in Figure 2.8.

38
The primary purpose in performing these comparisons was to check implementation. The
combined heat and moisture transport predictions were also compared with the original MOIST
3.0 predictions. Furthermore, the model for moisture transport by vapor diffusion has been
previously experimentally validated. Some limited validation of the model for liquid transport
has also been performed and additional more extreme validations will be performed for this
thesis.
105

Tout
Tsurf_Laplace
Tsurf_MOIST

95

Temperature (F)

85
75
650.00

12.00

24.00

36.00

48.00

55
45
35

Time (Hr)

Figure 2.8 Comparison for the inner surface temperature with result from Laplace Transform.

2.3. Example of Moisture Build up Caused by Solar Heating after Rain

Some homes in cold and mixed climates experience mold problems when employing waterabsorbent wood based siding or brick veneer systems. Rainwater is absorbed into the siding wood
during and after a rainstorm and then water vapor is transported from the rain-wetted wood
through the exterior sheathing into the exterior wall cavities by vapor diffusion. This water vapor
condenses on the cavity side of the interior polyethylene vapor barrier. When sufficient moisture
condenses on the surface of the polyethylene vapor barrier, it runs down the surface of the
polyethylene under the action of gravity where it accumulates on the bottom plate of the wall. If
sufficient water accumulates at the bottom plate, mold growth will occur ultimately followed by
wood decay.
Consider a wood-based building material exposed to ambient conditions. When the wood
gets wet after a rainstorm and then is warmed by the sun, the moisture within the wood is at a

39
higher concentration than its surroundings.. Some of the moisture will move out of the wood into
the ambient air because the ambient air has a low water vapor pressure.

However, not all the

vapor will escape from the exterior building material to ambient. Some vapor diffuses inward
because the inner part of the building envelope can be at a lower temperature and lower moisture
content leading to a lower water vapor pressure. Therefore, moisture will accumulate, even
though the inner part of the building envelope is not wetted by rain directly (capillary transfer
oriented). This moisture accumulation can be accelerated by solar heating of an exterior surface
that has been wetted by rain.
It is reasonable to regard the surface porous building material as a moisture "capacitor" that
is "charged" with moisture during a rainstorm and that subsequently "discharges" in the direction
of a temperature gradient and concentration gradient. Moisture flow by vapor diffusion occurs
from warm temperature to cold temperature and from high vapor concentration to low
concentration. When the wall cavity is colder than the surface porous building materials, and drier
than them, water vapor will migrate into the cavity by the process of vapor diffusion. If the zone
is air-conditioned, and the thermostat setting point is low, this can be even worse.
A demonstration case study was performed to show the effects of solar-driven vapor
diffusion following a rain event. First of all, the wall simulation was carried out for 4800 hours
with constant boundary conditions in order to erase the effects of initial wall conditions. This
was followed by 8 hours of heavy rain (beginning at hour 4801 and ending at hour 4808), which
was then followed by solar radiation for 10 hours. For comparison purposes, a reference case with
8 hours of rain without solar heating was considered. The following summarizes the simulation
parameters and conditions:
Step change for rain shower followed by solar radiation / no solar radiation;
Simulation Options:
Variable properties with T and moisture content: k. cp. .
Including moisture transfer/Capillary effects;
Including latent heat;
Time step: 30 minutes, run first 4800 hr to eliminate initial error;
Material selected: 2 inches sugar pine wood, single layer, 16 nodes (node 1 facing interior);
Surface condition: bare wood, no paint
Boundary condition: ho =2.008 Btu/ft2-F, hi =0.4579 Btu/ft2-F

40
Indoor condition: 20C (68F) 50% RH.
Ambient condition: 30C (96F) 80% RH except raining hours
Rain:

8 hr rain (3mm/hr, 4801- 4808 hr, 99% RH for ambient air during rain);

Solar Radiation:

10 hr Solar radiation (600W, =0.7) right after rain ends (4809-4818 hr)

if solar radiation is applicable;

Initial condition:

25C (77F), G=0.005(kg/kg);

It should be noted that the wood is bare without the protection of siding materials or surface
paint. For a real building assembly, either siding or oil paint would be employed to isolate the
wood material from rain exposure.
Figure 2.9 depicts the temperature within each node of the wall for a period including the
rain and solar heating. After 8 hours of rain with no solar heating, the wall temperatures were
relatively constant with no sharp changes. Only the exterior nodes had a slight temperature drop
due to the latent heat caused by evaporation. However, when there was 10 hours of solar
radiation heating after the rain, sharp temperature increases were found throughout the wall,
especially at the exterior node where the solar heating occurs.
The time variation in moisture content is shown in Figure 2.10 for the cases with and
without solar radiation. For both cases, the moisture content of the exterior surface node (node
16) increased sharply during the rain. However, the transport of moisture into the wall occurred
only for a relatively shallow depth during the time when rain occurred. Immediately after the rain
stopped, the exterior node moisture content dropped sharply, while the case with radiation is
much steeper. For the case without solar radiant heating, this drop was primarily due to
evaporation of liquid and the moisture content stabilized relatively quickly after the end of the
rain fall. For the case with solar radiation, the surface moisture content dropped to a much lower
value because the higher surface temperature increased evaporation and drove moisture inward to
the wall and outward to the ambient air at the same time. During the period of solar radiation, a
moisture wave moved through the wall and ultimately increased the moisture at the inner surface.
For the case without solar radiation, the indoor side moisture content increase was almost
negligible. In this case, most of the moisture captured during the rain shower just evaporated to
the ambient air, since there was not enough driving force to promote inward moisture migration.

41

90

Temperature(F)

85

80

indoor

75

node #01

70

Node 01
Node 08
Node 13
d

65
4800

4812

4824

outdoor

node #16

Node 04
Node 10
Node 15
Time(hr)

4836

4848

a. without solar radiation

130
Node 01
Node 08
Node 13
Node 16

120

Node 04
Node 10
Node 15

Temperature(F)

110
100
90

indoor

outdoor

80

node #01

70

node #16

Time(hr)

60
4800

4812

4824

4836

4848

b. followed 10 hours solar radiation heating

Figure 2.9 Comparison of wall temperatures after rain shower with/without solar heating.

42

0.60

Moisture Content (lbm/lbm)

0.50

Node 01

Node 04

Node 08

Node 10

Node 13

Node 15

Node 16

0.40

outdoor

indoor

0.30
0.20

node #01

node #16

0.10
Time(hr)

0.00
4800

4812

4824

4836

4848

a. without solar radiation

0.600
Node 01
Node 08
Node 13
Node 16

Moisture Content (lbm/lbm)

0.500

Node 04
Node 10
Node 15

0.400

indoor

0.300
0.200

node #01

outdoor

node #16

0.100
Time(hr)
0.000
4800

4812

4824

4836

4848

b. followed 10 hours solar radiation heating

Figure 2.10 Comparison of wall moisture contents after rain shower with/without solar heating.

2.4. Summary
In this chapter, the basic modeling approach for one-dimensional transient heat and moisture
transfer in building materials (including vapor diffusion and capillary transfer) was developed,
including a wetted surface boundary condition to consider rain effects. A numerical scheme was
presented to solve the model and sample results were presented that demonstrate the effect of
solar heating following a rain event.

43

CHAPTER 3. EXPERIMENTAL VALIDATION OF EXTERIOR WALL MODEL

Researchers at NIST have developed a salt-in-water solution method (Richards et al., 1992)
and cup method (Burch et al., 1992 and 1995) to test building material moisture properties.
Furthermore, a comprehensive laboratory study to verify the accuracy of MOIST for 12 different
wall specimens was done (Zarr et al., 1995). The heat flux predicted by MOIST was within 10%
of the values measured for steady state and within 15% of the values for transient operation for a
series of diurnal ambient temperature cycles. The moisture content predicted by MOIST was
within 1% of the measured values. However, the experiments were only designed to consider
vapor diffusion effects and did not consider wetted surface effects that would include capillary
liquid transfer.
NRCC conducted mid-scale laboratory measurements for the drying process of oriented
strand board (OSB) (Maref etc., 2002) and large-scale laboratory measurements for wood-based
wall assemblies (Maref etc., 2004) under controlled laboratory environment conditions. They
used an automated weighing system and moisture pin-pairs to benchmark 1D-hygIRC. The
overall agreement between experimental and simulated results was very good in terms of the
shape of the drying curve. In these tests, capillary transfer was considered. However, only the
bulk moisture content for the specimen was tested and no information for the moisture content
and moisture content gradient within the specimen was determined. In addition, surface heating
effects were not considered.
WUFI was validated for transient behavior of a stone faade (Knzel, 1995) and the drying
of bricks (Knzel, 1996). However, WUFI is commercial software and little detailed information
can be obtained. There has been no third party validation of WUFI.
In this chapter, experiments are presented that were designed to validate capillary liquid
water transfer, and at the same time, to develop a better understanding of the nature and
significance of solar-driven inward vapor diffusion. In order to focus on capillary transfer,

44
experiments were conducted on a one-layer building material which has large moisture content
(e.g., wood) when thoroughly soaked. After starting with a fully wetted initial condition, the
wood sample was subjected to varying boundary conditions (air flow velocity, radiation, etc).

3.1. Methodology for the Experimental Validation

An automated weighting and moisture sensor system was developed to conduct drying
experiments on wood based products.

The automated weighting system was used to monitor

weight changes so that these data could be used to benchmark the surface vapor flux and the
overall moisture content of a specimen. The application of a weight scale to track the weight
change is particularly necessary in the first stages of drying when the material moisture content is
very high and beyond the upper limit of the moisture sensors that were embedded within the
wood and used to characterize moisture distributions within the wood.
Specimens of sugar-pine wood board panel having nominal dimensions of 50.8cm by
45.7cm (20 X 18 inches) with 3.8cm (1.5 inch) thickness were tested. This size of sample was
readily available and was thought to be sufficiently large to ensure that the overall heat and
moisture transfer could be viewed as 1-dimensional with the edges sealed and insulated to
minimize edge moisture and heat transport. The system and experiments were designed to collect
data that could be used to compare results with those derived from the MOIST 1-dimensional
wall model under wetted conditions with the effects of radiant heating. Data were obtained
continuously over time to give overall and local moisture contents of the wood-based specimens.
The tests were performed by first soaking samples in liquid water to achieve a saturated
condition. The samples were then dried by exposing them to a dry environment with radiant
heating. The radiant heating was switched between the two sides of the specimen in a manner
that would create large moisture gradients and transfers within the sample. This was meant to
provide a rich data set for validation of the model.
Figure 3.1 shows a schematic for this experiment depicting the test specimen, weight
monitoring system, radiant heating, air flow, and data acquisition. A shelf was made to mount the
weight scale. The wood specimen was put under the weight scale, 4 steel wires were used to hang
the wood specimen so as to 1) keep the wood specimen in balance without swinging, and 2)
transfer the weight of the specimen to the beam bar or beam board above the weight scale.

45
Moisture content sensors were installed within the wood specimen. In order to eliminate the
weight effect of the moisture sensor cables, very thin and lightweight cables were fastened to the
weight scale, so that the weight of the cable would not change during the experiments after they
were installed. The radiant lamps were used to heat the surface of the wood and promote
moisture transmission into the wood and to the air. Air flow over the wood specimen was
controlled using a variable-speed fan. Details of the equipment utilized in the experiments are
provided in the next section.

Test Room

Weight Scale

Moisture pin cable

Shelf

MC
RH
T

R
V

Wood sample

Radiation Intensity
Control
Fan Speed Control

Data Logger

Ki-Mo-Trol
Plus system
Weight Scale
RS-232

Chamber Control

Figure 3.1 Experimental set up.

3.2. Equipment and Calibrations

Moisture content measurement and sensor selection

There are many approaches, either through direct or indirect ways, that have been developed
to test the moisture content within building materials or soil over the past few decades, such as

46
the Electrical Resistance Method (Gawandea et al., 2003), microwaves (Kaariainen et al., 2001),
Neutron probe (Yuen et al., 2000), and laser radiation (Klemma et al., 2002 I, II, III). Healy (2003)
published a summary of techniques for measuring moisture levels in building envelopes. A
comparison of the measurement technologies for moisture content for building materials, based
the summary of Healy and other researchers, is listed in Table 3.1 in terms of advantages and
disadvantages.
The Electrical Resistance Method was chosen for measuring moisture content in this project
because of its low cost, high sensitivity and the ability to test point moisture content. This
method is described in ASTM D 4444 (ASTM 1992). The Electrical Resistance Method is based
on the principle that a mediums moisture content can be determined from the value of the
electrical resistance measured between electrodes (pin pairs) inserted into the target medium. Dry
wood, for instance, has a very high electrical resistance. It is reported that at 7% moisture content
(MC), the resistance for red oak is about 15,000 megohms, and for hard maple 72,000 megohms.
While at 30% MC the resistance for red oak is about 0.50 megohms and for hard maple it is about
0.60 megohms. However, for very low (< 6%) or high (> 40-60%) moisture contents, the
electrical resistance has large variability and small sensitivity to moisture and therefore is not a
reliable method in this range. The resistance is also dependent on temperature, so temperature
corrections need to be included. In addition, the resistance will depend on the wood species and
the direction of grains (Simpson and Tenwolde, 1999) and the variability between sensors can be
quite large before calibration corrections are applied (Zarr et al., 1995).

Therefore, careful

calibrations are needed.

In order to realize acceptable accuracy, calibration was processed for the pin-pair based on
the selected wood as described later. In general, when the moisture content within the wood is
low (e.g., 6~8% for the equilibrium moisture content with common indoor air conditions), the
moisture pin-pair is very sensitive and accurate. However, when the moisture content is high
(e.g., >30%,), the reading from the moisture pin-pair has relatively large differences compared to
the true moisture content.

47
Table 3.1 Comparison of the measurement technologies for moisture content.
Measurement Technique

Advantages

Disadvantages

Gravimetric Plugs

1. Direct means of measuring moisture

content
2. Accurate and repeatable

1. Not automated, but labor intensive

2. Destructive measurement
3. Lack of continuity in wall

Thermal Capacitance

1. Automated measurement is
possible;
2. Sensors are simple and inexpensive
3. Simple installation

1. Questionable accuracy
2. Sensors exhibit sensitivity to
disturbances during installation;
3. Nonlinear relationship between moisture
and thermal conductivity
4. Results affected by changes in electrical
conductivity and temperature;
1. Destructive
2. Electrical interference
3. Sensors suffer from hysteresis at low
moisture contents;
4. Results affected by changes in electrical
conductivity and temperature;
5. Sensor to sensor variability;
1. Results affected by changes in electrical
conductivity;
2. Local heterogeneity affects the results;
3. Sensor to sensor variability;
4. Susceptibility to electrical noise

Probe

Electrical
Resistance Probe

Dielectric Properties Time

Domain Reffectometry
(TDR)

Infrared

Microwave

Neutron Probe

Nuclear Scattering

Humidity Sensing

1. Automated measurement is
possible;
2. Significantly sensitivity
3. Sensor installation is easy;
4. Low cost;

1.
2.
3.
4.
5.
1.
2.
3.
4.
1.
2.
3.

Automated measurement is possible

Good sensitivity
Results are reproducible;
Simple Electronics
Non-Destructive
Automated measurement is possible
Non-contact measurement
Accurate
Insensitive to electrical noise
Automated measurement is possible
Accuracy
Non-destructive

1. Accurate
2. Direct detection of water
3. Can differentiate between different
states of water
1. Direct detection of water

1. Existing technology
2. Small sensor
3. Indicator for mold growth potential

1.
2.
3.
4.
1.
2.

Surface measurement only

Aging of surface will change reflectance
Expensive
Challenge of multiplexing
Scattering may happen;
Relatively low accuracy when the
moisture levels are very high or low;
3. Distributed measurement
1. Automation is not possible;
2. High cost
3. Large in size
1.
4.
5.
6.
1.
2.
3.
4.

Safety issues
High cost
Resolution
Packaging for in-situ sensor
Drift
Uncertainty in sorption isotherm
Instability at high RH
Accuracysmall errors in RH lead to
large errors in moisture content

48
The Kil-Mo-Trol Plus moisture content measuring system made by Delmhorst was
selected to take moisture content measurements within wood specimens in the experiments.
Related pin-pairs, cable and matched wood drilling equipment, a data logger, etc. were also
purchased. Figure 3.2 shows photos of the moisture pin-pair and the Kil-Mo-Trol Plus system.

Figure 3.2 Delmhorst moisture content pin pairs and Kil-Mo-Trol Plus system.

Test specimen:

Sugar-pine wood board was chosen as the specimen for the experiments. The material
properties of sugar-pine and the geometry of the specimen are listed in Table 3.2.

Table 3.2 Material properties and geometry of the test specimen: Sugar-pine.
sat
(kg/kg)

dry
(kg/m3)

mdry
(kg)

msat
(kg)

kdry
(W/m-C)

Dimension
L x W x H (cm)

Exposed Area
(m2)

Volume
(m3)

81.0%

365.1

3.221

5.820

0.0865

50.8x45.7x3.9

0.232

0.0088

It should be noted here that:

1.

Sugar-pine was selected because its properties are well known and available within the
MOIST properties database and because it has a relatively large capacity for storing moisture
which is helpful in reducing experimental error;

2.

The dimensions were selected so that the length and width are larger than 10~15 times the
depth so as to minimize two-dimensional effects.

49
Sealant was applied to the edges of the test specimens so as to limit lateral transport of liquid
and water vapor through boundaries. In selecting the edge sealant for the moisture transfer
experiments, the following issues were considered (Svennberg and Segerholm, 2006): flexibility,
hydro-phobation, thermal durability, vapor permeability, hygro-scopicity, adhesion and
workability. Considering the availability of material, asphalt with aluminum foil was selected.
Asphalt is an adhesive that could fasten the foil to the specimen. Asphalt itself is a widely used
water sealant on roofs. It has low vapor permeability, low hygro-scopicity, and good thermal
durability for the experimental conditions that were encountered. Aluminum foils prevent vapor
from escaping, given that they could be attached securely around the edge outside of the asphalt.
In addition, aluminum has high reflection for radiation, which helps to enhance the insulation
effect with respect to the radiant heating. For the experiments, about 3mm asphalt was evenly
pasted on the edge of wood specimen. Then, a single layer of aluminum foil was attached on the
surface. The edge sealing treatment was done prior to pre-conditioning (soaking) of the wood
specimen.
For any depth within the wood, the moisture content is determined using measurements from
two moisture sensor pins embedded to that depth. For the single layer sugar-pine board (1.5
inches for thickness) depicted in Figure 3.3, three insulated moisture sensor pin pairs were
embedded within the wood at the centerline and inch below the surface on each face. Figure
3.3 and Table 3.3 summarize the locations for each pin. This arrangement was designed to test
both shell and core moisture contents. Pre-drilled holes for the pin pairs were completed before
the single layer sugar-pine panel was placed in a water bath for preconditioning.

1
2,4

Pass above weight

scale (cable weight)
To Kil-Mo-Trol Plus

15 cm

15 cm

Figure 3.3 Plan for the location of moisture pin pairs.

50
Table 3.3 The position of the moisture pin-pairs for the 1-layer specimen.
Pin-pair
#

Testing
Position

inch from
Right hand
side

inch from
Right hand
side

inch from
Right hand
side

inch from
Left hand side

inch from
Left hand side

inch from
Left hand side

In order to avoid poor pin pair contact for the moisture sensors caused by wood shape
changes during drying, a special lumber grain orientation called quarter-saw was selected from
the wood specimen candidates. Quarter-sawed lumber has relatively uniform shrinkage without
cracking, as shown in Figure 3.4 (Simpson and Tenwolde, 1999). The pre-drilled holes were
sized for a forced fit and the pin-pairs were installed using a hammer to ensure a tight fit.

Quarter-saw

Figure 3.4 Illustration of the effect of shrinkage and grain orientation on shape.

Data acquisition system

Three kinds of data acquisition systems were utilized in the experiments:

1.

HP 75000 (for temperature and relative humidity sensor) with HP VEE dataflow
programming software environment for real time data collection;

2.

Delmhorst KMT-1100 moisture content measurement system with real time data
collection using an RS232 port based on its instrument communication protocol;

3.

Weight scale with real time data collection through an RS232 port based on its
instrument communication protocol.

51

Weighing system

The weighing system was designed to monitor the moisture transfer rate at the surfaces due
to evaporation. This is a direct way to validate the moisture transfer model in terms of surface
moisture flux, especially for periods of the test when the moisture pin pairs do not give very
accurate measurement (high or low moisture content). The capacity of the weighing system was
selected to be higher than the maximum weight of the soaked wood specimen with edge sealant,
pin-pairs, cables, etc. A digital weight scale with an accuracy of 0.1 gram and sufficient
capacity was selected and installed.

Radiant heating lamps

Five 250W radiant heating lamps (Sylvania Br40 infrared heating lamp) were utilized to
simulate surface heating effects in the laboratory experiments. In order to promote uniform
radiation on the wood specimen, the surfaces of the bulbs were ground and polished with sand
paper to produce a frosted surface. In addition, reflective surfaces were installed on the top, sides,
and bottom of the space between the light source and test specimen.
Radiation intensity measurements were taken using a pyranometer (EPLab Model PSP).
The PSP (Precision Spectral Pyranometer) is rated as a first class radiometer by the World
Meteorological Organization. PSP has low temperature dependence (1% over a temperature
range of -20 to 40C), fast response time (1 second), very good linearity (0.5% from 0 to
2800W/m2). Even more, PSP is capable of measuring radiation from a source with a spectrum
length between 0.285 to 2.800m, which covers the near infrared radiation (0.70~1.30m)
emitted by the heating lamps. Pictures of the radiant heating lamps, shields, and pyranometer are
shown in Figure 3.5.
In order to evaluate the uniformity of the radiation at the wood surface, the surface was
evenly divided into 9 zones (3 grids horizontally and 3 grids vertically) and the radiation intensity
was measured using the pyranometer at the center of each zone Table 3.4 gives radiation intensity
in W/m2 obtained by converting V readings from the pyranometer for each zone (1 W/m2
radiation is equivalent to 7.01V). The average radiation on the specimen surface was about 640
W/m2 and all of the nine zones were within 5% of this average. This degree of uniformity was

52
thought to be adequate for the purposes of these tests. During the drying tests, the pyranometer
was fixed above the wood specimen in order to track any decay of the heating lamps output.
However, during the test period, there was no noticeable decay in the radiation levels (the heating
lamps have 5000+ hours life rating) and the radiation intensity could be viewed as constant
throughout the experiments.

Figure 3.5 Heating lamps, reflective shields and pyranometer.

Table 3.4 Radiation Measurements at Different Locations on the Wood Specimen (W/m2).
Vertical \ Horizontal
I
II
III

3
604.9
596.3
676.2

2
659.5
671.9
601.3

1
647.6
641.9
672.6

Air flow system

In order to generate a forced upstream air flow across the specimen surface, a blower with a
motor having a variable frequency drive was installed. To enhance uniformity of the air flow, the
duct outlet was filled with honeycomb cores. Pictures of the blower, duct, and honeycomb core
outlet are displayed in Fig. 3.6. A handheld vane wheel type anemometer with a testing range
from 0.4-20 m/s and accuracy of 3% of full scale was used to measure the air velocity.

53

Figure 3.6 Blower (driven by variable frequency drive) and honey-comb-core outlet.

Relative Humidity Sensors

In addition to radiant heating and air dry bulb temperature, air relative humidity is a very
important boundary condition for the validation of the hygrothermal model. Two Vaisala
relative humidity sensors and Vaisala transmitters HMP233 were chosen to measure the air
relative humidity. A chilled mirror dewpoint meter was employed to check calibration of the RH
sensors and transmitters. With known air dry bulb temperature, dewpoint temperature, and
atmosphere pressure, the air relative humidity can be found in tables or from software such as
EES. For the cases randomly picked for the conditions shown in Table 3.5, the readings from the
Vaisala RH sensor were within 1% of the results determined from measured dry bulb
temperature, dewpoint temperature, and atmosphere pressure. This is consistent with the 1%
RH uncertainty specified by the manufacturer. Therefore, the RH sensors and transmitters were
believed to provide accurate results for this experiment (relative humidity range from 5~40%)
without compensation.
Thermocouples

Laboratory made Copper-Constantan thermocouples type T were utilized to measure the

ambient air dry bulb temperature and specimen surface temperature on both sides. For ambient air
temperature, two thermocouples were applied to take average readings. To avoid direct radiation,
both thermocouples were mounted on the un-heated side, close to the honeycomb core outlet
together with the relative humidity sensors, as shown in Figure 3.7. The thermocouples for the
specimen surface temperature were pasted on the specimen using transparent tape with pores.
Tiny small aluminum foil was used between and tape and sensor as a radiation shield. In addition,

54
a thermocouple was applied to measure the air temperature around the heating lamps in case of
fire so that the power supply could be cut off for safety purposes. The thermocouples connected
to the data acquisition system were calibrated in an OMEGA calibration box at 0C and 100C.
Calibration data for the thermocouples are shown in Table 3.6. The thermocouple readings were
within 0.1C of the calibration device and it was not necessary to correct the default calibration
characteristics employed within the data acquisition system.

Air temperature
thermocouple 1
RH sensor 1
Air temperature
thermocouple 2
RH sensor 2

Figure 3.7 Relative humidity sensor and the ambient air temperature thermocouples.

Table 3.5 Calibration for the relative humidity sensor and transmitter.
Case #
1
2

Tair
(C)
22.1
23.2

Tdew
(C)
-4.0
1.3

Patm
(Pa)
100258
100525

RH (T,P,Td)
From EES
16.4%
23.6%

RH sensor
15.5%
22.8%

Moisture Content Sensors

The Delmhorst moisture content pin-pairs and Kil-Mo-Trol Plus system were calibrated for
the wood specimen through comparisons of moisture content outputs to baseline results
determined using a weight scale.

Each baseline result was determined by computing the

difference in weight of the specimen containing a uniform moisture content and the base dry
weight for the specimen. The dry weights of the wood specimen were determined using an oven-

55
drying technique based on the ASTM D4442-07 standard by the Wood Research Laboratory in
the Department of Forestry and Natural Resources, Purdue University.
Moisture contents for the calibration process were determined through drying of a wood
specimen that was initially soaked to a saturated condition. The soaked specimen was then placed
in room air to force surface evaporation to occur. In order to obtain each calibration point, the
specimen was placed in a small sealed tank for a long time and allowed to reach internal
equilibrium. Several points were obtained by alternating the specimen between the room air and
sealed tank. Desorption was chosen for the calibration process rather than absorption because it
was more straightforward than designing an experiment that would build up moisture with a
uniform distribution. The direct moisture content by weight change was considered to be the true
value for calibration of the moisture content sensors.
Comparisons between the moisture sensor readings and the actual moisture content are given
in Figure 3.8. There are relatively large differences between the two results. When the moisture
content is low, the reading from the moisture pin pair agrees with the actual moisture content
fairly well. However, when the moisture content is high, the reading is much less than the actual
moisture content, especially in the range where the wood is nearly saturated (50% compared to
80%). Based on these results, a linear correlation was generated and used to correct moisture
content readings from the moisture sensor system.

Table 3.6 Calibration for the thermal couples.

Sensor #

Calibration condition 1:
0 C

Calibration condition 2:
100C

Application

0.06 C

99.96 C

Fire Alarm

-0.04 C

100.01 C

Room air temperature

-0.06 C

99.92 C

Room air temperature

-0.05 C

99.94 C

Specimen Surface A temperature

0.03 C

99.95 C

Specimen Surface A temperature

-0.04 C

100.05 C

Specimen Surface B temperature

-0.03 C

99.94 C

Specimen Surface B temperature

56

90%
Test
Ideal
Linear (Test)

Real Moisture Content (kg/kg)

80%
70%
60%
50%

y = 1.6642x - 0.0767

40%
30%
20%
10%
0%
0%

10%

20%

30%

40%

50%

60%

Moisture Content Reading

Figure 3.8 Calibration for the moisture content pin pairs.

3.3. Experimental Procedure

Determination of hc, hr and

It is important to know the wood specimen radiation absorptance, convective heat transfer
coefficient and radiant heat transfer coefficient for the simulation. A methodology was developed
to estimate hc, hr and based on steady-state testing. Figure 3.9 depicts the experimental setup
and important measurements for this determination. The tests for determining these parameters
were performed on a one-inch thick wood specimen that was allowed to come into equilibrium
with a typical dry environment (normal dry condition for moisture content of about 6%). The
wood specimen was allowed to reach steady state when subjected to radiant heating of 640W/m2
on one side and the room ambient condition on the other side. Steady state data were collected
for a range of air velocities associated with laminar air flow. Then the data were used in
combination with a steady state energy balance to estimate hc, hr and using the following steps:
1: Estimate hmix, cold, the overall heat transfer coefficient for the un-heated side.
In order to determine convection and radiation coefficients, a combined coefficient was first
determined for the cold side of the wood from a steady-state energy balance using measured hot
and cold-side surface temperatures along with air temperature on the un-heated side.

57

I=640w/m2
Ts,hot

Ts,cold

Tair, RHair
Vs,hot

Vs,cold

Figure 3.9 Schematic of setup and measurements for estimating hc, hr and .

Ts ,hot Ts ,cold
= hmix ,cold *(Ts ,cold Tair )
d

(3.1)

The thermal conductivity of sugar pine from the MOIST program was used (k=0.0865 W/mC) in Eqn. (3.1). For comparison, an ORNL report (Tenwolde et al., 1988) gives a value of k for
sugar pine of 0.090 W/m-C.
Table 3.7 gives measurements and calculations of hmix,cold determined for different air
velocities. The air velocity on the heated surface was higher than that on the unheated side. This
is due to the close proximity of the duct elbow to the outlet. The flow was very uniform along the
width of the wood sample but varied from front to back along the depth of the sample.
2: Estimate hr,cold, hc,cold, and then hc,hot
For laminar flow, the convective portion of the combined heat transfer coefficient should be
proportional to the square root of the air velocity. Furthermore, the radiation coefficient and the
proportionality constant for convective heat transfer can be determined graphically or using
regression with data covering a range of air velocity.

Figure 3.10 shows the combined heat

58
transfer coefficient as a function of the square root of the air velocity for the cold side of the
wood sample. The slope of this line is the proportionally constant for the convective heat transfer
whereas the projection of the line to the intercept is an estimate of the radiation heat transfer
coefficient. Using the test data, the radiation heat transfer coefficient for the unheated side was
estimated to be 8.72 W/m2-C.

Table 3.7 Air velocity, Temperature and Re number for computing hr/hc.
VFD
12 Hz
14 Hz
16 Hz
18 Hz
20 Hz
22 Hz

Vs,hot
(m/s)
1.6
2.4
3.4
4.2
5.2
5.7

Vs,cold
(m/s)
1.0
1.5
2.1
2.6
3.4
4.3

Tair
(C)
22.24
20.66
21.79
22.42
22.01
19.42

Ts,cold
(C)
25.19
23.23
23.87
24.28
23.63
20.96

Ts,hot
(C)
39.23
36.33
36.25
35.89
34.81
31.63

hmix,cold
(W/m2-C)
16.2
17.3
20.3
21.3
22.6
23.6

Re
number*
4.8E+04
7.2E+04
1.0E+05
1.3E+05
1.5E+05
1.7E+05

* Only the heated side Reynolds number is shown in this table, since Re <5*105 ( the critical Re for
external flow) and the unheated side should have a lower Re due to a smaller air velocity. Therefore, the
flow is laminar for both sides.
25.0

20.0
0.5

hmix,cold = 7.59 V

+ 8.72

hmix_cold

hmix (W/m -C)

Linear (hmix_cold)
15.0

10.0

hr,cold
5.0

0.0
0.00

0.50

1.00

0.5

1.50

2.00

2.50

Figure 3.10 Specify the overall and the radiant heat transfer coefficient for the unheated side.

59
Figure 3.11 shows a comparison between the hc correlation determined from test data for the
unheated side and an empirical correlation for external laminar flow from a heat transfer text
book (Fundamentals of heat and mass transfer, F.P. Incropera and D.P. Dewitt, 4th edition). The hc
correlation from the tests was within 20% of the textbook correlation.
The correlation for convective heat transfer determined for the cold side was used for the
heated side. However, the air velocity is different so that the hot and cold side convection
coefficients are related using the following equation:
0.5
hc , hot
Re hot 0.5 Vs ,hot
=
=
hc , cold Re cold 0.5 Vs0.5
,cold

(3.2)

18.0
16.0
0.5

hc,cold = 7.59V

14.0

R = 0.9785

hc_cold (W/m -C)

12.0
10.0
0.5

hc,cold = 5.70V

8.0

R =1

6.0
hc_cold_test
hc_cold_cor
Linear (hc_cold_cor)
Linear (hc_cold_test)

4.0
2.0
0.0
0.50

1.00

1.50

2.00

2.50

0.5

Figure 3.11 Comparison of hc from tests and hc from a textbook correlation.

3: Estimate hr,hot
The coefficient for long-wave radiation from the hot side, hr,hot, is significantly different than
the cold side value because the wood surface is heated and it views surfaces that are heated to
some extent by the radiant lamps. The radiation coefficient was estimated using measured wood
surface temperatures, typical values for long-wave emissivity, and assuming that the surrounding
surfaces are at the measured ambient temperature for the hot side of sample. Then, the radiation
coefficient was estimated as:

60
hr ,hot = (1 Fs ,l ) (Ts ,hot + Tair )(Ts ,hot 2 + Tair 2 )

(3.3)

where Fs,l is the effective view factor for the specimen surface to the surroundings excluding the
heating lamps, is the emissivity of the wood surface, is the StefanBoltzmann constant, Ts,hot
is the specimen average surface temperature, and Tair is the room air average temperature. For
the experimental setup, Fs,l was computed using the dimension and geometry of the light bulbs
and wood specimen, so that Fs,l=0.148. The emissivity for the wood was taken as 0.9, whereas
the average surface temperature and air temperature on the headed side were assumed to be 35 C
and 22C, respectively. Using these numbers in Eqn. (3.3), the hot side radiation coefficient (hr,hot)
was estimated to be 4.8 W/m2-C.
4: Estimate surface absorptance
An overall steady-state energy balance on the wood sample was used to estimate the solar
absorptance, , for known incident radiation, surface temperatures, air temperature, and
convective and radiant heat transfer coefficients on both sides:

I = hmix ,hot (Ts ,hot Tair ) + k

Ts ,hot Ts ,cold
d

(3.4)

An value average =0.48, with a variance of 0.02, was determined using Eqn. (3.4) and data
from Table 3.7.
Preconditioning the wood specimen

The general procedures for the preparation of the wood specimen could be described in the
following steps:
1.

Edge sealing treatment - 3mm asphalt with aluminum foil pasted outside;

2.

Pre-drill the pin-pair holes;

3.

Precondition the test specimen in a water tank for a period sufficient to reach a fullysoaked initial condition (20-day immersion in a water bath followed by 10-day moisture
redistribution, as described below);

4.

Mount the pin-pairs, install the test specimen in the test room and monitor specimen
surface temperature, moisture content inside and weight change;

61
The preconditioning method used by NRCC (Maref etc., 2002) was applied here and is
depicted in Figure 3.12. After pre-drilling holes for the moisture pin pairs, the specimens was
completely immersed in water to achieve the soaked conditions. Aluminum angles of 0.5-m
length were used to separate the stacked specimens from one another and from the bottom of the
immersion bath to insure that water was in good contact with all surfaces of the sugar pine
specimens. This helped hasten the process of saturating the samples. Since the fully immersed
specimens had a tendency to float, some weights (metal blocks) were placed on top to keep them
submerged. The goal of the preconditioning was to generate uniform saturation moisture content
within the specimen so that a known soaked initial condition could be specified as an initial
condition in the simulation. The wetted wood specimen was weighed every few days to identify
when an equilibrium moisture condition was reached. After 20 days, most of the water in the bath
was drained. A relatively small amount of water was left in the bath at the bottom so that a 100%
RH condition within the air was established when the tank was sealed using adhesive tape. The
specimen remained in the bath under saturated air conditions for another 10 days to allow
moisture to be re-distributed evenly within the wood board. Following this 10-day re-distribution
period, the wood specimen was removed and moisture content measurements were taken with
Delmhorst moisture pin-pairs. Weights of the specimen were also recorded. The moisture pin-pair
readings were primarily used to check that the sensors were mounted tightly within the predrilled
holes in the wood specimen. In this test, after 30 days pre-conditioning, the average moisture
content was 81.4% by weight, and the average readings at the specified locations were: 75.4% for
1/4 beneath surface A (surface that receives radiation first), 76.9% at the center line for the wood
specimen and 82.0% for 1/4 beneath surface B (surface that receives radiation at a later time).

Figure 3.12 Preconditioning to soak the specimen.

62

3.4. Uncertainty Analysis

In order to evaluate the quality of the experimental measurements, an uncertainty analysis
was performed. The uncertainty of a directly measured value, UXi, is determined from laboratory
calibration or otherwise provided by the instrument manufacturer. The uncertainty of a calculated
quantity, UR, can be determined by the equation:
1/ 2

2
n R

U R =
U Xi

i =1 Xi

(3.5)

This equation assumes a functional relationship between the calculated quantity, R, and the
measured values, Xi, of the form:

R = f ( X 1 , X 2 , X 3 ,....., X n )

(3.6)

The uncertainty analysis can be used to estimate propagation of measurement errors on

calculated qualities and can determine the impact of uncertainty associated with each measured
quantity. Table 3.8 summarizes the uncertainties of the directly measured values for the
experiments. It should be noticed that the wood specimens thermal conductivity is included in
the uncertainty analysis, but it is not directly measured. The wood specimens thermal
conductivity under dry conditions was adopted from the property database from MOIST 3.0.
Furthermore, the difference between the MOIST value and other widely referenced values was
taken as the material property uncertainty.

Table 3.8 Uncertainties of directly measured values/material properties.

Measurement

Uncertainty

Temperature (C)

0.1 C

Relative Humidity (-)

1% reading

Radiation intensity (W/m2)

36 W/m2

Air velocity (m/s)

3% of full scale

Moisture content (-)

6%

Weight (g)

1 gram (includes estimate of uncertainty associated

external effects such as air flow, vibration, etc: scale
accuracy is 0.1 gram)

*Wood thermal conductivity (W/m-C)

0.035 W/m-C (*estimated)

63
Table 3.9 to 3.12 give uncertainty analysis results for mixed heat transfer coefficient on the
unheated side (based on temperature measurements), convective heat transfer coefficients (based
on the correlation with air flow velocity), wood specimen surface absorptance and specimen
surface vapor flux. Overall, the uncertainty for hmix_cold is less than 7%, while the uncertainty for
convective heat transfer coefficients are as high as 24%.

hmix_cold is estimated based on

temperature which uses thermocouples that have low uncertainty compared to the magnitude of
measure value. On the other hand, the convective heat transfer coefficient is estimated based on
air velocity flow across the specimen, which has a relatively high uncertainty especially at low
velocities. The material surface absorptance uncertainty is less than 15%.
The uncertainty in the surface vapor flux was estimated numerically using the basic equation
for surface vapor flux:
m = KM surf ,m * ( Ps Pair )

(3.7)

where Ps is the surface nodes vapor pressure (Ps is a function of the surface nodes moisture
content and temperature), Pair is the ambient air partial vapor pressure (Pair is a function of air
temperature and relative humidity), and KMsurf,m is the vapor transfer conductance:
KM surf ,m =

1
+
hm

1
x / 2

(3.8)

where hm is the surface convective mass transfer coefficient, x is the mesh size, and is
permeability the material property of vapor transport coefficient ( is a function of material
moisture content and temperature).

When the specimen is wet, the uncertainty from the

convective mass transfer coefficient is the major error source and the absolute uncertainty for the
surface vapor flux is high. However, the magnitude of surface flux is also high and the relative
uncertainty is about 8%. In the later stages of the test, the wood specimen is dry and the absolute
vapor flux is low with a relative uncertainty of about 10%. In this later stage, the key uncertainty
source is the moisture content.

64
Table 3.9 Uncertainty analysis for hmix_cold.
Case

Tair
(C)

Ts,cold
(C)

Ts,hot
(C)

k
(w/m-C)

hmix,cold
(W/m2-C)

22.240.1

25.190.1

39.230.1

0.08650.0035

16.200.79

20.660.1

23.230.1

36.330.1

0.08650.0035

17.300.90

21.790.1

23.870.1

36.250.1

0.08650.0035

20.301.19

22.420.1

24.280.1

35.890.1

0.08650.0035

21.301.35

22.010.1

23.630.1

34.810.1

0.08650.0035

22.601.66

19.420.1

20.960.1

31.630.1

0.08650.0035

23.601.72

Table 3.10 Uncertainty analysis for convective heat transfer coefficient.

Case

Vs,hot
(m/s)

Vs,cold
(m/s)

hc,hot
(W/m2-C)

hc,cold
(W/m2-C)

1.60.6

1.00.6

9.602.29

7.591.80

2.40.6

1.50.6

11.761.86

9.301.47

3.40.6

2.10.6

14.001.57

11.001.23

4.20.6

2.60.6

15.551.41

12.241.11

5.20.6

3.40.6

17.301.23

14.001.00

5.70.6

4.30.6

18.121.10

15.740.95

Table 3.11 Uncertainty analysis for wood specimen surface absorptance.

Case

Tair
(C)

Ts,cold
(C)

Ts,hot
(C)

k
(w/m-C)

hmix,hot
(W/m2-C)

I
(W/m2)

22.240.1

25.190.1

39.230.1

0.08650.0035

14.401.80 64036

0.460.06

20.660.1

23.230.1

36.330.1

0.08650.0035

16.561.47 64036

0.480.05

21.790.1

23.870.1

36.250.1

0.08650.0035

18.801.23 64036

0.490.04

22.420.1

24.280.1

35.890.1

0.08650.0035

20.351.11 64036

0.490.04

22.010.1

23.630.1

34.810.1

0.08650.0035

22.111.00 64036

0.500.04

19.420.1

20.960.1

31.630.1

0.08650.0035

22.920.95

0.49 0.03

64036

65
Table 3.12 Uncertainty analysis for specimen surface vapor flux.

Tair
(C)
21.70.1

air
(%)
0.160.01

Tsurface
(C)
30.10.1

M.C.
(kg/kg)
0.810.06

hm
(g/inhg-hr-m2)
171.813.7

Vapor flux
(g/m2-hr)
194.715.7

22.80.1

0.170.01

24.60.1

0.330.04

42.73.4

33.22.7

22.80.1

0.170.01

34.20.1

0.240.02

16.81.3

23.52.0

19.10.1

0.160.01

32.10.1

0.160.01

13.40.9

4.00.4

Case

3.5. Model Validation

The experimental conditions for the transient test are summarized in Table 3.13. In the first
48 hours of the test, the VFD was set to 12Hz, and the radiant heating was projected on surface A
of the soaked specimen. In the subsequent 48 hours, the VFD was adjusted to 18Hz to give a
higher air velocity and thus higher convective heat transfer and mass transfer coefficients. During
this time, the radiant heating was still applied to surface A. Then, at the 96th hour, the wood
specimen was rotated 180 degrees so that the radiant heating was projected on surface B, and the
VFD was lowered to 16Hz. This condition remained until the end of the experiment.

Table 3.13 Experimental condition summary.

Material

Sugar pine, 1.5 inches in thickness

Designed radiant intensity

640W/m2 / 202.9 BTU/ft2-hr

Material initial condition

15.7C / 60F with moisture content 81% (Soaked)

Test time span

19 days, from 02/14/08~03/04/08

Sampling rate

5 minutes

Test chamber dry-bulb temperature

21~24C / 70~75F

Test chamber relative humidity

8~38%

Test period 1:

Radiation on Surface A
1~48hr,

VA= 1.6 m/s

VFD=12Hz

VB= 1.0 m/s

Test period 2:

Radiation on Surface A

48~96hr,

VA= 4.2 m/s

VFD=18Hz

VB= 2.6 m/s

Test period 3:

Radiation on Surface B

96~456hr,

VA= 2.1 m/s

VFD=16Hz

VB= 3.4 m/s

66
The test bench was located in the west wing of the Herrick Lab. The air temperature and
relative humidity were not precisely controlled but were recorded and then used in the simulation
for the purposes of model validation. For much of the test period, the room air was very dry,
because it was early spring. The air dry bulb temperature and relative humidify during the period
of the experiment are shown in Figure 3.13 and Figure 3.14, respectively.

24.50
24.00

Room Air Temperture (C)

23.50
23.00
22.50
22.00
21.50
Tair

21.00
20.50
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00

2/29/08 0:00

3/5/08 0:00

Time

Figure 3.13 Room air dry bulb temperature during the experiment.

40.0%
35.0%

Relative Humidity (-)

30.0%
25.0%
20.0%
15.0%
10.0%
Rhair

5.0%
0.0%
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00

2/29/08 0:00

3/5/08 0:00

Time

Figure 3.14 Room air relative humidity during the experiment.

67
Validation model details

Based on the laboratory experiment condition, the validation model details can be
summarized in Table 3.14.

Table 3.14 Validation model condition summary.

Material

Sugar pine, 1.5 inches in thickness

Initial condition

15.7C / 60F with moisture content 81% (Soaked)

Boundary condition: temperature

Recorded temperature file as shown in Fig. 3.13

Boundary condition: radiation

640W/m2 / 202.9 BTU/ft2-hr

Boundary condition: humidity

Recorded relative humidity file as shown in Fig. 3.14

Mesh size

21 nodes, 0.0714 inch for each grid

Time step

5 minutes

Test period 1:

Radiation on Surface A
1~48hr,

hc,A=1.691 BTU/ft2-F-hr, hr,A=0.845 BTU/ft2-F-hr

VFD=12Hz

hc,B=1.337 BTU/ft2-F-hr, hr,B=1.532 BTU/ft2-F-hr

Test period 2:

Radiation on Surface A

48~96hr,

hc,A=2.739 BTU/ft2-F-hr, hr,A=0.845 BTU/ft2-F-hr

VFD=18Hz

hc,B=2.155 BTU/ft2-F-hr, hr,B=1.532 BTU/ft2-F-hr

Test period 3:

Radiation on Surface B

96~456hr,

hc,A= 1.937BTU/ft2-F-hr, hr,A=1.532 BTU/ft2-F-hr

VFD=16Hz

hc,B=2.465 BTU/ft2-F-hr, hr,B=0.845 BTU/ft2-F-hr

Comparison of Surface Temperatures

Measurements of surface temperature were obtained for the experiments and used for model
validation. Figure 3.15 shows comparisons of predictions with measurements over the entire
experimental period, where TS_A is the measured surface temperature (measured on the center of
the wood specimen) on surface A, and Tsim_A is the predicted value for surface A. In general,
the predictions are in close agreement with the experiments. Both model and experiments show
a slow rise in the temperature of the surface at the beginning of the experiments as the surface
dried out due to evaporation and moisture transfer into the wood. Once the surfaces were no

68
longer saturated, then the surface temperatures were relatively constant until radiant heating was
switched to the opposite side. Both the model and experiments show a nearly step change in
temperature for both surfaces at this point.
40. 00

Temperature ( C)

35. 00

30. 00

25. 00

20. 00

15. 00

TS_A

TS_B

Tsim_A

Tsim_B

10. 00
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00

2/29/08 0:00

3/5/08 0:00

T ime

Figure 3.15 Measured and prediction surface temperatures for the wood specimen.

Comparison of Local Wood Moisture Content

As previously described, moisture content was measured at the center of the wood specimen
and inch beneath both surfaces of the specimen. Within the Delmhorst data acquisition system,
the species was set to sugar pine, the testing compensation temperature was set to 23C before
the test, and the sampling rate was chosen to be 5 minutes. Figure 3.16 shows moisture content

measurements (the moisture output from the Delmhorst system was adjusted by using the
calibration curve shown in Figure 3.8) and predictions for 1/4 inch beneath surface B (unheated
for 96 hours and then heated). Both the test and model show a similar decay in moisture content
over the length of the test period. The rate of decay tended to decrease from beginning to end of
the experiment as the wood dried out and the driving potential for moisture transport was
reduced. On Feb. 18th at 17:00, the radiation was switched to surface B. At this point, there was a
somewhat sharper decline in moisture content at this location, because the radiant heating may
increase the water vapor pressure on this side, which leads to higher surface vapor flux plus vapor
incursion to the center of the specimen from the outside surface. This effect was greater for the

69
experiment than for the simulation. The differences in experimental and predicted values were
largest at the beginning and end of the experiment. Larger errors at the beginning were expected
since the moisture pin pairs have less accuracy when the moisture content is high. The errors at
the end of the experiment could be due to imperfect contact for the moisture pin-pair with the
wood specimen which would result in higher resistance and lower moisture content readings.
However, since the errors grew larger over time a better explanation may be that there was
additional moisture transfer from the edges of the sample that are not accounted for in the
simulation. Although the edges were sealed, the sealing is imperfect and even relatively small
water/vapor leakage could accumulate over time to cause the differences.
0.90
0.80

Moisture Content (kg/kg)

0.70
0.60

Switch radiation

0.50
0.40
G_sim_1/4" beneath Surf. B
0.30

G_test_1/4" beneath Surf. B

0.20
0.10
0.00
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00
Time

2/29/08 0:00

3/5/08 0:00

Figure 3.16 Moisture content at inch beneath surface B of the specimen.

Comparisons of test and simulation results for the moisture content at the center line of the
wood specimen are shown in Figure 3.17. In general, the centerline results look less noisy than
the measurements closer to the surface because the effects of boundary condition variations are
damped. The overall decays in moisture content are similar between the experiments and model.
However, the model is less affected by switching the radiation to the other side. Also, the model
tends to overestimate the moisture content at end of the experiments. This may be due to the fact
that the model does not consider edge effects.

70

0.90
0.80

Moisture Content (lb/lb)

0.70

Switch radiation

0.60
G_sim_center
G_test_center

0.50
0.40
0.30
0.20
0.10
0.00
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00

2/29/08 0:00

3/5/08 0:00

Time

Figure 3.17 Moisture content at the center line of the specimen.

Figure 3.18 illustrates the moisture content change over time for inch below surface A
(switched between heated and unheated) for the test and simulation. In this case, the switching of
the radiation had a much more significant effect on the moisture content transient. After radiation
was switched to the other side, then the moisture content for this location increased for a period of
time. This is because the moisture had been driven towards the center of and other side of the
wood and the direction of transport within the wood was reversed when the radiation was
switched.

Both the model and experimental results show this trend, although the model

predictions are less dramatic.

The change in moisture transport direction is due to a change in vapor pressure gradient.
After the radiant heating is removed from surface A, the temperature near the surface drops
leading to a lower local vapor pressure. At the same time, the temperature and vapor pressure
within the wood near surface B both increase. Consequently, moisture moves from side B to side
A due to the positive pressure gradient. In addition, the low vapor pressure at surface A can lead
to moisture diffusion into the surface from the ambient because this wood surface has been
previously dried from solar heating effects. As a result of these effects, there is a moisture build
up within the wood near surface A. Such moisture build up is very common for building

71
envelopes which get wetted (or partially wetted) after wind driven rain and then heated and
cooled due to solar radiation and diurnal effects. Conditioning of the indoor space can contribute
to these effects by providing a boundary condition of low vapor pressure to enhance moisture
transmission from the exterior to interior for rain events followed by solar heating.
0.90
0.80

Moisture Content (kg/kg)

0.70

Switch radiation
G_sim_1/4" beneath Surf. A

0.60

G_test_1/4" beneath Surf. A

0.50
0.40
0.30
0.20
0.10
0.00
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00
Time

2/29/08 0:00

3/5/08 0:00

Figure 3.18 Moisture content at inch beneath surface A of the specimen.

Specimen Weight Change

Tracking the specimen weight change is the most direct way to evaluate the moisture flux
for this hygrothermal transfer test since a weight scale is more accurate compared to moisture pin
pair. The total weight of the wood specimen is proportional to its moisture content. Furthermore,
predictions of vapor flux are important for determining vapor incursion from the building
envelope to the room air. A very precise digital weight scale with 0.1 gram accuracy was used in
the test with a sampling rate of 5 minutes. Transient changes in weight were used to determine the
overall moisture content of the wood and the net vapor flux from both sides to the ambient. The
results from the experiment were compared with predictions determined using the simulation tool
as shown in Figure 3.19.
The model predicts the overall moisture content very accurately during the first part of the
drying process when the entire wood specimen is wet. This provides a good validation of the

72
capillary transfer model for the MOIST code. However, later in the test period, especially after 8
days, differences in model and experimental results increase. At the end of the experiment, the
difference is about 230 gram (test data: 3331.5 and predicted data: 3564.4) or 7% error in overall
wood moisture content. The over prediction of the moisture content at the end of the experiment
as compared with the weight measurements is consistent with the results presented for local
moisture contents determined using the pin-pair sensors. The most likely reason for the test data
being lower than the predicted data is the imperfect edge sealing.
Over time, vapor transport from the edges can cause significant loss of moisture from the
wood that is not accounted for within the model.

A simple leakage model was developed that

assumes that 10% of the edge area was not perfectly sealed. Figure 3.20 shows comparisons for
the specimen weight change from the test data with results from the simulation model that
includes the edge vapor leakage. In this case, the model and data are in very good agreement.
6000

Specimen Total Weight (gram)

5500

Switch radiation

5000
4500

Weight from Simulation

Weight from Test

4000
3500
3000
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00
Time

2/29/08 0:00

3/5/08 0:00

Figure 3.19 The specimen weight change for the drying process.

73

6000.0

Specimen Total Weight (gram)

5500.0

Switch radiation

5000.0
Weight from Simulation
(consider vapor leakage)
4500.0

Weight from Test

4000.0

3500.0

3000.0
2/14/08 0:00

2/19/08 0:00

2/24/08 0:00
Time

2/29/08 0:00

3/5/08 0:00

Figure 3.20 The specimen weight change for the drying process (with edge vapor leakage
included in the model).

3.6. Summary
In this chapter, an experimental setup and procedures were developed for validating the wall
model under conditions where capillary liquid water transfer is an important mechanism. The
experiments involved a drying process for one layer sugar-pine panels with a soaked initial
condition. An automated weighing system was used to trace the overall wood moisture content
and moisture pin-pairs measured the local moisture content within the test specimen. A blower
driven by a variable frequency driver with a honeycomb core outlet was utilized to produce
uniform flow conditions along each surface. Convective/radiant heat transfer coefficients and the
wood specimen surface absorptance were determined using a dry steady-state test. During the
transient drying test, radiant heating was projected on one surface then switched to the other so as
to develop a better understanding of the nature and significance of solar-driven inward vapor
diffusion.
There was relatively good agreement between moisture content predictions and
measurements from the moisture pin-pair sensors and the overall moisture content from the

74
weighing system. These results provide a good validation of the MOIST model for heat/vapor
/capillary transfer.
Some of the differences between moisture content predictions and measurement could be
due to the use of a one-dimensional model that neglects edge effects. For the small size sample
tested, vapor diffusion through the edges could be significant even though the edges were sealed.
These edge losses could explain the fact that the predictions of moisture content were higher than
measured values at the end of the tests as determined with both the pin-pair measurements and the
weighing system. Other experimental factors that could cause errors for the moisture content test
include:
1.

The specimen used to calibrate the instrument may have different properties than the
wood panel used for the transient test. Even when measuring moisture content in a
single wood species, different results can occur for different locations of the pin-pair
due to different grain orientations. Even more importantly, wood from the lower trunk
and upper trunk of the same tree could have significantly different properties.

2.

The properties of the wood used in the simulation could also differ from the actual
properties. Properties from the MOIST database were used for the simulations.

3.

The moisture sensor system uses a single temperature compensation. However, the
surface temperature is not constant which could cause some errors in these
measurements.

4.

The moisture pin-pairs are calibrated to read moisture content of material that is in
equilibrium with no moisture gradients. Although each pin within a pin-pair is located
at the same depth, there may have been moisture gradients that influenced the readings.

5.

It is possible that the moisture pins had poor contact with the wood for part of the tests.
Poor contact can be created from bending and shrinkage of the wood specimen. This
would result in a higher resistance reading and therefore a lower moisture content
reading.

75

CHAPTER 4. A SIMPLIFIED MODEL FOR GROUND-COUPLED HEAT TRANSFER IN

FLOOR SLABS

4.1. Introduction
Ground-coupled heat transfer through concrete floor slabs is typically a significant
component of the total load for heating or cooling in low-rise buildings like residential buildings,
etc. It was estimated that earth-contact heat loss accounted for about 10% of the annual heat loss
for average homes in the US during the early 1970s (Labs et al. 1988). Since that time, buildings
have been constructed with more insulation, better windows, and more air tight envelopes. As a
result, buildings are generally more energy efficient today and ground-coupled heat loss may
account for 30~50% of the total heat loss for a well-built house (Deru, 2003; Cleasson and
Hagentoft, 1991 and Labs et al. 1988). Furthermore, transients associated with floor slabs can be
very significant and important to consider in estimating both peak loads for sizing of equipment
and total energy requirements for economic analyses. Recently, there has been considerable
interest in utilizing energy storage within floor slabs and other building materials to achieve load
shifting and reduction in on-peak cooling requirements and costs (e.g., Braun, 2003 and Braun
and Zhong, 2005). The motivation for the work presented in this thesis was the need for a simple
model for predicting transient heat transfer in ground-coupled floor slabs that could be integrated
within an hourly simulation program.
There are well-established models for steady-state, ground-coupled heat transfer from
basements and slab floors. For instance, ASHRAE (2001) presents a steady-state model for heat
loss from ground-coupled floor slabs that are typically used for modeling in residential
applications. The model assumes that the predominate path for heat loss from the slab occurs
between the slab perimeter and the ambient. The floor heat transfer is characterized in terms of a
perimeter heat loss factor (Fp). ASHRAE gives typical values of Fp for a limited number of
foundation insulation configurations and climatic conditions. The method and heat loss factors
were developed from measurements conducted for specific foundation sizes and soil properties

76
under steady-state heat transfer conditions.

No general approach has been presented for

estimating perimeter heat loss factors from material and soil properties.

EN ISO 13370

(European Committee for Standardization, 1998) presents a similar method for estimating steadystate ground floor heat flux. Medved and Cerne (2002) used this general formulation and
developed simplified method to calculate parameters of the model for different sizes and shapes
of buildings with and without basements with a variety of properties.
Krarti et al. (1988) developed the ITPE (Interzone Temperature Profile Estimation) method
that combines analytical and numerical techniques to obtain two- and three-dimensional solutions
of the heat conduction equation for slab-on-grade floors and basements (Claridge et al. 1993).
The ITPE method provides a general solution for the temperature field within soil under steady
conditions for a slab-on-grade and slab-on-ground floor. The ITPE formalism has been applied to
develop a simplified method for seasonal foundation heat loss (Krarti and Choi, 1996), evaluate
the thermal bridge effects of floor foundation (Al-Anzi and Krarti, 1997), optimize insulation Rvalue by life-cycle cost analysis (Chuangchid and Krarti, 2000) and develop simplified tool for
the calculation of foundation heat gain (Chuangchid and Krarti, 2000).
One approach for modeling ground-couple floor slabs involves solution of two- or threedimensional transient conduction equations using finite-difference or finite-element methods.
The solution generally requires the numerical solution of a large number of coupled equations.
Because of the computation required, studies have been performed to develop fast computational
schemes or modeling simplifications for this problem.

For instance, Seem and Klein et al.

(1989) developed a method for calculating transfer functions in multi-dimensional conduction

heat transfer problems that is useful for modeling ground-coupled floor slabs. The approach
utilizes the following steps: 1) convert a finite-difference representation of the partial differential
equations to a state-space formation, 2) utilize the analytical solution to the state-space formation
with a ramp or step input to determine a transfer function representation, 3) utilize model-orderreduction techniques to obtain a simplified transfer function representation. The transfer function
representation is determined prior to or at the beginning of a simulation and is then used at each
timestep to calculate energy flows at the boundaries of the domain (e.g., at the floor-interior
interface).

For annual simulations, the transfer function representation requires much less

computation than that associated with numerical solutions to the partial differential equations at
each time step.

77
Adjali et al. (2000) performed numerical simulations for a ground-coupled floor slab and
compared predictions with measurements. Two- and three-dimensional conductive heat transfer
models based on a finite-volume method for the floor and soil were added to APACHE, a whole
building thermal simulation program widely used in Europe. In-situ measurements were made at
the Cardiff School of Engineering, University of Wales, over a period of 18 months.
Temperatures (internal and external air, at different depths in the slab and the soil down to 4 m
beneath the floor surface), heat fluxes through the concrete floor slab, the thermal conductivity of
the ground and the water table and soil moisture content were measured. The results showed that
the purely conductive earth-contact model accurately predicted the thermal behavior beneath the
building. A two-dimensional analysis was found to be adequate. Three-dimensional effects were
only important near the corners of the slab.
Even with simplifications, the computational requirements for solving two or threedimensional heat transfer within floor slabs and ground soil is very significant. Simulation tools
available for estimating cooling and heating loads in commercial and residential buildings
typically do not employ detailed multi-dimensional models for ground-coupled heat transfer. For
instance, TRNSYS (2000) considers one-dimensional transient conduction in the floor and
materials immediately beneath the floor. A portion of the soil can be considered with a specified
underfloor boundary condition. However, no guidance is given for the amount of soil to be
included and the temperature of underfloor boundary condition to employ.
Some simulation models (e.g., Energy10, 1995) consider two paths for heat transfer from a
ground-coupled slab floor: 1) steady-state heat loss from the perimeter to the ambient and 2) onedimensional transient conduction in the floor and materials immediately beneath the floor with an
adiabatic underfloor boundary condition. However, no literature could be found that validates this
modeling approach. Furthermore, its not obvious how to specify perimeter heat loss factors and
underfloor soil depths and the number and location of nodes for the simplified thermal circuit.
The current work starts with the general two-path modeling approach for transient slab heat
transfer and develops heuristics and correlations for general application of the method.

In

particular, results from detailed two-dimensional finite-element models for typical floor
constructions and soil properties were used to identify 1) locations for nodes within the slab and
soil, 2) correlations for soil depth as a function of soil properties associated with the underfloor

78
adiabatic boundary condition, and 3) correlations for perimeter heat loss factor as a function of
soil properties and edge insulation levels for different constructions.

4.2. Methodology

4.2.1. Case Study Descriptions

A number of ground-coupled floor slabs were considered in the model development and
evaluation study. The basic geometry and some of the parameters are shown in Figure 4.1. The
finite-element modeling was applied to a two-dimensional slice including the slab floor and
surrounding soil. The construction and dimensions are typical of a small commercial building.
Many of the results were determined using a slab length of 20 m. It was determined that the
spatially-averaged heat flux at the indoor floor surface is insensitive to slab length for values
greater than about 10 m. The dimensions of the soil (depth of 10 m and width of 50 m) were
chosen so that the bottom and side boundaries were nearly adiabatic corresponding to the far
field assumption (Adjali et al. 2000). Due to symmetry, only half of the floor and soil shown in
Figure 4.1 was modeled.

The internal zone boundary condition was convection to a zone

temperature that varies within a relatively narrow range. The boundary condition at the soil
surface considered both convection and radiation using an equivalent sol-air temperature, which
is an equivalent outdoor air temperature that accounts for the effects of incident solar radiation
and radiant energy exchange with the sky and other surrounding surfaces (ASHRAE, 2005).
Typical convective heat transfer coefficients and a solar surface absorptance for summer
conditions were specified for the upper surfaces of the floor and ground. Two widely used
constructions were selected for floor slabs from the Builders Foundation Handbook (Carmody et
al. 1991). Configuration A used a 15 cm (6 inch) heavy weight concrete slab laying directly on
ground soil. Configuration B had a 10 cm (4 inch) heavy weight concrete slab with 4 inches of
sand and pea gravel above the ground soil. A heavy weight concrete foundation (20 cm in width
and 1 m in depth) with various foundation insulation levels was considered.
For a given floor construction and set of boundary conditions, the primary factors that affect
slab heat flux are the level of insulation on the foundation and the soil properties. If insulation is
employed on the foundation, then typically about 1 (2.54 cm) to 2 (5.1 cm) of styrofoam is

79
attached to the outer foundation wall below ground. In this study, outer foundation insulation
was assumed to extend from the top surface of the ground to a 1 m depth. Base case results were
determined for insulation having a thermal resistance corresponding to 1 (2.54 cm) of styrofoam
insulation (R=0.95 K-m2/W or R=5.4 ft2-F-hr/Btu). A range of different insulation levels were
considered (R=1.425K-m2/W, R=0.95 K-m2/W, R= 0.475 K-m2/W and R=0 K-m2/W) in
developing correlations for perimeter heat loss factors.
Table 4.1 gives thermal properties of floor, underfloor materials, and soil used for base case
results. Soil properties can vary considerably according to soil type and moisture content. It is
extremely difficult to consider the time dependence of soil properties due to changes in moisture
content. Typically, average properties are assumed for a particular site. The base case soil
properties in Table 4.1 are representative of saturated soil of medium sand with fine gravel (Henk
et al. 2002). In addition, soil properties were varied over a wide range to develop correlations for
perimeter heat loss and soil depth for the simplified model.

Tz, Internal zone temperature

2

hi=8.48 W/m -K

10 m

Tsol-air, Sol-air temperature

a=0.7, ho=18.6 W/m2-K

10 m

15 m

10 m

Figure 4.1 Schematic of the slab-on-ground geometry.

For steady-state conduction, the only soil property of interest is the thermal conductivity.
For transient conduction problems, thermal diffusivity (ratio of thermal conductivity to the
product of the density and specific heat) is the relevant parameter. Thermal diffusivity, , is a
measure of the ability of the material to diffuse thermal energy through conduction. The higher
the thermal diffusivity the faster a material will respond to changes in the boundary conditions.
In addition to the base case, soil thermal conductivity was varied in the range of 0.6 to 3.5 W/mK. This corresponds to thermal diffusivities in the range of 2.99E-6 m2/s to 1.74E-6 m2/s.

80
Table 4.1 Base case thermal properties of floor and ground materials.
Material

(kg/m3)

k (w/m-K)

cp (J/kg-K)

(m2/s)

Heavy weight concrete

1.80

2245

837

9.58E-7

Pea Gravel

2.52

1800

1000

1.40E-6

Ground Soil

1.73

1844

1090

8.61E-7

For the simulation cases considered in this study, the zone and ambient sol-air temperature
boundary conditions were varied using 24-hour periodic functions of the following form.
T(t) =E 0 + E1sin(

12

t+1 ) + E 2sin(

t+ 2 ) + E3sin(

(4.1)

t+ 3 )

For the ambient sol-air temperatures, different sets of coefficients were determined through
regression using hourly weather data from different locations and different times of year. Hourly
ambient temperatures and horizontal radiation were used along with the solar absorptance and
convection heat transfer coefficient given in Figure 4.1 to compute sol-air temperatures for the
different cases. For instance, Figure 4.2a shows a sample 24-hour variation in sol-air temperature
for Arcata, CA (California climate zone 1) on August 1st based on TMY2 data. The zone
temperature variation was chosen based upon typical results associated with conventional night
setup control in small commercial buildings (Braun and Zhong, 2005), as shown in Figure 4.2b.
In this study, Figure 4.2a and 4.2b were used as the base case to build the correlations only.
Ambient sol-air temperatures and zone temperatures for small commercial buildings from other
locations in different seasons under various controls were used in validating the model.

60.0
50.0

Tz for Arcata, CA
Aug-1st

24.0
23.5

40.0

Tz (C)

Tsol-air (C)

24.5

Tsol-air for Arcata, CA

Aug-1st

30.0

23.0

20.0

22.5

10.0

22.0

0.0

21.5

0:00

6:00

12:00

Time

18:00

a. Ambient sol-air temperature

0:00

0:00

6:00

12:00

Time

18:00

b. Zone temperature

Figure 4.2 Example variations in ambient sol-air and zone temperatures.

0:00

81
4.2.2. Finite-Element Model
A commercially available finite-element program, FEHT (F-Chart Software, 2006) was
employed to solve the two-dimensional transient heat conduction problem for the geometry
shown in Figure 4.1. The slab floor slab and surrounding soil, footer and concrete wall structure
were modeled using approximately 3,000 nodes. Figure 4.3 depicts the non-uniform mesh
arrangement employed. The element sizes were smaller in areas of larger temperature gradients,
such as near the top surface and at the edge of the floor slab. Adiabatic boundary conditions were
specified at 15m from the edge of the slab and at a depth of 10 m from the top surface. The
internal zone (slab and wall) and ground surfaces were subjected to convective heat transfer using
the boundary conditions described in the previous section.
For the transient modeling, the initial temperatures of the floor slab and soil directly beneath
the slab were chosen as the average of the zone air temperatures over the 24-hour period. The
initial temperatures of the surrounding soil were determined using the model of Baggs given in
Eqn. (4.2) (Popiel et al. 2001), which was developed for naturally occurring ground temperatures
away from building structures.
T(y,t)=Tm - 1.07k v As Exp(-0.00031552 y -0.5 ) cos[

2
( t-t 0 +0.018335 y -0.5 )]
365

(4.2)

where the vegetation shade factor, kv, was taken to be 1.0 for bare ground like a parking lot or
side walk outside a retail store.
A one-hour time step was used for all simulations. The model was run for many identical
days in a row until the solution approached a steady-periodic condition. This required simulation
lengths of up to 12000 hours.

82

10 meters

15 meters
2

convective, hi=8.48 W/m -K

adiabatic

10 meters

adiabatic

convective, ho=18.60 W/m2-K

Adiabatic

Figure 4.3 Schematic of the two-dimensional finite-element modeling mesh.

4.2.3. Simplified Ground-Coupled Floor Modeling

For estimating zone cooling and heating loads, the quantity of interest for modeling a
ground-coupled floor is the heat transfer rate at the top surface of the floor. The simplified
modeling approach considers two primary one-dimensional paths for heat transfer from a groundcoupled floor slab: 1) one-dimensional steady-state heat transfer from the perimeter of the slab to
the ambient and 2) one-dimensional transient heat transfer between the slab interior surface and a
portion of the soil beneath the slab. At any time t, the total heat transfer rate from the zone air to
the floor is determined as

Q& t = Fp Pm (Tz Tsol air ) + Af q"f ,t

(4.3)

where Fp is the slab perimeter heat loss factor, P is the slab perimeter, Tz and Tsol air are 24-hour
average zone and sol-air temperatures (updated continuously), Af is the total floor surface area,
and q "f is the heat flux (heat transfer rate per unit area) at the floor surface for a one-dimensional
transient analysis. The one-dimensional heat flux is essentially the heat flux that occurs at the
centerline of the slab (i.e., the left-most boundary of Figure 4.3).
The one-dimensional transient analysis uses a simple thermal circuit employing 3 nodes with
an adiabatic boundary condition at a specified depth within the soil underneath the slab as

83
depicted in Figure 4.4. The number and location of for the nodes was determined through
analysis of and comparisons with finite-element modeling results. The resulting model can be
represented with three continuous, linear, time-invariant differential equations as:
C1

dT1 Tz T1 T2 T1
=
+
d
R1
R2

C2

dT2 T1 T2 T3 T2
=
+
d
R2
R3

C3

dT3 T2 T3 Tg T3
=
+
d
R3
R4

(4.4)

Tz(t) = Room Temperature

R1=Rin+ R6HWC/2

T1

C1= C6HWC

T2

C2= Csoil_1

T3

C3= Csoil_2

R2= R6HWC/2+Rsoil_1
R3= Rsoil_2
R4= Rsoil_3 +
Adiabatic Bottom

a. Floor Configuration A
Tz(t) = Room Temperature

R1=Rin+ R4HWC/2

T1

C1= C4HWC

T2

C2= Cgravel

T3

C3= Csoil

R2= R4HWC/2+Rgravel/2

R3= Rgravel/2+Rsoil/2

R4= Rsoil/2+
Adiabatic Bottom

b. Floor Configuration B
Figure 4.4 Simplified 3-node models for a one-dimensional ground-coupled floor.

84
The three differential equations can be put in matrix form as:
dT1 1 1
dt R1C1 R2C1

1
dT2 =
dt
R2C2

dT

dt

1
R2C1

1
1

R2C2 R3C2
1
R3C3

T1 R1C1
1
T2 + 0
R3C2
T3 0
1

R3C3
0

0 0
T
z
0 0 0

0 0 Tg

(4.5)

Eqn. (4.5) is a state-space representation for the simplified one-dimensional conduction heat
transfer problem. It is solved analytically and converted to a transfer function using the method
of Seem et al. (1989). The transfer function solution gives the current one-dimensional heat flux
at the floor surface in terms of current and past hourly values of the zone temperature and past
hourly values of the heat flux.

q f ,t " = s0Tz ,t + s1Tz ,t 1 + s2Tz ,t 2 + s3Tz ,t 3 + e1q f ,t 1" + e2 q f ,t 2" + e3q f ,t 3"

(4.6)

The s and e transfer function coefficients are determined from the solution to the state-space
problem with linear changes in zone temperature over each hour as outlined by Seem et al.
(1989). The transfer function representation is determined prior to a simulation and is then used
at each hourly time step to calculate the floor heat transfer.
The initial temperatures of the floor slab and soil directly beneath the slab were chosen as
the average of the zone air temperatures over the 24-hour period. A one-hour time step was used
for all cases and the model was run to a steady-periodic condition for each day type considered.

4.2.4. Process for Correlating Perimeter Heat Loss Factor and Soil Depth
In order to apply the simple model described in the last section, it is necessary to specify a
perimeter heat loss factor and depth of soil under the floor. The perimeter heat loss depends on
the floor construction, amount of foundation insulation, and soil properties. The appropriate
depth of soil to include in the analysis depends on the floor materials and soil properties. Results
from finite-element modeling were used to develop correlations for these two parameters. The
following steps were followed to develop and evaluate the correlations and simplified model.

85
1. The finite-element model was used to generate one-dimensional heat fluxes at the
adiabatic center line of the floor slab (i.e., the left-most boundary in Figure 4.3) for floor
configurations A and B with different soil properties and the baseline ambient and zone
temperature variations.
2. A simplified one-dimensional transient model was also developed for each configuration
and soil type in terms of the unknown soil depth.
3. For a given floor configuration and soil properties, errors in predictions of hourly
centerline floor surface heat fluxes were minimized with respect to soil depth for the
simplified model. The problem involved minimizing the following cost function:
J=

24

(q
t =1

"
f ,t

"
qFE
,t )

(4.7)

with respect to the soil depth, D, where q"f ,t was determined from the simplified model
"
and qFE
,t was from the finite-element model.

4. Correlations of optimized soil depth D in terms of soil conductivity, k, were developed

for estimating centerline floor heat transfer.
5. The two-dimensional finite-element model was used to generate transient heat transfer
results over the entire domain depicted in Figure 4.3 for floor configurations A and B
with different soil properties and the baseline ambient and zone temperature variations.
The total heat transfer rates at the top floor surface were determined for each hour from
the two-dimensional results.
6. For a given floor configuration, foundation insulation, and soil properties, errors in
predictions of hourly floor surface heat transfer rates were minimized with respect to the
perimeter heat loss factor for the simplified model. The problem involved minimizing
the following cost function:
J=

24

( Q&
t =1

f ,t

Q& FE ,t

(4.8)

with respect to perimeter heat loss factor, Fp.

7. Correlations of optimized perimeter heat loss factor in terms of soil conductivity and
foundation insulation R value were developed for the different floor configurations.
8. Predictions from the resulting simplified model were compared with two-dimensional
finite-element modeling results for a range of ambient conditions (winter and summer in
different locations), indoor temperature variations, soil conductivities, and foundation
insulation levels.

86

4.3. Results and Discussion

4.3.1. Floor Configuration A: 4heavy-weight concrete on soil

Figure 4.5 gives an example of the effect of soil depth on the performance of the simplified
one-dimensional heat flux model (Figure 4.4a) for a soil thermal conductivity of 1.73 w/m-K.
The optimal soil depth to use in this simplified model is about 45 cm for this case. Figure 4.6
compares the centerline heat flux determined with the one-dimensional finite-element model and
simplified model with the optimized soil depth.

Mean Square Error (W/m 2)

0.30
0.28
0.26
0.24
0.22
0.20
0.18
10

20

30

40
Depth (cm)

50

60

70

Figure 4.5 Example optimization of soil depth for centerline heat flux.

12.0
k*=1, R*=0

Heat Flux (W/m 2)

8.0

Finite-Element
Simp lified (CTF)

4.0
0.0
0:00
-4.0

6:00

12:00

18:00

0:00
Time

-8.0
-12.0

Figure 4.6 Comparison of centerline heat flux for finite-element and simplified models
(Configuration A, base case).

87
Optimized soil depths for a range of soil thermal conductivities were determined and used to
develop a correlation. For floors without gravel below the slab (Configuration A), the resulting
correlation is:
D* = -0.0489k* 2 + 0.5692k* + 0.4709

(4.9)

where k* is the ratio of the thermal conductivity to the baseline thermal conductivity of 1.73 w/mK and D* is the ratio of the soil depth to the soil depth associated with the baseline conductivity
(45 cm). Figure 4.7 shows comparisons between the curve-fit and optimized soil depths.
1.60
1.40

D*

1.20
1.00
0.80

Finite-Element
Curve Fit

0.60
0.40
0.00

0.50

1.00

1.50

2.00

k*

Figure 4.7 Curve fit for optimized soil depth for centerline heat flux (Configuration A).

The overall simplified model utilizes the simplified centerline heat flux model with the soil
depth correlation and an edge heat loss model as given in Eqn. (4.3). Errors between model
predictions from Eqn. (4.3) and two-dimensional finite-element results were minimized to obtain
perimeter heat loss factors for various edge insulation levels and soil conductivities. Figure 4.8
gives example hourly comparisons for average floor surface heat flux obtained using an
optimized soil depth and perimeter heat loss factor for baseline parameters and conditions. The
simplified model can accurately characterize the transient heat transfer at the floor surface.

88
10.0

k*=1, R*=0

Average Heat Flux (W/m 2)

8.0

Finite-Element
Simplified (CTF+Fp)

6.0
4.0
2.0
0.0
-2.00:00

6:00

12:00

18:00

-4.0

0:00
Time

-6.0
-8.0

-10.0

Figure 4.8 Comparison of average floor surface heat flow flux for finite-element and simplified
models (Configuration A, base case).

The perimeter heat loss factor depends strongly on the edge insulation and soil conductivity
and weakly on the dimensions of the zone. Figure 4.9 shows the effect of dimensionless soil
conductivity and half-floor length (distance from centerline to edge) on the optimized perimeter
heat loss factor for baseline parameters and conditions, except with no foundation insulation.
Without edge insulation, the impact of floor length is negligible for overall floor lengths greater
than about 10 m. Adding edge insulation leads to even lower impact of the zone dimension on
the edge heat loss. Therefore, it is reasonable to neglect the effect of zone dimensions on edge
heat loss for most buildings.
4.00
3.50

Fp (W/m-K)

3.00
2.50
2.00

L=10m
L=7.5m
L=5.0m
L=2.5m

1.50
1.00
0.50

0.00

0.50

1.00

k*

1.50

2.00

Figure 4.9 Effect of soil conductivity and half-floor length on optimized perimeter heat loss factor
(Configuration A, no edge insulation).

89
Figure 4.9 indicates that perimeter heat loss factor is linear with soil conductivity. However,
it also depends upon the edge insulation. The following correlation was developed for floor
configuration A using two-dimensional finite-element results for the baseline geometry and
conditions over a range of soil conductivities and edge insulation levels.
Fp=ak*+b

(4.10)

where the slope a and intercept b are correlated as

a = -0.1935R*3 + 0.4594R*2 - 0.5029R* + 1.4911
b = -0.6539 R*3 + 1.5876R*2 - 1.4163R* + 0.5628
and where R* is the ratio of the edge insulation R-value to the base case edge insulation R=0.95
W/m2-K.
Figure 4.10 shows comparisons between the optimized perimeter heat loss factors and values
determined from the correlation of Eqn. (4.10).
4.00
3.50

Fp (W/m-K)

3.00

R*=0.0
R*=0.5
R*=1.0
R*=1.5

2.50
2.00
1.50
1.00
0.50
0.00
0.00

0.50

1.00

k*

1.50

2.00

Figure 4.10 Curve-fit results for edge heat loss factor (Configuration A).

90
4.3.2. Floor Configuration B: 10 cm (4 Inches) HWC Slab with 4 inches of Gravel
Optimized soil depths for a range of soil thermal conductivities were determined for
configuration B (Figure 4.5b) and used to develop the correlation in Eqn. (4.11). Figure 4.11
shows comparisons between the curve fit and optimized soil depths.
D* = -0.1106k*2 + 0.7305 k* + 0.3809
1.60

(4.11)

Finite-Element
Curve Fit

1.40

D*

1.20
1.00
0.80
0.60
0.40
0.00

0.50

1.00

1.50

2.00

k*

Figure 4.11 Curve fit for optimized soil depth for centerline heat flux (Configuration B).

For floor configuration B, the linear form given in Eqn. (4.10) was used to correlate
perimeter heat loss factor with soil conductivity and edge insulation. The resulting expressions
for the slope a and intercept b in Eqn. (4.10) are
a = -0.118R*3 + 0.3928R*2 - 0.4419R* + 1.4463
b = -0.2937R*3 + 1.0164R*2 - 1.2702R* + 0.5897

Figure 4.12 shows comparisons between the optimized perimeter heat loss factors and
values determined from the correlation of Eqn. (4.10).

91
4.00
3.50

Fp (W/m-K)

3.00
2.50
2.00
1.50

R*=0.0
R*=0.5
R*=1.0
R*=1.5

1.00
0.50
0.00
0.00

0.50

1.00

1.50

2.00

k*

Figure 4.12 Curve-fit results for edge heat loss factor (Configuration B).

4.3.3. Numerical Validation

The correlations were developed using a single profile for ambient temperature variations
with an associated deep ground temperature that was representative of summer in California and a
single profile for zone temperature variation that was associated with night ventilation precooling.
However, the resulting model was tested over a wide range of ambient and zone temperature
conditions, including summer and winter conditions in a variety of locations. Figure 4.13 shows
example results in New York City in summer (Aug-1st) under night-ventilation control for floor
configuration A. Figure 4.14 gives results in Chicago for winter (Feb-1st) with conventional night
setup control for floor configuration B. These results are typical of the accuracy of the simplified
model in predicting slab heat transfer rate as compared with the finite-element results. Part of the
explanation for the good agreement between the simplified model and the numerical simulations
is the fact that the correlations for the simplified model were developed using the same software
and assumptions used for the validation. Agreement with measurements would not be expected
to be as good because of un-modeled effects, such as time variation in soil moisture content.

92

6.0

Finite-Element

Average Heat Flux (W/m 2 )

5.0

Simp lified (CTF+Fp)

k*=1.15, R*=0.70

4.0
3.0
2.0
1.0
T ime

0.0
-1.0

0:00

6:00

12:00

18:00

0:00

-2.0

Figure 4.13 Example simplified model accuracy for New York in summer (Configuration A).

10.0

Finite-Element

Average Heat Flux (W/m 2)

8.0

Simplified (CTF+Fp)

6.0

k*=0.85, R*=1.20

4.0
2.0

Time

0.0
-2.00:00

6:00

12:00

18:00

0:00

-4.0
-6.0
-8.0
-10.0

Figure 4.14 Example simplified model accuracy for Chicago in winter (Configuration B).

4.4. Summary
A simple model was developed for ground-coupled concrete floor slabs that can predict
spatially-averaged transient heat flux at the interior surface and could be integrated with existing
building load simulation tools. The model incorporates correlations for the effects of ground soil
properties and edge insulation that were developed for two common floor configurations. Hourly
heat flux predictions compared very well with predictions from a two-dimensional finite-element
program for these two geometries.

93

CHAPTER 5. WHOLE BUILDING HEAT AND MOISTURE MODELING

Stand-alone building envelope hygrothermal simulation does not consider the effects of
changing indoor temperature and humidity conditions on wall heat and moisture transfer and
doesnt include the evaluation of occupant comfort conditions.

This chapter presents the

development of a whole building heat and moisture balance model and along with some results of
the application of this model.
There are many software tools available for whole building energy simulation, such as DOE
(Sullivan, 1998), TRNSYS (McDowell et al., 2003), etc. However, most of them are designed to
determine energy usage and dont include detailed models for moisture transfer. HAM-Tools
(Sasic Kalagasidis, 2003) has a whole building moisture simulation model, but it does not include
the ability to compute the liquid water capillary transfer. UMIDUS has a wetted surface model
and Mendes et al. have coupled this model with DOE-2.1E to investigate moisture effects on
building heat conduction loads (2003I). They then developed DOMUS 2.0 based on UMIDUS
(Mendes etc., 2003II) to realize a whole building heat and moisture analysis. However, this
model does not include a detailed infiltration model, does not consider variations in humidity
conditions for different parts of the building, and doesnt include moisture removal due to HVAC
equipment. Therefore, a whole building heat and moisture balance model was developed that
includes heat/moisture transfer from the building envelope, heat/moisture internal generation, air
infiltration, HVAC equipment heat/moisture removal, and moisture buffering.

5.1. Description of the Whole Building Heat-Moisture Balance Model

Figure 5.1 depicts several factors that can influence indoor moisture conditions, including
building envelope moisture transfer, air infiltration and ex-filtration, moisture buffer effects
caused by indoor soft materials (such as curtains, bedding) and hard materials (furniture, floors,
etc), indoor moisture generation caused by indoor occupant activities (such as cleaning, bathing,

94
cooking, etc) and moisture removal caused by condensation on the coil for the cooling system.
The model developed in this study considers residential geometries of one to three conditioned
floors with a slab on grade ground floor (no basement or crawl space). An unconditioned attic is
also included and heat and moisture transfer from ceiling to the interior surfaces for the top floor
is considered. Transfer of moisture from outdoors (including ambient and attic) to the interior air
through infiltration includes the effects of the wind pressure field around the building. Moisture
can also be added to or removed from individual zones by internal sources such as indoor
moisture generation, inter-zonal air exchange, an exhaust fan, and the HVAC equipment. For
simplicity of the simulation, the air within each floor is considered to be fully mixed at a single
temperature and humidity. For some of the cases considered in this thesis, the whole building is
treated as a one-zone model, which means the air within the entire building is assumed to be fully
mixed. The use of a one-zone building model leads to dramatic reductions in computational
requirements and is shown to work well in terms of tracking moisture content within the walls
due to moisture migration due to external effects. However, it should be noted that the use of a
fully-mixed air assumption for the building or for each floor does not allow the evaluation of
moisture problems within walls that could be caused by local internal effects, such as persistent
inadequate ventilation of bathrooms during showers. Multiple zones within a single floor could
be added in future versions of the model but with additional computational cost.
A moisture balance on the air for each floor over a time step t at time t is expressed as:
m a,i (i,t -i,t-t ) =

m&
f

a,f

t(f,t -i,t ) +

wall

& t+ G
& t- G
& t
A wall t +G
buff
gen
remv

(5.1)

where ma,i is the mass of indoor air with subscript a for air and i for indoor, i,t is the indoor
humidity ratio at time t, m& a , f is the air flow rate from zone f (from another floor, infiltration from
a building opening, or ventilation from ambient to the zone caused by an exhaust fan, etc.), f,t is
the air humidity for zone f, gwall is the vapor flux from each wall with a surface area of Awall,

G& gen is a moisture generation term caused by indoor moisture sources due to human activities,
G& buff is a moisture buffering term associated with internal buffer materials, such as curtains or
beddings, and G& remv is the moisture removal due to the condensation on the coil of the cooling
system.

95

Infiltration
moisture
removal

Exfiltration

moisture
transmission
moisture
buffer
moisture
generation

Figure 5.1 Schematic of moisture room paths and driving forces.

The determination of moisture conditions is strongly coupled to the energy analysis because
of the coupled heat and mass transfer within the envelope and the role of HVAC equipment in
removing and adding moisture. A sensible energy balance on the air within each floor over a
time step t at time t is expressed as

& a,f t c p (Tf,t -Ti,t ) +

ma ,i c p (Ti,t -Ti,t-t )= m
f

wall

Awall t

& t +Q
&
&

Awin t+ Q
qwin,cond
gen
floor t-Q remv t

(5.2)

where Ti,t is the indoor temperature at time t, Tf,t is the air temperature for zone f, qwall is the heat
flux for each wall (for top floor, including ceiling) having a surface area Awall, qwin,cond is
conduction at the windows with surface area Awin that is transferred to the air, Q& gen is heat

96
generation caused by indoor heat sources, Q& floor is the heat transfer at the slab floor for the ground
level only, and Q& remv stands for the heat removal caused by air conditioning equipment (or heat
addition in negative sign because of heating devices).

Models for the terms within the moisture and energy balance and external boundary
conditions are presented in the following subsections along with approaches for numerical
solution of the whole building model.

5.1.1. Infiltration, Inter-zonal, and Ventilation Air Flows

Air flow within a residential building can have a very important impact on moisture
conditions.

A simplified multi-zone air flow model has been developed that is based on

CONTAMW (Walton, 1989). Appendix 1 presents a detailed description of the model and
numerical solution. The model estimates the net infiltration flow for each floor of the building
and the air flow between floors. Quasi-steady air flow through openings in the building envelope
(infiltration or exfiltration) and between floors are modeled using a power law relationship
(Q=C(P)n). The flow coefficient C can be calculated from the ELA (effective leakage area)
based on the method adopted by the ASHRAE Handbook of Fundamentals (ASHRAE, 2005,
P27.13). A value of the flow exponent n of 0.65 is typical for residential buildings and was used
for this study. Pressure differences between the ambient air and air on different floors are driven
by wind and stack effects. A set of non-linear equations are solved at each simulation timestep
using a multi-dimensional Newton-Raphson method. In order to check implementation, results
from the simplified multi-zone air flow model have been numerically validated with results from
COMTAMW for a range of different wind and thermal conditions.

5.1.2. Internal Moisture and Heat Generation

Internal moisture and heat generation are driven primarily by occupants and vary with time
of day and time of week. The simulation tool allows scheduling of moisture gains and heat using
basic elements, such as lights, people, plants, showers, etc. Each occupant is viewed as having
67W of sensible heat gain and a moisture gain of 1.9 lbm/person/day. A sensible utility load of

97
470W is suggested by ASHRAE (ASHRAE, 2005) for kitchens and laundry rooms that contain
continuously operating appliances such as refrigerators and/or freezers. For lighting, a gain of 5
W/m2 is assumed with a 70% /30% split between radiation and convection. For a computer, 270
W of heat generation with 30% /70% radiation/ convection split was adopted as specified by Zhai
and Chen (2006). Default schedules for moisture generation are available that are based on
surveys made in Hong Kong (Yik etc., 2003) and in the U.S. (Aoki-Kramer and Kragiozis, 2004).
These schedules include the effects of the following:
1. Transpiration from humans (respiration, perspiration)
2. Personal hygiene activities (bathing)
3. Cleaning for dwelling
4. Wash up
5. Cooking and dinning;
Figure 5.2 demonstrates the daily heat and moisture generation for a family of four members
(two adults and two kids).

The internal gains are due to the human body, illumination,

refrigerators, televisions, a personal computer, and electrical home appliances (e.g., coffee
machine, cooking oven, toaster and microwave, etc). The time variation is based on a typical
work day schedule. For the heat generation shown in Fig 5.2a, the daily average is about 525W.
The daily peak usually happens during the early evening hours because of cooking, dining, and
entertainment. A 30/70% radiation/convection split is assumed for the overall indoor heat gain.
For moisture generation, human respiration, cooking, dinning, bathing, and house cleaning are
considered with a time schedule as shown in Fig. 5.2b. The peak usually happens during the
morning and evening due to bathing and cooking/dinning. The daily average is about 0.323 kg/hr
which agrees well with the section on Internal Moisture Gains in Chapter 20 of the 2004
ASHRAE HandbookHVAC Systems and Equipment. This chapter states that a family of four
typically produces about 320 g/hr of moisture. The scheduled exhaust fan schedule is also
depicted in Figure 5.2c and assumes that a 100 CFM exhaust fan is switched on during morning,
noon, and evening due to cooking and bathing.
Unless otherwise state, the data and schedules presented in this section were used to generate
results for whole-building analysis in this thesis.

98

1500

Heat Generation
Average

Heat Generation (W)

1250
1000
750
500
250
0
0:00

6:00

12:00

18:00

0:00

Time

a.) indoor overall heat generation

1.80
Moisture Generation
Average

Moisture Generation (kg)

1.50
1.20
0.90
0.60
0.30
0.00
0:00

6:00

12:00

18:00

0:00

Time

b.) indoor moisture generation

120
Exhaust Fan

Exhaust Fan (CFM)

100
80
60
40
20
0
0:00

6:00

12:00

18:00

0:00

Time

c.) Exhaust fan schedule

Figure 5.2 Example for typical daily indoor heat and moisture generation for a family of four.

99
5.1.3. Wall, Ceiling, and Floor Heat and Moisture Fluxes
The wall model presented in Chapter 2 for building envelope hygrothermal analysis was
implemented within the whole-building model. The wall model calculates moisture and heat
fluxes on the interior surfaces of building envelopes for the indoor air heat/moisture balance.
The default building considered in this study includes a vented attic space. Figure 5.3
presents a common example for a single family house with a ventilated attic. The attic has many
air flow components, including vents for the eaves, gable, and ridge. Attics sometimes include a
roof vent driven by a powered ventilation fan. However, there are also many cases where the
attic is sealed and un-vented, which have much lower air change rates compared to the vented
cases.
Based on a literature review for attic ventilation published by the Florida Solar Energy
Center (Parker, 2005), 1 to 4 air changes per hour (ACH) is the common range for attic
ventilation rates. While for un-vented attics, the review showed that the natural ventilation rate is
about 0.40 ACH. This later result was obtained from field data for a large sample of homes in
Massachusetts and Rhode Island (466 homes with fiberglass insulated ceilings). For the current
study, a default value of 1.0 ACH was utilized to represent a vented attic.
The moisture and energy balance models for attic air are treated in the same manner as for
the building interior zones. However, the attic is an unconditioned space, so the temperature
floats at all times and there is no heat and moisture removal due to HVAC equipment. In
addition, there is no heat generation or moisture generation for the attic and no windows because
it is unoccupied. However, the attic is affected by solar radiation absorbed on the roof and is
coupled to the conditioned space through moisture and heat transfer through the ceiling and air
leakage.
Moisture transfer within the floor slab is neglected, whereas heat transfer at the surface of
the floor slab for the ground floor is calculated using the transient model presented in Chapter 4.

100

a. Building with attic and ceiling

b. Infiltration of vented attic with ambient air

Figure 5.3 Schematic for attic and attic ventilation.

5.1.4. Windows
Windows have a large influence on building energy usage because they transmit solar
energy and have a relatively large conductance for heat loss or gain compared to walls. Transient
effects for windows are typically neglected because they have low mass. A simplified model for
energy transfer presented by ASHRAE (2005) has been adopted within the current work. At any
time, the net energy transfer from the window to the interior air is determined as
Q& win = Q& win ,cond +Q& win , Rad =U Awin (To,t -Ti,t ) + SHGC Awin H sol

(5.3)

where Q& win is the overall heat transfer, Q& win ,cond is the heat conduction part that acts on the indoor
air, Q& win , Rad is the solar gain through windows that acts on the inner surface of the building
envelope and floor, U is the conductance (U-factor) for the window, SHGC is the solar heat gain
coefficient, and Hsol is the incident solar radiation. The solar heat gain coefficient, SHGC,
considers the direct effect of transmitted solar radiation and the effect of absorbed radiation on
conduction through the window. The solar heat gain is assumed to be evenly distributed as a
radiation heat flux to the inner surfaces of the building, including the floor, walls, and ceiling.
ASHRAE (2005) provides tabulated data for window U and SHGC values for a number of
commonly used window types. Table 5.1 gives some typical values. For results presented in this
thesis, windows with double glazing and 12.7 mm air space were employed. In addition, the ratio
of window to wall area (Awin/Awall) was assumed to be 0.12 for all exterior walls for the default
case.

101
Table 5.1 U-factor and SHGC for windows.
U-factor
Windows type

SHGC-incident angle

(W/m2-C)
Alum. Wood.

40

50

60

70

80

Single glazing 3.2 mm glass

6.12

5.05

0.86 0.84 0.82 0.78 0.67 0.42

Double glazing 12.7mm airspace

3.42

2.87

0.76 0.74 0.71 0.64 0.50 0.26

Double glazing 12.7mm airspace Low-E

2.89

2.39

0.65 0.64 0.61 0.56 0.43 0.23

5.1.5. Moisture Buffer

There are many materials in a building that are porous and can store a significant amount of
moisture, such as furniture and other furnishings (curtains, bedding, carpet and paper). Typically,
furnishings have relatively large contact area with the indoor air and relatively large moisture
storage capacity.

Therefore, they can have a significant influence on the transient indoor

moisture balance model.

Wood furnishings and internal wall constructions have very slow response in terms of
changes in moisture and are treated as internal building components using the modeling approach
applied to external walls but subjected to indoor boundary conditions.
Soft furnishings, such as curtains and beddings, have faster response to moisture changes
compared to the building envelope and need to be treated separately. A relatively simple semiempirical model presented by Persily (1998) is utilized for moisture storage effects. The model
incorporates some fundamental mass transport principles (Axley, 1990) and considers diffusion at
the boundary of a moisture storage element having a linear adsorption isotherm. At any time, the
rate of moisture diffusion to the air from the storage element is:
G& buff =h m,buff A furnishing (s,t /K buff i,t )

(5.4)

where, hm,buff is the surface-average mass transfer coefficient (estimated to be 0.72 m/h), is the
film density of the air (taken as 1.2 kg/m3), Afurnishing is the surface area of the storage material
(m2), i,t is the humidity ratio for the zone air at time t, s,t is the equilibrium moisture
concentration of the storage material at the current room air conditions, and Kbuff is an empirical

102
coefficient termed the partition coefficient. Persily (1998) estimated a value for Kbuff of 5.
Simulation results obtained with this buffer model have been shown to give good agreement with
test data obtained for a residential test building (Emmerich, Persily and Nabiner, 2002).
Svennberg (2006) developed adsorption isotherms for materials found in common household
furnishings using the form:

s,t =

i ,t
A1 +A 2 i ,t -A 3 i ,t 2

(5.5)

where A1, A2 and A3 are adsorption isotherm constants for the selected materials, and i,t is the
indoor air relative humidity at time t. The adsorption isotherm constants for common materials
found in curtains and bedding are listed in Table 5.2.

Table 5.2 Adsorption isotherm constants for curtain and bedding.

Material
100% cotton curtain
100% cotton bedding
58% cotton - 42% flax
59% cotton 41% polyester
100% polyester
50% viscose - 50% wool
15% polyamide - 85% wool
100% wool (plain)
100% wool (felted)

A1

Absorption
A2

A3

2.74
2.53
1.98
4.69
29.6
0.849
1.25
0.727
0.886

30.9
22.5
28.9
44.1
360
15.7
16.5
18.2
16.9

28.9
19.7
26.5
41.6
313
13.7
13.3
15.8
14.5

A1

Desorption
A2

A3

3.31
3.04
1.95
4.73
68.4
1.01
1.58
1.04
1.37

16.9
13.4
22.4
27.6
212
9.93
8.21
9.85
7.33

14.9
10.9
19.5
24.3
199
7.94
5.03
7.36
4.90

5.1.6. HVAC Equipment and Thermostat

The thermostat that controls the heating and cooling equipment is assumed to be ideal with a
dead-band between cooling and heating setpoints. For any timestep, the zone temperature where
the thermostat is located is determined assuming that the equipment is off. If the temperature
falls below the setpoint for heating then it is assumed that a furnace or heat pump turns on and
cycles as necessary to maintain the temperature at the setpoint for the entire timestep.
Conversely, if the temperature floats above the cooling setpoint with the equipment off, then it is

103
assumed that the air conditioner turns on and cycles as necessary to maintain the cooling setpoint
for the entire timestep. For a multi-floor building, the zone temperature and humidity are treated
separately for each floor and the thermostat is assumed to be located on the ground floor.
Furthermore, the supply air for the HVAC equipment is split evenly between different floors.
The furnace and air conditioner are modeled using an approach from ASHRAEs HVAC
Toolkit (Brandemuehl et al., 1993).

These models utilize prototypical performance

characteristics, which are scaled according to the capacity requirements and efficiency at design
conditions. The air conditioner model includes the effects of moisture removal using a bypass
factor (BF) approach. The bypass factor is defined as the ratio of flow that is unaffected by the
coil (bypass flow) to the total air flow. The bypass factor is determined from rating information
for the unit and then used to estimate outlet humidity conditions at any operating condition. In
addition, power related outputs, such as the power consumption of the compressor, fan power,
and COP, are determined using rated information for capacity and efficiency and empirical
correlations for part-load characteristics.
In order to size the HVAC equipment for a given building and location, the whole-building
simulation program is executed for selected months in winter (Nov 1st Feb 28th) and summer
(May 1st Aug 31st) with constant heating and cooling setpoints. The maximum heating load and
maximum sensible cooling load are identified along with their associated indoor / ambient
temperature and humidity ratio conditions. For cooling, the total cooling load is greater than the
sensible cooling load because of the latent load. The HVAC equipment is oversized by 20%
compared to the peak cooling and heating loads.

5.1.7. Weather Processing and Boundary Conditions

TMY2 data includes ambient air temperature, dew point temperature, direct normal and total
horizontal solar radiation, sky cover, wind speed and direction, hourly precipitation data and
snow depth on the ground (if it is applicable) for hundreds of locations in the U.S. and was used
in this study.
For exterior surfaces (walls, roofs and windows) at given tilts and azimuth angles, incident
solar radiation is determined as a function of sun position using the model suggested by Duffie

104
and Beckman (1980). The solar incident on the building exterior surfaces includes direct beam
from the sun, diffuse solar radiation from the sky, and diffuse reflection from the ground. The
snow depth from the weather data is used as factor to determine the reflectance of the ground.
The ground reflectance is considered to be 0.2 if there is no snow and 0.7 when there is a snow
cover.
Long-wave sky radiation is incorporated in this model using an effective sky temperature.
The sky temperature is estimated using a method presented by Walton (1983) based on an earlier
method of Clark et al, (1978). The model incorporates a correlation for sky emissivity that is
determined using sky cover and dew point data available from the TMY2 data.
In order to calculate the infiltration rate, the model from CONTAMW is utilized for
determining wind pressure on each exterior wall as a function of wind speed, wind direction,
surface orientation, and the buildings wind pressure profile.
The rain intensity on exterior wall surfaces is determined using a model by Lacy (1977) as a
function of precipitation data, wind speed, and surface orientation. The rainfall intensity in units
of mm/hr on any building exterior vertical surface is estimated as:
Rdr = 0.222 V Rh0.88 cos

(5.6)

where 0.222 is a driving rain coefficient (average value) that results from adopted empirical
relations, V is the wind speed in units of m/s, Rh is the rain intensity on the horizontal surface in
units of mm/hr, and is the angle between the wind direction and the line normal to the wall.

5.1.8. Model Implementation and Numerical Solution

Figure 5.4 depicts the modular elements that form the whole building heat and moisture
model and their interactions.

Hourly weather data are used to generate ambient boundary

conditions (temperature, water vapor pressure, wind pressure, solar radiation, rain intensity,
effective sky temperature) for determining transfer processes at the external surfaces (capillary
transfer, evaporation, vapor diffusion, conduction, radiation exchange, infiltration). Building
material properties used for the transport processes can be read from the expandable building
material database. The whole building moisture and energy balances use the energy and moisture

105
transfers from building components, air flow, internal sources (such as generation) and sinks
(such as storage, buffers and removal). The zone and attic temperatures and humidity are
computed at each time step through integration of the moisture and energy balances.
Weather Database
Sky Temperature
Model

Solar Radiation
Model

Wind-driven Rain
Intensity Model

Window Heat
Transfer Model

Material Properties
Database

Ground Floor Heat

Transfer Model

Enhanced MOIST
Envelope Model
HVAC equipment Model

Indoor Heat/Moisture
Generation Model

Multi-zone Air Flow

Model

Indoor Furniture / Textile

Buffer Model

Whole building heat and moisture balance model

Attic/Indoor T

Attic/Indoor

Figure 5.4 The modular elements that form the whole building simulation.

The couplings between individual elements and the whole building energy and moisture
balances can be solved using either explicit or implicit schemes. The moisture and heat fluxes
from interior surfaces of the walls to the air within each zone depend on the zone temperature and
humidity. The explicit solution utilizes values of the zone conditions at the end of the previous
time step to evaluate these fluxes, whereas the implicit method utilizes end values over the
current step. The latter approach requires an iterative solution for the zone conditions at each
time step and was found to have greater computation requirements than the explicit method for
the same level of accuracy and appropriately chosen time steps. Using numerical experiments, it
was found that a time step of 15 minutes is adequate for the conditioned spaces when utilizing the
explicit integration method.

The attic space requires a smaller time step because it is an un-

106
conditioned space with a relatively small air volume and can have relatively high fluxes due to
solar and high ventilation. A time step of 3 minutes was employed for the attic in this study.
Figure 5.5 provides the basic information for the explicit integration scheme for the whole
building heat and moisture simulation with HVAC equipment. The ambient boundary condition
data (To, PPo, Tsky, Hsol, etc) for the given orientation is called first from the weather data
treatment subroutine. After backing up the information (temperature, moisture content) for each
node, the tri-diagonal matrices for temperature, water vapor pressure and capillary pressure are
built and then solved to obtain updated node temperatures, water vapor pressures and capillary
pressures.

This allows calculation of the heat and vapor fluxes on the surface of each

wall/roof/ceiling. Then, the subroutine for computation of the air flow is called using inputs of
room air temperature, attic air temperature, ambient air temperature, wind speed and wind
direction (so that wind pressure around the building can be specified), and exhaust fan operating
condition.

This routine determines the air infiltration rate for the rooms and attic. This is

followed by determination of internal heat and moisture generation and, window and ground floor
heat transfer, and moisture transfer due to the moisture buffer model. The total convective heat
gain to each zone is used within the overall energy balance to determine zone temperature and
equipment operation. A floating zone temperature is compared with the setpoints for cooling and
heating to decide whether HVAC equipment model is needed based on the method presented
previously. If cooling is required, then moisture is removed and the amount of moisture removal
is calculated based on the supply air flow rate and return air humidity ratio from the previous
timestep. The moisture gains and equipment moisture removal are used within the zone air
moisture balance model to determine new indoor humidity ratios for the time step. Attic air
temperature and humidity ratio are then updated using energy and moisture balances. The
program loops through all of the time steps from the start to the end time of the simulation.

107

Backup necessary data

Get ambient B.C. (To,PPo,Tsky, Hsol, etc.) for the 4 orientation
Build each building envelope Heat/Moisture/Capillary Matrix
Compute each walls node T/ Vapor Pressure/Cap. Pressure
Compute surface vapor flux/heat flux
Compute infiltration/exfiltration
Compute internal heat and moisture generation
Compute windows/ground floor heat transfer
Compute moisture buffer
Energy balance model for room

Troom |Tset,heating, Tset,cooling |

H
Equipment model for Heating
Troom = Tset,heating
Equipment model for Cooling
Troom = Tset,cooling
Equipment moisture removal

Moisture balance model for room

Heat/Moisture balance model for attic
Completed for the simulation?

N
IT = IT + 1

END

Figure 5.5 Flow diagram for the explicit solution scheme.

108
5.1.9. Comparisons with Other Models
Validations of individual parts of the whole-building model were performed through
comparisons with other simulation tools and experimental results. The air flow model was
extensively compared with the results of CONTAMW and gave nearly identical results for typical
residential conditions. Results from the updated wall model developed in this thesis gave the
same results as MOIST 3.0 for conditions where the original model was applicable. Furthermore,
predictions for the effects of solar heating following wetting were compared with experimental
results in Chapter 4.
In order to verify proper implementation of the whole building energy analysis, predictions
were compared with results from TRNSYS for a simple case study having the same boundary
conditions (ambient dry bulb temperature, solar irradiation and sky temperature exported from
TRNSYS) under the assumption of no space conditioning and no moisture transport. TRNSYS
includes a detailed analysis of energy flows within the building envelope but does not consider
detailed moisture transport. Parameters of the specific case study considered are as follows:

Selected 1-floor simple residential model (16m*8m*3m);

External walls with single layer brick (30 cm) with = 0.7, hmix,out=17.8 w/m2-K,
hmix,in=8.3w/m2-K;

Adiabatic roof and floor;

Constant material properties with no moisture transfer;

No indoor heat and moisture gains

No infiltration

No windows

No heating or cooling equipment;

Figure 5.6 shows comparisons between the whole building model and TRNSYS results for a
two-day period. Under the simplified conditions, the models predict nearly identical results.

109

32.0
30.0
28.0

Temperature (C)

26.0
24.0
22.0
20.0
18.0
16.0
14.0
12.0
10-1 0:00

Tamb
Ti (TRNSYS)
Ti
10-1 8:00 10-1 16:00 10-2 0:00

Time
10-2 8:00 10-2 16:00 10-3 0:00

Figure 5.6 Comparison of zone air temperatures with results from TRNSYS.

5.2. Case Study Results

5.2.1. Case Study Description

A case study is presented to show results from the whole building simulation with the
following parameters and considerations:

Single floor with four family members having schedules with internal heat and moisture
generation and exhaust fan usage presented in the previous section;

Weather data: Indianapolis, TMY2 data;

Vented attic with a pitch angle of 30 and a constant ACH of 1.0 from the ambient;

Simple residential geometry (16*8*3m, or 1385 ft2 with about 10 ft high walls);

Including ceiling heat and mass transfer from the attic;

Including floor heat loss: the slab-on-grade is a 4 heavy weight concrete slab with 4 of
sand and pea gravel above the ground soil, with (16+8)*2=48m peripheral edge with a
perimeter heat loss coefficient =1.31 W/m-C ;

Longer wall lengths (16m) facing north and south and shorter lengths (8m) facing east
and west;

Exterior surface absorptances: 0.60 for walls and 0.70 for roofs;

110

Double glazing low E glass for windows with window to wall area ratio of 0.12 for all
four walls with U=2.87W/ m2-C;

Effective Leakage Areas: 160 cm2 (North and South), 65 cm2 (East and West) , 260 cm2
(Ceiling)

Air conditioner and furnace having a setpoint for heating (Tset,

setpoint for cooling (Tset, cooling) of 75F

Moisture transport within exterior walls and ceiling including both vapor diffusion and
capillary transfer;

Buffer materials: 34.6 m2 of 100% cotton curtain and 6 m2 of 100% cotton-bedding

(material properties as shown in Table 5.2);

HVAC Equipment rated parameters: EER=9.5, SHR=0.7, supply fan air flow rate= 350
CFM/ton, supply fan power=0.40 W/CFM, heating furnace Efficiency=0.75, heat pump
COP=3.0.

heating)

of 70F and a

The building materials used in the case study are listed in Table 5.3. For surface paint, the
unit of measurement typically used is the perm, where one perm stands for 1 grain (1/7000
lbm) of water vapor transfer per square foot of area per hour for a vapor pressure difference of
one inHg. Materials with a perm less than 10 are viewed as vapor retarders. Typical perm values
for some layers that have negligible storage capacity are: Latex paint: 10 perm; 3.5 inches glass
fiber insulation: 33.1 perm; 6-mil glue: 1 perm; Vinyl wallpaper: 0.5 perm.

Table 5.3 Parameters for building structure.

Structure
Type

surface film
interior
hi=4.2W/m2-C
Exterior
hr=5.2W/m2-C
Wall
Latex Paint:
10 Perm
hi=5.7W/m2-C
Roof
hr=5.2W/m2-C
no paint
hi=2.7 W/m2-C
Ceiling
hr=5.2W/m2-C
no paint
h =3.5W/m2-C
Indoor Floor/ i
hr=5.2W/m2-C
Furniture
Paint: 10 perm

Layer 1
Gypsum
board
0.5

Layer2

Layer 3

Asphalt
US R-19
vapor
insulation
retarder
23.9 Perm
0.1
OSB board Roofing Paper
1
0.1
Gypsum board
0.5
Sugar Pine
2

Layer 4
OSB
Sheathing
1

Layer 5

Wafer
Board
Siding
0.5
Asphalt Roofing
Shingles
0.5
US R-30 insulation
11.0 Perm

surface film
exterior
hi=11.4W/m2-C
hr=5.2W/m2-C
Latex Paint:
5 Perm
hi=11.4W/m2-C
hr=5.2W/m2-C
no paint
hi=4.0W/m2-C
hr=5.2W/m2-C
no paint
hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

111
Yearly weather conditions for Indianapolis in terms of air temperature and relative humidity
variations are shown in Figure 5.7. These data demonstrate that Indianapolis is cold during
winter while hot in summer.

In addition, the ambient relative humidity is relatively high

throughout the year.

100

1.00

Tamb
80
Relative Humidity

Temperature (F)

0.80
60
40
20

0.60
0.40

0.20

-20

0.00

RHamb
0

1000 2000 3000 4000 5000 6000 7000 8000

Time (hr)

a. Ambient air temperature

1000 2000 3000 4000 5000 6000 7000 8000

Time (hr)

b. Ambient air relative humidity

Figure 5.7 Ambient air temperature and relative humidity for Indianapolis.

5.2.2. Results for Indoor/Attic Conditions and Equipment

Figure 5.8a shows the attic temperature variation over the year for this case study. The
temperature range is much larger than the ambient temperature variation because of solar effects.
The attic has a minimum temperature of about 2F that occurs during the early morning hours in
winter and a high temperature of about 125F during summer in the afternoon. Furthermore, the
attic air temperature has a strong daily temperature swing (up to more than 55F), which leads to
huge daily attic air relative humidity changes, as shown in Figure 5.8b.
The sizing results and design conditions for the HVAC equipment are listed in Table 5.4.

Table 5.4 Sizing results for the HVAC equipment.

Season

Sensible load
(W)

Latent
Load
(W)

Cooling

4233

2180

Heating

Total
Capacity

Design Tin
(F)

Design
RHin (-)

2.2 tons
41500
BTU/hr

75

60%

70

Design Design
Tamb
RHamb
(F)
(-)
89.0
38.0%
-9.1

75.7%

Temperature (F)

112

130
120
110
100
90
80
70
60
50
40
30
20
10
0

Time (hr#)
0

2000

4000

6000

8000

a. Attic air temperature

1.0
0.9

Relative Humidity (-

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

Time (hr#)

0.0
0

2000

4000

6000

8000

b. Attic air relative humidity

Figure 5.8 The yearly attic air temperature and relative humidity.

The maximum predicted heating load was 41500 BTU/hr and occurred at hour # 388. For
two days that include this hour, Figure 5.9 shows ambient, indoor, and attic temperatures and
relative humidities along with ACH for the conditioned space.

Relatively high ACH for the

conditioned space occurred at the beginning of this two-day sequence and previous day (not

113
shown), since it was windy (wind speed up to 14 m/s with an average of more than 12 m/s for 20
continuous hours until hour # 375). The attic temperature remained higher than the cold ambient
air over this two-day period due to solar effects and heat transfer from the conditioned space. The
relative humidities within the attic and ambient are high because of the low temperatures.
However, the humidity ratios are very low (less than 0.001 kg/kg). On the other hand, the relative
humidity within the indoor air is relatively low because of low ambient moisture gains and since
the indoor air temperature is fixed at the heating setpoint. The fluctuations in indoor RH are
primarily due to the indoor moisture generation schedule coupled with fluctuations in ventilation
due to wind speed variations. Moisture buffering within materials damps the effects of moisture
generation and ventilation variations on indoor humidity.
The predicted maximum sensible cooling load occurred at hour # 4266 (June 27th at 18:00)
and the resulting overall cooling capacity for the air conditioner at this time was 2.2 tons. Figure
5.10 depicts some of the conditions and model predictions for two days that include this hour. For
both days, the skies were clear with significant solar radiation during the day time and almost no
wind. There was a relatively large ambient air temperature swing because of the clear conditions
and relatively low relative humidity except early in the morning when the temperature was
lowest. The attic air temperature reached a maximum of about 125F during the early afternoon
hours and dropped to 65F for early morning hours. This large temperature variation resulted in
huge relative humidity changes (roughly 10-90% range) that were out of phase with the
temperature swing (largest RH at lowest temperature and lowest RH at highest temperature). The
indoor air temperature was constant at 75F because the air conditioner ran continuously during
this period. The indoor relative humidity varies as a complex function of the individual terms
within the moisture balance, including indoor moisture generation, air infiltration, moisture
removal by the air conditioner coil, and moisture buffer effects. Fig. 5.10c shows the air
infiltration for this building varies between about 0.15 and 0.4 ACH. The peaks in infiltration
rate are due to scheduled operation of the 100 CFM exhaust fan.

114

80
70

Temperature (F)

60

T_Attic

50

Tin

Tamb

40
30
20
10
0
-10
-20
372

384

396
Time (hr#)

408

420

a. Air temperature
1.0
0.9

Relative Humidity (-)

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

RH_Attic

0.0
372

384

396
Time (hr#)

RHin

RHamb

408

420

b. Air relative humidity

0.80
ACH
0.70

Air Change per Hour (-)

0.60
0.50
0.40
0.30
0.20
0.10
372

384

396
Time (hr#)

408

420

c. Air change rate per hour (ACH)

Figure 5.9 Whole building simulation result for winter case.

115

125

Temperature (F)

115
105
95
85
75
65
T_attic
55
4248

4260

Tin

4272
Time (hr#)

Tamb

4284

4296

a. Air temperature
1.0
0.9

Relative Humidity (-)

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
4248

RH_attic
4260

RH_amb

4272
Time (hr#)

RH_in

4284

4296

b. Air relative humidity

0.45
0.40

Air Change per Hour (-)

0.35
0.30
0.25
0.20
0.15
0.10
4248

ACH
4260

4272
Time (hr#)

4284

4296

c. Air change rate per hour (ACH)

Figure 5.10 Whole building simulation result for summer case.

116
Figure 5.11 shows the equipment cooling loads and performance during this period. In the
early morning, there is significant latent cooling load but small sensible cooling load, because at
that time the indoor moisture generation (e.g., cooking and bathing) is high, but the ambient air is
still cool. As the result, a low SHR (<0.5) occurs during this time. Peak sensible cooling loads
usually occur during the afternoon, where the ambient air is hot with plenty of solar radiation.
This results in a high SHR (>0.8). COP depends on several factors, such as the ambient air
temperature and humidity, sensible heat load ratio (SHR) and part-load ratio. High COPs occur
with relatively low ambient temperature but when the load is still significant. During the
afternoon, the COP dropped due to high ambient temperature. For late night, COP dropped below
1.5 due to a low part-load ratio (i.e., cycling losses).

5.2.3. Results for Envelope Moisture

Predictions of detailed temperature and moisture content profiles within building materials
can be used to evaluate designs for potential danger of mold growth due to favorable air
temperature and relative humidity or structural damage related to high moisture content. There is
no general agreement on an exact criterion that is most appropriate for mold growth. However,
relative humidity over 80% with air temperatures from 40~100F for a continuous 30 day time
span is suggested by the ASHRAE Handbook of Fundamentals for potential mold growth
(ASHRAE, 2005). Structural failures caused by wood decay are rare but have occurred (Merrill
and TenWolde 1989). Decay generally requires wood moisture content at fiber saturation (usually
about 30% moisture content) or higher and temperatures between 10 and 40C. Therefore, the
fiber saturation moisture content can be view as the threshold to evaluate potential material
damage.
Some examples of detailed temperature and moisture content information within the
envelope structure of the case study building presented in the previous section are shown in
Figure 5.12. Fig. 5.12a depicts the daily average temperature of the interior and exterior surface
nodes (gypsum wall board layer and the OSB sheathing layer) for the west wall selected from the
whole building analysis. T[1] is the interior surface node, so its temperature has a relatively small
variation and is a little higher than the indoor cooling setpoint in summer and slightly below the
heating setpoint in winter. T[11] is the inner surface node of the OSB sheathing layer that is in
contact with the R-19 insulation and is close to the ambient condition. Therefore, it undergoes

117

5000

Latent Load
Sensible Load

3000

2000

1000

0
hr#4254

hr#4260

hr#4266

hr#4272

hr#4278

hr#4284

hr#4290

a. Cooling load condition (sensible, latent cooling load)

0.80

0.60
SHR

hr#4248

0.40
SHR
0.20

0.00
4248

4260

4272

4284

4296

Time (hr#)

b. Sensible Heat Load Ratio (SHR)

3.50
3.00
2.50

COP

Cooling load (W)

4000

2.00
1.50
COP

1.00
0.50
0.00
4248

4260

4272
Time (hr#)

4284

4296

c. COP for the compressor

Figure 5.11 Cooling load condition for the equipment.

hr#4296

118
large variations throughout the year. Fig. 5.12b depicts the daily average moisture content for
three layers, defined as the ratio of the weight of water vapor within the material to the weight of
dry building material. For this example, the moisture content of the wood-based outer layers is
between about 0.06 and 0.08 lb/lb moisture content. However, very small moisture content exists
within the gypsum wallboard.

100

Gypsum Wallboard
Asphalt Paper
R-19 Insulation

90

Node Temperature (F)

80
70

T[1]

60

T[11]

T[1]
T[11]

50
40
30

OSB Sheathing

20

Wafer Board Siding

10
0

60

120

180

240

300

360

Time (Day#)

a. node temperature

0.09

Gypsum Wallboard
Asphalt Paper
R-19 Insulation

Layer average moisture content (lb/lb)

0.08
0.07
0.06
0.05

OSB Sheathing

0.04

Wafer Board Siding

0.03

Gypsum Wallboard

0.02

OSB Sheathing

0.01

Wafer Board Siding

0
0

60

120

180
240
Time (day#)

300

360

b. layer moisture content

Figure 5.12 Whole building simulation result for summer case.

Moisture content is not a particularly good way to describe the wetness of a building
material because different materials have different capabilities to absorb and hold moisture within

119
their porous cavities. A better indicator is the relative humidity of air that would be in equilibrium
with the wood at its moisture content. At any time, the equilibrium relative humidity can be
determined from the isothermal equilibrium absorption curve for any wall layer using the current
moisture content. Figure 5.13 shows the equilibrium absorption curves for the three materials
considered for the results of Figure 5.12. In comparing Figure 5.12b and Figure 5.13, the
equilibrium relative humidities are less than 60% for the outer wood layers of this wall. The
moisture content would need to be more than 0.10 lb/lb to realize an equilibrium RH of 80% that
might promote mold growth.

For the case study considered in this section, none of the walls

experienced conditions that would lead to mold growth or material damage.

0.25

Gypsum Wallboard

Moisture Content (lb/lb)

0.20

OSB Sheathing
Wafer Board Siding

0.15

0.10

0.05

0.00
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Air Relative humidity

Figure 5.13 Isothermal absorption curve for selected building materials.

5.2.4. Importance of Whole-building Analysis for Wall Moisture Performance

There is a relatively large computational overhead associated with performing a wholebuilding analysis that may be unnecessary if the only goal is to evaluate the moisture performance
of exterior walls. An alternative approach, adopted by MOIST 3.0, is to simulate an individual
exterior wall subjected to constant indoor conditions but with time varying ambient boundary
conditions associated with the particular wall under consideration. In order to evaluate tradeoffs
in these two approaches, the west wall that was considered for the results of the previous section
was simulated alone using the annual average room temperature and relative humidity as an
indoor boundary condition. Figure 5.14 shows results for the stand-alone wall analysis (denoted

120
with symbols _W) compared with results determined for the whole-building analysis (denoted
with symbols _B). The plots show average daily equilibrium relative humidity for materials at
the interior and exterior surfaces.
There is very little difference in the moisture results at the exterior surface of the siding
because of the close coupling to the ambient and weak coupling to the indoor condition
(separated by insulation and a vapor retarder). However, much larger differences occur at the
interior surface for the wall and whole building analyses. The stand-alone wall analysis results in
relatively stable moisture content within the gypsum because of the constant indoor boundary
condition. In contrast, the gypsum undergoes large fluctuations in moisture content for the
whole-building analysis because of varying indoor air humidity conditions.

Although the

moisture levels would not cause mold or material damage problems for this case study, the standalone wall analysis would not identify potential problems that could occur due to more significant
indoor moisture gains that could potentially occur, such as moisture gains within a bathroom not
having an exhaust fan.
The moisture content within materials near the indoor air can be sensitive to the location of
the vapor retarder. Figure 5.15 shows example results for an alternative wall construction with
plywood as the sheathing and the vapor retarder located next to the exterior siding. For this case,
average daily values of equilibrium relative humidity for the interior layer and an internal layer
are presented. Again, for the interior layer, there are some significant differences in the time
variation of moisture contents predicted by the whole-building and stand-alone wall models.
However, for internal layer, much smaller differences were found. For this case, the interior layer
dampened (though it could not insulate) the indoor temperature and moisture swing. Compared to
the result from wall simulation, lower layer RH occurred in winter and higher layer RH resulted
in summer since the indoor air RH for building simulation is usually low in winter and high in
summer compared to the yearly average indoor air RH (which is used as a fixed indoor air RH for
the wall simulations indoor boundary condition).

121

0.75

RH_Gypsum Board_B
RH_Gypsum Board_W

0.70

Gypsum Wallboard
Asphalt Paper
R-19 Insulation

Equilibrium RH (-)

0.65
0.60
0.55

Interior
Layer

0.50
0.45
0.40
0.35

OSB Sheathing

0.30

Wafer Board Siding

0.25
0

60

120

180

240

300

360

Time (day#)

a. Interior layer

0.80

Equilibrium RH (-)

0.75

Gypsum Wallboard
Asphalt Paper
R-19 Insulation

RH_Wafer Siding_B
RH_Wafer Siding_W

0.70

Exterior
Layer

0.65
0.60
0.55

OSB Sheathing
Wafer Board Siding

0.50
0

60

120

180
Time (day#)

240

300

360

b. Exterior layer

Figure 5.14 Comparison of interior/exterior layer equilibrium RH for whole-building and standalone wall modeling (vapor barrier on inside of insulation).

122

Gypsum Wallboard

0.8

Plywood Sheathing
R-19 Insulation

0.7

Equilibrium RH (-)

0.6

Interior
Layer

0.5
0.4
RH_Gypsum Board_W
0.3

RH_Gypsum Board_B

0.2

Asphalt Paper

0.1
0

60

120

180
Time (day#)

240

300

360

Wafer Board Siding

a. interior layer
Gypsum Wallboard

0.46

Plywood Sheathing
R-19 Insulation

0.44

Equilibrium RH (-)

0.42

Internal
Layer

0.40
RH_Plywood_W
RH_Plywood_B

0.38
0.36
0.34
0.32

Asphalt Paper

0.30

Wafer Board Siding

0

60

120

180

240

300

360

Time (day#)

b. internal layer

Figure 5.15 Comparison of interior/exterior layer equilibrium RH for whole-building and standalone wall modeling (vapor barrier on outside of insulation).

5.3. Summary
This chapter presented a whole building energy and moisture model that integrates separate
models for exterior walls, ceiling, roofs, windows, slab floor heat transfer, furnishings moisture
buffer, HVAC equipment, internal heat/moisture generation, infiltration and inter-zonal air flow,
and an attic. Infiltration and inter-zonal air flow in the building are determined using a quasisteady model that is coupled to the overall energy and moisture balance equations. The models
are driven with hourly weather data that includes ambient temperature and humidity, wind speed
and direction, and beam and diffuse solar radiation. The weather data is processed to determine

123
radiation incident upon all surfaces, effective sky temperature, and pressures and wind driven rain
intensity acting on the building exterior surfaces.
The whole building model is designed for residential buildings and performs both energy
and moisture analyses. Previous tools for detailed analysis of moisture transport in walls have
considered individual walls with specified indoor conditions. A case study was performed to
evaluate the impact on wall moisture performance of performing a whole building analysis as
compared with a single wall analysis. At the exterior of the wall, there is very little difference in
the moisture contents for the two approaches. However, the differences can be significant within
materials that are close to the interior surface. The stand-alone wall analysis would not identify
potential problems that could occur due to more significant indoor moisture gains that could
potentially occur, such as moisture gains within a bathroom not having an exhaust fan. The
whole-building model has a greater potential for evaluating moisture problems caused by indoor
conditions and also can present the impacts of design choices on indoor air moisture levels.

124

CHAPTER 6. CASE STUDIES FOR BUILDING HEAT AND MOISTURE

PERFORMANCE

In this chapter, some case studies for building heat and moisture performance are presented
that analyze the moisture performance of building envelopes in some featured climates with
representative building constructions. The whole building energy and moisture model developed
in Chapter 5 was utilized to perform these analyses. While the building structures (building
envelope, ceiling, and roof) were changed for each location, the other parameters (e.g., building
geometry, window/wall ratio, indoor heat/moisture generation rate, moisture buffer, ground heat
loss factor, etc) were kept the same as those used in Chapter 5. Four featured locations were
selected to include cold, mixed, hot/dry and hot/humid conditions and are listed in Table 6.1.

Table 6.1 Locations for the parametric studies.

Location

Latitude

Longitude

City

State

Minneapolis

MN

4453"

Nashville

TN

Phoenix
Houston

Annual Average

Climates zone

T (F)

RH(-)

9313"

45.1

66%

7: cold

367"

8641"

58.6

68%

4A: mixed and humid

AZ

3326"

1121"

72.5

36%

2B: hot and dry

TX

2959"

9532"

68.1

75%

2A: hot and humid

6.1. Heating Climate-Minneapolis, MN

In a cold weather area, the temperature and humidity of the air within a heated, occupied
residence is considerably higher than that of the outdoor air. In this situation, moisture from the
indoor environment may permeate walls by means of diffusion to the building construction, and
then moisture is partially adsorbed and accumulates within exterior layers of the construction.
Duff (1968) observed that the moisture content in outer layers of the wall increases during cold

125
winter periods and subsequently decreases during warm summer periods. Therefore, an interior
vapor retarder and air barrier in the construction is strongly recommended to keep the moisture
content of the sheathing and siding materials from approaching and rising above damage criteria.
Otherwise, the outer layers are susceptible to warping, paint peeling, and fungal deterioration. In
particular, local moisture may accumulate in the partial walls close to kitchens and bath rooms.
This is because these local areas tend to have higher indoor water-vapor pressure compared to the
other bulk living areas and bed rooms due to their smaller volumes and higher moisture
generation rates.
Minneapolis was picked as a typical site for cold climates. The weekly maximum, average
and minimum outdoor air temperature and relative humidity are displayed in Figure 6.1.
Minneapolis is very cold during winter (lows of about -20 F) but is relatively mild in summer
(highs of about 90 F). Furthermore, the air is on the humid side year-round with average weekly
relatively humidities typically in the range of 60 to 80%.

The building structures used for the cold climate case study are listed in Table 6.2. For the
exterior walls, vinyl wallpaper was selected as an interior vapor retarder and located on the inner
surface of the gypsum wallboard. To meet the local energy code requirement, R-19 and R-38
insulation was chosen for the exterior wall and ceiling, respectively.
Sizing results for the HVAC equipment (including the design conditions) are summarized in
Table 6.3. The cooling capacity of 2.1 tons is moderate due to the mild weather during summer,
whereas significant heating capacity (43660BTU/hr) is needed because of the low ambient air
temperature. The cooling and heating capacities for Minneapolis are not much different than the
cooling and heating capacities determined in Chapter 5 for Indianapolis (2.2 tons and 41500
BTU/hr).

Even though Indianapolis has a significantly higher design ambient temperature in

winter, greater insulation was employed for the building in Minneapolis. Furthermore, infiltration
was not a major factor in the cooling and heating loads for this relatively air-tight structure (less
than around 0.40 ACH).

ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek 9
ee
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

Outdoor RH (-)

ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek
ee 9
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

Outdoor temperature (F)

126

100

80

60

40

20

-20
Max Hourly
Average Weekly
Min Hourly

-40

a. Ambient air temperature

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1
Max Hourly
Average Weekly
Min Hourly

0.0

b. Ambient air relative humidity

Figure 6.1 Ambient air temperature and relative humidity for Minneapolis.

127
Table 6.2 Parameters for building structure in Minneapolis.
Structure
Type

surface film
interior

Exterior
Wall

hi=4.2W/m2-C Gypsum
hr=5.2W/m2-C board
vinyl wallpaper: 0.5
0.5 Perm

Roof

hi=5.7W/m2-C
hr=5.2W/m2-C
no paint

OSB
1

Ceiling

hi=2.7 W/m2-C
hr=5.2W/m2-C
Latex paint:
10 Perm

Gypsum
board
0.5

Indoor
Floor/
Furniture

hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Layer 1

Layer2

Layer 3

Layer 4

US R-19
insulation
23.9 Perm

Fiberboard 1 air space

sheathing R-1.0
1
160 Perm

Exterior grade
plywood:
0.5

surface film
exterior

Face brick hi=11.4W/m2-C

3.5
hr=5.2W/m2-C
Paint:
10 Perm
hi=11.4W/m2-C
hr=5.2W/m2-C
no paint

Asphalt Roofing
Shingles
0.25

Roofing paper
0.1
Polyethylene
vapor retarder:
1.0 perm

Layer 5

US R-38
insulation
10.0 Perm

hi=4.0W/m2-C
hr=5.2W/m2-C
no paint
hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Sugar Pine
2

Table 6.3 Sizing result for the HVAC equipment in Minneapolis.

Design Design
RHamb
Tamb
(F)
(-)

Season

Sensible
load
(W)

Latent
Load
(W)

Total
Capacity

Design Tin
(F)

Design
RHin (-)

Cooling

4133

2120

2.1 tons

75

60%

91.0

42.3%

Heating

43660
BTU/hr

70

-20.9

62.2%

Good moisture performance can be found for both ceiling and roof as shown in Figure 6.2
for the north facing element. This plot gives average weekly equilibrium relative humidities as a
function of week of the year. During the summer season, the ceiling structures (plywood and
gypsum wallboard) endure peak moisture contents that are close to 70% equilibrium RH for about
10 weeks, which is much higher than that during winter (20% equilibrium RH). However, this is
still within the safety criteria (80% equilibrium RH).
Since an interior vapor retarder was employed that can prevent indoor moisture from
penetrating outward to the building envelope outer layers and accumulating there, satisfactory
moisture content (below 80% equilibrium RH) within the walls was found throughout the year.
Figure 6.3a shows equilibrium RH results for different layers within the worst moisture condition
wall for the 4 orientations (west in this case). During the winter time, the fiberboard sheathing

128
layer reached moisture contents of about RH=70%, but dropped below 60% during summer. The
layer average temperatures for the fiberboard sheathing and exterior surface face brick were very
low (below the freezing point of water) during the coldest winter times, as depicted in Figure
6.3b. It is important that the moisture content for these two layers is far below the fiber saturation
moisture content (RH=97%) in order to avoid structural damage due to ice formation.

0.75
0.7

Asphalt shingle

Layer average RH(-)

0.65

Roofing Paper

0.6
0.55

OSB

0.5
0.45
0.4

OSB
Roofing Paper
Asphalt Shingle

0.35
0.3
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. roof

0.8

R-38 insulation

0.7

Plywood board

Layer average RH(-)

0.6
0.5
0.4
0.3

Gypsum wallboard

0.2

6-mil glue

plywood board

Gypsum wallboard

0.1
0
1

13

17

21

25

29

Time (week #)

33

37

41

45

49

b. ceiling

Figure 6.2 Roof and ceiling equilibrium RH in Minneapolis.

129

0.8

Gypsum Wallboard
R-19 Insulation
1" air space

0.7

Layer average RH(-)

0.6
0.5

Vinyl
wallpaper

0.4
0.3
Gypsum wallboard
Fiberboard sheathing
Face brick

0.2
0.1

Fiberboard Sheathing

0
1

13

17

21

25

29

33

37

41

45

a. Moisture

100

Gypsum Wallboard
R-19 Insulation
1" air space

90
Layer average temperature (F)

Face Brick

49

Time (week #)

80
70
60

Vinyl
wallpaper

50
40
Gypsum wallboard
Fiberboard sheathing
Face brick

30
20

Fiberboard Sheathing

10
0
1

13

17

21

25 29 33
Time (week #)

37

41

45

Face Brick

49

b. Temperature

Figure 6.3 Wall structures internal equilibrium RH and temperature in Minneapolis (north).

An example of solar driven moisture incursion is shown in Figure 6.4, where average
equilibrium RH for the exterior layer (face brick) and the interior layer (gypsum wallboard) are
compared for south and north orientations. For the exterior layer, there was 3 hours of rain from
hours 5355 to 5357, followed by clear (hours 5363 to 5369), cloudy (hours 5370 to 5406), and
then clear (hours 5407 to 5448) weather conditions. Before the rain, both the materials within the
south wall (gypsum wallboard and face brick) had lower moisture contents than for the north wall
because of solar heating effects.

After the rain, the exterior face brick layer facing south

maintained lower moisture contents than the north brick because of its lower initial moisture
content and because of solar heating. Both exterior surfaces gained comparable moisture during
the rain. Subsequently, the exterior moisture was driven into the wall for both the north and south

130
walls. However, the moisture build up in the gypsum layer was considerably higher for the south
wall than for the north wall because of the greater incident solar radiation. In fact, the moisture
content of gypsum wallboard layer on the south wall became significantly higher than that for the
north wall during the clear days that followed the rain and cloudy weather.

Layer Average RH(-)

0.8

Gypsum Wall Board (South)

Gypsum Wall Board (North)

clear

Brick (South)
Brick (North)

0.7

rain
0.6

0.5
Time (hr#)
0.4
5352

5364

5376

5388

5400

5412

5424

5436

5448

Figure 6.4 Solar driven rain moisture build up for building structure

In order to show the importance of the location of the vapor retarder, the wall structure
described Table 6.2 was modified to remove the interior surface vinyl wallpaper but add a layer
of polyethylene vapor retarder on the exterior surface of the fiberboard sheathing. A comparison
of fiberboard sheathing layer average equilibrium RH for the two cases is shown in Figure 6.5.
The model predicts that the fiberboard sheathing would stay above the fiber saturation moisture
content throughout the whole year if the vapor retarder is located outside of the fiberboard
sheathing. During winter time, moisture builds up close to the fiber saturation moisture content
within this layer due to its low local water vapor pressure (caused by low temperature). The air
space acts as a capillary break, which blocks capillary transfer from the sheathing wood to the
exterior face brick layer. Furthermore, the polyethylene vapor retarder is designed to help prevent
vapor incursion from outside. However, in this case it also traps internal vapor from diffusing to
the exterior. As the result, the moisture content within the sheathing layer remains very high, and
this structure is susceptible to ice freezing/thaw damage during winter and mold grow in summer.
It is not an acceptable design for this climate.

131

Gypsum Wallboard
R-19 Insulation
1" air space

Plywood layer average RH(-)

1.00
0.90
Internal vapor retarder (outside
fiberboard sheathing)

0.80

Latex paint

Interior surface vapor retarder

0.70
0.60
0.50
1

13

17

21

25

29

33

37

41

45

49

Fiberboard Sheathing
Polyethylene vapor retarder
Face Brick

Time (week #)

Figure 6.5 Effect of vapor retarder location for cold climate condition.

6.2. Mixed Climate-Nashville, TN

Mixed climates have significant heating or cooling loads but generally have less extreme
conditions than cold or hot climates. Buildings in mixed climates may encounter high interior
humidity during winter and high exterior humidity during summer. Summer cooling or winter
heating for comfort in mixed climates usually does not lead to serious moisture problems in
exterior walls and ceilings. However, if a vapor retarder is necessary in a mixed climate, its
placement presents somewhat of a complication. Usually for climates dominated by cooling
loads, a vapor retarder should be located on the outside of the insulation. However, for conditions
dominated by heating, it would be better to locate the vapor barrier on the inner side of the
insulation.
Nashville is a typical mixed climate. The weekly maximum, average and minimum outdoor
air temperature and relative humidity are revealed in Figure 6.6.

Nashville is relatively cold

during winter and hot in summer, with lows around 0 F during winter and highs of about 90F in
summer. It is relatively humid all year long with an annual average at 68% RH and weekly
maximums of more than 90% RH.

132

110

Outdoor temperature (F)

90
70
50
30
Max Hourly
Average Weekly
Min Hourly

10

ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek 9
ee
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

-10

a. Ambient air temperature

1.0
0.9

Outdoor RH (-)

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

Max Hourly
Average Weekly
Min Hourly

w
ee
w k1
ee
w k3
ee
w k5
ee
w k7
ee
w k
ee 9
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

0.0

b. Ambient air relative humidity

Figure 6.6 Ambient air temperature and relative humidity for Nashville.

If the same building construction design used in Minneapolis were applied in Nashville, it
would not perform well as shown in Figure 6.7. During summer, because of the high ambient
temperature/humidity and low indoor temperature, moisture migrates inward and builds up on the
gypsum wallboard. The equilibrium relative humidity remains very high (about 70% RH) for a
period of about 8 weeks, which makes the gypsum wallboard layer highly susceptible to mold
growth and mildew. A vapor barrier close to ambient side is necessary to stop the ambient vapor
from transferring inward.

133

0.9

Gypsum Wallboard
R-19 Insulation
1" air space

0.8

Layer average RH(-)

0.7
0.6
0.5

Vinyl
wallpaper

0.4
0.3
Gypsum wallboard
Fiberboard sheathing
Face brick

0.2
0.1

Fiberboard Sheathing

0
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

Face Brick

Figure 6.7 Internal equilibrium RH and temperature for cold climate wall structure in Nashville.

In order to balance moisture concerns for both heating and cooling conditions, an alternative
wall structure, called a sandwich panel wall with low-permeability insulation as suggested for
manufactured houses by Burch (19952), was considered in the mixed climates. In this structure,
3.5 inches extruded polystyrene insulation was bounded by 6-mil glue attached with 0.5 inch
OSB on both sides. This extruded polystyrene insulation has an R-11 value, but its permanence is
as low as 1.1 perm, so it is called low-permeability insulation. In other words, this layer of
polystyrene insulates both heat and vapor transfer. The roof, ceiling and floor share the same
design as the building simulated in Minneapolis. The building structures used in this mixed
climate case study are listed in Table 6.4.

It should be noted that this type of wall has the

potential to remain dry under a variety of conditions. However, if moisture were trapped between
the two vapor barriers then the wall not readily allow it to escape. The problem of drying a
sandwich construction could be solved by applying smart vapor retarders (such as MemBrain
made by Certainteed) which have low permeability when the building material is dry but offer
high permeability when the building material is wet. In this way, drying is possible to either side,
and the design is robust.
Good moisture performance was found for both the ceiling and roof. In addition, due to the
use of the low-permeability insulation, the layers within the exterior walls also had excellent heat
and moisture performance as shown in Figure 6.8. The sandwich wall panel performs well for
mixed climates because it can block moisture incursion/excursion from either side.

134
Table 6.4 Parameters for sandwich building structure used in Nashville.
Structure
Type

surface film
interior

Exterior
Wall

hi=4.2W/m2-C Gypsum
hr=5.2W/m2-C board:
0.3
Latex paint:
10 Perm

Roof

hi=5.7W/m2-C
hr=5.2W/m2-C
no paint

OSB
1

Roofing paper
0.1

Ceiling

hi=2.7 W/m2-C
hr=5.2W/m2-C
Latex paint:
10 Perm

Gypsum
board
0.5

Polyethylene
Exterior grade
vapor
plywood:
retarder:
0.5
1.0 perm

Indoor
Floor/
Furniture

hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Layer 1

surface film
exterior

Layer2

Layer 3

Layer 4

Layer 5

OSB: 0.5
+
6-mil glue

Polystyrene
extruded:
3.5

6-mil glue
+
OSB: 0.5

Sugar
hi=11.4W/m2-C
pine siding hr=5.2W/m2-C
Paint:
0.5
5 Perm

Asphalt Roofing
Shingles
0.25

hi=11.4W/m2-C
hr=5.2W/m2-C
no paint

US R-38 insulation
10.0 Perm

hi=4.0W/m2-C
hr=5.2W/m2-C
no paint
hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Sugar Pine
2

Sizing results for the HVAC system are summarized in Table 6.5, which has lower heating
capacity (35700 BTU/hr) and higher cooling capacity (2.4 tons), compared to the Minneapolis
case.

Table 6.5 Sizing results for the HVAC equipment in Nashville.

Design Design
RHamb
Tamb
(F)
(-)

Season

Sensible
load
(W)

Latent
Load
(W)

Total
Capacity

Design Tin
(F)

Design
RHin (-)

Cooling

4750

2350

2.4 tons

75

60%

100.9

28.2%

Heating

35700
BTU/hr

70

-2.9

84.9%

135
Polystyrene
extruded

0.60
Gypsum wallboard
OSB inner side
Extruded polystyrene
OSB outer side
Sugar pine siding

Layer average RH(-)

0.58

Gypsum
Wallboard

0.56

Sugar Pine
Siding

0.54

6-mil glue

0.52

OSB Sheathing

0.50
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. Moisture
Polystyrene
extruded

90

Layer average temperature (F)

80
70

Gypsum
Wallboard

60

Sugar Pine
Siding

50
Gypsum wallboard
OSB inner side
Extruded polystyrene
OSB outer side
Sugar pine siding

40
30
20

6-mil glue
OSB Sheathing

10
1

13

17

21

25 29 33
Time (week #)

37

41

45

49

b. Temperature

Figure 6.8 Sandwich wall structure internal equilibrium RH and temperature for Nashville.

Although the use of vapor retarders on both sides of a wall can reduce moisture incursions, it
also creates the possibility of moisture getting trapped when the vapor seals are not perfect.
Therefore, the sandwich type wall structure may not always be the preferred option for mixed
climates. In many cases, a conventional external wall structure with no vapor retarder may be a
good choice for residential buildings in mixed climates. Table 6.6 provides two example external
wall structures that were considered in simulation that use vinyl siding with either plywood or
fiberboard as sheathing. Detailed moisture comparisons for these two cases in Nashville are
shown in Figure 6.9. The choice of sheathing is important here: as far as moisture is concerned,
plywood is acceptable in this case, but fiber board is not adequate because fiber board has higher
permeability.

136
Table 6.6 Parameters for conventional wall structures used in Nashville
surface film
interior

Structure
Type

surface film
exterior

Layer 1

Layer2

Layer 3

Layer 4

hi=4.2W/m2-C
Ext. Wall A:
hr=5.2W/m2-C
without vapor
Oil paint:
retarder
5 Perm

Gypsum
board:
0.5

Fiberglass
insulation:
R-13,
28.1 Perm

Plywood
Sheathing:
1

Vinyl siding hi=11.4W/m2-C

hr=5.2W/m2-C
0.5
Paint:
5 Perm

hi=4.2W/m2-C
Ext. Wall B:
hr=5.2W/m2-C
without vapor
Oil paint:
retarder
5 Perm

Gypsum
board:
0.5

Fiberglass
insulation:
R-13,
28.1 Perm

Fiber Board
Sheathing:
1

Vinyl siding hi=11.4W/m2-C

hr=5.2W/m2-C
0.5
Paint:
5 Perm

0.8

R13 insulation

0.7

Layer average RH(-)

0.6

Vinyl
Siding

Gypsum
Wallboard

0.5
0.4
Gypsum Wallboard
Plywood Sheathing
Vinyl Siding

0.3

Plywood Sheathing

0.2
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. plywood sheathing

0.9

R13 insulation

0.8

Layer average RH(-)

0.7

Gypsum
Wallboard

0.6

Vinyl
Siding

0.5
0.4

Fiber Board Sheathing

Gypsum Wallboard
Fiber Board
Vinyl Siding

0.3
0.2
1

13

17

21

25

29

Time (week #)

33

37

41

45

49

b. fiber board sheathing

Figure 6.9 Conventional wall structures equilibrium RH for Nashville (no vapor retarder).

137

6.3. Hot and Dry Cooling Climate- Phoenix, AZ

The weather in Phoenix is hot and dry: the yearly average air temperature is 72.5F, while
the average relative humidity is only 36%. Figure 6.10 depicts the weekly ambient air
temperature and RH for Phoenix. It is rare for the temperature to drop below the water freezing
point during winter, while it is common to find temperatures higher than 100 F during summer.
It is so dry that almost every week the minimum weekly RH is between 10 and 20%.
120
110
Outdoor temperature (F)

100
90
80
70
60
Max Hourly
Average Weekly
Min Hourly

50
40

ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek 9
ee
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

30

a. Ambient air temperature

1.0
0.9

Max Hourly
Average Weekly
Min Hourly

0.8
Outdoor RH (-)

0.7
0.6
0.5
0.4
0.3
0.2
0.1

ee
w k1
ee
w k3
ee
w k5
ee
w k7
ee
w k9
ee
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

0.0

b. Ambient air relative humidity

Figure 6.10 Ambient air temperature and relative humidity in Phoenix.

138
Traditional construction was employed: 0.3 inch gypsum wallboard with Kraft paper, R-11
glass fiber insulation, 0.5 inch fiberboard sheathing and 0.5 inch exterior sugar pine siding, as
listed in Table 6.7.

Table 6.7 Parameters for building structure in Phoenix.

Structure
Type

surface film
interior

Exterior
Wall

hi=4.2W/m2-C
hr=5.2W/m2-C
Latex paint:
10 Perm

Roof

surface film
exterior

Layer 1

Layer2

Layer 3

Layer 4

Layer 5

Gypsum
board:
0.3

Kraft paper
0.1

Fiberboard
sheathing: 0.5

Fiberglass
insulation:
R-11,
33.1 Perm

Sugar
pine siding
0.5

hi=5.7W/m2-C
hr=5.2W/m2-C
no paint

OSB
1

Roofing paper
0.1

Ceiling

hi=2.7 W/m2-C
hr=5.2W/m2-C
Latex paint:
10 Perm

Gypsum
board
0.5

Polyethylene Exterior grade

US R-38 insulation
vapor retarder: plywood:
10.0 Perm
0.5
1.0 perm

Indoor
Floor/
Furniture

hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Sugar Pine
2

hi=11.4W/m2-C
hr=5.2W/m2-C
Paint:
10 Perm
hi=11.4W/m2-C
hr=5.2W/m2-C
no paint

Asphalt Roofing
Shingles
0.25

hi=4.0W/m2-C
hr=5.2W/m2-C
no paint
hi=3.5W/m2-C
hr=5.2W/m2-C
Paint: 10 perm

Sizing results for the HVAC equipment is listed in Table 6.8, which shows a low heating
capacity (close to half of Minneapoliss heating capacity) and high cooling capacity (almost
double that of Minneapolis).

Table 6.8 Sizing result for the HVAC equipment in Phoenix.

Design Design
RHamb
Tamb
(F)
(-)

Season

Sensible
load
(W)

Latent
Load
(W)

Total
Capacity

Design Tin
(F)

Design
RHin (-)

Cooling

7760

3540

3.9 ton

75

60%

114.1

11.1%

Heating

23000
BTU/hr

70

27.0

48.7%

Because of the dry ambient conditions, good moisture performance can be found for both the
ceiling and roof in Phoenix, as shown in Figure 6.11 for the north facing structure (the poorest
roof orientation for this case).

139
Although Phoenix is located in a cooling climate where the vapor retarder is typically
located on the exterior side, the hot and dry conditions do not cause moisture incursions from
outside. Therefore, placement of the vapor retarder close to the interior side also gives very good
moisture performance. As shown in Figure 6.12 for the north wall construction, none of the wall
layers experienced high moisture contents. In such very hot and dry climate conditions, moisture
is not typically a problem. It is more important to select the building envelope based on building
energy performance.

0.55
OSB
Roofing paper
Asphalt Shingle

0.50

Layer average RH(-)

0.45

Asphalt shingle

0.40

Roofing Paper

0.35
0.30

OSB

0.25
0.20
0.15
0.10
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. roof

0.7
Gypsum wallboard
0.6

Layer average RH(-)

R-38 Insulation

plywood board

Plywood board

0.5

0.4

6-mil glue

0.3

Gypsum Wallboard
0.2
1

13

17

21

25

29

Time (week #)

33

37

41

45

49

b. ceiling

Figure 6.11 Roof and ceiling equilibrium RH for Phoenix.

140

0.6

Layer average RH(-)

0.5

Fiberboard Sheathing
R11 insulation

0.4

Sugar Pine
Siding

Gypsum
Wallboard

0.3
Gypsum wallboard
Kraft paper
Fiberboard Sheathing
Sugar pine siding

0.2

0.1
1

13

17

21

25

29

33

37

41

45

kraft paper

49

Time (week #)

a. Moisture

110

Layer average temperature (F)

100

Fiberboard Sheathing
R11 insulation

90
80

Sugar Pine
Siding

Gypsum
Wallboard

70
Gypsum wallboard
Kraft paper
Fiberboard Sheathing
Sugar pine siding

60
50

kraft paper

40
1

13

17

21

25 29 33
Time (week #)

37

41

45

49

b. Temperature

Figure 6.12 Wall structure internal equilibrium RH and temperature for Phoenix.

Figure 6.13 depicts the moisture information for an external wall structure which does not
employ a vapor retarder, based on the external wall structure described in Table 6.7. Almost
identical average layer moisture content can be found, which implies it is not necessary to apply a
vapor retarder in this type of climate.

141

0.6

Fiberboard Sheathing

0.5

Layer average RH(-)

R11 insulation
0.4

Gypsum
Wallboard

0.3

Sugar Pine
Siding

Gypsum wallboard
Fiberboard Sheathing
Sugar pine siding

0.2

0.1
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

Figure 6.13 Wall structure equilibrium RH and temperature for Phoenix (no vapor retarder).

6.4. Hot and Humid Cooling Climate- Houston, TX

In contrast to Phoenix, Houston is warm but humid. Houstons average year-round
temperature is 68.1F but the average year-round RH is about 75% with an annual precipitation
rate of more than 45 inches. Compared to Phoenix, the temperature in Houston is mild: highs in
the summer can reach the 90s, but the lows in winter rarely dip below freezing. The weekly
average ambient air temperature and RH are exhibited in Figure 6.14.
The same building construction used in the Phoenix case was tested for the Houston climate.
Sizing results are given in Table 6.9 Sizing result for the HVAC equipment in Houston.and reveal
that the total cooling capacity is comparable to the cooling load for Nashville. However, the
sensible load is slightly lower but the latent load is a little bit higher than Nashville because of the
wet weather. The heating capacity for Houston is smaller than for Nashville due to the mild
winter weather.

142

100

Outdoor temperature (F)

90
80
70
60
50
40
Max Hourly
Average Weekly
Min Hourly

30
20

w
ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek
ee 9
w k1
ee 1
w k1
ee 3
w k1
ee 5
w k1
ee 7
w k1
ee 9
w k2
ee 1
w k2
ee 3
w k2
ee 5
w k2
ee 7
w k2
ee 9
w k3
ee 1
w k3
ee 3
w k3
ee 5
w k3
ee 7
w k3
ee 9
w k4
ee 1
w k4
ee 3
w k4
ee 5
w k4
ee 7
w k4
ee 9
k
51

10

a. Ambient air temperature

1.0
0.9
0.8
Outdoor RH (-)

0.7
0.6
0.5
0.4
0.3

Max Hourly
Average Weekly
Min Hourly

0.2
0.1

ee
w k1
ee
w k3
ee
w k5
ee
w k7
e
w ek 9
ee
w k 11
ee
w k 13
ee
w k 15
ee
w k 17
ee
w k 19
ee
w k 21
ee
w k 23
ee
w k 25
ee
w k 27
ee
w k 29
ee
w k 31
ee
w k 33
ee
w k 35
ee
w k 37
ee
w k 39
ee
w k 41
ee
w k 43
ee
w k 45
ee
w k 47
ee
w k 49
ee
k
51

0.0

b. Ambient air relative humidity

Figure 6.14 Ambient air temperature and relative humidity in Houston.

Table 6.9 Sizing result for the HVAC equipment in Houston.

Season

Sensible
load
(W)

Latent
Load
(W)

Cooling

4550

Heating

Design Design
RHamb
Tamb
(F)
(-)

Total
Capacity

Design Tin
(F)

Design
RHin (-)

2410

2.4 ton

75

60%

93.0

52.4%

29680
BTU/hr

70

15.1

51.5%

143
The roof design does not perform as well in Houston as in other locations (see Figure 6.15a).
The OSBs equilibrium RH is high since it is exposed directly to attic air. A latex paint would be
beneficial in protecting this material from humid air. Figure 6.15b shows that the moisture
performance for the ceiling is not acceptable because moisture builds up near the indoor surface
during summer. This is due to the relatively low local water vapor pressure caused by lower local
temperature at the indoor surface and high attic water vapor pressure caused by high attic air
temperatures and relative humidities. As a result, the equilibrium RH can exceed 80% for the
plywood board, which could lead to mold and mildew growth. A vapor retarder should be
employed to separate the plywood board and the attic air in order to block the vapor incursion.

0.75

Asphalt shingle

OSB
Roofing paper
Asphalt Shingle

0.70

Roofing Paper

Layer average RH(-)

0.65
0.60

OSB

0.55
0.50
0.45
0.40
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. roof (facing north)

0.9
0.8

Layer average RH(-)

R-38 Insulation

Gypsum wallboard
plywood board

Plywood board

0.7
0.6
0.5

6-mil glue

0.4

Gypsum Wallboard

Mold growth

0.3
1

13

17

21

25

29

Time (week #)

33

37

41

45

49

b. ceiling

Figure 6.15 Roof and ceiling equilibrium RH for Houston.

144
Similar moisture performance results occur for the exterior wall structures as shown in
Figure 6.16. There is a build up of moisture within the fiberboard during the summer season that
produces conditions that are very conductive to mold and mildew growth. In this situation, the
vapor retarder should be placed outside of the insulation and some low-permeability siding
material is strongly recommended, such as vinyl siding.
Fiberboard Sheathing
0.8

R11 insulation

Layer average RH(-)

0.7

Sugar Pine
Siding

Gypsum
Wallboard

0.6
0.5
Gypsum wallboard
Kraft paper
Fiberboard Sheathing
Sugar pine siding

0.4

kraft paper

0.3
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

Figure 6.16 Moisture performance for the original wall structure (facing north) in Houston.

The building structure was modified to improve the moisture performance and the improved
building was simulated. The modifications include:
1. Roof: Latex paint (10 perm) was added on the OSB surface exposed to the attic air;
2. Ceiling: A 6-mil glue layer (1 perm) was placed between the plywood layer and
insulation, rather than between the plywood and gypsum wallboard;
3. Wall: A kraft paper vapor retarder was placed between the siding and insulation, rather
than close to the interior surface;
4. Wall: the sugar pine siding was changed to a vinyl siding.
These modifications result in significant enhancement for the building structure moisture
performance, as shown in Figure 6.17. The building material is safe from mold and mildew
growth. Therefore, in hot and humid climates, a vapor retarder should be located near the exterior
of building elements to block vapor incursion.

145

0.65
OSB
Building paper
Asphalt Shingle

Layer average RH(-)

0.60

Asphalt shingle
Roofing Paper

0.55

OSB

0.50

Latex paint
0.45

0.40
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

a. roof (facing north)

R-38 Insulation

0.8

Plywood board

Layer average RH(-)

0.7

0.6

0.5

6-mil glue

Gypsum wallboard
plywood board

0.4

Gypsum Wallboard

0.3
1

13

17

21

25

29

33

37

41

45

49

Time (week #)

b. ceiling

0.8

Fiberboard Sheathing
R11 insulation

Layer average RH(-)

0.7

Vinyl
Siding

Gypsum
Wallboard

0.6

0.5

Gypsum wallboard
Kraft paper
Fiberboard Sheathing
Sugar pine siding

0.4

kraft paper

0.3
1

13

17

21

25
29
33
Time (week #)

37

41

45

49

c. wall (facing north)

Figure 6.17 Modified building structure moisture performance for Houston.

146

6.5. Summary
This chapter presented case studies to evaluate the impact of different wall and ceiling
constructions on moisture performance for typical climates (heating climate, mixed climate, and
dry or humid cooling climate).

The whole building energy and moisture model developed in

Chapter 5 was used for this analysis and the results showed that:
1. For heating dominated climates, the vapor retarder should be placed close to the interior
surface so as to prevent vapor excursion from the relatively warm and moist indoor air;
2. For mixed climates, a conventional wall structure without a vapor retarder can work
acceptably.

A sandwiched structure with vapor retarders on both sides of a low-

permeability insulation is an option that can provide better performance, because it can
handle either vapor incursion from the ambient during summer or vapor excursion from
the indoor air during winter;
3. For dry cooling dominated climates, there are typically no moisture performance
problems and performance for common building structures is very good. The building
structure performs well even without a vapor retarder because of the dry ambient
condition;
4. For humid cooling dominated climates, a vapor retarder should be positioned close to the
exterior surface to stop vapor incursions. In addition, a surface paint could be applied to
the surface above the ceiling in the attic and low-permeability siding is recommended for
humid and rainy areas.

147

CHAPTER 7. CONCLUSIONS AND FUTURE WORK

7.1. Conclusions

Residential buildings are meant to provide a safe, healthy, and comfortable indoor
environment for occupants. However, many residential buildings suffer from a variety of
moisture problems. First of all, unfavorable indoor relative humidity (either too high or too low)
can make occupants uncomfortable. More over, excess moisture may cause building material
damage, such as discoloration of surface finishes, changing of volume (warping and shrinkage,
etc.), degradation of insulation, and corrosion of metal components. Even worse, moisture
incursion during winter time may cause freeze-thaw deterioration damage of concrete and
masonry for buildings in cold areas. In summer, mold and mildew growth may occur if the
building materials equilibrium relative humidity is higher than about 70 to 80% for a significant
period of time with favorable temperature conditions. Mold and mildew may decay the building
materials, and can cause some serious health problems for humans.
The research described in this thesis addressed modeling of building envelope transient heat
and moisture transfer for structures used in residential buildings. The wall model development
was based on a previously developed one-dimensional model called MOIST 3.0, which
incorporates water vapor diffusion and water liquid capillary transfer. An important aspect of this
research was the development of a wetted surface model that allows consideration of the moisture
transfer caused by wind-driven rain, where capillary liquid water transfer is an important
mechanism.
One of the major contributions was the experimental validation of the improved MOIST
model for heat/vapor/capillary transfer, especially for the saturation moisture content regime.
Previous work did not include a detailed validation for these conditions.

The validation

experiment involved a drying process for a sugar-pine panel that was initially soaked with water.

148
The drying process was carried with surface radiant heating and controlled air flow.

An

automated weighing system was used to trace the overall wood moisture content and moisture
pin-pairs were employed to measure the local moisture content within the test specimen. During
the transient drying test, radiant heating was projected on the one surface then switched to the
other so as to develop a better understanding of the nature and significance of solar-driven inward
vapor diffusion. There was good agreement between moisture content predictions and
measurements from the moisture pin-pair sensors and very good agreement for overall moisture
content from the weighing system.
This thesis also presented a simplified method for calculating transient slab-on-ground heat
transfer that can be incorporated within hourly simulation programs. The method assumes that
there are two primary one-dimensional paths for heat transfer from a ground-coupled floor slab:
1) one-dimensional heat transfer from the perimeter of the slab to the ambient and 2) onedimensional heat transfer between the slab interior surface and a portion of the soil beneath the
slab. The perimeter heat transfer is assumed to occur at quasi-steady state and is characterized in
terms of a perimeter heat loss factor (Fp). Transient heat transfer within the slab and ground are
modeled using a simple thermal circuit employing three nodes with an adiabatic boundary
condition at a specified depth within the soil underneath the slab. Although some previous
simulation models considered this type of two-path model, there was no validation of this
approach and no guidance for specifying perimeter heat loss factors and underfloor soil depths
and node locations for the thermal circuit. In this thesis results from detailed two-dimensional
finite-element models for typical floor constructions and soil properties were used to identify 1)
locations for nodes within the slab and soil, 2) correlations for soil depth as a function of soil
properties associated with the underfloor adiabatic boundary condition, and 3) correlations for
perimeter heat loss factor as a function of soil properties and edge insulation levels for different
constructions. Transient heat transfer results from the simple model compared well with results
from the finite-element program for different floor constructions, edge insulation, soil properties,
locations, and times of year.
Another significant contribution of the research was the coupling of the detailed envelope
model with a whole building model that will allow investigation of the impacts of envelope
design on occupant comfort, energy use, and wall material conditions that can lead to mold
growth, insulation degradation, etc. The whole building model provides the wall model with

149
more realistic and variable indoor moisture boundary conditions, which could be affected by
several factors, such as building envelope moisture transfer, indoor moisture generation, coil
moisture removal by mechanical cooling, moisture buffering by soft furnishing (bedding and
curtains), and inter-zonal air flow. In addition to the external wall model, the whole building
model incorporated several individual models, including:
1.

Weather data treatment including wind driven rain, sky temperature and solar radiation,

2.

Indoor heat and moisture generation models,

3.

HVAC equipment (moisture removal by coils and energy consumption by compressor).

4.

Simplified multi-zone air flow model to provide air flow from ambient to the zone and
between zones,

5.

Ground floor heat transfer model,

6.

Indoor moisture buffering model associated with soft furnishing,

It is believed that the whole building energy/moisture model is the most comprehensive tool
that has been developed for analyzing moisture problems in residential buildings.

Most of the

previous tools for detailed analysis of moisture transport in walls have considered individual
walls with specified indoor conditions. A case study was performed to evaluate the impact on
wall moisture performance of performing a whole building analysis as compared with a single
wall analysis. At the exterior of the wall, there is very little difference in the moisture contents
for the two approaches. However, the differences can be significant within materials that are
close to the interior surface. The stand-alone wall analysis would not identify potential problems
that could occur due to more significant indoor moisture gains that could potentially occur, such
as moisture gains within a bathroom not having an exhaust fan. The whole-building model has a
greater potential for evaluating moisture problems caused by indoor conditions and also can
present the impacts of design choices on indoor air moisture levels.
Case studies for building heat and moisture performance were performed to investigate the
moisture performance of building envelopes in featured climates (heating climate, mixed climate,
and dry or humid cooling climate) with representative building constructions. For heating
climates, the vapor retarder should be placed close to the interior surface so as to prevent vapor
excursion from the relatively warm and moist indoor air; while for humid cooling dominated
climates, a vapor retarder should be positioned close to the exterior surface to stop vapor

150
incursions. For dry cooling climates, the building structure performs well even without a vapor
retarder because of the dry ambient condition. For mixed climates, a conventional wall structure
without a vapor retarder can work adequately. However, a sandwiched structure with vapor
retarders on both sides of a low-permeability insulation is an option that can provide better
performance, because it can handle either vapor incursion from the ambient during summer or
vapor excursion from the indoor air during winter.

7.2. Challenges for Future Studies

For the research of residential building heat and moisture transfer, some challenges remain
unsolved for this project at the current stage. This section briefly outlines some of these
challenges, which could be the basis for future studies.
Local surface condensation and humidity problems
Condensation may happen during the winter on surfaces inside of the building when their
temperature drops below the indoor air dew point. This is particularly a problem in rooms having
large internal moisture gains, such as bathrooms and kitchens, and can occur on exterior walls,
windows, concrete floors, etc. Local surface condensation problems can be eliminated through
air change exchange with drier parts of the building, local heating, or better insulation for exterior
elements. Even in the absence of indoor condensation effects, local variations in humidity within
a residence can influence occupant comfort and moisture buildup within materials.
Modifications to the model are necessary to consider local internal surface condensation and
humidity effects and their remedies. The model needs to be adapted to handle multiple air zones
within each floor (i.e., eliminate the fully-mixed assumption).

All of the model elements

necessary to accomplish this goal are included in the current building model. However, a
significant amount of additional input data and computational effort is required to consider
multiple zones for each floor. It is recommended that future work be carried out to modify the
existing model to include multiple zones. This work should include an investigation of numerical
methods with the goal of improving computational efficiency.

151
Two-dimensional effects
In many cases, two-dimensional or even three-dimensional heat and moisture effects may be
important to consider for building structures. Some situations where two-dimensional effects
might be important include shading effects for solar radiation and wind-driven-rain because of the
existence of eaves, water capillary incursion from the ground, and water condensation on the
interior surface of windows and window frames.

All of these situations result in non-uniform

boundary conditions for the heat and moisture transfer for the building structure. Futhermore, the
building structure is made of wall studs, which make the heat and moisture transfer multidimensional. Therefore, it is important to study the circumstances where two-dimensional effects
might be important to consider in wall structures.
Diffuse air movement through porous building structure
Indoor or ambient air may pass through porous building materials such as fiberglass or
uncoated concrete blocks when there is significant air pressure difference. This is an alternative
path for moisture incursion into building materials that is not considered in the current model.
Future work should modification of the whole building model to include these effects.
Parametric studies could also be performed to evaluate situations where air diffusion through
porous structures is detrimental or beneficial.
Effect of snow for the attic heat/moisture performance
Snow accumulation on roofs during winter is common in the northern part of the U.S.A. The
effect of snow accumulation on attic temperatures and humidities is not known and could be the
subject of future studies. Under some certain conditions, snow accumulator might be beneficial
since it can act as an insulation layer when the ambient temperature is below the freezing point.
Furthermore, it can reduce night long wave radiation. On the other hand, snow melting can cause
problems under certain conditions. For example, with inadequate attic insulation and ventilation
a warm attic could cause snow melting under conditions where the ambient is above freezing.
The melted water flows down the roof, under the blanket of snow, onto the eaves and into the
gutter where it then can refreeze. Eventually, ice accumulates along the eave and in the gutter,
creating an ice dam that doesnt allow additional snow melt to drain properly. The existing
model could be modified to allow investigation of these effects

152
Effect of the attic ventilation on attic heat/moisture performance
Extreme temperatures and humidities can occur within an attic because it is not conditioned.
In particular, high attic air relative humidity combined with high temperature during summer can
cause moisture to migrate into roofs and ceiling. Ventilation or desired infiltration is one effective
way to improve the heat and moisture performance of an attic. Some controlled mechanical
ventilation system with thermostats (e.g., a ventilation fan that turns on when the attic
temperature reaches 125F) may work well, but they require electrical energy to operate and add
cost. A vented attic usually has fixed gables for infiltration and in many cases this natural
ventilation may be sufficient during summer. Although natural air ventilation may provide
adequate heat and moisture removal under favorable ambient condition, there are situations where
natural ventilation heats and humidifies that attic. Two plausible alternatives for a better system
that could be studied using the whole building modeling tool:
1. Attic ventilation system powered by a photovoltaic panel and controlled based on
temperature and humidity,
2. A thermally activated roof ventilation damper that opens in response to heating from the
sun and ambient.
Importance of considering wall moisture transport on indoor humidity levels
Vapor flux from the building envelope to the indoor air has some influence on indoor
humidity levels. However, typical building energy simulation tools do not include wall moisture
transport and the computation cost for whole building heat/moisture simulation (including
moisture transfer from building envelope) is high. It would be useful to use the simulation tool to
evaluate situations where it is important to consider moisture transport in walls for determining
indoor humidity levels, and situations where these effects can be neglected if the primary goals
are energy analysis and indoor comfort conditions.
Construction defects and smart building materials
There is a general opinion that moisture problems arise less because of vapor transfer due to
appropriate vapor retarder design and more because of bad construction details that allow bulk
water to enter the structure. Examples include poor installation of vapor barriers and unwanted
water and moisture infiltration from rain showers and underground water. Therefore, better
drying properties are necessary for the building envelope assemblies. One possibility would be to
employ smart vapor retarders, which have low permeability when the structure is dry and high

153
permeability for wet conditions. There are some commercial smart vapor retarders available on
the market already, such as MemBrain SMART Vapor Retarder made by Certainteed. This
product reacts to relative humidity by altering pore size, and results in less than 1 perm when the
air is dry and more than 25 perm when the RH exceeds 95%. Future work could investigate the
performance of combinations of construction defects with the use of smart vapor retarders.

LIST OF REFERENCES

154

LIST OF REFERENCES

Adjali M.H., M. Davies, S.W. Rees and J. Litter, 2000, Temperature in and under a slab-onground floor: two- and three dimensional numerical simulations and comparison with experiment
data, Building and Environment, vol.35, pp.655-662.
Al-Anzi A., and M. Krarti, 1997, Evaluation of the Thermal Bridging Effects on the Thermal
Performance of Slab-on-Grade Floor Foundation, vol.103 part 1.
Andreas Holm and Hartwig M. Knzel, 1999, Combined Effect of Temperature and Humidity on
the Deterioration Process of Insulation Materials in ETICS, Building Physics in the Nordic
Countries, August 1999, Gothenburg.
Aoki-Kramer, M. and Karagiozis, A.N., 2004, A new look at residential interior environment
loads, ASHRAE Thermal IX Conference Clearwater Beach, FL, December, 2004.
ASHRAE, 2005, ASHRAE Handbook - Fundamentals (2005), Atlanta: American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Inc.
ASTM, 1992, ASTM D 4442-92, Standard Test Method for Direct Moisture Content
Measurement of Wood and Wood-Based-Material, West Conshohocken, PA: American Society
of Testing and Materials.
Axley, J.W., 1990, Adsorption Modeling for Macroscopic Contaminant Dispersal Analysis,
NIST-GCR-90-573, National Institute of Standards and Technology.
Balcomb J.D., and R.S. Crowder (NREL), 1995, ENERGY-10, A Design Tool Computer
Program for Buildings, Proc. Solar '95 (20th Passive Solar Conference), American Solar Energy
Society, July 15-20, 1995, Minneapolis, MN.
Bholah, R. and A.H. Subratty, 2003, Humid Building Problems in Mauritius, Indoor Built
Environ, vol.12, pp.221-225.
Braun, J.E., 2003, Load Control using Building Thermal Mass, ASME Journal of Solar Energy
Engineering, vol.125, pp.292-301.

155
Braun, J.E. and Z. Zhong, 2005, Development and Evaluation of a Night Ventilation Precooling
Algorithm, International Journal of Heating, Ventilating, Air-Conditioning and Refrigeration
Research, 2005.
Brandemuehl, M.J., Gabel, S., and Andresen, I. 1993. HVAC2 Toolkit: Algorithms and
Subroutines for Secondary HVAC System Energy Calculations. American Society of Heating,
Refrigerating, and Air Conditioning Engineers, Inc., Atlanta, GA.
Burch, D. M., W.C. Thomas and A.H. Fanney, 1992, Water Vapor Permeability Measurements of
Common Building Materials, ASHRAE Transactions, vol.98, part 2. pp.486-494.
Burch D.M. and C.A. Saunders, 19951, A Computer Analysis of Wall Constructions in the
Moisture Control Handbook, NISTIR 5627.
Burch D.M., C.A. Saunders, and A. TenWolde, 19952, Manufactured Housing Walls that Provide
Satisfactory Moisture Performance in all Climates, NISTIR 5558.
Burch D.M. and A.O. Desjarlais, 1995, Water-Vapor Measurement of Low-Slope Roofing
Materials, NISTIR5681.
Burch D.M., G.A. Tsongas, and G.N. Walton, 1996, Mathematical Analysis of Practices to
Control Moisture in the Roof Cavities of Manufactured Houses, NISTIR 5880.
Burch, D.M. and Chi, J., 1997, MoistA PC Program for Predicting Heat and Moisture Transfer
in Building Envelopes Release 3.0, NIST Special Publication 917.
Burch, D.M. and J. Choi, 1997, MOIST- A PC Program for Predicting Heat and Moisture
Transfer in Building Envelopes, NIST Special Publication 917.
Carmody J., J. Christian and K. Labs, 1991, Builders foundation hand book, ORNL/CON-295,
ORNL Report Sub/86-72143/1, Oak Ridge National Laboratory.
Chuangchid P., and M. Krarti, 2000, Economic Analysis of Foundation Heat Gain for
Refrigerated Warehouses, ASHRAE Transaction, vol.106, part 2.
Chuangchid P., and M. Krarti, 2000, Parametric Analysis and Development of a Design Tool for
Foundation Heat Gain for Coolers, ASHRAE Transaction, vol.106, part 2.
Claesson J. and C.E. Hagentoft, 1991, Heat Loss to the Ground from Building - I. General
Theory, Building and Environment, vol.26 no.2, pp.195-208.

156
Claridge D.E., M. Krarti, and J.F. Kreider, 1993, Energy Calculations for Basements, Slabs, and
Crawl Spaces, Final Report, ASHRAE 666-TR, Atlanta: American Society of Heating,
Refrigerating and Air-Conditioning Engineers, Inc.
Clark G. and Allen C., 1978, The Estimation of Atmospheric Radiation for Clear and Cloudy
Skies, Proc. 2nd National Passive Solar Conference (AS/ISES), pp.675-678.
Clarkin, Michael E. and Brennan, Terry M., 1998, Stack-Driven Moisture Problems in a MultiFamily Residential Building, ASHREA 1998 winter meeting, TO-98-25-3.
Conte, S.D., de Boor, C. 1972. Elementary Numerical Analysis, McGraw-Hill, New York, NY.
Deru M., 2003, A Model for Ground-Coupled Heat and Moisture Transfer from Buildings,
NREL/TP-550-33954, National Renewable Energy Laboratory Technical Report;
Dhatt, G., Touzot, G., Catin, G. 1984. The Finite Element Method Displayed, John Wiley &
Sons, New York.
DIN 4108-2: 2001-03: Heat insulation and energy saving in buildings Part 2: Minimum
requirements on heat insulation, German standard DIN (Deutsches Institut fr Normung).
DIN EN ISO 13788: 2001-11: Calculation of surface temperature to avoid critical surface
humidity and calculation of condensation within building parts, German standard DIN (Deutsches
Institut fr Normung).
DIN 68800: 1996-05: Wood protection Part 2: Preventive building measures in building
construction, German standard DIN (Deutsches Institut fr Normung).
Dols W.S. and Walton, G.N., 2002, CONTAMW 2.0 User Manual, Multizone Airflow and
Contaminant Transport Analysis Software, NISTIR 6912.
Duff, J.E., 1968, Moisture Distribution in Wood-frame Walls in Winter, Forest Products Journal,
vol. 18, No.1, pp.60-64.
Duffie, J.A and Beckman, W.A., 1980, Solar Engineering of Thermal Process, John Wiley &
Sons, ISBN 0-471-05066-0.
Emmerich, S.J., Persily, A.K., and Nabinger. S.J., 2002, Modeling moisture in residential
building with a multizone IAQ program, Proceedings of Indoor Air 2002, Monterey, California,
2002.
European Committee for Standardization, 1998, Thermal Performance of Buildings Heat
Transfer Via the Ground Calculation Methods, European Standard EN ISO 13370.

157
F-Chart Software, 2006, EES Engineering Equation Solver, Madison, WI, USA.
F-Chart Software, 2005, FEHT A Finite Element Analysis Program, Madison, WI, USA.
Fagerlund G., 1977, The Critical Degree of Saturation Method Assessing the Freeze/Thaw
Resistance of Concrete, Materials and Structures, vol.58, pp.217-229.
Fang L., G. Clausen and P.O. Fanger, 2000, Temperature and Humidity: Important Factors for
Perception of Air Quality and for Ventilation Requirements, ASHRAE Transaction, vol.106, part
2.
Fanger, P.O., 1972, Thermal comfort, New York: McGraw-Hill.
Gawandea N.A., D.R. Reinharta, P.A.Thomasa, P.T. McCreanorb, T.G. Townsendc, 2003,
Municipal Solid Waste in Situ Moisture content measurement using an electrical resistance
sensor, Waste Management, vol.23, pp.667674.
Haverinen U., M.Vahteristo, D. Moschandreas, A. Nevalainen, T. Husman and J.Pekkanen, 2003,
Knowledge-based and statistically modeled relationships between residential moisture damage
and occupant reported health symptoms, Atmospheric Environment, vol.27. pp.577-585.
Healy M.H., 2003, Moisture Sensor Technology-A Summary of Techniques for Measuring
Moisture Levels in Building Envelopes, ASHRAE Transactions, vol.109, part 1.
Henk J.L., and G.J. van Gelder, 2002, In Situ Measurement of Ground Thermal Conductivity: A
Dutch Perspective, ASHRAE Transactions vol.108, part 1.
Herrlin, M.K. 1992. Air-flow studies in multizone buildings models and applications, Royal
Institute of Technology.
Jacques Rousseau, 1982, Rain Penetration and Moisture Damage in Residential Construction,
Building Science Insight, National Research Council Canada / Institute for Research in
Construction (NRC-IRC), HTTP://IRC.NRC-CNRC.GC.CA/BSI/83-1.CMHC .
Kaariainen, M. Rudolph, D. Schaurich, K. Tulla, H. Wiggenhauser, 2001, Moisture Measure in
Building Materials with Building Materials with Microwaves, H., NDT&E International, vol.34,
pp. 389-394.
Karagiozis, A., Knzel, H.M., Holm A., 2001, WUFI-ORNL/IBP - A North American
Hygrothermal Model. Contribution to "Performance of Exterior Envelopes of Whole Buildings
VIII", Dec. 2-7 2001, Clearwater Beach, Florida.

158
Klaus, Sedlbauer, 2000, Prediction of mould fungus formation on the surface of and inside
building components, Ph.D. Thesis of Fraunhofer Institute for Building Physics.
Klemma A.J., P. Klemmb, K. Rozniakowskib and G. Galbraitha, 2002, Non-contact methods of
measuring moisture concentration in external layers of building partitions. IThe in)uence of
geometrical microstructure on the kinetics of moisture condensation on glass surfaces, Building
and Environment, vol.37, pp.1215-1220.
Klemma A.J., P. Klemmb, K. Rozniakowskib and T.W. Wojtatowicz, 2002, Non-contact
methods of measuring moisture concentration in external layers of building partitions. II
Monitoring of the water vapour condensation on porous surfaces, Building and Environment,
vol.37, pp.1221-1232.
Klemma A.J., P. Klemmb, K. Rozniakowskib and T.W. Wojtatowicz, 2002, Non-contact methods
of measuring moisture concentration in external layers of building partitions. IIIWater vapor
condensation on the cementitious surfaces, Building and Environment, vol.37, pp.1233-1240.
Krarti M., D.E. Claridge, and J.F. Kreider, 1988, ITPE Method Applications to Time-varying
Two-dimensional Ground-coupling Problems, Int. J. Heat Mass Transfer, vol.31, pp.1899-1911.
Krarti M., and S. Choi, 1996, Simplified Method for Foundation Heat Loss Calculation,
ASHRAE Transaction, vol.102, part 1.
Kumaran, M.K. et al., 2004, An Integrated methodology to develop moisture management
strategies for exterior wall systems, NRCC-45987.
Kumaran, M.K. et al., 2002, A Methodology to develop moisture management strategies for
wood-frame walls in North America: application to stucco-clad walls, NRCC-45694.
Knzel H.M., 1995, Simultaneous heat and moisture transport in building componentsOne- and
two-dimensional calculation using simple parameters, IRB Verlag Suttgart 1995, ISBN 3-81674103-7.
Knzel H.M. and K. Kiebl, 1996. Drying of Brick Walls after Impregnation, Bauinstandsetzen Ed
2, pp.87-100;
Labs K., J. Carmody, R. Sterling, L. Shen, Y. Huang, and D. Parker, 1988, Building Foundation
Design Handbook, ORNL Report Sub/86-72143/1, Oak Ridge National Laboratory.
Lacy, R.E., 1977, Climate and Building in Britain, Her Majestys Stationery Office, London, UK;
Lappalainen, S. et al., 2001, Evaluation of priorities for repairing in moisture-damaged school
buildings in Finland, Building and Environment, vol.36, pp.981-986.

159
Sullivan, R., 1998, Validation studies of the DOE-2 building energy simulation program, LBNL42241.;
Loureno, P.B. et al, 2006, Defects and moisture problems in buildings from historical city
centres: a case study in Portugal, Building and Environment, vol.41, Issue 2, pp.223-234.
Lstiburek, J. and J. Carmody, 1991, Moisture Control Handbook: New, Low-rise, Residential
Construction, Prepared for the U.S. Department of Energy Conservation and Renewable Energy.
Maref, W. et al., 2002, Benchmarking of the advanced hygrothermal model hygIRC with mid
scale experiments, NRCC-43970;
Maref, W. et al., 2004, An Advanced hygrothermal design tool 1-D hygIRC, NRCC-46902.
Maref, W. et al., 2004, Large-scale laboratory measurements and benchmarking of an advanced
hygrothermal model, NRCC-46784.
Maref W., M. A. Lacasse, D. Booth, M. Nicholls and T. O'Connor, 2002, Automated weighing
and moisture sensor system to assess the hygrothermal response of wood sheathing and combined
membrane-sheathing wall components, 11th Symposium for Building Physics Proceedings,
Dresden, Germany, Sept. 26-30, 2002, pp.595-604.
McDowell, Timothy P.; Emmerich, Steven; Thornton, Jeff W.; Walton, George, 2003, Integration
of airflow and energy simulation using CONTAM and TRNSYS, ASHRAE Transactions,
vol.109, No.2, pp.757-770.
Medved S. and B. Cerne , 2002, A simplified method for calculating heat losses to the ground
according to the EN ISO 13370 standard, vol.34, pp.523-528.
Mendes, N. et al., 1999, UMIDUS: a pc program for the prediction of heat and moisture transfer
in porous building elements, Building Simulation Conference IBPSA 99, pp.277-283.
Mendes, N. etc., 2003I, Moisture effects on conduction loads, Energy and Buildings, vol.35
pp.631-644.
Mendes, N. etc., 2003II, DOMUS 2.0: A Whole-Building Hygrothermal Simulation program,
Eighth International IBPSA Conference, Eindhoven, Netherlands, 2003.
Merrill, J.L. and TenWolde A., 1989, Overview of moisture-related damage in one group of
Wisconsin manufactured homes. ASHRAE Transactions, vol.95, part 1, pp.405-414.
MEWS report: Task2, 2002, Report from Task 2 of MEWS Project Description of 17 Large Scale
Wall Specimens Built for Water Entry Investigation in IRC Dynamic Wall Testing Facility, IRCRR-111.

160
MEWS report: Task3, 2002, Summary Report from Task 3 of MEWS Project at the Institute for
Research in Construction Hygrothermal Properties of Several Building Materials, IRC-RR-110.
MEWS report: Task4, 2002, Report from Task 4 of MEWS Project Task 4- Environmental
Conditions Final Report, IRC-RR-113.
MEWS report: Task6, 2003, Report from Task 6 of MEWS Project Experimental Assessment of
Water Penetration and Entry into Wood-Frame Wall Specimens Final Report, IRC-RR-133.
MEWS report: Task7, 2003, Report from Task 6 of MEWS Project Performance of Wall Systems
as a Function of Climate, Material Properties, etc. Through Mathematical Modeling, IRC-RR132.
MEWS report: Task8, 2002, Report from Task 8 of MEWS Project MEWS Methodology for
Developing Moisture Management Strategies Application to Stucco-Clad Wood-Frame Walls
in North America, IRC-RR-112.
Musser, A. 2001. An analysis of combined CFD and multi-zone IAQ model assembly issues,
ASHRAE Transactions, vol.107, part 1.
Newport Partners Report, 2004, Building moisture and durability- Past, Present and Future work,
Report Prepared for: U.S. Department of Housing and Urban Development.
Mukhopadhyaya, P.; M.K. Kumaran,; D. van Reenen, 2004, Vapor Barrier and Moisture
Response of Wood-frame Stucco Wall - Results from Hygrothermal Simulation, NRCC-46864.
Parker, D.S., 2005, Literature Review of Attic Ventilation Impacts on Florida homes for the
Florida Department of Community Affairs, Florida Solar Energy Center, FSEC-CR-1496-05.
Pedersen, C.R., 1990, PhD Thesis: Combined heat and moisture transfer in building
constructions, Thermal Insulation Laboratory, Technical University of Denmark.
Persily, A.K., 1998, A modeling study of ventilation, IAQ and energy impacts of residential
mechanical ventilation, NISTIR 6162, Gaithersburg, MD.
Popiel C.O., J. Wojtkowiak and B. Biernacka, 2001, Measurement of Temperature Distribution in
Ground, Experimental thermal and fluid science, vol.25, pp.301-309.
Richards, R. F., Burch, D. M. and Thomas, W.C., 1992, Water Vapor Sorption Measurements of
Common Building Materials, ASHRAE Transactions, vol.98, part 2.

161
Rengin Unver, N.Y. Akdag, G.Z. Gedik, L.D. Ozturk, and Z. Karabiber, 2004, Prediction of
building envelope performance in the design stage: an application for office buildings, Building
and Environment, vol.39, pp.143-152.
Restivo, A. 1979. Turbulent flow in ventilated rooms. University of London, Mechanical
Engineering Dept.
Stuart Dols, W. and Walton, G., 2002, CONTAMW 2.0 user manual, Gaithersburg, MD, NISTIR
6921.
Rousseau, M.Z., 2003, Heat, Air and Moisture Strategies for Managing Condensation in Walls,
National Research Council Canada NRCC-46734.
Rousseau, M.Z., 2004, Moisture Management in Exterior Walls: Case Study, National Research
Council Canada NRCC-47024.
Sasic Kalagasidis, A., 2003, The Whole Model Validation for HAM-Tools. Case Study: HygroThermal Conditions in the Cold Attic under Different Ventilation Regimes and Different
Insulating Materials, Report R: 03-6, Department of Building Technology, Chalmers University
of Technology, Gothenburg, Sweden, 2003.
Seem J.E., S.A. Klein, W.A. Beckman and J.W. Mitchell, 1989, Transfer Function for Efficient
Calculation of Multidimensional Transient Heat Transfer, ASME Journal of Heat Transfer,
vol.111, pp.5-12.
Shoemaker, R.C. and D.E. House, 2004, A Time-series Study of Sick Building Syndrome:
Chronic, Biotoxin-Associate Illness from Exposure to Water-damaged Buildings,
Neurotoxicology and Teratology, vol.27, pp.29-46.
Simpson, W. and Tenwolde, A., 1999, Wood Handbook: wood as an engineering material, Chap.
3: Physical Properties and Moisture Relations of Wood, FS-FPL-4714. USDA Forest Service,
Forest Products Laboratory, Madison, WI.
Sipes J.M. and M.H. Hosni, 2000, Computer Analysis and Comparison of Experimental Data for
Moisture Accumulation in Concrete Masonry Walls, ASHRAE Transaction, vol.106, part 2.
Solar Energy Laboratory (SEL), 2000, TRNSYS 15 A Transient System Simulation Program,
University of Wisconsin Madison, USA.
Straube, J.F., 2002, Moisture in Buildings, ASHRAE Journal, Jan 2002, vol.44, No.1, pp.15-19.
Straube, J.F. and E.P. Burnett, 2000, simplified prediction of driving rain deposition, proceedings
of International Building Physics Conference, Eindhoven, 2000.

162
Svennberg, K. and I. Segerholm, 2006, Performance of edge sealing systems used in moisture
transport experiments, Research in Building and Building engineering, Taylor and Francis Group,
London, ISNB 0-145-41675-2, pp.193-199.
Svennberg, K., 2006, Moisture Buffering in the Indoor Environment, Doctoral Thesis, Lund
University, Sweden.
Tenwolde A., McNatt J.D. and Krahn L., 1988, thermal properties of wood and wood panel
products for use in buildings, ORNL/Sub/87-21697/1.
TenWolde A. and W.B. Rose, 1994, Moisture Control Strategies for the Building, Envelope,
Wood Design Focus, vol.5, No.4, pp.7-10.
Terhi Yli-Pirila, Jaana Kusnetsov, 2004, Amoebae and other protozoa in material samples from
moisture-damaged buildings, Environmental Research, vol.96, pp.250-256.
Toftum J., A.S. Jorgensen and P.O. Fanger, 19981, Upper Limits for Indoor Air Humidity to
Avoid Uncomfortably Humid Skin, Energy and Buildings, vol.28, pp.1-13.
Toftum J., A.S. Jorgensen and P.O. Fanger, 19982, Upper Limits of Air Humidity for Preventing
Warm Respiratory Discomfort, Energy and Buildings, vol.28, pp.15-23.
Tsongas G.A., D.M. Burch, Carolyn Roos, and M. Cunningham, 1995, Parametric Study of Wall
Moisture Contents Using a Revised Variable Indoor relative humidity Version of the MOIST
Transient Heat and Moisture Transfer Model, Thermal Performance of the Exterior Envelopes of
Building VI Conference Proceedings, Dec 1995. Clearwater Beach, FL.
Tsongas G.A., B.A. Thornton, D.M. Burch and G.N. Walton, 1996, A Computer Analysis of the
Moisture Performance of Roof Constructions in the U.S. DOE Moisture Control Handbook,
NISTIR 5919.
Tsongas G., 2000, Case Studies of Moisture Problems in Residences, Proceeding of the 2nd
Annual Conference on Durability and Disaster Mitigation in Wood-Frame Housing, November
2000, Madison, Wisconsin.
U.S. Department of Housing and Urban Development, 1994, Manufactured Home Construction
and safety Standards, Code for Federal Regulations, Title 24, Part 3280.
Van der Wel, G.K. and Adan, O.C.G., 1999, Moisture in organic coatings a review, Progress in
Organic Coatings, vol.37, No. 1, pp.1-14.
Walton G.N., 1983, Thermal Analysis Research Program Reference Manual, NBSIR 83-2655.

163
Walton, G.N. 1989. Airflow network models for element-based building airflow modeling,
ASHRAE Transactions, 89-6-5, pp.611-620.
Yik, F.W.H., P.S.K. Sat and J.L. Niu, Moisture generation through Chinese household actives,
Indoor Built Environ, vol.15, pp.115-131.
Yuen, S.T.S., McMahon, T.A., Styles, J.R., 2000, Monitoring In Situ Moisture Content of
Municipal Solid Waste Landfills, ASCE Journal of Environmental Engineering, vol.126,
pp.10881095.
Zarr R.R., D.M. Burch and A.H. Fanney, 1995, Heat and Moisture Transfer in Wood-based Wall
construction: Measured versus Predicted, NIST Building Science Ser173, Gaithersburg, MD.
Zhai, Zhiqiang and Chen, Qingyan, 2006, Sensitivity analysis and application guides for
integrated building energy and CFD Simulation, Energy and Buildings, vol.38, No. 9, pp.10601068.
Zhang, J.S., 2005, Combined Heat, Air, Moisture, and Pollutants Transport in Building
Environmental Systems, JSME International Journal Series B, vol. 48, No. 2, pp.182-190.

APPENDIX

164

APPENDIX . INTRODUCTION OF CONTAMW

It is very important to know the air flow pattern and the moisture dispersion mechanism for
the study of building heat and moister simulation. When detailed analysis is needed, CFD
(computational fluid dynamics) models are available. Such models divide a single zone into
numerous small cells. By solving discretized mass, momentum, energy and species conservation
equations, CFD generates detailed temperature, concentration and airflow fields in a zone. The
CFD approach is very powerful, and currently many successful CFD software tools like
FLUENT, POENICS are available to study the detailed air flow inside buildings. However, this
level of detail has traditionally come with high cost: the CFD software is expensive to obtain, and
the computing consumes considerable computational resources, and a fairly sophisticated user is
needed.
On the other hand, many multi-zone air flow models that can provide designers gross
information regarding indoor air quality in a building have been developed. The multi-zone
models simulate airflow patterns from one zone to another in an entire building with only coarse
structural information and assume uniform air parameters in a zone. These models have been
used in the past 20 years primarily to support air quality and ventilation research efforts, but more
recently their practical application has increased. For example, they have been used to design
stairwell pressurization systems for smoke control, in indoor air quality forensic investigations,
and to understand airflows and pressures in complex building systems. It is anticipated that these
models will become increasing useful in the design of buildings and ventilation systems in
response to more demanding ventilation, energy efficiency and indoor air quality requirements.
Compared to CFD, multi-zone models can provide fast estimation for gross information with
acceptable accuracy (Musser, 2001).
Many multi-zone models were developed as in-house research tools, while only a few have
been made available to the public. CONTAM (Stuart Dols, et al. 2000) is among these that are
available or public access and has become increasingly popular. CONTAM takes a macroscopic

165
view of air flow and contaminant migration, which enable the analysis of whole building airflow
patterns and airborne contamination levels. CONTAM includes graphical input and output
interfaces. It also has the capacity to allow the user to specify schedules for occupants,
contaminants, weather conditions, and HVAC system operation.

A.1. Background on the Multi-zone Model CONTAM

CONTAM is a multi-zone indoor air quality and ventilation analysis computer program
designed to determine general airflow patterns, contaminant concentrations and personal
exposure if needed. The model can be applied to a variety of applications, such as assessing the
adequacy of ventilation rates in a building and obtaining the distribution of ventilated air
throughout a building. It can also be used to predict contaminant concentrations so as to
determine the indoor air quality performance for a building in its design stage or for an existing
building, and to evaluate the impacts of various design decisions related to the ventilation system
construction. Predicted contaminant concentration can also be used to estimate personal exposure
based on occupancy patterns within the building.
The perfectly mixed zone approximation and the lack of dependence of airflow on
contaminant properties can limit the applicability of multi-zone models for some types of
problems. Water vapor is an example of a non-trace contaminant whose properties might
influence building airflow in some situations. The more common problematic case, however, is
that of the imperfectly mixed room. Two examples of imperfect mixing are displacement
ventilation systems and thermally stratified spaces. For these types of zones, contaminant transfer
through an opening will be to a large degree determined by the location of the opening in the
room and may differ significantly from that predicted by the multi-zone model.
The multi-zone network airflow model approach has been extensively evaluated using both
analytical solutions and experimental data. The past validation work and comparison of the model
predictions with tracer gas measurements taken in a five-story building has been done by
researchers where it was found the predicted tracer gas concentrations by CONTAM were within
approximately 10% of true values. Deviations of this order of magnitude are commonplace in
infiltration and contaminant prediction and have been deemed acceptable for many types of
analysis (Musser, 2001).

166

A.2. Building Representation in CONTAM

To represent a building using the CONTAM model, the building must be simplified into a
set of zones that represent individual rooms and spaces. CONTAM provides a macroscopic model
for a building, that is, zones are treated as perfectly mixed volumes in the simulation. Figure A.1
illustrates the process of building idealization and its working input interface of CONTAM 2.1. In
many applications, each room of a building can be represented as a single zone. These zones are
connected to one another or to ambient by airflow paths that represent the cracks, openings, fans,
etc. The temperatures of each zone and weather condition need to be specified in order to perform
simulation. Optional inputs include contaminant sources and sinks, occupant schedules, and
HVAC systems with ducts, filters, and recirculation.

Figure A.1 Illustration of building idealization in CONTAM.

CONTAM requires the use of various assumptions in order to implement mathematical

relationships to model airflow and contaminant dispersion. Firstly, each zone representing a
certain building space is treated as a single node, where the assumption of uniform (well-mixed)
condition is employed. The uniform assumption treats zone temperature, pressure and
contaminant concentrations as single values whereas the localized effects within a given zone are
overlooked. Secondly, CONTAM does not treat heat transfer automatically. Users are responsible
to manually set the temperatures in all zones. The model can only determine airflows induced by
temperature differences between zones including ambient caused by stack effect. Moreover,
empirical nonlinear mathematical models are utilized to represent the airflow paths to which the

167
pressure drop relates. Other major assumptions include quasi-steady airflows, trace contaminants
and source/sink models, etc.
A variety of air movement models have been developed for estimating airflows in buildings.
These flows include infiltration, natural ventilation, inter-room airflows through various openings
including doorways and flows through the HVAC system (Walton, 1989). Infiltration is the result
of air flowing through openings. These opening could be large or small, intentional or accidental
in the building envelope. Infiltration is driven by pressure difference, p, across the opening. The
relationship between the airflow through an opening in the building envelope and the pressure
difference across it can be modeled in several ways in CONTAM.
It is assumed that Bernoullis equation governs the flow within each airflow element. So the
pressure different between zone j and k is:

V j2
V 2
P = Pj +
Pk + k + g ( z j zk )

2
2

(A.1)

In CONTAM, the pressure terms are rearranged and a possible wind pressure term for a
building envelope opening is added:

P = Pj Pk + Ps + Pw

(A.2)

where Ps is pressure difference due to density and elevation differences:

Ps =

j + k
2

g H

(A.3)

and wind pressure difference Pw is defined in the following equation:

Pw =

VH2
Wp
2

(A.4)

where Wp is wind pressure profile around the building.

Most infiltration models are based on the power law relationship between the flow and the
pressure difference across a crack or an opening in the building envelope. In CONTAM, three
basic variations of power law relationship for turbulent flow are included:
Q = Cq (P) n

(A.5)

168
F = C f (P) n

Q = Cd A

(A.6)

2P

(A.7)

Theoretically, the value of the flow exponent constant should lie between 0.5 and 1.0. A
variety of researches indicate that n=0.5 characterizes large openings well, while n=0.65 can be
used to describe crack-like openings.
Besides the power law flow elements, CONTAM also enables calculation based on a
quadratic relationship ( P = AQ + BQ 2 , for Q, P > 0 and P = AQ BQ 2 , for Q, P < 0 ).
It was indicated that infiltration openings could be more accurately modeled by a quadratic
relationship. Duct flow is treated based on 1997 ASHRAE Handbook of Fundamentals, and
CONTAM also includes special treatment of large openings for possible two-way flow. In the
present study, only one-way flow through the openings was consider, most of which can be
represented by power law relationship. Therefore, for the current stage of the coupling, only
power law airflow paths are considered.
To account temperature dependence, a correction factor to the base condition is used,
according to the following formulae for computing air density () and dynamic viscosity ():
= P / (RT)

(A.8)

=3.714310-6 + 4.928610-8T /

(A.9)

where R=287.055J/kgK is the gas constant of air. The base conditions refers to standard
atmospheric pressure and 20C, where 0 = 1.2041 kg/m3 and 0 = 1.508310-5 m2/s.

A.3. Solution Methods

CONTAM computes the infiltration and ventilation rates in a building by solving a system
of non-linear equations for all zones. An iterative method can be used in which a linear system of
equation is solved in each step of the process. The Newton-Raphson method is often used for this
kind of problem, which will be talked about in detail later.

169
An airflow network consists basically of a set of pressure nodes (zones) connected by links,
which are called airflow elements in CONTAM. The zones may represent rooms, connection
points in ductwork, or the ambient environment. The airflow elements correspond to discrete
airflow passages such as doorways, construction cracks, ducts, fans, and other openings. The
airflow rate from zone j to zone k, Fj,k [kg/s], is some function (f) of the pressure drop along the
flow path, Pj Pk:
Fk,j = f (Pj Pk)

(A.10)

The mass of air, mk [kg], in zone k is given by the ideal gas law:
mk = kVk = PkVk / (RTk)

(A.11)

For a transient solution the principle of conservation of mass states that

mk
= Fj ,k + Fk
t
j

(A.12)

where:
Fj,k = airflow rate [kg/s] between zones j and zone k: positive values indicate flows from j to k and

negative values indicate flows from k to j, and

Fk = non-flow processes (sources or sinks) that could add or remove significant quantities of air

from the zone.

Sources and sinks are not considered in CONTAM and flows are evaluated by assuming
quasi-steady conditions, mk /t = 0, which leads to:

j ,k

=0

(A.13)

The steady-state airflow analysis of multiple zones requires simultaneous solution of Eqn.
(A.13) for all zones. Since the function in Eqn. (A.10) is usually nonlinear, a method is needed to
obtain the solution of simultaneous nonlinear algebraic equations. The Newton-Raphson (N-R)
method (Conte, and de Boor 1972) is chosen in CONTAM to solve the nonlinear problem by
iteration. In the N-R method a new estimation of the vector for all zone pressures, {P}*, is
computed from the current estimate of pressures, {P}, by
{P}* = {P} {C}

where the correction vector, {C}, is computed by the matrix relationship

(A.14)

170
[J] {C} = {B}

(A.15)

and {B} is a column vector with each element given by

Bi =

j ,k

(A.16)

And [J] is the square Jacobian matrix whose elements are given by
Ji,j =

Fj ,k

Pj

(A.17)

In Eqn. (A.16) and (A.17), Fj,k and Fj,k/ Pj are evaluated using the current estimate of
pressure {P}. The CONTAM program contains subroutines for each airflow element, which
returns the mass flow rates and the partial derivative values for a given pressure difference input.
Eqn. (A.15) represents a set of linear equations, which must be set up and solved in each
iteration until a convergent solution of the set of zone pressures is achieved. In its full form [J]
requires computer memory for N2 values, and a standard Gauss elimination solution has
execution time proportional to N3. Sparse matrix methods can be used to reduce both the storage
and execution time requirements. A skyline solution process following the method presented by
Dhatt (1984) was utilized. This method can be applied to solve equations with symmetric or
nonsymmetrical matrices. In this case, the Jacobian matrix is symmetric.
Analysis of the element model will show that
|Jk,j| =

| Jk,j|

(A.18)

j +k

This condition allows a solution without pivoting, although scaling may be useful. Note that
the degree of sparsity of the Jacobian matrix after factoring is dependent on the arrangement of
the zones. The CONTAM user interface ensures the correct interconnection the airflow elements
in the network.
CONTAM allows zones with either known or unknown pressures. The constant pressure
zones are included in the system of Eqn. (A.15), which is processed so as not to change the
pressure of the chosen zone. This gives flexibility in defining the airflow network while
maintaining the symmetric set of equations. A sufficient condition for the Jacobian to be

171
nonsingular is that the entire unknown pressure zones being linked, either directly or indirectly,
by pressure dependent flow paths to a constant pressure zone. In CONTAM, the ambient (or
outdoor) air is treated as a constant pressure zone. The pressure difference due to wind effect is
considered in Eqn. (A.2) separately. The ambient zone pressure is assumed to be zero for the flow
calculation causing the computed zone pressures to be values relative to the true ambient pressure
and helping to maintain numerical significance in calculating P.
Conservation of mass at each zone provides convergence criterion for the N-R iterations.
That is, when Eqn. (A.12) is satisfied for all zones for the current system pressure estimate, the
solution has converged. Testing for relative convergence at each zone attains sufficient accuracy:
| j Fj ,k |

| Fj ,k |

<

(A.19)

with a test of |Fj,k| < 1 to prevent division by zero. The magnitude of can be established by
considering the use of the calculated airflows, such as in the situation of an energy balance. In
any case, round-off errors may prevent perfect convergence ( = 0).

To achieve faster and reliable numerical convergence, a simple constant under-relaxation

coefficient is used. Eqn. (A.14) for the iteration becomes:
{P}* = {P} {C}

(A.20)

where is the relaxation coefficient. A relaxation coefficient of 0.75 has been found to be usable
for a broad range of airflow networks. This value is not a true optimum but appears to work quite
well without the computational cost of finding the theoretically optimum value.

When Convergence is progressing rapidly, under-relaxation ( < 1) slows convergence

compared to no relaxation. To prevent this, a global convergence value is computed:

| F
|F
k
k

j ,k

j ,k

(A.21)

when * < , is set to 1. Currently, CONTAM uses = 30%. This often reduces the number of
iterations.

172
Newton-Raphsons method requires an initial set of values for the zone pressures. These
may be obtained by including in each airflow element model a linear approximation relating the
flow to the pressure drop:

Fj ,k = c j , k + b j ,k ( Pj Pk )

(A.22)

Conservation of mass at each zone leads to a set of linear equations of the form
(A.23)

[A] {P} = {B}

Matrix [A] in Eqn. (A.23) has the same sparsity pattern as [J] in Eqn. (A.15) allowing use of
the same sparse matrix solution process for both equations. This initialization handles stack
effects very well and tends to establish the proper directions of the element models used by
CONTAM. When solving a set of similar problems, such as when approximating a transient
solution by successive steady-state solutions, it tends to be preferable to use the previous solution
for the zone pressure as the initial values for the new problem.

A.4. Boundary Conditions

The determination of airflow pattern and contaminant dispersion in an entire building
requires boundary conditions to be provided. The boundary conditions of CONTAM include the
weather data and wind pressure information on the building envelope.
CONTAM enables the user to incorporate the effects of weather on a building. Weather
parameters include ambient temperature, barometric pressure, wind speed, wind direction and
outdoor contaminant concentrations. Depending on different simulation purposes different
boundary conditions are needed. Steady state weather information is provided to CONTAM if
only steady state simulation is used. During the simulation, CONTAM keeps the ambient
temperature, wind speed and direction, and ambient contaminant concentration unchanged.
Transient weather data is used to simulate the changing outdoor weather and wind conditions
when performing a transient simulation. Such data are stored in a weather file that has a special
format.
Wind pressure can be a significant driving force for air infiltration through a building
envelope. It is a function of wind speed, wind direction, building configuration, and local terrain

173
effects. CONTAM enables the user to account for the effects of wind pressure on the flow paths
across the building envelope (external airflow paths). It also provides general approaches to
handle the variable effects of wind on the building envelope, through which the local wind
pressure coefficient for the building surface is determined. The detailed information can be found
in the user manual of CONTAMW 2.0.

A.5. Simplified Model Applied in Whole Building Model

CONTAM (current published version CONTAM 2.4) is a multi-zone indoor air quality and
ventilation analysis program. In the whole building heat and moisture analysis, air flow rates
(ambientzone, zonezone, atticzone) are needed for energy and moisture balance purpose.
Thus, a simplified model for multi-zone air flow model, which can only treat some typical
specified cases for multi-floor air flow, was developed based on CONTAMWs theoretical
background. The simplifications include:
1.

Only considers quasi-steady infiltration, exfiltration, and floor-to-floor airflows and pressure
differences in a building driven wind pressures acting on the exterior of the building,
buoyancy effects induced by temperature differences between the building and the outside,
and exhaust fans, but does not consider other mechanical ventilation;
A. Typical residential building geometry have a rectangular shape with a length that is
twice the width, (deonoted as S&C_2_Long in CONTAM) was adopted as the default
typical residential building with the wind pressure profile shown in Figure A.2;
B. Only limited zones were considered: up to 4 floors including attic;
C. The pressure in the attic assumed to be equal to the ambient pressure at the same height;

2.

Does not consider contaminant concentrations because the building heat and moisture
balance model determines the moisture balance;

Pressure Coefficient (-)

174

Wind angle ()

Figure A.2 Default wind pressure profile in MOIST whole building model (S&C_2_Long).

The simplified air flow model was validated through comparisons of results with
CONTAMW 2.4 for a number of cases. In all cases, the simplified model predicted nearly the
same results as CONTAMW.