United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-95/103 July 1995
Project Summary
Laboratory Assessment of the
Permeability and Diffusion
Characteristics of Florida
Concretes: Phase II. Field
Samples and Analyses
R. Snoddy
The report gives results of a study to
establish the capability to measure
concrete's permeability and diffusivity;
to measure these parameters in a small
sampling of the typical types of Florida
concrete; and, if possible, to correlate
the physical parameters of the concrete
(mix design, porosity, etc.) to the mea-
sured diffusion and permeability coef-
ficients.
Sample permeability was measured
and analyzed with a device developed
for the project. The diffusion coefficient
was determined using a system also
developed for the project.
The statistical analysis consisted
of 1) a stepwise search using an
Efromyson-type search methodology,
2) a standard linear-modeling, least-
squares approach to evaluate indi-
vidual variables and sets (up to sets
of three), and 3) robust or least-me-
dian-square methods to further evalu-
ate possible correlations and examine
the dataset for outliers.
For both permeability and diffusivity,
the amount of water added to the mix
at the site was directly and positively
related and identified as a possible
major factor. The data indicated that
the total amount of sand and stone in
the mix was possibly correlated to the
permeability. A third possible correla-
tion involved the amount of fly ash or
the cement to fly ash ratio.
This Project Summary was developed
by EPA's National Risk Management
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).
Overview
Much of Florida's natural soil, as well
as the disturbed overburden and the sand
tailings from the phosphate mining/
beneficiation process, contains significant
quantities of radium. Buildings constructed
on these high-radium soils contain elevated
radon gas levels. Elevated indoor radon
gas levels can cause lung cancer in hu-
mans.
To decrease elevated radon gas levels,
Florida's legislature instructed its Depart-
ment of Community Affairs to develop new
construction standards for radon-resistant
buildings, primarily slab-on-grade construc-
tion.
It is well-known that the concrete slab is
the primary barrier to radon entry. How-
ever, the extent of its ability to permit air
flow under pressure (permeability) and to
permit the passage of radon gas without
any pressure difference (diffusivity) has
not been well-determined. To establish a
standard concrete mix and its maximum
radon-resistant placement, these param-
eters needed to be quantified and their
Printed on Recycled Paper
-------
relationship to concrete's physical proper-
ties evaluated.
Primary research objectives were to es-
tablish the capability to measure concrete's
permeability and diffusiv'rty; to measure
these parameters in a small sampling of
the typical types of Florida concrete; and,
if possible, to correlate the physical pa-
rameters of the concrete (mix design, po-
rosity, surface finish, etc.) to the measured
diffusion and permeability coefficients.
Sample permeability was measured with
a device developed for this project. The
sample holder's open end was sealed air-
tight into the permeability test fixture us-
ing Mortite, a non-hardening, clay-type
sealant. Atop plate kept the sample holder
sealed to the fixture. Compressed air at
25 psi (172 kPa) (nominal) pressurized
the space enclosed by the test fixture and
the bottom side of the concrete. The pres-
surizing valve was shut, and a pressure-
sending unit measured the pressure in
the sealed volume. As the air escaped
through the concrete, the pressure de-
creased.
The diffusion coefficient was determined
with a system developed by and purchased
from Rogers and Associates Engineering
Corporation. The method uses uranium
mill tailings as a strong emitter of radon
gas. The tailings are in a 30-gal (114 L)
drum the lid of which has a fitting into
which the sample holder is mounted, and
the detector assembly is mounted on top
of the holder. After the background count
rate is measured, the valve between the
drum of radon gas and the bottom surface
of the concrete sample is opened. The
scalar/ratemeter counts and produces a
paper record of the number of counts per
interval. When the count rate stops in-
creasing, the radon gas in the drum and
in the space above the concrete has
reached equilibrium. The valve is shut,
arvdvho samp\e holder and detector appa-~
ratus are disassembled.
The permeability time-versus-pressure
data were analyzed by software written
for this permeability determination method.
The software also provided the automatic
data collection system. Usually, data were
collected from the pressure sender every
10s. Then, six of the data points were
averaged to produce a raw data point
every 1 min. This could be varied; for
some of the high permeability concrete
samples, raw data points were saved each
second with no averaging. The sampling
technique allowed more data to be col-
lected and improved the standard devia-
tion.
The physical parameters of the test (air
temperature, sample thickness, sample
diameter, and volume under pressure)
were used to calculate the permeability
coefficient. The software requested the
data as measured, then converted it to
the appropriate units. The errors in each
parameter were used to determine an es-
timated standard deviation for the calcu-
lated permeability coefficient. The results
were written to a data file in a format
suitable for printing.
Rogers software was used to determine
the diffusion coefficient. The software uses
10 pairs of data points from the break-
through region of the alpha activity data.
For this work, 10% offsets from the
baseline and the equilibrium level were
used to determine the breakthrough re-
gion.
Within this region, 10 data points spaced
evenly in log time were selected. The next
highest adjacent data point was used as
the second data point of the pair. The first
data point of the pair was used to calcu-
late the diffusion coefficient. The second
data point was used to estimate the stan-
dard deviation of the diffusion coefficient.
Proper function of the diffusion test
system was verified by testing four ma-
terials with different diffusion coefficients:
air, 4 mm glass beads, coarse sand,
and fine sand. Because the permeabil-
ity test was a new method, there were
no available data on the permeability of
concrete, and there was no test stan-
dard for calibration purposes. It is not
possible to calculate the percent differ-
ence between the actual and measured
permeability values. The data collected
in these tests compare favorably (order
of magnitude) with the expected perme-
ability coefficients. The limiting factor in
the permeability test is the system leak
rate. This leak rate imposes a minimum
detectable limit of 10'20 m2 which is a
factor of 10,000 lower than the concrete
—samples- measured:—• — —r-.r—T-T-^~
A statistical analysis of the data was
performed using the S-Plus statistical
analysis package for desktop computers.
A three-step process was used to evalu-
ate possible correlations of the mix design
and other factors to the permeability and
diffusion coefficients. The first step used a
stepwise search of the possible explana-
tory variables using an Efromyson-type
search methodology. The second step
used a standard linear-modeling, least-
squares approach to evaluate individual
explanatory variables and sets of explana-
tory variables (up to sets of three). The
third step was to use robust or least-me-
dian-square methods to further evaluate
possible correlations and examine the
dataset for outliers. The correlation analy-
sis has provided some indications of the
components of mix design and placement
which may have the greatest effect on the
permeability and diffusivity of the concrete.
The primary .correlation to the diffusion
coefficient was a direct relationship to the
amount of water added to the mix at the
site before the concrete was placed. Other
possible correlations were the cement to
fly ash ratio and the total amount of fly
ash in the mix. The permeability coeffi-
cient was most correlated to the total
amount of aggregates (sand and stone) in
the mix. Other possible correlations to the
permeability coefficient were the amount
of water added at the site, the ratio of the
total amount of aggregates to the total
amount of cementitious materials, and the
amount of-fly^ash-in the mix. ,.
The objectives of Phase II of this project
have been met. Concrete samples from
typical residential slabs from throughout
Florida have been tested. The mix design
and placement data have been examined
for correlations to the permeability and
•diffusion coefficients.
For both permeability and diffusivity, the
amount of water added to the mix at the
site was directly and positively related and
identified as a possible major factor. The
data indicated that the total amount of
sand and stone in the mix was possibly
correlated to the permeability. A third pos-
sible correlation involved the amount of fly
ash or the cement to fly ash ratio. Fly ash
is commonly used as a cementitious ma-
terial in addition to cement powder. It has
good properties as a binding agent but is
more porous than cement powder. This
porosity may be negated somewhat by
the small size of the fly ash particles and
the ability of the cement powder to seal
these small porosities and reduce their
effect. The data for the admixtures were
not included in the correlations since they
•^tended-to.-be-unavailable-or-unreliabJe——-
It is also interesting to note that samples
c002a, c002b, and c018, which had de-
fects, all have high permeability and aver-
age diffusivity. This indicates that small
defects with high tortuosity have small ef-
fects on the diffusivity but are very good
channels for air flow.
First and foremost, great care should
be taken to ensure that the concrete slab
remains a fully intact barrier. The effects
of a very tight concrete mix are nullified
by cracking and unsealed penetrations.
Proper grading/compaction will help to
eliminate settlement cracking. Drying
shrinkage cracks can be reduced by en-
suring that the slab wet cures for an ex-
tended time to allow the concrete to gain
strength before stresses due to drying are
-------
placed on the slab. Reducing the amount
of water in the mix may also reduce the
amount of drying shrinkage cracks. In-
creasing the amount of cement in the mix
will reduce radon transport and produce a
better quality of concrete but will also in-
crease the cost.
Additional barrier measures should also
be considered. A plastic sheet or layer of
asphaltic cement will reduce both trans-
port and the impact of defects in the con-
crete. Coating the top of the slab (e.g.,
painting) could also be used to seal im-
perfections and small cracks in the slab. It
is important that cracks are sealed to depth
to prevent the soil gas from bypassing a
thin surface seal. Resealing could be done
when the floor coverings are replaced.
All of these options support the primary
goal: to provide an intact primary barrier
with reduced transport properties.
-------
R Snoddyls with Acurex Environmental Corp., Research Triangle Park, NC 27709.
David C. Sanchez is the EPA Project Officer (see below).
The complete report, entitled "Laboratory Assessment of the Permeability and
Diffusion Characteristics of Florida Concretes: Phase II. Field Samples and
Analyses,'(Order No. PB95-243168 Cost: $19.50, subject to change) will be
available only from
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at
National Risk Management Research Laboratory
(formerly Air and Energy Engineering Research Laboratory)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
National Risk Management
Research Laboratory (G-72)
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use
$300
EPA/600/SR-95/103
------- |