HomeMy WebLinkAboutinfo on exterior insulation/filling material, etc.-1999 3 av•s.�.
' National Research Conseil national
Council Canada de recherches Canada
mcwcmc
Cllent Deport
A-3132.1
In-situ Performance Evaluation of Exterior
Insulation Basement System (EIBS) —
EPS Specimens
for
Canadian Plastics Industry Association
365 Bloor Street E., Suite 1600
Toronto, Ontario
M4W 31_4
22 March 1999
In-situ performance Evaluation of Exterior
Insulation Basement system (EIBS) - EPS
Specimens
Author(s)
N. Normandin R. Marchandl
\\
Dr. W. Maref L,Dr. M-T: Bomberg
M.C. Swinton
Quality
Assurance
Dr. M.K:.Ku—mg ran -
Approved G tom,
R. M. Paroli, Director
Building Envelope and Structure
Report No: A3132.1
Report Date: March 22, 1999
Contract No: A3132
Reference: Joint Research Agreement
Laboratory: Building Envelope and Structure
40 pages +6 appendices
Copy No.1 of 4 copies
This report may not be reproduced in whole or in part without the written consent of
both the client and the National Research Council Canada
Table of Contents
EXECUTIVE SUMMARY...................
.............................................................................................
INTRODUCTION..................................
............................................................................
OBJECTIVES....................................................
PROJECT DELIVERABLES.................................
PART 1-DOCUMENTATION OF EXPERIMENTAL APPROACH..................................................3
APPROACH....................
SPECIMENS&INSTALLATION......................................................................................................................3
CONTROL OF INTERIOR CONDITIONS..........................................................................................................4
CHARACTERIZATION OF SPECIMEN PROPERTIES 5
TASKS ...................................................................5
DURATION OF THE .
• FIFI-D ASSESSMENT...........................
INSTRUMENTA ......................................
TION................... 7
PART2-EXPERIMENTAL DATA..........................................................................................................7
8
INDOOR&OUTDOOR AIR TEMPERATURE PROF7I.FSINDOOR PH
...........................
8
SOIL TEMPERATURES................................................................................................................................ 8
....
CoNTNr ....:
TEMPERATHE WALL ATMID-HEIGHT.t.............................................
TURE PROFILES THROUGH 9
...................
EFFECT OF GRADING9
..........................................
EFFECT OF POLYETHYLENE ................................... ................. 10
_ AT THE SOIL I ....."'.'
TEMPERA INTERFACE......................
TEMPERATURE PRO ..................................................
FILES AT ALL VERTICAL LOCATIONS 10
HEAT FLUX PROFRM .............. 10
PART 3-ANALYSIS AND DISCUSSION.............................................................................................. 1 I
METHOD TO MEASURE THERMAL RESISTANCE IN SITU. """•'•••••••11
2-DIMENSION ....................... 11
AL ANALYSTS.................................. ..............................
COMPARISON OFINSTALLATION ........................"" ....... 12
IMPROVEMENT IN THERMAL PERFORMANCE...............................................................
........................................................................................... 13
ASSESSMENT .................... 13
ENT OF 3-DIMENSIONAL HEAT Flow
LAB MEASURED PROPER 14
............................
TSS AFTER EXPOSURE........................................................
DISCUSSION 15
CONCLUSIONS..................... 16
...................................................................................................... .............
...........17
REFERENCES... ...
APPENDIX A-ADDITIONAL DETAILS OF THE TEST SET-UP.......................................
APPENDIX B-INSTRUMENTATION REPORT...............................
APPENDIX C-SOIL CHARACTERIZATION REPORT......................................................
APPENDIX D- MONITORED TEMPERATURE PROFILES FOR EACH SPECIMEN.................. 1
� J
APPENDIX E- OBSERVATIONS MADE DURING REMOVAL OF THE SPECIMENS................1
APPENDIX F- TABLES AND GRAPHS-POST-EXPOSURE MOISTURE CONTENTS AND
MATERIAL PROPERTIES BEFORE AND AFTER EXPOSURE........................................................1
Report A3132.1
Page 1
In-situ Performance Evaluation of Exterior Insulation
Basement System (EIBS) - EPS
Executive Summary
In 1995, the Expanded Polystyrene Association of Canada and the Canadian
Plastics Industry Association established a joint research project with the
Institute for Research in Construction to assess the in-situ thermal performance
of a number of insulation products used as exterior basement insulation in
contact with the ground.
In October 1995, sixteen insulation specimens measuring 610 mm and 1220 mm
wide were installed on the exterior basement wall at Test Hut #1 at the NRC
Campus on Montreal Road in Ottawa. These specimens were instrumented
prior to backfilling and their thermal performance was monitored over two full
years. Soiltemperatures and moisture content were monitored concurrently.
Weather events were recorded on a daily basis.
Through analysis of the surface temperatures of the specimens, the presence of
water was detected at their outer surface through various periods of heavy rain
and major thaws throughout the two-year period. Over the same periods, the
surface of the concrete on the inside of the insulation showed no evidence of
water penetration through the insulation layer through most of the'height of the
basement wall.
For the conditions recorded over two years of monitoring, the thermal
performance of each insulation specimen was found to remain stable. The
thermal performance appeared not to be significantly affected by water
movement at the exterior face of the insulation. Furthermore, many specimens
showed improvement of thermal performance in the second year. The dry
summer and dryer soils noted during the second year may have contributed to
this apparent marginal improvement.
The insulation specimens were retrieved after 30 months of exposure in the soil.
Samples were taken from these exposed specimens. Thermal, moisture and
mechanical properties were tested in the lab and compared to initial properties.
No significant change in properties of the insulating materials was found.
Report A3132.1
Page 2
In-situ Performance Evaluation of Exterior Insulation
Basement System (FIBS) - EPS
Introduction
Exterior basement insulation performs many functions. It not only provides
thermal resistance between the soil and the interior, it also protects the structure
from a challenging environment (e.g. moisture from wet soils, heaving and
adhering soils due to frost action). It provides a means of water management at
the interface of the soil and insulation, while promoting drying out of the
structure. In this context, it is important to set performance parameters for
setting appropriate materials.
The Expanded Polystyrene Association of Canada and the Canadian Plastics
Industry Association established a joint research project with.the Institute for -
Research in Construction to asses$ the in-situ thermal performance of a number
of insulation products used as exterior basement insulation in contact with the
ground.
This project investigates the in-situ performance of exterior basement insulation
systems. By examining the thermal performance of EPS in-ground over an
extended period, the characteristics of what constitutes acceptable performance
was established.
This project was steered by the Expanded Polystyrene Association of Canada (EPAC),
a sub-committee of the joint Canadian Plastics Industry Association / National
Research Council (EPAC/CPIA/NRC) Research Steering Committee:
James Whalen, Chairman, Plasti-Fab, Division of PFB Corporation
Brent Barnes current representative of CPIA
Rob McInnes past representative of CPIA
Andra St-Michel BASF Canada
Jim Shannon Huntsman Chemical
Fred Sonnenberg Styrochem International
Bob Vasseur Beaver Plastics
Shveta Ba'gade NOVA Chemicals
Mike Swinton project manager, IRC/NRC
Mark Bomberg technical advisor and initial project manager, IRC/NRC
Kumar Kumaran IRC/NRC
Nicole Normandin IRC/NRC
r
Report A3132.1
Page 3
Objectives
This project was undertaken to record the effect of exposure for several types of
EPS insulation specimens used as exterior basement insulation. The insulation
was exposed to the Ottawa climate, including below grade conditions, over a two
and a half year period from October 1995 to June 1998. The official monitoring
began in June 1996. The following key performance factors of the EPS
insulation were investigated: in-situ thermal resistance, laboratory tested thermal
conductivity and compressive strength. The performance factors were
investigated in the context of:
• a prolonged exposure to the below grade environment
• measured changes in local environmental conditions; i.e., changes in soil
temperatures, soil moisture content, surrounding air temperatures inside and
out, and,
• measured laboratory properties before and after exposure.
A number of types of EPS were investigated to establish whether'type' has a
bearing on these performance factors.
Finally, an equally important objective of this study was to determine the.
relationship between changes in EPS material properties after 31 months of
exposure in the field and changes in EPS properties due to exposure to -
environmental cycling performed in a controlled laboratory test.
Project Deliverables
The project deliverables are:
1. Test results of insulation material properties before exposure.
2. Test results of insulation material properties after exposure
3. Field performance of the material under environmental loading conditions
4. Laboratory evaluation of durability under environmental cycling (separate
report)
Part 1 - Documentation of Experimental.Approach
Approach
Different types of EPS boards were installed over the full height of the exterior
basement wall of Test Hut #1, at NRC's main Campus in Ottawa. Their
performance was monitored using a strategy developed to monitor the in-situ
thermal performance of roof insulation 2,3. A thermally calibrated, 25 mm layer of
expanded polystyrene board was installed over the entire surface of the interior
of the_basement wall. Thermocouples were systematically placed at the surface
of each element in the wall, in a vertical array consisting of 16 points per
specimen. A schematic diagram of this arrangement is shown in Figure 1. As
well, heat flux transducers were installed at three vertical locations.
L
Report A3132.1
Page 4
The monitored temperature difference across the calibrated insulation
i
used to calculate the heat flux profile into the wall on a continuous basis.yer was
Detailed analysis of heat transfer through the wall was used to assess the
resulting heat flux into each exterior insulation specimen. Using this heat flux
and temperature difference across the specimens, the apparent in-situ thermal
resistance of the specimens was deduced.
Boundary conditions, including soil temperatures and moisture content were
recorded, as well as observations of weather extremes. Four separate soil
analyses were performed to characterize the soil environment, including vertical
profiles of moisture content. One soil characterization report is included in
Appendix C. This information was used to qualify differences in observed
thermal performance of the specimens.
Specimens & Installation
The field experiment consisted of placing 5 expanded polystyrene (EPS)
specimens on each of the east and west basement wall of Test Hut #1, for a total
Of 10 specimens. There were three types of EPS insulations used, labelled A, B,
and C, as listed in Table 1. On the east wall, specimens E2, E3 and E5 were fully
instrumented as shown in Figure 1, whereas specimens E1 and E4 only had a
mid-height set of sensors. On the west wall, W2, W3, and W5 were fully
instrumented. Specimens W1 only had one set of sensors at mid-height. W4
had thermocouples at the 4 vertical locations but only one heat flux transducer at
mid-height.
Two different installation methods were used on the east and west wall
configurations, labelled System 1 and System 2 respectively. System 1 on the
west wall (Fig. 2) featured two horizontal rows of metal z-bars, separated by a
wood spacer, all fastened to the header. Once the insulation was in place , the
cementitious covering boards were fastened to the z-bars and wood space. No
other fasteners were used, so the cementitious board was effectively
'cantilevered' over the insulation specimens. The soil was sloped at 5% grade
towards the wall, to simulate a settled condition.
System 2 on the east wall (Fig. 3) featured metal z-bar supports placed vertical)
between each insulation specimen. The z-bars were fastened directly to the v
concrete wall and wood header on the inside, and fastened to the cementitious
board on the outside. Each metal z-bar was therefore a thermal bridge around
the insulation. System 2 also featured an initial 5% sloped grade away from he
basement wall.
A plan view of the arrangement of the specimens is shown in Figure 4. The EP
samples are numbered 1-5 on each wall. Other specimens consisting of different
insulation types were also installed farther down the walls, as shown.
Report A3132.1
Page 5
Control of Interior Conditions
The test but was heated in the winter and air-conditioned in the summer. The
indoor temperature was initially set at 21°C. After an initial monitoring period
through the first summer, this temperature was reset to 23°C, to increase the
accuracy of the monitoring in the shoulder seasons. Figure 5 shows the relatively
steady indoor temperature control through the monitoring period.
The indoor RH was not controlled, although some summertime dehumidification
probably occurred as a by-product of air-conditioning. The indoor RH profile is
show in Fig 6.
The drainage system featured a sump pump. The level of water in the sump
was observed to be quite high (at footing level) for the first 200 days of
monitoring. At day 215, the sump pump controls were reset to maintain the
water level in the sump at about 300 mm below the footings.
Characterization of Specimen Properties
Samples and insulation products were labelled upon receipt from the manufacturers
and the following properties were assessed before and after the field exposure:
• density
• thermal conductivity
• compressive strength
Average pre-exposure properties of the specimens exposed at the test but are
listed in Table 1, along with other physical characteristics of the specimens.
r
i
Report A3132.1
Page 6
Table 1 Specimens Evaluated In the EIBS Project
(All specimens installed in the exterior of the concrete,over the full 2. 4 m height of t
• 4 9 he basement wall.)
ID # Material IDPost-Ex osure Pro ertles
Code Mean Mean Mean
Density Thermal Compressive Comments
ftftn ) ConductivityWidth Nominal Date
wpm oC It mm Thickness Received
East face. Grade initialI slo ed 5%awa from wall.S stem 2 su orts kPa mm
E-1 EPS"C"
" " 22.5
E-2 EPS B 17 1 0.0343 136 moved face Iain ed e
E-3 EpS"C° 0.0351 104
E-4 EPS"C" 20'5 0.0352 moved face Iain ed a 610
20.0 120 76 Oct 3/95
Iain face shi la 610 76 Oct 3/95
E-5 EPS"A" 0:0359 117.6 rooved face shi la 610 76.
'rest face. Grade initial) slo ed 5%towards x0448 Na 610 Se 29/95
�. EPS wra ed in of eth ene 76 Se 29/95
W-, EPS"B" 16.4 II. S stem 1 su orts 1220 76
0.0365 Oct 3/95
W-2 EPS"B" 15.6 96.6 moved face shi a
0.0362
W_3 FPS C. 20.1 91.5 moved face shi la ' 610 76 Oct 3/95
�q°C° 0.0354 117
W-4 EF,
23.1 0.0354 moved face shi la 610 76 Oct 3/95
W-5 3' EP6 `B Na Na 133 Iain face shi la 610 76 Se 29/95
Na Three 25 mm EPS layers wrapped in 610 76 Se 29/95
of , y pp 1220
76 Oct 3/95
Report A3132.1
Page 7
Tasks
The field assessment of exterior insulation performance at Test Hut #1 consisted
of 17 tasks:
1. Acquisition of materials
2. Characterization of material properties
3. Preparation of site for installation of insulation specimens
4. Installation of interior reference insulation and instrumentation
5. Installation of specimens
6. Instrumentation of specimens and surrounding soil
7. Covering and backfilling
8. Final soil instrumentation and grading
9. Commissioning of monitoring system, replacement of faulty sensors
10. Monitoring and on-going quality review
11. Soil characterization and moisture content monitoring
12. Documentation of major weather events
13. Development of analysis methods
14. 2-D Analysis and presentation of results
15. 3-D Corrections
16. Uncertainty Analysis --
17. Excavation of the specimens, physical observation and testing
Duration of the Field Assessment
The specimens were installed on the outer face of the basement wall in the 3`d
and 4 week of October 1995. Thermocouples were then placed on the outer
surface of the specimens. Backfilling took place on November 30, 1995, after all
specimens had been installed and instrumented. Between installation and
backfilling, the specimens were exposed to ambient conditions for about 4 to 6
weeks.
The data acquisition system was then installed along with final instrumentation in
the soil and surroundings. The data acquisition system was commissioned in
tray..of. 1996. Official monitoring started on June 5, 1996 and ended on June 5,
1998:
Instrumentation
The instrumentation packWvnsisted of approximately 145 thermocouples, 2,
RH sen+sors,.21 EMF sere site-built and calibrated heat flux transducers, -
several junctimn boxes, a date, isition unit and computer. Monitoring of soil
moisture content•was parformed mft" separate data acquisition unit. A detailed
description of the instrumentation tpack,;ie is included in Appendix A.
Report A3132.1
Page 8
Part 2 — Experimental Data
The experimental data set contains a number of time-series records, including
indoor and outdoor air temperatures
• indoor and outdoor RH
• soil temperatures
• soil moisture content
• the temperatures at each material interface in the test wall, at four levels from 50
mm above the slab to 270 mm below grade (see Figure Al in Appendix A)
• the temperature differences in the reference insulation at three levels measured
by lab-assembled heat flux transducers (Figure A2 in Appendix A)
Indoor & Outdoor Air Temperature Profiles
The indoor and outdoor air temperature profiles over the two heating seasons
are shown in Figure 5. Data was recorded every half hour. Plotted is every 8"'
reading, i.e., every 4 hours.
The indoor basement air temperature was held relatively steady at 23°C, with
some loss of control experienced in the shoulder seasons. The switch from
heating to air-conditioning on the heat pump was manual, resulting in slight
overheating of the basement on warm days in spring and fall when the heat
pump was set to heating.
The outdoor temperatures show both the diumal variations and seasonal trends
following a rough sine wave. The winter of 1997 featured one very cold night
only, with temperatures dipping to about -30°C. Neither winters showed a
sustained cold snap—each very cold evening was often followed by a thaw
(temperatures climbing above 0°C). This unexpected weather pattern may have
resulted in low frost penetration, as documented further in the report.
Indoor RH
The outdoor RH was also monitored. The outdoor RH varies from dry conditions
(e.g. 40%) to wet (e.g. near 100%) on a regular basis, with no significant change"
from season to season (plot not shown).
The indoor RH profile is shown in Figure 6 for the monitored period. The
basement air was significantly dryer during the second summer than during the
first.
Soil Temperatures d
Soil temperatures were measured at a distance of 2 m. from the wall. The
results for the instrument set nearest specimen W6 are shown for three depths
in Figure 7. These were measured at 150 mm, 740 mm and 1840 mm below
grade, over two year period. The sensors nearer the soil surface shows the
greatest variation from summer to winter, and diumal effects can be observed.
Report A3132.1
Page 9
No diumal effects can be seen at the two lower depths. The temperatures at the
lowest depth approximates a sine wave each year.
It should be noted that in both winters, the frost penetrated approximately 150
mm. The frost depth temporarily reached 270 mm in the second winter (not
shown in graph). Deep snow cover, and repeated thaws in both heating seasons
may be the cause of shallow frost penetration. An independent temperature
measurement in the second winter, 10 m from the house showed the same
result. This indicates that there is an undetectable effect because of basement
heat loss on the soil temperature at 2 m from the house.
Soil Moisture Content
A single TDR probe (time domain reflectometry) was deployed about 2 m from the east
wall, and 1 m down into the ground, to monitor soil moisture content on an ongoing
basis. Figure 8 presents the results of the soil moisture content monitoring over the
period from October 1996 to June 1998. It shows that the soil was wet throughout the
first heating season and dried substantially in the summer of 1997. Throughout most of
the second heating season, the soil stayed dryer than the first, until the spring thaw in
1998.
Temperature Profiles Through the Wall at Mid-Height
Figure 9 shows the two-year temperature record at four locations through the
wall at specimen W2: the interior surface, both sides of the concrete, and the
exterior surface of the specimen in,contact with the soil. The inner surface of the
wall is kept near 21°C, with small variations throughout the two years. Main
control events such as power outages and changes from heating to air
conditioning and back are evident from these temperature readings. The
temperature at both sides of the concrete are quite close to one another
(concrete being a poor thermal insulator), and these vary from 15°C to 20°C,
from winter to summer.
The lowest curve in the graph is the temperature record for the insulation/soil
interface. These vary between about 5°C in winter up to a maximum of about
20°C in summer.
The periodic 'spikes' in this curve correspondYorded events of heavy
precipitation or winter thaws. The August 8, 1996 rain was a 1 in 75 year event
for Ottawa, which caused local flooding around the test hut.. During this storm,
the temperature at the insulation/soil interface deflected upwards, apparently due
to warm rainwater moving down the wall. Such deflections were observed at the
mid, low and bottom thermocouple positions during the same period, tracing the
path of the water. These deflections were much less noticeable at the high
position, where the soil temperature would be closer to the temperature of the
moving water.
L
Report A3132.1
Page 10
The temperature deflections in the winter at the soil/insulation interface are
downward because the melt water temperature is initially 0°C, which would cool
the soil and insulation at the interface.
Effect of Grading
Figure 10 shows temperature measurements on the east wall of the test but for
the same insulation product that was shown on the west side in Figure 9. In the
first year, these periodic temperature deflections were often smaller or absent on
the east wall where the ground surface was properly graded outw
relative absence of temperature deflections can be observed in Fiard. This
gure 10 for the
first winter.
In the second year however, the differences between east and west were less
noticeable, and the temperature deflections were then quite noticeable. A final
review of soil slopes near the wall revealed that by the end of the second year,
most of the slopes had settled on the east wall. These slopes were now mostly
inward, as recorded in-Appendix A, Figures A3 and A4.
Effect of Polyethylene at the Soil Interface
Figure 11 shows the temperature profile at the polyethylene/soil interface for the
specimen at position W5 on the west wall. The temperature depressions during _
rain and thaw intervals are relatively large compared to those shown in Figures
and 10. ures 9
g
Temperature Profiles at all Verticaf Locations
Temperature profiles similar to the ones presented in Figure 9 are presented in
Appendix D for all vertical locations. In these graphs, it can be noted that the
vertical deflections in the temperature profile at the insulation/soil interface occur
at the same time. This strongly suggests the downward movement of water
along this interface. These temperature deflections are less apparent or non-
existent at the high position. The thermocouples in this location are covered by
the cementitious board (see Figure 1), so that water is not likely in direct contact
with the insulation specimen at this high vertical location.
With the.possible exception of the lowest position (50 cm up from the slab), the
concrete, which is behind:thessPecimen, generally do not show corresponding
temperature deflections during these events. This suggests that the concrete
wall is remaining dry behind the insulation.
s
Report A3132.1
Page 11
Heat Flux Profiles
Temperature differences across the thermally calibrated insulation layer at the
inner face of the wall were calculated at 4 vertical locations with thermocouple
and three vertical locations with heat flux meters shown in Figures Al and A2.
Example profiles of heat flux into the basement wall are shown in Figure 12.
The smoothness or regularity of these curves attest to the fact that the seven
vertical readings are consistent with one another. These weekly averaged heat
flux profiles have the highest level of accuracy associated with them. Each heat
flux meter reading is the result of the average of 5 EMF readings.
Part 3 - Analysis and Discussion
Method to Measure Thermal Resistance In Situ
A test method, developed at NRC during previous project (Muzychka, 1992;
Bomberg and Kumaran 1994)2'3, involves testing two materials placed in contact
With each other- a reference material whose thermal conductivity and specific
heat are known as a function of temperature, and a test specimen whose _.
thermal properties are unknown. Because of its comparative character, this
method has been called a heat flow comparator(HFC). In a previous experiment
involving roof insulation, the reference and tested specimens were placed in the
exposure box representing a conventional roofing assembly. Thermocouples
were placed on each surface of the standard and reference materials to measyre
temperatures, which then were used as the boundary conditions in the heat flow
calculations.
The beat flux across the boundary surface between reference and tested
specimen is calculated using a numerical algorithm to solve the heat transfer
equation through the reference material. Imposing the requirement of heat flux
continuity at the contact boundary between test and reference materials,
corresponding values of thermal conductivity and heat capacity of the tested
specimens are found with an iterative technigwe.- Performing these calculations
for each subsequent data averaging period Wirt"result in a set of thermal
properties of the test material which, over the period of measurements, give the
best match with its boundary conditions (temperatures and heat flux).
Since thermal conductivity of the specimen is a function of its temperature, the
solution of the heat transfer equation is based on central finite difference
calculations that include Kirchoff's potential function (integral of thermal
conductivity over the range of temperature) and uses a Taylor's series to
calculate heat flux through the surface. Subsequent developments improved the
stability of the numerical solution and produced a user friendly computer code
Lj
Report A3132.1
Page 12
that includes optimization routines. This method was documented and applied
for determining the in-situ thermal resistance of roof insulations (Gomberg and
Kumaran, 1994)
2-Dimensional Analysis
The analysis technique for EIBS study was adapted to account for two significant
differences in the test set-up:
1. The 200 mm concrete wall was interposed between the reference insulation layer
and the test specimen. Analysis showed that the concrete layer modifies the heat
flux leaving the reference specimen through heat storage and through heat flow up
and along the concrete wall.
2; Temperatures in the soil did not vary significantly on a diumal basis, so that a
statistically valid relationship between specimen conductivity and temperature was
not needed to the-extent considered for the roofing specimens.
The analysis was therefore adapted to assess the heat storage effects and two
dimensional heat loss through and up the wall. A second, more elaborate method was
developed to assess the heat loss in the third dimension, along the wall, and to suggest
what corrections to the 2-D results would be needed.
The 2-D analysis consisted of calculating the horizontal heat flux (inside to outside) and
vertical heat flux (boftom to top) through all materialsinthe control volume defined in
Figure 13. A finite difference technique was used to solve the heat transfer equations
for dynamic heat flow through solids with known boundary conditions, at each point of a
nodal network used to represent the materials in the control volume (see Figure 14).
Using this analysis technique, the temperature differences across the specimen and the
concrete were calculated for each measurement interval. The temperature difference
across the reference insulation determines the heat flux into the concrete. Finite
difference analysis was used to assess the direction and magnitude of heat flux in the
concrete as well as the amount of heat stored or released by the concrete. After these
quantities are evaluated through the concrete, the resulting net heat flux into the
specimen was assessed. Using this heat flux and a postulated thermal conductivity of
the specimen, the resultant temperature differences across the specimen were
calculated. The calculated temperature differences are then compared to measured
results, every ten minutes. Mean errors were calculated on a weekly basis. An iterative
technique was devised to minimize the mean error between calculated temperature
differences and measured, by adjusting the postulated conductivity of the insulated
specimen on a weekly basis.
The factor by which the conductivity was adjusted relative to lab-determine
conductivities of the specimens was labelled 'conductivity adjustment factor'. These
were recorded and plotted on a weekly basis. As well as the reciprocal—the thermal
resistance adjustment factors was plotted. As a final step in the analysis process, the
Report A3132.1
Page 13 '
adjustment in thermal resistance of the specimen as normalized to an initial average
adjustment for October 1996 —the first period of cold weather in the monitoring period.
In-Situ Thermal Performance
Figures 15 a-g show the resulting thermal performance plots for the specimens on the
west face (a-d) and the east face (e-g). The change in thermal resistance adjustment is
shown on a weekly basis over two heating seasons. Key observations are as follows:
• all specimens show relatively steady performance through the heating seasons
• the second heating season shows equal or improved performance for all specimens
• the results for the warm periods are unreliable, since the temperature difference
across the specimens are very small (< 0.50C). During such periods, thermocouple
errors can be as large as the actual temperature difference.
• On day 215, modifications were made to faulty controls on the sump pump. Water
levels around the footing were lowered over time. The sump pump connection is
nearest specimens W5 and W4. These specimens appeared to fiave been
temporarily affected by this modification.
• Major rain and thaw periods (shown in Figures 9, 10 and 11) do not appear to
significantly affect the thermal performance of the specimens during these episodes.
Comparison of Installation Systems
The seasonally averaged thermal performance of the insulating systems on the east
wall installed using System 2 (vertical Z-bars) were compared to those of the
corresponding systems on the west wall, installed with System 1 (horizontal z-bars at
the top) Table 2 lists the reduction in seasonally averaged thermal resistance due to
installation Systems 2 (with vertical k-bars) relative to System 1 (horizontal z-bars).
System 2 was less thermally effective than System 1 by about 13%, on average.
Table 2. Thermal Performance: Installation S stem2 Relative to System 1
Specimen Position 2 3 5 Average
Year 1 2 1 2 1 2
(System2— Systeml) 15% 19% 14% 22% (4%) 7% 13%
increase
Improvement in Thermal Performance
The seasonally averaged thermal performance of the insulating systems for heating
seasons 1 and 2 were compared to those at the start of the first heating season. Table
3 records the percent improvements for each specimen. Most specimens showed
sustained or improved performance in both years, with more improvement in the
second year, especially for specimens on the west wall.
Li
Report A3132.1
Page 14
Table 3.- Change in Thermal Performance – Yearl and Year 2 Relative to
Start
S ecimen Position 2 3 4
Yearl 5 Avera e
Systeml (West) 1.2% 0% -3.4%*
S stem 2 East 2.9% 1.3% -0.2%
Avera 0% --- 8.0% 3.6%
Year2 a 2.1% 0% -3.4% 4.6% 1.4%
Systeml (West) 6.0%
S stem 2 Imimon
Avera e 4.5% 2 9% -
' Modified sum pp 2.2% 10.4% 5.4%
p pump controls appear to have affected this specimen in the first year.
Assessment of 3-Dimensional Heat Flow
Over the course of the 2-dimensional analysis, it was noted that the temperature of the
concrete (behind the specimens) differed from specimen to specimen. This raised
Possibility that heat could flow through the concrete from one specimen to another, the
resulting in possible uncertainty in the orders of magnitude of performance assessed
using the 2-dimensional analysis.
In response to this, a detailed 3-dimensional heat transfer analysis_ of the both the east
wall and the west wall was undertaken. The same control volume depicted in Figure 14
was used for the 3-D analysis, but this Was extended in the third dimension along the
wall, to encompass all specimens. The objective of this much more detailed analysis
was to determine the order of magnitude of lateral heat flow in the concrete from one
specimen to another, and to provide a correction on the 2-dimensional results for in-situ
R-value of each specimen, where needed.
To illustrate the result of this analysis, an example result is shown in Figure 16 for the
west wall. The angle between the actual heat flux vector and the normal direction is
defined here as the heat flux angle; e.g, a zero angle denotes heat flux normal to the
wall —no lateral heat flow. The heat flow angle is plotted for the 'top' location (top of
the control volume) in the concrete, at one point in time. A positive angle means lateral
heat flow towards one end of the wall, and a negative heafflow means lateral heat flow
towards the other.
The significance of this diagram is as follows. If the flow angle in the concrete behind
the specimen is small, and in the same direction, then the lateral heat flow has little
effect on the results. This was the case for W2, W3, W4, W5, the EPS specimens
The 3-D analysis thus confirmed the order of magnitude of performance analyzed usin
the 2-6 method for the EPS specimens. g
Report A3132.1
Page 15
The current report presents the 2-D analysis results, reported in terms of thermal
resistance relative to an initial condition
Details of the 3-dimensional analysis are published in a separate document4.
Physical Observations during Removal of Specimens
The 10 EPS specimens tested during the EIBS project were removed from the soil on
Jun 23t 1998. Appendix E records observations made during removal. In general, the
specimens appeared to be in good condition. The protected portions of the insulation
(protected by cementitious board) showed no signs of sedimentation (soil deposition at
the surface), nor did the majority of the surface of the specimens in contact with the
concrete. Sedimentation was visible but limited for the most part to portions of the
insulation directly in contact with the ground, and at the lowest level, about 100 mm
above the footing. -
Lab Measured Properties after Exposure
Appendix F records the changes in thermal, mechanical and vapour diffusion properties
of the specimens after exposure. As well, moisture contents of the specimens are
reported.
Table 4 summarizes the average thermal resistance of the insulating materials tested
initially and after removal from the Test Hut wall.
Table 4. Average Thermal Resistance of the Insulation Tested Before and After
Exposure at the Test Hut (m2 C/W at 25.4 mm thickness)
Type Initial Post-exposure Difference
m2 CNV m2 C/W %
@ 25.4 mm @ 25.4 mm
EPS Type "A" 0.604 0.605 <1%
EPS Type "B° 0.680 0.683 <1%
EPS Type "C" 0.710 0.711 <1%
Note: all results normalized to 25.4 mm thickness using the DI PAC model .
'The DIPAC model, used to calculate thermal conductivity of cellular plastics as a function of their
polymeric composition, blowing agents (if used instead of air), aging period as well as the specimen
,temperature and thickness has been experimentally verified. For details consult:Bomberg,M.T.;
Kumaran,M.K. "Roofing exposure of cellular plastics manufactured with alternative blowing agents to verify
methods for predicting their long-term thermal performance"ASTM Special Technical Publication pp. 1-
25.(NRCC-36898)(IRC-P-3444)
In general, the results confirmed that there is no significant change in any of the
measured properties—thermal resistance, compressive strength, and water vapour
permeability. The measured moisture contents of the excavated EPS specimens were
low.
Report A3132.1
Page 16
Discussion
The following general observations are made as a result of these experiments.
1. The in-situ tracking of thermal performance of the specimens indicated stable
thermal performance over the two years of monitoring. In most cases there was an
improvement in specimen thermal performance in the second heating season.
correlates with dryer prevailing soil conditions in the second year. This
2. Based on the temperature profiles at the specimen/soil interface, and corres ondin
observations of heavy rainfall or thaw periods, the specimens are apparently g
'handling' moving water at the specimen surface. This appears to have negligible
effect on thermal performance of the specimen. There is also independent evidence
that the EPS insulation protected the concrete structure during these events (no
temperature deflegtions on the inside face, and clean interior surfaces observed on
removal of the insulation).
3. The samples wrapped with a polyethylene cover appeared to have similar water
handling and thermal characteristics as the samples without thisrotectio
data suggests that, by whatever mechanism, there may be even more water.
The
movement at the soiVpolyethylene interface than at the interfaces of the other
specimens.
4. Type B and Type C EPS specimens equally maintained stable thermal performance
through the monitoring period.
5. The Type °A" EPS specimen was wrapped in polyethylene. It maintained thermal
Performance throughout the monitoring period as well.
6. The effect of grooves in the insulating board was also indistinguishable within
margin of error of the method. No evidence of water movement down the back of
the board was recorded.
7. The effect of shiplap joints was indistinguishable from butt-joints - none of the EPS
specimens showed evidence of moisture movement behind the specimens.
8. The following parameters of EPS products used on the exterior of basement walls
appeared to have little or no effect on the observed thermal performance of
specimens within the scope of this experiment:
• duration of exposure
• mean temperature of the specimen
• water movement at the outer surface
• density of product
• freezing cycles
i
Report A3132.1
Page 17
Conclusions
For the conditions recorded over the two year monitoring period in this
experiment, the EPS insulation specimens showed stable thermal performance
in the soil.
The EPS Type B and EPS Type C specimens showed sustained thermal performance
over the two-year monitoring period. On average, there was little to differentiate these
two types of EPS products in their ability to sustain thermal performance in the ground
over the two-year monitoring period.
Installation System #1 (Horizontal z-bars attached to header) yielded consistently
superior thermal performance of the system compared to Installation System #2
(vertical z-bars attached to concrete).
When tested in the lab after recovery and drying of the specimens, the
compressive strengths of the EPS samples were the same as those of samples
tested at the beginning of the test, within the margin of error of the test method.
This was consistent with results of the environmental cycling tests5.
When tested in the lab after recovery and drying of the specimens, the measured
thermal conductivities showed no significant change over the pre-exposed
samples. This was consistent with results of the environmental cycling tests.
We conclude that the key performance factors of thermal conductivity and
compressive strength of the EPS specimens were not affected by the 31-month
in-situ exposure.
The laboratory environmental cycling is not intended to duplicate field exposure.
The lab test procedures exposed the product to extreme environmental cycling
with no measurable change in properties. Details of the lab exposure tests are
recorded in a separate report—A3132.25
PAGE: 1 of 7
Prepared by:
Arnis L. Kurmis
TWIN CITY TESTING CORPORATION
662 Cromwell Avenue
St. Paul, Minnesota 55114
Phone: (612) 659-7309
Expanded Polystyrene Thermal Insulation Performance
in a Below-Grade Application
FINAL TEST RESULTS
Prepared for:
The Society of the Plastics Industry
Attn: Susan Herrenbruck
1275 K Street NW
Suite 400
Washington, DC 20005
Twin City Testing Corporation Project Number: 4140 92-2757
Date: July 23, 1993
TWIN CITY TESTING CORPORATION Reviewed by:
Arnis L. Kurmis Derrick J. Swan on, P.E.
Mechanical Engineering Supervisor Manager
Mechanical/Metallurgical Department Mechanical/Metallurgical Department
The test results contained in this report pertain only to the samples submitted for testing and not
necessarily to all similar products.
TutAn city testinqc
PROJECT NO: 4140 92-2757 DATE: July 23, 1993
PAGE: 2
FINAL TEST RESULTS
EXPAINDED POLYSTYRENE THERMAL INSULATION PERFORMANCE
IN A BELOW-GRADE APPLICATION
INTRODUCTION•
This report presents the results of a three year evaluation of the thermal performance of expanded
polystyrene (EPS) insulation in a below-grade foundation application in a cold climate environment. The
EPS was manufactured to meet the physical property requirements of ASTM C578 (Standard Specification
for Preformed, Cellular Polystyrene Thermal Insulation).
The scope of the study was limited to:
1. Evaluating EPS insulation applied to the below-grade foundation of Twin City Testing's Cromwell
Ave building in St. Paul, Minnesota.
2. Conducting tests on control samples to determine density, thermal resistance reference values, a!
well as to document compressive and flexural strength (see Appendix A: TCT Report # 414189.
659, dated 9-27-89).
3. Subsequently, for three years beginning October, 1990, removing a set of samples and conducting
tests to determine density, thermal resistance, and moisture content of the EPS foundation insulation
(see Appendix A: TCT Report #414190-0022, dated 11-1-90, TCT Report#414191-0878, dates
10-31-91, and TCT Report # 4140 92-2757 data supplement, dated 5-10-93).
4. Documenting temperatures, annual precipitation, and soil moisture content (see Appendix B).
AS A MUTUAL PROTECTION TO CLIENTS THE PUBLIC AND OURSELVES,ALL TWIN CITY TESTING CORPORATION REPORTS ARE SUBMFT.TED AS THE CONFIDENTIAL PROPERTY OF
AND AUTHORIZATION FOR PUBLICATION OF STATEMENTS CONCLUSIONS OR EXTRACTIONS FROM OR REGARDING OUR REPORTS IS RESERVED PENDING OUR PRIOR WRITTEN APPPCA
DATE: July 23, 1993
PROJECT NO: 4140 92-2757 PAGE: 3
SAMPLE IDENTIFICATION:
EPS thermal insulation molded to meet the requirements of ASTM C578, Type I and Type II classification,
were submitted to our laboratory by Diversifoam Products, Rockford, Minnesota
in August
sheets each of 89 for ere
Society of the Plastics Industry EPS Block Molders p Two 4' by 9'
identified as follows:
Type I: 2.73 inches thick, white in color
Type 1I: 2.52 inches thick, white with green beads
le thicknesses were designed to provide reference thermal resistance values of approximately 10
The sam p
h-fe-'F/Btu. The tested R-values and densities were as follows:
T_Xl�
Aver Density lb/ft' Avera e R-Value h-fF-'F/Btu
1.01 10.2
1 1.46 10.6
II
OBSERVATIONS• '
The general condition of test samples was documented upon removal from the foundation each year:
First
Year: Scattered minor surface cracks and indentations were obser ed brushing). amount
significant
ficant
impregnation in the surface was apparent (that which could nopbarent deterioration of the EPS;boards due
distortion, shrinkage or swelling ha ma was
haveevident.
occurred There was no a p
to any freeze-thaw cycling y
Year: Scattered minor surface cracks and indentations were observed. A greater amount of soil
Second y g No significant
impregnation in the surface was apparent (that which could not be removed b brushing). g
ion shrinkage or swelling was evident. There was no apparent deterioration of the EPS boards due
distort ,
to any freeze-thaw cycling that may have occurred.
Year: Apart from some slight damage during the removal, the two
na ion in the surface s were intact. was sdl
Third Yea psoil
minor surface cracks and indentations were observed. Some significant distortion, shrinkage or swelling
apparent (that which could not be removed by brushing). No sig
was evident. There was no apparent deterioration of the EPS boards due to any freeze-thaw cycling that
may have occurred.
Refer to Appendix C for photographs of the samples.
CORPORATION REPORTS ARE SUBMITTED AS THE CONFIDENTIAL PROPERTY OF CL:E%T
AS A MUTUAL PROTECTION TO CLIE»TS.THE PUBLIC AND OURSELVES All TWIN CITY TESTING
A»0 AUTMCRIZATION FOR PUBLICATION OF STATEMENTS.CONCLUSIONS OR EXTRACTIONS FROM OR REGARDING OUR REPORTS IS RESERVED PENDING OUR PRIOR'N R'^Tvv�a?Pa0\r�
PROJECT NO: 4140 92-2757 DATE: July 23, 1993
PAGE: 6
TEST METHODS (cont.):
Sale Removal
One 2' by 8' piece of each ASTM C578 type was removed from the foundation each fall. The average soil
moisture content at the time of removal varied from 6.8 - 12.4 % by weight (this corresponds to 0.91 -
1.67 gallons of water per cubic foot of soil). Test specimens were removed from each sample, tested,
conditioned at 73.4 ±3.6°F and 50 ±5 % RH to moisture equilibrium, and tested again. The conditioning
atmosphere was considered to be"standard laboratory atmosphere" and moisture equilibrium was considered
to be the point where the weight of the sample did not change by more than 0.1% over a 24 hour period.
Moisture Content -
The moisture content was determined by drying the thermal resistance test specimens and recording the
weight loss. The specimens were dried for 24 hours at 122'F in a circulating air oven (uncontrolled
humidity). The dried specimens are considered to be without moisture and therefore the difference in
weights is attributable to the amount of moisture that was present in the samples. The moisture content was
determined at the time of removal from the foundation and after reconditioning at 73.4 ± 3.6°F and 50
± 5 % RH to moisture.equilibrium. The moisture content was calculated as the volume of water present
in a given volume of sample, on a percentage basis.
Density Test -
The density test was conducted in accordance with ASTM:D1622-88, "Standard Test Method for Apparent
Density of Rigid Cellular Plastics." Three, 12" by 12" specimens were conditioned at 73.4 ± 3.6°F and
50 ± 5 % RH to moisture equilibrium. The length, width, and thickness of each specimen was determined
as the average of three measurements. Each specimen was weighed on a model 1364MP Sartorius digital
balance, SIN 3202200. The density was calculated for each specimen as the weight divided by fhe volume.
Thermal Resistance -
The thermal resistance test was conducted according to ASTM: C518-91, "Standard Test Method for
Steady-State Heat Flux Measurements and Thermal Transmission Properties by Means of the Heat Flow
Meter Apparatus." Three 12" x 12" specimens were removed from each sample, tested, conditioned to
steady-state at 73.4 ± 3.6°F and 50 ± 5 % RH to moisture equilibrium, and tested again. Each specimen
was placed in a thermal conductivity instrument, model Rapid-K, manufactured by Holometrix, Inc.
Steady-state heat flux measurements were made at a mean temperature of 75°F using a hot face tempera4ure
of 100°F and a cold face temperature of 50°F. Thermal conductivity and thermal resistance 'Were
determined by comparing the heat flux measurements of each specimen to measurements made on a
standard reference material of known thermal resistance and thermal conductivity.
DATA:
See Appendix A of this report for the test data for each year.
AS A MUTUAL PROTECTION TO CLIENTS.THE PUBLIC AND OURSELVES.ALL TWIN CITY TESTING CORPORATION REPORTS ARE SUBMITTED AS THE CONFIDENTIAL PROPERTY OF CL EN'S
AND AUTHORIZATION FOR PUBLICATION OF STATEMENTS.CONCLUSIONS OR EXTRACTIONS FROM OR REGARDING OUR REPORTS IS RESERVED PENDING OUR PRIOR WRITTEN APPQCVAL
PROJECT NO: 4140 92-2757 DATE: July 23, 1993
PAGE: 7
CONCLUSIONS:
Based on the test results and observations summarized in this report, the following conclusions can be made
regarding the performance of the EPS samples in the study:
The amount of water absorbed in the samples at the times of removal was less than 3 % (by volume).
The 3 year moisture content averages were 1.18% for Type I and 1.01% for Type H. The EPS
insulation absorbed limited moisture; however it's thermal performance remained within 90% of the
original installed R-value.
A statistical analysis on the individual R-value data for Type I material suggests that the differences
observed between the averages over the 3 years are not statistically significant at a 95% confidence
level. Some of the differences in the Type II material results are considered statistically significant.
These differences are detected primarily due to the fact that there is less variability between specimens
in this data set than in the Type I data set. See Appendix D for the analysis of variance results.
The thermal performance (R-Value) after 3 years was at least 90% of the original tested values.
Therefore, the absorption of moisture had a limited effect on the thermal performance of EPS in this
exterior foundation application.
No significant physical degradation was evident in the samples throughout the three year exposure.
Therefore, the limited moisture absorption and any freeze-thaw that may have occurred had limited
effect on the thermal performance of the material, and did not cause any significant swelling,
cracking, deformation or other visible degradation of the material.
REMARKS:
The test samples will be retained for ninety days from the date of this report, then discarded unless
otherwise notified.
ALK\Thermal\92-2757.R2
AS A MUTUAL PROTECTION TO CLIENTS.THE PUBLIC AND OURSELVES.ALL TWIN CITY TESTING CORPORATION REPORTS ARE SUBMITTED AS THE CONFIDENTIAL PROPERTY OF CL,:E%V
AND AUTHORIZATION FOR PUBLICATION OF STATEMENTS.CONCLUSIONS OR EXTRACTIONS FROM OR REGARDING OUR REPORTS IS RESERVED PENDING OUR PRIOR WRITTEN APoROVA
EPS Facts Regarding
Below-Grade Applications
BELOW—GRADEJ .
Recently a report was presented from Huntingdon Moisture Content
Laboratories with the results of a moisture content Sample Specimen %by Volume
analysis conducted on several samples of expanded •3 Year"(EPs) 1 (Bottom) 2.1
and extruded polystyrene. (exposed approx. 2(Top) 1.5
3 years) 3(Bottom) 5.0
EPS: Observed to be a 2"nominal thickness sub- 4(Bottom) 4.2
( 5(Top) 2.0
mitted to the laboratory on September 20, Average 3.0
1991 and August 3, 1993). std.Deviation 1.5
XPS: Observed to be a 2"nominal thickness(sub-
mitted to the laboratory in May of 1993).
Moisture Content
Sample Specimen %by Volume
"I Year" (EPS) 1 (Bottom) 3.9
(exposed approx. 2(Top) 6.1
The "3 Year" EPS sample was applied to the be- 1 year) 3 (Top) 3.6
low-grade foundation of Huntingdon/Twin City 4(Top) 2.3
5(Bottom) 6.5
Testing's Cromwell Avenue Building on Septem- 5(Top) 7.5
ber 25, 1991. In May of 1993, the XPS sample Average 5.0
was applied to the foundation adjacent to the EPS. Std.Deviation 2.0
The"1 year"EPS sample was applied to the same
foundation on August 10, 1993.
The air temperature on the inside of this founda- Moisture Content
tion wall was maintained at 70+5 F. The samples Sample Specimen %by Volume
were combined in an area 4'wide by 8' deep. The Extruded(XPS) 1 (Top) 7.2
top of the samples were placed just below ground (exposed approx. 2(Top) 11.5
level and the samples were held in place by back 15 months) 3(Bottom) 17.2
4(Bottom) 18.2
fill. Average 13.5
Std.Deviation 5.1
The samples were removed from the foundation
on August 24, 1994. Approximately five speci-
mens from each sample were removed from vari-
ous depths, for moisture content analysis. _.
Moisture content was determined by weighing the Styrotech Incorporated
specimens after removal, then oven drying at 122 8800 Wyoming Avenue North
Brooklyn Park,Minnesota 55445
F reweighing and calculating %by volume. Telephone(612)425-4001 Fax (612)425-8994
LONG TERM PERFORMANCE AND DURABILITY OF EPS AS A LIGHTWEIGHT
FILLING MATERIAL
Tor Erik Frydenlund' and Roald Aabee'
ABSTRACT
Some 30 years of experience with Expanded Polystyrene(EPS)as a lightweight filling material in Norway has
brought about both a wider use on a global scale and the introduction of a number of different design applications.In
addition to reduced vertical loads,advantages from using EPS may also include reduced horizontal loads,simplified
designs,foundations placed directly on EPS blocks and increased speed and ease of performing construction
activities.This document describes practical experiences in Norway with long term performance and durability of
EPS as a fill material based on observations and recorded data from monitoring programmes.
I Chief Engineer,Soil Mechanics Division,Road Technology Department(NRRL),Directorate of Public Roads,
Norwegian Public Roads Administration,Gaustadalleen 25,P.O.Box 8142 Dep,N-0033 Oslo
2 Senior Engineer,Soil Mechanics Division,Road Technology Department(NRRL),Directorate of Public Roads
Norwegian Public Roads Administration,Gaustadalleen 25,P.O.Box 8142 Dep,N-0033 Oslo
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 1
INTRODUCTION
In Norway the Public Roads Administration has a long tradition in applying various types of lightweight filling
materials for road construction purposes.During the last 50 years wooden materials like sawdust and bark residue
from the timber industry have been applied for such purposes.Also waste materials form the production of cellular
concrete blocks and Leca(Light Expanded Clay Aggregate)have been widely used.In this connection also
monitoring programmes were initiated in order to investigate the long term performance of these materials.Presently
a new option is being investigated involving the use of granulated foamed glass produced by re-circulating waste
glass.
When a major research project on frost action in soils was carried out in Norway during the period 1965 to 1973
this included the investigation of various insulation materials for frost protection of roads like 50 to 100 mm thick
boards of foamed glass,extruded polystyrene(XPS)and expanded polystyrene(EPS).In this connection also fatigue
tests were performed.It was then concluded that EPS material could sustain the repetitive stresses occurring in a road
structure and the idea of applying EPS in greater layer thickness than boards emerged.
In 1972 the Norwegian Public Roads Authorities adopted the use of EPS as a super light filling material in road
embankments.The first project involved the successful reconstruction of road fills adjacent to a bridge founded on
piles to firm ground.Prior to reconstruction the fills,resting on a 3 in thick layer of peat above 10 in of soft marine
clay,experienced a settlement rate of more than 200 mm per year_By replacing 1 in of ordinary fill material with
two layers of EPS blocks,each layer with 0,5 in thickness,the settlements were successfully halted.When placed the
EPS blocks had a density nearly 100 times lighter than the replaced materials.
Since then authorities in several countries have also found the method advantageous for building roads across soft
ground and for other construction purposes where low loads are essential.In addition to reduced vertical loads,
advantages from using EPS may also include reduced horizontal loads,simplified designs,foundations on EPS and
increased speed and ease of performing construction activities.The method is now in common use in several
countries in Europe,Asia and North America.At present more than 350 road projects involving EPS fills have been
completed in Norway with a volume of totalling some 500,000 mi of EPS blocks.
EPS does of course not represent the only solution to bearing capacity and settlement problems.Other lightweight
filling materials should also be considered together with other alternatives such as replacement or displacement of
the weak soil or soil improvement,piled foundations etc.Availability and cost are important factors in this
connection,but in many cases the use of EPS will prove advantageous and in some cases represents the only
practical solution.In a book published(1997)by PIARC,the World Road Association,lightweight filling materials
in common use are presented together with case histories.
MONITORING PROGRAMME
Expanded polystyrene is a very stable compound chemically and no material decay should be expected when placed
in the ground and protected according to the present design guidelines.Still,since the first road insulation project
with EPS was performed in 1965 and the first EPS light weight embankment was constructed in 1972,EPS fills have
also been monitored for long term performance along the lines followed for other lightweight filling materials used
in road construction in Norway.
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 2
a 3 �31
4
y p mom. •s s�t �: `°
K.
Figure 1.An EPS embankment for a city tramline in Oslo with vertical walls as an alternative to a bridge.
The monitoring programme has focused on the following material qualities:
• Material behaviour
-Compressive strength
- Water absorption
-Decay
• Deformation
-Total fill deformation and deformation in EPS layers
-Creep effects
• Stress distribution
• Reduced lateral pressure
• Bearing capacity
� a
I
TESTING.FREQUENCIES ��
Since 1972,several tests have been carried out in order to monitor
possible material changes.In this connection test samples have been � ���
retrieved from existing fills to be checked for possible changes in
strength and unit density.Also variations in water absorption for blocks
placed in drained,submerged or semi-submerged positions have been
observed.In order to determine the stress distribution within blocks and
fills both laboratory and field tests have been performed.Finally load
creep effects have been observed both in the laboratory and on existing
fills. Figure 2.Excavation of the first EPS
embankment at Flom bridge
In Norway test samples have been retrieved from a total of five fills.
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 3
The testing frequency is shown in table 1.
Fill location Constructed Test samples retrieved
Year No.of years after construction
National road 159 Flom bridges 1972/73 0 7 12 24
National road 154 Solbotmoan 1975 4 9 21
County road 91 Lenken 1978 6
County road 261anghus 1977 7
National road 610 Sande-Osen 1982 9
Table 1 Testing frequencies of EPS embankments.
MATERIAL BEHAVIOUR
Material strength
According to Norwegian specifications the design compressive strength of EPS blocks have been set to be at least
is= 100 kPa when not otherwise specified In actual practice a shipment of blocks may be accepted if the average
strength of tested blocks is 6�:100 kPa.The average value for test specimens from one block(6 tests)should be
is z 90 kPa and no single tests should show values 6<SO kPa.
One major indicator of possible deterioration of blocks with time would be a decrease in the material strength.The
strength tests performed on retrieved samples from fills having been in the ground for up to 24 years are shown in
figure 4.as a fimction of dry unit density and compressive strength.Bearing in mind the criteria mentioned above for
accepting blocks to be placed in a fill,all test results give values of compressive strength above a=100 kPa except
for one test.
This one test was performed on samples taken from the
' first fill shortly after it was completed in 1972,and is
more an indication of variations in material quality of
EPS with the production process used at that time.Still
` +
F � � the observed value is within the accepted statistical
variations in material strength.
hE From figure 4 it may also be observed that the majority
of tests show values of compressive strength in relation
j; to unit density above that of a normal quality material.
Although it is of course impossible to make exact
u comparisons between material strength at the time of
construction and some time afterwards since tests cannot
be performed on the same specimen twice,the results
Figure 3.Excavation of a 24 years old EPS block indicate clearly that there are no signs of material
from the first EPS embankment at Flom bridge deterioration over the total time span of 24 years.If a
change tendency is to be noted,this would go towards a
slight increase in material strength.If this is the case,such an increase could be explained by a continued chemical
reaction leading to material hardening during the first few weeks after production.There is also a tendency that the
material strength is slightly higher in the middle of the block than towards the outer sides.Furthermore there is no
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001, 3rd International Conference, Salt Lake City,December 2001 4
sign of variation in material strength whether the retrieved specimens are tested wet or dry.This indicates that water
pickup over years in the ground will not affect material strength.
UNIT DENSITY
The only change in design rules that have been introduced in Norway since the first fill in 1972 is that the design
unit weight for EPS blocks placed in a drained position is reduced from y=1.0 kN/m3(p=100 kg/m3)toy=0.5
kN/m3(p=50 kg/m)when stability and settlement calculations are performed.For blocks placed in a submerged or
semi-submerged position the value of y=1.0 Mrnm (p=100 kg/m3)is maintained.The change mentioned above is
based on tests data from existing fills.
EPS placed in the ground will absorb water in two ways.One is by water entering possible voids between spheres
due to water pressure or capillary rise.Since water vapour may diffuse through the polystyrene when there is a
temperature gradient,the
150 water vapour will condense
"; in the spheres if there is a
N s drop in temperature below
140 the dew point.However,in
Z gg an EPS block of 500 mm
4 3E� �'iuz i �R {
130 lt� AT thickness or an EPS fill of
�. ♦ Soibotmoan greater thickness the
ff
= 120 'a * lr ■ Flom temperature difference over
r �f the block or fill will be very
H =i'�� • Lan hus small:Possible water
m
11U le,
absorption due to water
H } ♦ Lenken vapour diffusion is therefore
y
100Sande expected to be small.
CL
s U
YY
E
,f �' Normal qual.
V 90 Water absorption due to
�l 7
f
waxer pressure or capillary
80 rise depends on unit density
18 20 22 24 26 28 30 and how well the spheres are
welded together.A number
Density dry(kg/m3) ( )Year after construction of tests,mainly on small
samples in laboratories,have
Figure 4.Compressive strength on retrieved samples from EPS fills been peed in several
countries in order to study
water absorption effects.
Both the quality of these tests and the results vary somewhat.Tests performed on samples retrieved form existing
fills in Norway are in agreement with some of the laboratory tests.
Tests performed on samples retrieved from three EPS fills placed in a drained position,i.e.blocks are located above
the highest groundwater or flood level,all show water contents below 1 %by volume after more than 20 years in the
ground(fig.5).
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 5
Furthermore there is hardly any change in the water
content with time.Samples retrieved from the outer
parts of blocks facing the surrounding soil,may have
a higher water content as maybe seen from figure 6.
But only 500 mm further into the block the water
content is again below 1%by volume.So the average x
density of drained fills therefore has values of p<30 1 '
k m3.This is well below the ' -�-Solbotmoan
g/ specified design value s t Langhus ''
for such fills.
-�Lenken Q
sus ' ga
In blocks,which are periodically submerged,water
contents of up to 4 /o by volume have been measured
r *� £
In permanently submerged blocks measured water
contents have reached values close to 10%by volume
with some increase over the years fig.7.Further zs
increases above 10%by volume are,however,not to
be expected.For submerged fills the average densityQ
is therefore of the order of p=90-95 kg/m3 after
some 20 years in the ground.The water content Figure 5.Typical drained situation from 3 EPS fills
decreases rapidly above the water table and show
values for drained conditions only some 200 mm
above the highest water level.
M
IM
�N10
�
r r f z t fr a
Or
Q 3 Ham" > df
IN
4.�
srs �P�e •'��, � ' � y ,* s �,¢ �t ..fir �� �5
Figure 7.Typical water content in submerged EPS
Maur 6.Horizontal gradient of water in EPS. Figure
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 6
DEFORMATION AND CREEP EFFECTS IN EPS STRUCTURES
Both full scale and laboratory tests have been performed related to material creep and stress distribution in the
material.In general only about 30%of the material strength is utilised for supporting dead loads,i.e.qaa,<30 kPa
for normal strength blocks(a=100 kPa).In some special cases higher stress related to dead loads have been used.
{r�
3
G
A 3? Y
EE€ Orn
Profile A A 10,5 t
zx s
2mRn
Earth pressure cells
Figure 8.EPS test fill at the Norwegian Road Research Laboratory(now the Road Technology Department).
In a laboratory test at the Norwegian Road
Research Laboratory(now the Road
Technology Department)a test fill of height 2 m
& ' ,
with normal size blocks and a compressive SOD air A �`�
strength 6=100 kPa has been loaded to a valuea
a #�
of q"=52.5 kPa and the resulting deformations �� � 9
observed over a period of 3 ears(figure �� t � €a
p Y 8).The
results are shown in fig.9 together withii, 3
calculated deformations to be expectedAll
k3
according to the theories introduced by Magnan 3
&Serratrice. s
k i Y �
As may be seen the observed deformations are 0
was
only about half of the calculated values and
i �3�' E
creep deformations with time are also much �
smaller.
Figure 9.Deformation/creep in the test fill
Full scale test at Lokkeberg bridge founded on two EPS embankments.Long term monitoring of deformation,
creep and stress distribution.
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001, 3rd International Conference, Salt Lake City,December 2001 7
The L okkeberg bridge is a single lane Acrow steel bridge with a single span of 36,8 m crossing road E6 close to the
Swedish border.The bridge was built in 1989 in order to temporary(3-5 years)improve traffic safety until the
completion of a new motorway between Norway and Sweden.
Due to low bearing capacity and expected large settlements,light weight fill materials(EPS)were considered in the
r s k embankments adjoining the bridge.The project provided
an o ortuni
_ pp ty to place the bridge foundation directly on
:_ V top of the EPS fills(height 4,5 and 5m)on both sides as
an alternative to placing the bridge abutment on piled
foundations.Since the bridge was a temporary solution
and possible deformations could be adjusted during the
period of operation,it was decided to carry out the
project as a full scale test.
v
., �
�- Three different g�
qualifies of EPS material strep have
? s been used with design strengths of a=240 kPa in the
�F T = upper layer directly below the bridge abutment,a=180
kPa in the remaining layers halfway down the fill and 6
=100 kPa in the bottom half.In the upper layer only 25
%of the material strength has been utilised while in the
bottom layer the corresponding figure is 60%.
Construction details are shown in longitudinal profiles in-
figures 11 and 12.
Figure 10.Construction of one abutment on the EPS The bridge is today still in operation 12 years after
embankments at Lekkeberg bridge. completion.No signs of cracks or uneven deformation
36,8 m
€�� Shotcrete
E6 � F
Sand/gravel
M '
' � o
I
\\ i a avLIGG FOR B
have been observed.The bridge support has been lifted 30 cm on one side due to subsoil settlements in accordance
with the theoretical calculation.
L okkeberg bridge has provided a good opportunity for monitoring long time performance such as creep and stress
distribution of an EPS embankment.
Long term performance and durability of EPS as a lightweight filling material
to 5 rence, Salt Lake City,December 2001 8
24okP8 After 12 years in service measurements show only
small deformations 6 cm(1,3%of the EPS
acm �1ao�a height)in the EPS embankment.Most of the
EPS s
6,2crr4 deformation occurred during the construction
o=tookPa period and only minor creep effects have been
s.7om measured.Creep deformations as an average and
-------------------------- --- --wnd -- creep deformations for the lowest EPS layer(6,5
Gravel/sand %of the layer thickness)are shown in fig. 13 for a
day period of 10 years.
Figure 12.Deformations in EPS embankment at Lokkeberg. Observed deformations after 10 years in
operation are plotted in fig. 14 together with data
Time from the laboratory test and theoretical values
according to Magnan&Serratrice calculated for
1989 1991 1993 1995 1997 1999 various stress levels
0
The figure clearly shows that the average
5 deformation at the Lokkeberg bridge is small and
10 —TOS somme1t slightly over 1 %of the total fill height.Also
,a --U-.seW in ciay observed creep effects are almost negligible for
M 15 Ddwmfflm M 4,5 m� the total fill although deformations in the bottom
block layer was 4%initially and later creep
20 -°�"`�°""' eas�' effects amount to further 2.5%.Creep -
deformations in the bottom layer correspond with
v, 25 the theoretical values the first 5 years but has later
30
slowed down to almost zero.
Two similar structures, a 3 span bridge at
Yy7 '' t 1a AT? - ff �,'"' _, frit
Hjelmungen and a 3 span pedestrian bridge
�� �r has been built in 1994 and 1995 with
Fi � � uqq
"'� abutments founded on EPS fills.
gam ,rWOW
� Observations from these bridges
Y 2.
', x correspond with the measurements at
0_55 � s + gg Lokkeberg and shows that this can be a
j promising method for supporting bridge
� � abutments.
r
�Wk A,
,.0
z
3
00,
!�
re �
x �
...3x3 ��' Y
" RA
r 3�
r � F
WF` y X51 = af3yIi` Y
y #� 3 Awl,
� '
a din_
Figure 14.Creep deformations in EPS.
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001, 3rd International Conference, Salt Lake City,December 2001 9
STRESS DISTRIBUTION
In order to observe the stress distribution in the EPS material below the bridge abutment at Lokkeberg bridge
during construction and on a long term basis, 10 hydraulic earth pressure cells have been placed in different levels in
the fill including 3 cells in the sand layer below the EPS fill.In fig. 15 the measured stress level after 10 years in
— - Settlement tube
Earth pressure cells
Abutment
10cm concrete slab
_.79 Figures in red
7m 1 indicates stress
EPS 83 52 50 level in kPa
26
m
* "
E
28
4 E
F 30 a
c�
Figure 15.Lekkeberg bridge.Observed stress distribution and settlements in the cross
section after 10 years in operation.
service is indicated with red figures.
Observations indicate that cell
boundary effects may have
1989 1991 1993 1995 1997 1999 influenced the stress results,
120 especially in the first loading
CL stage,probably due to poor
m100 interaction between EPS and
the steel casings for the earth
m
80 pressure cells.
C80 Long term measurements from
.e 1 3 earth pressure cells below the
cc
fill and one situated 2 m higher
40 —Vertical load up have been plotted in fig. 16.
Earth pressure 2 m above the bottom During the first year of
41
20 —a Earth pressure below embankment,middle operation a stress decrease of
In
0 0 —Earth pressure below embankment, left 15-30%was observed.Later
—Earth pressure below embankment,right only small variations with time
have been observed. Measured
stresses corresponds well with
the theoretical vertical load in
Figure 16.Long term measurements of earth pressure at Lekkeberg bridge the lower part of the EPS fill.
In the upper part of the fill(with a higher material strength)lower stress than expected has been measured in a zone
under the central part of the embankment.One explanation could be some kind of arching effect due to large subsoil
settlements in the middle of the embanlanent.This can clearly be seen in fig.IS where the measurements from the
settlement tube are shown.In the lower part of the fill it is difficult to explain load concentration and deformation in
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001, 3rd International Conference, Salt Lake City,December 2001 10
the lowest EPS layer and likewise the low stress values in the upper part of the EPS fill.It may therefore be
concluded that there still may be load distribution mechanisms in EPS fills that are not fully understood.
Another test was performed to check the stress result.A dumper with a weight of 33 tons was placed at different
distances from the abutment.In the case when it was placed directly upon the abutment an increase of 6 kPa was
expected.The measured stress increases
in the various fill levels with the
Load=33 ton additional load from the dumper
correspond well with the stress
distribution without the dumper.The
increase in stress levels from the dumper
is shown in fig.17.The same tendency to
reduced vertical pressure in a zone
______ beneath the abutment was also observed
here.After unloading the abutment,the
3KN`�_ e.. ..a pressure cells immediately returned to
7 the initial stress levels.
Earth pressure cells 4kN t7?UWffr 4m le
EPS 2kN m
5 m ,Sm Attempts have been made to evaluate
6 the stress distribution in EPS blocks
based on stress observations from the
_-(SMM2 4m fightt�6kN/m 7kN/m ;4mTeft�• test hall experiments and stress
observations at the Lokkeberg bridge.
(gravel/sand) Stresses between blocks are of course
Cell 1 2 difficult to measure and the results
Yobtained vary quite a bit.In general
Cell 5 15 measured stresses are,however,relatively low
and indicate that the outer perimeter of the
rn
Stressdistribution 2:1 10 stress bulb may lie within a slope with a
Fid ss 'on from addi ' jpa gradient 2:1 measured from the outer edge of
0 5 the loading area.However,depending on the
w CII 284 0 stiffness of the loading area and possible load
eccentricities,local stress concentrations may
0 500 1000 occur..In figure 18 the stress level at the
bottom of the fill with distribution gradient 2:1
Time (days) is indicated with a black line and may be
compared to the stress levels of cells 2,3,4 and
Figure 18.Stress distribution measured in the test hall(fig.8) 5.
REPAIR OF HJELMUNGEN BRIDGE
Hjelmungen bridge is a three span,54 in long continuous concrete deck bridge completed in 1992 with abutments
and pillars founded on concrete piles to firm ground.The 5 in high fills adjoining the bridge consisted partly of
ordinary filling materials partly of waste material from the production of Leca building blocks.The fills rested on
subsoil consisting of some 10—14 m of soft sensitive marine clay,partly quick and with a high water content.
Some 2 years after completion it became evident that the bearing capacity of the soil beneath the abutments had
been exceeded as excessive settlements occurred and the abutments inflicted damage to the bridge deck.
Deformation monitoring was initiated and it soon became clear that immediate repair measures had to be initiated.
Since settlement caused by the approach fills was the main problem,it was decided to reduce the load on the subsoil
by some 30—40 kN/m2 in order to re-establish the initial subsoil stress conditions.This involved replacing parts of
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 11
the fills with EPS blocks and supporting new bridge abutments directly on the EPS.The repair design is indicated in
figure 19.
20 cm apron 10 cm.concrete
slab
aK
it
Figure 19.Supporting bridge abutments directly on EPS,Hjelmungen bridge,Norway
Repair works were initiated in December 1995 and completed in the spring of 1996.One abutment was treated at a
time while the bridge deck was provided with a temporary support as shown in figure 20.Thickness and densities of
the original filling materials were recorded as they were removed in order to have accurate data for control of load
and settlement calculations.After removing the old abutments,the concrete piles were inspected regarding possible
damage before being cut at ground level.No pile damage was observed.Construction of the EPS fills could then
start.In this case three different qualities of EPS were utilized.In the zone directly beneath the bridge abutment,as
indicated by the trapezoidal shaped lines in figure 19,a material quality of 6=235 kPa was specified for the first
three block layers beneath the bottom slab of the abutment.Further down a material
quality of a=180 kPa was
77
i
a>
k3 u
Figure 20.Temporary support of abutments at Hjelmungen
specified within the indicated zone.For the rest of the EPS fill a material quality of 6=100 kPa was used.These
quality requirements have been decided based on evaluation of stress distribution in the material in order to keep the
stress level for dead loads within 30%of the material strength.Stricter geometric requirements than normal were
also enforced related to block dimensions in order to obtain an even fill and reduce initial deformations when the
load from the bridge deck was transferred to the new abutment.
Behind both abutments a 10 m long and 200 mm thick concrete apron was specified to be cast above the EPS fill as
a friction plate in order to take up horizontal forces on the abutment.On the rest of the EPS fill a concrete slab of 100
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 12
mm thickness was specified.To complete the road pavement 400 mm of pavement material was placed on top of the
concrete slab.
Bridge abutment
0 x
, �
1 =Telescopic rods 3 = Settlement tube
A - E - Pressure cells
Figure 21. Cross section indicating location of monitoring equipment
In order to monitor the behaviour of the reconstructed bridge both settlement and stress gauges have been installed.
The different types of gauges and their locations in relation to the bridge abutment are indicated on the cross section
in figure 21.Prior to reconstruction the settlement rates of the adjoining fills were observed to be 100 mm/year and
constant.After reconstruction the settlements have nearly been halted as shown on figure 22.
Observed stresses beneath the EPS fill indicate a higher stress under the central part of the abutment than under the
edges as shown in figure 23.Calculated loads on the abutment are indicated by the heavy line drawn in the diagram.
Problems associated with providing enough lifting force when jacking up the bridge deck may,however,indicate
that reaction forces from the bridge deck are somewhat higher than calculated.
Hjelmungen bridge Axis 1 80
Settlements in 5 m EPS
70
0 +— 4 60
-5
E -10 .Z 50
E -15 .-.—Right
position m
c -20 t Left 40 -
E -25 position m
2 -30 a 30
m
N -35 t
20
-40 to
ic
Loading situation —CellA
�-45 10 Cell D -- -Cell C
jan.96 Jan.97 jan.98 jan.99 jan.00 Jan.01 eil B —6 Cell E
0
Figure 22. Measured settlements at Hj elmungen Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001,3rd International Conference, Salt Lake City,December 2001 13
Figure 23.Earth pressure below EPS layer at Hjelmungen
DURABILITY
Although polystyrene is a stable chemical compound it may dissolve when exposed to petrol agents.When placed
in a road fill the EPS blocks are therefore protected either by a concrete slab on top of the blocks also serving as a
load distributing layer,or a high density polypropylene sheet.Although it is possible that a petrol tanker may
overturn and spill petrol on the road surface at the location of an EPS fill,the statistical likelihood that such an event
should occur is very small and the precautions mentioned above should be sufficient to protect the EPS.Also it will
take some time for the petrol fluid to percolate through the soil on the side slopes,allowing time for corrective
measures.Still,if a petrol spill should find its way to the EPS,only the outer blocks are likely to be effected and
repair should be easy to perform.During the nearly 30 years that have passed since the first EPS fill was placed and
the accelerated use of the EPS method on a worldwide scale in later years,no such spill incident has been reported.
FAILURES
Of the many EPS projects now completed in many parts of the world,only five known failures have been reported.
Two failures are associated with water fluctuations and buoyancy forces.The other three are caused by fires.
On the 16s'of October 1987 Northern Europe experienced exceptionally strong storms with high wind velocities
and high rainfall intensities.Norway was also exposed to major floods,and in the Oslo area the first EPS fill built in
1972-floated off_as did an adjacent section of motorway constructed-somt years later.What was wrong?Had the
dangers of buoyancy forces not been considered?Yes,such calculations had been performed,but the highest
possible flood level predicted at the design stage in 1972 was 0,85 in lower than the flood level that occurred in
October 1987.So it was rainfall and flood level predictions in 1972 that were misleading.
Also the second failure reported from Thailand involved an unexpected high water level causing a completed road
fill to be washed away.So it should be duly noted that the dangers of buoyancy forces should be carefully studied
when considering the design of an EPS fill.Often soft subsoils are located in lowland areas subjected to flooding.In
such cases accurate predictions of the highest possible water level are essential in order to obtain a safe and lasting
road structure.
Ordinary polystyrene is a combustible material and will burn when set on fire.For this reason some precautions
should be taken when constructing EPS fills using normal quality material.Such precautions may include fencing in
any stockpiles at the construction site and provide guards round the clock,or place the blocks directly in the fill when
they arrive on site,working round the clock if necessary.Alternatively a self-extinguishing quality of EPS may be
used at approximately 5%increase in production costs.However,once the EPS is covered by the pavement material
on top,and soil on the slopes,there will not be sufficient oxygen available to sustain a fire.
Two failures due to fires have occurred in Norway,and both were caused by welding activities on bridge abutments
adjacent to EPS fills,during the construction phase.In the first case 1500 m3 of EPS were transformed into black
smoke in a matter of some 10 minutes.The concrete bridge abutment was also damaged due to the heat developed
with concrete spalding from the reinforcing bars.Since the fire was initiated by sparks from welding activities on the
bridge,the contractor responsible for the welding had both to repair the bridge abutment and replace the EPS fill at
his own expense.A similar incident occurred in 1995 and again the repair costs had to be covered by the contractor
responsible for the welding activities.So the fire potential should not be overlooked and in some counties in Norway
the local highway offices are only using self-extinguishing material at a somewhat higher cost in order to exclude
fire hazards.
Long term performance and durability of EPS as a lightweight filling material
EPS Geofoam 2001, 3rd International Conference, Salt Lake City,December 2001 14
A third fire incident is reported from Japan.
CONCLUSIONS
From the observations discussed above it may be fair to conclude that no deficiency effects are to be expected from
EPS fills placed in the ground for a normal life cycle of 100 years.This should hold true provided possible buoyancy
forces resulting from fluctuating water levels are properly accounted for,the blocks are properly protected from
accidental spills of dissolving agents and the applied stress level from dead loads is kept below 30-50%of the
material strength.The observed performance of the many projects designed and constructed on these principles
around the world so far supports this conclusion.
REFERENCES
Frydenlund T.E,Aaboe R.Expanded Polystyrene The Light Solution.International symposium on EPS
construction method.Tokyo 1996.
Aaboe,R.Deformasjonsegenskaper og spenningsforhold i fyllinger av EPS(Deformation characteristics and stress
conditions in fills of EPS)Internrapport 1645.Public Road adm. 1993.
Skuggedal H,Aaboe-R.Temporary overpass bridge founded on expanded polystyrene.Proceedings XECSMFE,
Florence May 1991,Volume 2.
Magnan&Serratrice,Propriete mechanique du polystyrene expanse pour ses application en remblais routier.
Bulletin 1CPC,France, 1989.
Long term performance and durability of EPS as a lightweight filling material