United States
            Environmental Protection
Office of
Radiation Programs
Washington, DC 20460
EPA 520/5-83/021
September 1983
            International Meeting on
            Radon-Radon Progeny

                   August 27-28, 1981
                      Sponsored By

          U.S.  Environmental Protection Agency
              Office of Radiation Programs
                    Washington, D.C.
                       Prepared By
                 Dingle Associates, Inc.
            1625 Eye Street, N.W.,  Suite 915
                Washington, D.C.  20006

                               TABLE OF CONTENTS

                                                                     ii ji

Introduction	1

Agenda	2


    The EWLMII, A Measurement and Data Acquisition System for
    Radon Daughters
         D. J. Keefe, W. P. McDowell, and P.  G.  Groer	5

    The Measurement of Radon-222 in Air by Direct Extraction
    Into a Liquid Scintillator
         Howard M. Prichard	20

    Application of Surface-Barrier Detectors  for the Measurement
    of Environmental Radon and Radon Daughters in Air
         Andreas Wicke and Justin Porstendorfer	34

    Recent Advances in the Design and Use of  Passive Radon Monitors
         Lyle Rathbun	47

    Track Etch® Radon Detector Calibrations and Field  Results
         H. Ward Alter and James E. Gingrich	55

    Assessment of Radon Measurement Techniques Used for  Uranium
         John C. Pacer	78

    A Review of Some Exact Methods for 222Rn-Daughter  Measurement
         Peter G. Groer	96

    Track Etch Calibration:  Design and Results of a Pilot Study
         Susan M. Hinkins	97

    Measuring Radon Source Magnitude in Residential Buildings
         W. M. Nazaroff, M. L. Boegel, and A. V. Nero	101

    Field and Laboratory Measurements of Radon Concentrations
    in Maine Houses
         C. T. Hess, C. V. Weiffenbach, H.  R. Prichard,  and
         T. F. Gesell	125

    Surveys and Decision Making in a Remedial Action Program
         A. G. Scott	138

    Airborne Rn-222 in Buildings Constructed With High-Radium-
    Content Concrete Blocks
         Bernd Kahn, Marcia Wilson,  and John T.  Gasper	147

    Environmental Radon Investigations  in  Sweden
         Gustav Akerblom	171

    Effect of Air Circulation and Dust  Removal on Indoor  Radon
    Decay Product Concentration
         S. N. Rudnick, W.  C.  Hinds, E.  F. Maher, J. M. Price,
         and M. W. First	189

    Some Measurements of the Equilibrium Factor  for  Radon
    Daughters in Houses
         R. E. Toohey, M. A. Essling, H. Wang, and J.  Rundo	201

    Regional Environmental Documentation of  Natural  Radiation
    in Sweden
         Carole Wilson	209

    An Apparatus for Calibrating Passive Integrating Radon
    Monitors  (PRIMS)
         Robert A. Washington	234

Open Discussions

    I.  Instrumentation and Measurement Methods	235

    II.  Measurements and Related Topics	262


    With the emergence of worldwide interest in radon-radon progeny levels  in
structures, many different methods and instruments  for  determining these
levels have been developed.  The Office of Radiation  Programs  (ORP), of the
U.S. Environmental Protection Agency (EPA), has been  developing a program to
determine the significance of exposure from this radiation source.  It has
been ORP's objective to use appropriate measurement methods to assure that  the
results of studies are meaningful and to interpret  the  results of previous
studies so that the limitations of the measurement  methods are understood.

    With this objective in mind, EPA sponsored a two-day  international meeting
on radon and radon progeny measurements.   The purpose of  the meeting was to
promote an interchange of information between international experts.  The
meeting was held August 27-28, 1981, in Montgomery, Alabama, at the Holiday
Inn - State Capitol.  Mr. Charles R. Phillips, Chief, Technical Support Branch
of the Eastern Environmental Radiation Facility (EERF)  was the meeting

    In addition to the 48 participants who represented  Federal and State
government offices  (both U.S. and abroad)  and private industry, some 15
attendees included personnel from the Eastern Environmental Radiation Facility
(the host site).  The countries of Australia, Canada, Federal Republic of
Germany, Great Britain, and Sweden were represented.

    The meeting format included formal presentations  and  directed
discussions.  Eighteen papers were presented on such  topics as measurement
methods and instrumentation and surveys and measurement programs using such
instruments and methods.  Participants also toured  the  EERF on the last
afternoon of the meeting.  EERF personnel were on hand  to host the tour and
answer questions from the group.

    An agenda and list of participants are included in  the succeeding pages.
The proceedings that follow include the complete texts  of 15 papers, abstracts
of two papers, and edited transcripts of the two open discussions:
"Instrumentation and Measurement Methods" and "Measurements and Related

                      U.S. Environmental Protection  Rgency

                                  mEETinc on  Rpoon-Rpoon
                           PROGGDV  mensuRemenis
                                flugu/t 27-28, 1981

                             Holiday Inn- State Capitol
                              fTlontgomery. filabama

WEDNESDAY, August 26, 1981

       6:00 - 8:00 p.m.

THURSDAY, August 27. 1981

       8:00 - 9:00 a.m.

       9:00 - 9:20 a.m.
        9:20 - 12:25 p.m.

        9:20 - 10:00 a.m.

        10:00 - 10:25 a.m.

        10:25 - 10:45 a.m.

        10:45 - 11:10 a.m.



Introduction of Alabama Officials
Welcome from the State of Alabama
Charles R. Phillips, Meeting Coordinator

Opening Remarks
   David M. Rosenbaum,  Ph.D. Office Director
   Office of Radiation Programs
   U.S. Environmental Protection Agency

   Session I
   Chair:  Peter G. Groer

"Environmental Measurements Laboratory Developed
 Instrumentation" *
   by Andsi&aA Geo/uje

"The EWLM II,  A Measurement and Data Acquisition
 System for Radon-Daughters"
   by VonaJLd J. Keefje


"The Measurement of Radon-222 in Air by Direct
 Extraction into a Liquid Scintillator"
   by  Howcuid M.
 *Invited paper

        11:10 - 11:35 a.m.

        11:35 - 12:00 noon

        12:00 - 12:25 p.m.

        12:25 -  1:45 p.m.

        1:45 - 3:00 p.m.

        1:45 -  2:10  p.m.

        2: 10 -  2:35  p.m.

        2:35 -  3:00  p.m.

        3:00    3:20  p.m.

        3:20 -  5:20  p.m.

        7:00 - 11:00  p.m.
"Application of Surface Barrier Detectors for
 the Measurement of Environmental Radon and
 Radon Daughters in Air"
   by hndA&cu,  WxLcfee

"Recent Advances in the Design and Use of Passive
 Radon Monitors"
   by Lytt A.  Ratkban

"Track Etch Detector Calibrations and Field Results"
   by Jam£4 E.  GJ-ngsvick


   Session II
   Chair:  Ron Colle

"Assessment of Radon Measurement Techniques Used
 for  Uranium Exploration"
   by Jokn C.  PO.C.&I

"A Review of Some Exact Methods for Rn-222 Daughter
   by P&tzA G.  G/toeA

"Track Etch Calibration Design and Results of a Pilot
   by Sttion H^Lnk^n^


                   MEASUREMENT METHODS
Moderator:       Peter G. Groer
Panel:            Speakers for Sessions I and II

Riverboat Cruise and Dinner
FRIDAY, August 28, 1981
        5:30 - 10:55 a.m.
        3:30 -  8:55 a.m.
   Chair:  Andreas George

"Measuring Radon Source Magnitude in Residential
   by W-LLtiam  W.
*Invited paper

       8:55 -  9:20 a.m.          "Field and Laboratory Measurements of Radon Concen-
                                 trations in Maine Houses"
                                   by Chanter T.

       9:20 -  9:45 a.m.          "Applications of Single Grab Sample Measurements to
                                 the Decision-Making Process Involved in Choosing
                                 Houses for Remedial Action" *
                                   by M£huSL Scott

       9:45 -  10:05 a.m.         Break

       10:05 - 10:30 a.m.         "Airborne Rn-222 in Buildings  Constructed with High
                                 Radium-Content Concrete Blocks"
                                   by Behind Kakn

        10:30 - 10:55 a.m.        "Environmental  Radon Investigations in Sweden"
        10:55- 1:00 p.m.        RELATED TOPICS
                                   Chair:  Bernd Kahn

        10:55 - 11:20 a.m.        "A Simple Procedure for Calibration of Continuous
                                 Flow-Through Radon  Monitors" **
                                   by EnLin.0, S&ia.nde.n

        11:20 - 11:45 a.m.        "Effect of Air Circulation and Dust Removal on Indoor
                                 Radon Decay Product Concentrations"
                                   by MeJLvsLn W.  TJJtbt

        11:45 - 12:10 p.m.        "Some Measurements of the Equilibrium Factors in
                                   by Hic-ka/id E.  Toonuj

        12:10 - 12:35 p.m.        "Regional Environmental Documentation of Natural
                                 Radiation in Sweden"
                                   by CaAotz. WxXion

        12:35 - 1:00 p.m.        "An  Apparatus for Calibrating Passive Integrating
                                 Radon Monitors"
                                   by Rob&it A. WaAhsLngton

        1:00 -  2:00 p.m.         Lunch

        2:00-  4:00 p.m.         OPEN DISCUSSION:  MEASUREMENTS AND  RELATED
                                Moderator: Arthur Scott
                                Panel:  Speakers for Measurements and Related Topics

        4:00 p.m.                 Closing

*Invited paper       **Paper  not presented


                   D.J.  Keefe, W.P. McDowell and P.G. Groer*

                             Electronics Division
                          Argonne National  Laboratory
                              Argonne,  IL  60439

    Historical Background1•2'3

    In 1972, the Radiological  and Environmental Research  Division of Argonne
National Laboratory contracted with the United States  Bureau of Mines for the
development of three identical instruments to measure  Rn-daughter concentra-
tions and Working Level (WL),  in uranium mine atmospheres.  This project,
under the direction of Dr. Peter Groer, led to the  development of the Instant
Working Level Meter (IWLM).   The basic principles for  the Groer method of
measurement of Rn-daughter concentration were developed and proven with these
instruments.  These IWLM's were small portable units uitlizing a calculator
chip to automatically make the mathematical calculations  required to compute
the WL.  CMOS circuitry and a  small 12 liter/minute pump  were used to permit
battery powered operation.

    This development let to another contract between the  Electronics Division
of Argonne National Laboratory and the Bureau of Mines for the development of
the Remote Working Level Monitor (RWLM).  The RWLM's electronics package was
micro-processor based to provide automatic operation.  The monitor was pro-
grammable via a keyboard to start and stop at selected times and to take meas-
urements at desired time intervals, providing a fully  automatic operation.
This instrument differed from  the IWLM in several ways:

    It was a stationary unit with two remote monitoring stations automatically
controlled from a central micro-processor.  It was  similar in its measurement
sensitivity, (.01 WL)  and utilized a flow rate of 12 liters per minute.  The
RWLM, however,  contained an automatic filter advance mechanism where the IWLM
was manually operated.  The data were printed out on a Texas Instruments Si-
lent 700 terminal.

    The Environmental Working  Level Monitors (EWLM) were  developed under a
contract between the EPA and ANL.  Under this contract, four instruments were
built.  Although the method of measurement was the  same for all instruments
developed, these units differed considerably from the  previously constructed
*Institute for Energy Analysis, Oak Ridge,  Tenn.

 instruments.  The first difference between the  EWLM and the RWLM was measure-
ment sensitivity.  The EWLM had a sensitivity of  (.001 WL) and a sampling
flowrate of approximately 40 liters per minute  compared to  (.01 WL) and I2
liters per minute for the RWLM.  The second significant difference was that
each EWLM was self contained.  That is, it contained  a complete system, in-
cluding the micro-computer in one portable aluminum case.   The EWLM's were
designed for environmental measurements in non-mine atmospheres.

    As experience with the field operation of the EWLM's  accumulated, the ease
with which these instruments automatically monitored  environmental Rn-daughter
concentrations and WL became more and more evident.   This field experience led
to the development of additional EWLM's for two of the  scientific  divisions  at
Argonne National Laboratory.  The new EWLM's incorporated improvements de-
signed to correct some maintenance problems which occurred in the  original
EWLM design.  These new instruments  (EWLMII) and their  differences with the
EWLM's are discussed in detail below.


    A brief review of the Groer method for the measurement of the  Rn-daughter
concentrations and the WL is beneficial in the understanding of the  operation
and principles of the EWLM and EWLMII.

    The EWLM and the EWLMII use identical timing intervals to accumulate the
data for the Groer method of Rn-daughter measurement.  They first  take a back-
ground count for three minutes.  They then collect an air sample onto a filter
membrane for three minutes and then move the filter membrane to a  counting
station.  This air sample is then counted for three minutes starting 13 sec-
onds after the end of the sampling period.  The instrument has three counting
channels; the lower-energy alpha channel, the upper-energy alpha channel, and
a beta sensitive channel.  The RaA counts observed are  accumulated in the
lower-energy alpha channel; the RaC' counts are stored  in the upper-energy
alpha channel, which the total beta counts from RaB and RaC are recorded in
the beta channel.  The counts in these three channels are functions  of NA,
NB' NO the unknown concentrations of RaA, RaB and RaC  in the ambient air
 (units are atoms/liter).  The following relationships hold:

         TA        = 1.023403

         B + C     = (0.110393E2 + 0.007653E3)VNA +  (0.214298E2
                     + 0.023559)VNB + 0.283042E3VNC

         TC1       = (0.007653NA + 0.023559NB + 0.283042NC)EjV      (1)


         TA       = alpha counts from RaA accumulated during the  3 rain count-
                    ing period starting 13 sec after  the  end of sampling.

         B + C    = beta counts from RaB and RaC in beta  channel accumulated
                    during the same time interval as  above.

         TC1       = alpha counts from RaC1  (same  period of  accumulation  as

         V        = flowrate (liters/min).

         EI       = detection efficiency from RaA and RaC1.

         £2       = detection efficiency from RaB.

         £3       = detection efficiency from RaC.

         NA       = RaA concentration in units of atoms/liter.

         Ng       = RaB concentration in units of atoms/liter.

         NQ       = RaC concentration in units of atoms/liter.

The numerical coefficients in (1) follow from the laws of radioactive  series
decay.  The half-lives used are:

                        Nuclide             Half-life

                        RaA (218Po)           3.05 min

                        RaB (214pb)          26.8   min

                        RaC (2i-4Bi)          19.7   min

To show the principle of calculation of these numerical coefficients,  a  sam-
ple calculation is given below:

         EXAMPLE:  Calculate the numerical  coefficient for  the  first of  (1).

                   This coefficient  is the  product of three factors:

                   1)   Buildup factor = 1  - exp  (-xAfcB)

                        fcB = buildup time = 3 min.

                   2)   Delay factor = exp  (~XAD)

                        D = Delay before start of the counting  interval

                          = 13 sec or 13/60 min.

                   3)   Decay factor = [1 - exp (-xAfcD)]x/A

                        tD = decay time = 3 min.

                        XA = RaA decay constant = 0.2272614.

                   Multiplying all these factors  one obtains:

[1 - exp (-XAtB)]exp (-XAD) [1-exp (~xAtD)]/xA = 1.023403.

    The derivations of the analogous coefficients  for the other equations are
more complex but easily obtainable by the computer programs.  If all the effi"
ciencies and the flow rate are known, (1)  contains only numerical coefficients
and the unknowns NA, NB and NQ.   Inverting (1)  results in another set of
equations which give a NA, NB, and NC for every observed set of TA,
 (B+C) and TC'.  The WL can be calculated easily from the resulting NA, Ng,
and NC-

    The alpha efficiency E^ is easily measured in  the laboratory, but to
determine the beta efficiencies,  £2 and £3,  a  calibration run must first
be made.  Once EI is known, £2 and £3 can be determined by  the following
procedure.  Knowledge of E^ allows a complete  analysis of any air sample by
a method using total alpha counts.  This method allows calculation of NA,
NB and NC for this particular air sample. NA,  NB, and NC can then
be used to calculate the beta disintegrations  from RaB and  RaC.  By comparing
the calculated beta disintegrations with the observed total beta counts, the
values for E2 and £3 are obtained.  The necessary  calculations are all
performed by the "EWLM CALIBRATION PROGRAM."

    To calculate the beta efficiencies £2 and  £3,  the Rn-daughter concen-
trations (atoms liter) NA, NB and NC must first be determined from the
alpha counts during several counting periods using the following equations:

         NA = 0.711318*A5/E1*V

         NB = (-.689775*A5 - 7.54536*C5 + 1.86689*C3)/E!*V

         Nc = (0.0441761*A5 + 3.01566*C5 - 0.203315*C3)/£i*V       (2)


         A5 = RaA    counts observed during  5  min  starting  13 sec after the
                     end of the 3 min sampling time.

         €5 = RaC'    counts observed during  the same time interval as above.

         C3 = RaC1    counts observed during  30 min.  starting at the same time
                     as above.

         The numerical coefficients in (2) are again derived from the laws of
         radioactive-series decay.  This derivation  is straight-forward but
         lengthy and will therefore be omitted.

         With NA/ NB and NC known, £2 and £3 can be  determined from
         the following equations:
+The overlap corrections must be incorporated into  the alpha counts  (A5,
C5 and C3).

         Ql = (0.212378*NA + 0.348201*NB)*V*E2 + (0.019829*NA

                + 0.0496395*NB + 0. 455804*NC) *V*Es

         Q2 = (1.482571*NA + 1.549219*NB) *V*E2 + (0.599567*NA

                + 0.736276*NB + 1.842208*NC) *V*E3                    (3)


         Ql =   total beta counts observed during  5  min starting  13  sec after
                the end of the 3 min sampling time.

         Q2 =   total beta counts observed during  30 min same time as above.

    Inverting equations (3) , the beta efficiencies E2 and £3  can  be  deter-
mined, then  (1) can be inverted and properly scaled.  When properly  scaled,
this yields a set of equations which gives the Rn-daughter concentrations in
pCi/liter.  The WL can also be expressed as a linear combination  of  A,  (B + C)
and C1.  These counts are net counts, i.e., the background has been  sub-
tracted.  The equations are of the following form:

         WL             = C!(A) + c2(B + o + C3(c')

         RaA(pCi/liter) = C4(A) + C5(B + C) + C6(C')

         RaB(pCi/liter) = C7(A) + CS(B + C) + C9(C')

         RaC(pCi/liter) = C10(A) + CH(B + C)  + C12(C')                (4)
C]_ through C^2 are the derived weighting coefficients which  are  stored  in
the memory of the EWLM.  It is clear from this description of  the calibration,
that NA, NB and NC are treated as independent unknowns (i.e.,  no a^
priori relationship between these quantities other than radioactive  series
decay is assumed).  The EWLM determines, therefore, the Rn-daughter  concentra-
tions and WL, without any assumptions about the Rn-daughter  equilibrium.

    Calibrations , 6

    To be able to measure the WL and the Rn-daughter concentrations,  the  coef-
ficients GI through Ci2 of equations (4) must be contained in the EWLMII 's
permanent memory.  A calibration program automates the solution of the  equa-
tions for radioactive-series decay, as applied to the EWLMII, and calculates
the required efficiencies and coefficients.   However, to provide this external
program with input data, a calibration run by the EWLMII is required  first.

    An internal calibration program is requested via the keyboard which auto-
matically takes a series of 20, 41-minute calibration runs.  In each  41-minute
run a normal measurement is made using a single 3-minute sample but the

counting time is not terminated after  3 minutes as in a regular run.  Data is
stored in memory and printed out on  the printer for several time periods.  Tne
data printed out is the WL followed  by the Rn-daughter concentrations for a
standard run as well as the counts for all the counting channels for 3 min-
utes, 5 minutes, 30 minutes, 35 minutes and  the counts during the time inter-
val between 39 and 41 minutes after  sampling.

    This data can then be used as input for  a "Calibration Program" running on
a separate microcomputer.  After entering the data generated by the calibra-
tion runs of the EWLMII into this external program, it will perform the fol-
lowing functions:

         1)   It calculates the Rn-daughter  concentrations by three different
    methods and the WL by four different methods  from the input data of each
    individual run.  The four different methods are, the alpha-spectroscopic
    method, the total-alpha method,  the Kusnetz method and the Groer method.

         2)   It calculates the RaB  and RaC  efficiencies using the Rn-daughter
    concentrations calculated by the total-alpha  method, and the beta counts
    at two different periods of time.

         3)   It calculates the 12 coefficients needed for the Groer method.

         4)   It calculates the Rn-daughter  concentrations from 9 of these
    calculated coefficients, in units of atoms/liter.

         5)   It calculates the WL  from the  Rn-daughter concentrations, and
    with the use of the three remaining coefficients it also determines the WL
    directly from the measured counts.

    This calibration program has two branches,  one which calculates  the beta
efficiencies from the input data, and another which  uses the mean value of the
beta efficiencies previously calculated to derive the  final weighted coeffi-
cients.  A tabulation of the beta efficiency values  calculated in 20 separate
calibration runs and the final weighted coefficients are shown in the Table of
Efficiencies and Coefficients.

    Inputting the mean value of the  beta efficiencies  into the second branch
of this program, as well as the input count  data  asked for by the program,
results in the calculated values of  the Rn-daughter  concentrations and WL.
The calculated concentrations and WL's are  for  a  single 3 minute sample of air.

    A summary of the calculated values is  shown in the Table of Calibration
Data.  The units used in this table  for the  Rn-daughters are Atoms/Liter.  The
constants used in these calculations for the specific  Unit calibrated are also
shown, with, EA the alpha efficiency,  EB and EC the  RaB and RaC beta efficien-
cies respectively, V the flow-rate in liters/minute  and OL the overlap correc-
tion factor.  You will notice good agreement between the values of  the
Rn-daughter concentrations and the WL's calculated by the various independent
methods for each individual run.

    Electronic and Mechanical Improvements

    After extended use of the EWLM's in the field, certain areas of high main-
tenance began to appear.   To eliminate these problems modifications were
incorporated into the EWLMII and are listed below:

    The Bomar printer suffered from excessive mechanical failures because its
plastic parts were apparently too fragile for continuous operation.  There-
fore, the printer and its interface card were replaced with a more reliable
system.  A Texas Instruments model EPN 9120 thermal printer and a standardized
RS-232 serial interface card were substituted.

    The take-up gear assembly on the automatic filter advance mechanism re-
quired proper lubrication and adjustment and was  prone to failure if not main-
tained properly.  This caused failures in the filter advance cycle.  To solve
this problem, a torque motor was substituted for  the take-up gears to provide
the desired filter advance operation.

    The paper position fiber-optic sensor caused  failures in the filter
advance cycle because the optical sensitivity of  the sensor was too great.
This occasionally caused the filter advance to prematurely stop due to light
reflections received 180 degrees out of phase with the positioning mechanism.
The solution to this problem was to use a different optical sensing method
which eliminated the fiber-optics and used a less expensive, less sensitive,
directly coupled optical sensor.

    The solenoid clutch-brake occasionally failed due to binding of its cy-
lindrical shaft which prevented full closure of the brake assembly.  To
eliminate this problem a complete redesign of the clutch-brake assembly was
made.  This included a longer cylindrical shaft on the clutch assembly and
additional windings in the coil assembly.

    The scotch yoke assembly of the filter transport mechanism showed exces-
sive wear due to the lack of proper lubrication.  The solution was to install
a sealed bearing on the driver shaft.

    To adjust the length of movement for precise  filter membrane positioning
in the EWLM's, shimming of the top plate of the filter transport mechanism was
required.  A more flexible positioning method was needed.  Therefore, we rede-
signed the upper roller arm of the paper drive assembly incorporating a locked
cam adjustment in order to more conveniently adjust the filter membrane posi-

    In the process of servicing the EWLM's, we found that they frequently
needed to be removed from the outer case.  This process was found to be time
consuming and costly as all sub-assemblies were bolted separately to the outer
case.  Therefore, we decided to repackage the EWLMII to provide easy access to
all parts without requiring disassembly.  An inner mounting frame was designed
to which all components could be fastened and which could easily be removed
from the outer case.  This allowed easy access to components for troubleshoot-
ing and repair.

    On extended monitoring periods we found that  the printer paper was occa-
sionally blocking the airflow into the sample port.  Therefore, the filter
advance mechanism was repositioned with respect to the printer so as to elim-
inate any possibility of obstruction of the air sampling port.

    In addition to the modifications for the EWLMII listed above, certain
other improvements were incorporated.  The standard  (STD) bus was selected to
implement the electronics package.  This allowed  us to purchase commercially
most of the required micro-processor cards such as the CPU, the serial and
parallel I/O cards and the memory card.  We updated our CPU from the 8080 to
the 8085 which reduced power requirements and doubled the processing speed.
The PROM memory capacity was doubled by using 2k  memory chips rather than the
original Ik memory chips.

    A 16 channel analog to digital converter card has been incorporated in
order to be able to measure additional parameters.  This card, along with the
addition of a digital data recorder allows us to  transfer data more efficient-
ly from the EWLMII to external computers for data analysis.

    Software Changes

    The original operating programs for the RWLM  and EWLM were written in
PL/M, a high level language developed by the Intel corporation.  At that time,
Intel's software consisted of a compiler and cross assembler residing exter-
nally to their development system.  This meant  that  the compiling and loading
of the hex files had to be made via a phone linked into TSO, a time sharing
program used by our central computer facility.  Although the compiler allowed
structured programming it had no capability to  link assembly language programs
and provided non-relocatable addressing.  With  these  restrictions we were
forced to resort to tricks to use assembly language modules such as a floating
point mathematics package.  To make a small change in  the program, the entire
program had to be recompiled.  This was time consuming both in the compiling
operation as well as the resulting debugging that was often required.

    However, with the introduction of a resident  software PL/M the Intel pack-
age now provides linking and locating capabilities.   This allowed us to modul-
arize the programming of the EWLMII, linking and  locating modules in our local
development systems.  By building libraries of  these modules linking was fur-
ther simplified.  Now minor changes to the program can be made at the module
level, and only those modules changed required  recompiling.  This also means
that new modules could quickly be written and linked  into the main program.
With the use of this new software package, the  new or modified modules could
then be linked and located in minutes rather than the days taken in the past
to recompile and debug the program.

    Updated PROM programmers further reduced the  time required to change mem-
ory.   These new instruments and software allow  us to down line load several
modules at a time into the EWLMII's RAM memory  for debugging.  After all
modules have been thoroughly checked out, the entire program is re-linked and
loaded into the PROM programmers buffer.  Each  PROM for the EWLMII's operating
system is then programmed and inserted into the memory board.  This process
greatly accelerates the time needed to debug, check out and program a new
operating program.


    The increased memory capacity and the addition of  an  Analog  I/O subsystem
in the EWLMII allows us to measure additional parameters.   An  Analog Devices
Model RT-1225 STD bus compatible card was used.   It provides the system with
16 channels of analog input and 2 channels of analog output.   It contains a
standard sample and hold amplifier with a monolithic 10 bit A/D  converter with
the flexibility of 16 single ended or 8 differential inputs.   It operates off
a single +5 volt power supply and contains a memory mapped  I/O.

    Through the use of this card additional parameters such as,  inside and
outside temperatures, atmospheric pressure, wind velocity and  direction, hu-
midity and control switches are interfaced to the EWLMII, making this instru-
ment a true data acquisition system.  These external DC inputs and others can
be added to the system to output additional parameters on either a predeter-
mined or programmable schedule.  Additional program modules for  this added
capability can easily be linked into the main EWLMII program.  This relinked
program can then be burned into the PROM's to provide  this  additional monitor-
ing ability.

    Data Storage, Display and Analysis

    The EWLM and EWLMII temporarily store their  data in active memory  (RAM)
and this information is then transferred to the  printer at  the completion of
the measurement cycle.  Subsequent measurements  use the same data area causing
the loss of previous data.  The only storage media used in  the EWLM was the
printed data stored on a continuous paper tape.   This  method of  data readout
required that these recorded values be entered into an external  computer for

    To streamline the data processing, a TEAC MT-2 digital  data  recorder was
added to the EWLMII.  Immediately after the data has been printed out, it is
then recorded on the digital cassette recorder.   The EWLM's print out a short
report onto the 20 character per line printer.  There  is  additonal data stored
in memory which is not printed out in this short report.  The  digital cassette
records the complete memory image, therefore all data  is  permanently stored on
the cassette.  This allows us to use any part of or all of  this  data for our
analysis.  The data from each sample is continuously stored on the  tape.  When
the series of runs is complete an "E" is entered into  the keyboard which is
then removed from the EWLMII and is ready for a data transfer.

    To display and analyze this data stored on the cassette,  we  used a micro-
computer with 32k of memory configured with an identical  digital data  record-
er, an eight inch floppy disk and a dual serial port.

    Software was written which loads the tape, transfers  the  cassette  tape's
data into the micro-computers memory and then rewinds  the tape.   After  the
data have been transferred to the computer's memory the cassette can  be
removed for reuse in the EWLMII.  Another program was  written to display a
complete report of the data on the terminal or line printer.   Additional
programs are to be written for data analysis.  This stored data can be
transferred to the floppy disk and data  from additional tapes can be  added to

 this extended data file.  This library of data can  then  be  used for data
 analysis.  Programs can be written to either  tabulate or plot  the WL and
 Rn-daughter concentrations versus time.  Other manipulations of the data can
 be printed such as Rn-daughter ratios, statistical  evaluations and comparisons
 of the daughter concentrations with:  barometric pressure, humidity,
 temperature, inside-outside temperature differences, air circulation rates,
 air filter effects with air flow-rates and time and other parameters.


    Development of new methods and instruments for  the measurement of  WL and
 Rn-daughter concentrations has been in progress for several years.  However,
 developing efficient methods of data  acquisition has not been  the primary
 thrust of this research.  If large volumes of data  from  automated instruments
 are to be efficiently analyzed, a built in data acquisition system is  needed.
 One such instrument, the EWLMII, has  been described in this paper.  The need
 for accelerated research for the understanding and  control  of  Rn-daughters can
be enhanced by such an instrument.

1.  P.G. Groer, D.J. Keefe, W.P. McDowell and R.F. Selman.  "The Instant
         Working Level  Meter-"   U.S.  Bureau of Mines, Department  of Interior,
         Final Report (August, 1974).

2.  D.J. Keefe, W.P. McDowell and P.G. Groer.  "The Remote Working Level
         Monitor."  U.S. Bureau  of Mines, Department  of  Interior,  Final Report
         (November 18, 1977).

3.  D.J. Keefe, W.P. McDowell and P.G. Groer-  "The Environmental Working
         Level Monitor."   United States  Environmental Protection Agency, Final
         Report (September 29, 1978).

4.  P.G. Groer, D.J. Keefe, W.P. McDowell and R.F. Selman.  "An Instant
         Working  Level  Meter with Automatic Individual  Radon Daughter Readout
         for Uranium Mines."  Presented  at  the  Third  International Congress of
         the  International Radiation Protection  Association,  Washington,  D.C.
         (September  9-14  1973).  Published  in  Proceedings (Part  II:  950-956,
         National Technical Information Service).

5.  W.P. McDowell, D.J. Keefe, P.G. Groer and R.T. Witek.  "A Microprocessor-
         Assisted  Calibration  for  a  Remote   Working  Level  Monitor".   IEEE
         Trans. Nucl. Sci. NS-24, No. 1, 639-644  (February, 1977).

6.  D.J. Keefe, W.P- McDowell, P.G. Groer and R.T. Witek.  "Optimizing Techni-
         ques  for  the  Measurement of Environmental Levels of  Rn-Daughter  Con-
         centrations."  IEEE  Trans. Nucl.  Sci.  NS-25, 787-789 (February, 1978)
         Presented at  the  1977  IEEE Nuclear Science  Symposium,  San Francisco,
         CA (October 19-21, 1977).


                                 EWLMII UNIT 2
                   2-9-81, TWILIGHT MINE   URAVAN, COLORADO
AVERAGE EB = .1852

AVERAGE EC = .2984
                  STANDARD DEVIATION =  .0211

                  STANDARD DEVIATION =  .0250

-1.71403658529E-05   WL
-1.65513410988E-03   RaA
-6.92698319685E-03   RaB
 5.28141146336E-03   RaC

             EWLMII UNIT 2
 EB=.1852        EC=.2984       V=39.01






















































































RUN 16
RUN 17
RUN 18
RUN 19
RUN 20





















                             A LIQUID SCINTILLATOR

                              Howard M. Prichard

                The University of Texas School of Public Health
                                P.O. Box 20186
                             Houston, Texas   77025

    Radon is extracted  from  ca.  10 liter ambient  air  samples directly into a
solution of  hexane  and toluene held  at -78  degrees C  in a  dry ice-acetone
bath.  The solution is  sealed  in  a 22 ml glass liquid scintillation vial con-
taining 2 ml of concentrated fluor.   The vials are counted on  a liquid scin-
tillation counting system with  a  narrow counting  window  set  around the alpha
peaks of radon and its short-lived progeny.   A large number of  samples can be
prepared in  a  relatively short time  with  minimal equipment, and counting is
performed automatically.  The  large  sample  volume  more than compensates  for
the  relatively  high  alpha  background  of  liquid  scintillators.   Alternative
liquid scintillation sampling strategies  are discussed.


    Limitations  imposed by counting statistics  require  that accurate measure-
ments of radon-222 in  air at ambient concentrations (less than IpCi/l) involve
air  samples  of  at least one  liter.   To meet  this requirement,  one  must  use
either a large volume  counting  device or employ a concentration technique of
some sort.   The technique described  here makes use of  the high solubility of
radon in certain organic solvents  to extract  radon  from an air stream.  If the
solvent chosen is suitable  for liquid scintillation counting,  then  the radon
may be counted directly  upon addition of appropriate fluors  to the solution.
The lower the temperature of the solvent at the time of extraction, the great-
er will be the degree  of radon concentration and hence  the sensitivity of the
process.   The  temperature  recommended  for  most  applications  is  -78  C,  the
temperature  of a dry  ice  and acetone  bath.   Several appropriate solvents  are
liquid at this temperature  and dry ice  is easily  carried to remote locations
in an ordinary ice chest.  While this extraction process  does not produce the
complete radon retention obtained  in  techniques  involving solid sorbents held
at  liquid  nitrogen temperatures   (e.g.  Da73), the  degree  of   extraction  is
reproducible and in good agreement  with  theory.

    The distribution of  radon  between  air  and a solvent in a closed system is
described by L, the Ostwald coefficient,  according  to  the equation:
                           Cs va                                (Eg. 1)
where  Cs  and  Ca are  the  equilibrium  concentrations of  gas  in  the solvent
and  air  phases and  Vs  and  Va  are  the  volumes  of  the  two  phases.   The
temperature dependence of the  Ostwald coefficient  for radon in  various organic
solvents is of the general form:

                   L = A + B exp(- u6)                           (Eq. 2)  (We78)

where  8  is the  temperature as  expressed  in  corresponding degrees centigrade
and  A, B,  and  u  are  experimentally  determined  parameters.    Degrees  of the
"corresponding  centigrade"  scale  are  defined  as  one  hundredth  part  of the
interval  between the  melting  and  boiling  points of the solvent in questions.
The  values  of  A,  B,  and u  for a  number  of  solvents are  listed  in  Table 1,
and  a graphical  representation of  L  vs  temperature is  shown in Figure 1.
Carbon disulfide, which  has the highest affinity  for radon in  the temperature
range  shown,  is a strong quenching agent  and  is incompatible with the liquid
scintillation   process.    Normal  hexane,  however,  is   a  workable  liquid
scintillation  solvent  (Ha53)   and appears  to  be  a  better   extractant   than
toluene,  the more commonly  used scintillation solvent.  While the light output
of   a  hexane-based  scintillation  solution  is only  a   quarter  that   of  a
toluene-based  solution,  the high  photon  production  associated  with  the  alpha
and  hard  beta emissions of radon and its daughters assures that ample light is
available for detection.

     If  a  sample of air with volume  Vg  and a  radon concentration (: is brought
into equilibrium with  a  volume Vs  of solvent  having  an Ostwald coefficient
of L at the equilibrium temperature, then  the activity residing in the solvent
phase will be given by:

                   As = LVSC/(1  + LVs/Va)                        (Eq. 3)

When the  volume  of   air  Va   is  large  compared   to LVS,  As  is relatively
insensitive  to  variations  in  Va,  tending to  a  steady  value  of CLVS  as Va
tends  to  infinity.  In  this  range, details of  the  contact process such as air
flow rate and  bubble  size become  unimportant.    Figure 2  shows  predicted  radon
retentions for  three  values of L and  data from  a  calibration run  which in-
volved 20 ml of a solution  consisting  of  90% n-hexane and 10% toluene, by vol-
ume.   At  high air volumes the  points fit  the  L = 300 curve, but fall below it
as  Va  becomes  small  compared  to  LVS.    For  sample sizes of LVS or great-
er,  no flow  rate dependence was observed  between  0.5 and 3.0 liters  per  min-
ute,  and  only a small  reduction in radon retention was noted  when an orifice
that produced large bubbles was substituted for  the  regular frit  that produced
small bubbles.

     At  -78  C,  the  solubility  of  oxygen  in organic solvents  is  considerably
higher than at  room temperature.   Because  oxygen reduces  the  light output of  a

scintillation solution, some consideration must  be  given to the effects of the
additional oxygen  acquired  during radon extraction.  While  reduced  light out-
put  itself  is  easily  dealt with  by a gain  adjustment, uneven  oxygen uptake
from sample to  sample  could produce  variations  in counting  efficiency.  If the
counting window is set to  include the radon and polonium  alpha  peaks  over a
reasonable range of quench factors this problem  can be  minimized, as beta con-
tinuum  counts  lost on  one  side of  the  alpha peaks are approximately compen-
sated for by the counts gained  on the other  as  the  peak centroids are shifted
by varying degrees of  quench.   The external  standard feature of most commerci-
al liquid scintillation units can  readily be  used to determine whether a given
sample  lies  within  the acceptable  quench range.   Figure   3  shows  two  radon
spectra near the  extremes  of  quench that we  have noted in  practice.   In this
range,  the variation  in  counting  efficiency,  as  measured by  the  ratio  of
counts in the window to the total  counts, is  acceptably small.   The quench due
to  absorbed oxygen  can  in  fact be  exploited  to detect improperly  sealed
vials.  The  loss  of oxygen  due to  a  slow leak would   decrease  the quenching
effect and raise  the  external  standard ratio.   A quick recount of the samples
for  external  standard  ratio  performed 48  hours  or more  after  the  original
count would  be sufficient to indicate  any  vial  that had lost oxygen,  and  in
all probability, radon.  In many  cases,  a recount of the "leakers"  could then
establish an effective half-life  of  radon in the vial,  permitting the estima-
tion of radon content at the time the vial was sealed.


    The  choice of  extraction  temperature  and  primary  solvent   represents  a
trade-off between convenience and  completeness of  radon recovery.   As figure 1
demonstrates,  the  lower  the  temperature,  the   more  efficient the  extraction
will  be.   The   higher  the  efficiency, the  larger  the  air   sample that  it  is
practical to process,  as  Figure  2 shows.   In effect then,  it is  the size  of
the  sample  required  for the desired degree  of  sensitivity that  controls  the
maximum  practical  extraction  temperature.   For  the measurement  of  radon  at
ambient outdoor concentrations, the  -78  C extraction into  hexane as  shown  in
Figure  2  represents  an acceptable  balance  between  sensitivity  and  conveni-
ence.  The dry ice required for  the low temperature bath  is  inexpensive  and
easily  carried  in ordinary  styrofoam or plastic containers.  The  extraction
train shown  in Figure 4  (not  to scale)  is  composed  of  standard  laboratory
equipment.   Collapsible 20  liter  sample  bags (A)  are  filled at  locations  of
interest and returned  to a central field  station for processing.   If  the time
between sampling and  processing is  more  than an hour  or so,  as  would  be  the
case in a 24 hour  integrated  bag  sample, care  must be  taken  that  the  bag  is
composed of material  that  has  been  shown to be  impervious to radon.  At  the
extraction station, air is drawn from the bags  by a pump (B)  at  approximately
2 liters per minute  through a  dessicator or  cold trap   (C)   to  remove moisture
that might condense  and block  portions  of the  sample   train.   The dried  air
passes through an impinger  tube (E)  containing 20 ml of solvent  held  at  -78 C
in the dry ice  and  acetone  bath (D).  The airstream is broken into  fine bub-
bles by the fritted orifice of  the  impinger, thus  assuring good  phase contact
between  the air and the solvent.   After warming to near ambient temperature,
the exhaust air  passes through  a  flow meter  (F).  When the desired  volume  of
air has been  passed through  the   solvent, the  pump is  stopped,   the  impinger

disconnected/ and its contents  gently  poured into a pre-chilled vial contain-
ing 2 ml of concentrated fluors (40 g/1 PPO and 5 g/1 dimethyl POPOP, or simi-
lar commercial  concentrate.)   The vial is  then  tightly  capped with a plastic
cone-style cap and placed in an ice bath to warm to 0 C.   The impinger tube is
then rinsed  and 2O ml of cooled  solvent  is added in preparation for the next
extraction.   (Note - the thermal  expansion  of hexane over  a  1OO C interval is
not trivial.)   After a  batch of vials  has warmed to O C  in the ice bath, each
cap is briefly  (ca. 1 second) loosened to relieve excess pressure, then tight-
ened again.   This  step  is  a precaution against vial breakage due to overpres-
sure and causes less  than  1% radon loss.   The vials are  then  allowed to warm
to room  temperature,  and are ready for counting  at any  time later than three
hours after initial sealing.


    Each picocurie  of  radon-222  in equilibrium  with  its short-lived progeny
produces on the average 11.1 alpha or hard  beta disintegrations per minute.  A
window set as shown in Figure  3  wil accept approximately 80% of these events,
or 8.9 cpm.   Equation 2 predicts  that  37.5% of  the  radon  in  1O liter air sam-
ple will be  retained  in a 20 ml  solvent phase with L=300, so that 33.4 counts
per pCi/1  of radon-222 are  expected in the counting window at full equilib-

    Because  alpha  particles are relatively  inefficient  light  producers  in
liquid scintillators,  energy discrimination between alphas  and betas  is  not
possible and compton electrons  from environmental  gamma radiation contribute
heavily  to  the background  in  the  region   of  the  radon  and  polonium  alpha
peaks.   In  the window shown  in Figure 3 the background will be approximately
12  cpm,  depending  on  the  exact  window  settings  and  the performance  of  the
counting system.   If,  for example, the  concentration of  radon in air is 0.25
pCi/1,  a 100  minute  count  on  a  10  1 sample  will produce 825  expected  net
counts against  an  expected background  of 1200 counts.   The minimum detectable
true activity  (A163) would  be 0.04  pCi/1  if the lack of independence of radon
daughter  counts can  be neglected.   Although the  background  is considerably
higher than  would be obtained  from  a Zn:S phosphor,  it  is seen that the large
effective  sample volume provides  sufficient compensaton  for radon concentra-
tions  encountered  in continental  atmospheres  and in dwellings.   In  a recent
intercomparison  at  the  Department of Energy's Environmental Measurements Lab-
oratory  (Ge81) ,  four  sample bags were  filled with  air  from  the radon chamber
and shipped  to our Houston Laboratory for  processing.  The mean  of the four
samples  was  46.2 pCi/1  with a standard  deviation  of 0.8 pCi/1,  as compared
with the nominal value of 44.6 =1.1 pCi/1.


    A  great  deal of  latitude  exists in  the choice of solvent, the  operating
temperature,  sample volume,  and solvent volume.   (Twenty milliliters is, how-
ever,  an effective upper  limit to the solvent  volume  if a commercial liquid
scintillation  counting  system  is  to   be   used.)   Some  common  solvents have
liquid ranges that permit  their use at  temperatures  well  below -78  C.  Pentane
is liquid  in the range  -13O to  36 C, and isopentane is liquid  in the interval
-160  to  28  C.   If nearly  complete  radon  recovery  from  large air samples is
desired, a low  freezing point  solvent  cooled by an appropriate bath  or device
could  be used  in  place of  the -78 C  system  described  above.   If the sample

volume were less  than  LVS,  however, it is probable that  the  details of phase
contact would  influence  the  degree of sample  retention  much  more  seriously
than in the -78 C  case.

    In environments with  radon concentrations appreciably  above  normal back-
ground it  would  be  possible  to  operate effectively  at  O C  or at  ambient
temperature.   The quantity  LVS would  then  be relatively small,  on  the order
of 500 ml.   At air sample volumes appreciably above that,  the concentration of
radon  in solvent  would tend to LC.   Weigel  (We78)  states  that olive  oil and
other  hydrocarbons of  vegetable  origin have Ostwald coefficients  of about 28
at room temperature, which  implies that 20 ml of  solvent in equilibrium with
air would contain  about the  same  amount of  radon  as a conventional 5OO ml grab
sample.  Whether or not ambient temperature extraction is a desirable alterna-
tive  to  conventional  grab  sampling would depend  upon the logistics  of the
individual sampling  problem.

    The high  affinity  of organic  solvents for  radon leads  to a  simple and
sensitive method for detecting radon  that  has been entrained on  an activated
carbon trap.   Activated carbon at the temperature of dry ice is easily used as
a radon trap,  but  the  subsequent  removal of the radon is somewhat more compli-
cated, and requires elevated  temperatures,  vacuum,  and pressurized gas.  We
have  noted that 98% of  the radon  can  be  removed from  the carbon  simply by
sealing the carbon  in  several volumes of a solvent such  as toluene or hexane
for a few  hours,  after which  the   radon can  be  counted by liquid  scintilla-
tion.  To  confirm the  degree  of  desorption,  one  gram of activated carbon upon
which  radon had been entrained at -78 C was divided  into  equal  portions after
thorough mixing.  One  portion was   immediately suspended in 2O  ml of scintil-
lating gel and  the  other  portion  was placed in 15 ml  of  toluene  for 3 hours.
The toluene was then decanted and the carbon rinsed with an additional 5 ml of
toluene, after which the  carbon was suspended in a gel in  the  same  manner as
the first  portion.  While alpha particles emitted  by  radon and progeny within
the carbon granules could not be  expected to  contribute  effectively  to the
count  rate, the energetic betas emitted by Pb-214 and Bi-214 did register well
in the gel suspension.   It  was found  that  the carbon fraction that had been
treated with toluene displayed less than  2%  of the count rate of the fraction
that had been placed directly in  the gel.

    The potential problem of radon  loss during  the initial second  or  two of
desorption was solved  by  placing  the cold carbon granules directly into a vial
containing 10  ml  of cyclohexane  that had been frozen by placing  the  vial in
the dry ice container.  The  vial  was then tightly capped and allowed  to come
to ambient temperature.  At  some convenient  time after the carbon had sunk to
the bottom of  the liquid cyclohexane, the vials  were placed back  in  the dry
ice  container  until  the  cyclohexane  was again  frozen.    Ten  ml  of  toluene
scintillation fluid  was then added  to the vial and the vials were recapped and
allowed to warm to room  temperature.  At  this stage,  the solvent could be
withdrawn  for  counting,  or  counting could proceed  with the carbon granules
resting on the bottom of the vial.  When  1.5 g carbon traps  were  used, the
counts lost due to  the presence of the  carbon were found  to be  a  small and
reproducible fraction of  the total.
                                        O A


    The  high  solubility of  radon in  various  liquid scintillation compatible
solvents leads to a set of simple and  sensitive concentration techniques.  The
choice of  solvent,  operating temperature, and sample volume  can be varied in
response to a wide  range of  sampling requirements.   The relatively high back-
ground associated  with liquid  scintillators  is overcome  by the  high sample
volume in  most  applications.   The   liquid  scintillator  extraction techniques
seem most  promising for applications  in which a large  number  of samples are
routinely processed, such as a  survey of indoor radon  levels  in a community.
Table 2 shows data  obtained  by  this  technique  from dwellings in Maine compared
with data from commercial alpha track  detectors (c.f. Pr81).  The track detec-
tors provide  a  measure of  the  integrated radon  "exposure" over  periods of
three to six months, while the grab  samples indicate only the radon concentra-
tion at  the time of the alpha track detectors' deployment.  Nevertheless, the
descriptions of  the radon levels  in  the  community obtained by the two methods
are quite similar.  The means of  the indoor measurements are not statistically
different from the mean of the grab samples, as shown by the p-values of a two
sample t-test, and  the grab  and alpha  track measurements are correlated at the
p = .01 level.   The outdoor and basement alpha track data do not show the same
agreement  with   the indoor grab samples, of  course.   We do  not  suggest that
grab samples  by  this  or any  other method are  an  acceptable  substitute for
integrated  measurements in  all  cases.    If   due  consideration  is given  to
meteorological variables, however,  reasonable estimates of  a  community radon
profile can be made with grab sampling techniques which offer the investigator
the advantage of low cost along with prompt and controlled  access to the data.


A163   Altshuler, B.  and Pasternack,  B,  1963,  "Statistical Measures  of  the
       Lower Limit of Detection of a Radioactivity Counter", Health  Physics 9,

Da73   Darrall,  K.G., Richardson  P.J.,  and  Tyler,  J.F.C., 1973,  "An  Emanation
       Method for  Determining  Radium  Using  Liquid  Scintillation  Counting",
       Analyst 98,  pp.  610-615.

Ge81   George, Andreas,  and  Fisenne,  Isabel,  personal  communication,  1981.

Ha53   Hayes, F.N.,  Rogers, B.S. ,  and  Sanders, P.C.,  1953,  "Importance  of
       Solvent in Liquid Scintillators"  Nucleonics 13, pp.  46-48.

Lo28   Loomis, A.G. , 1928, "Solubilities of Gases in  Liquids" in  International
       Critical  Tables  of Numerical  Data Vol.  Ill,  pp.  254-271,  McGraw-Hill,
       New York.

Pr81   Prichard,  H.M.,  Gesell, T.F., Hess,  C.T., Weiffenbach,  C. , and  Nyberg,
       P., 1981,  "Integrating Radon Detector  Data from Dwellings in Maine  and
       Texas", submitted to  Health Physics.

We78   Weigel, F.,  1978, "Radon",  Chemiker Zeitung 102, pp. 287-299.

                             Table 1

                Solubility Coefficients for Radon
                       in Various  Solvents*
Ethyl Ether
Point (°C)
Point (°C)
*Values taken from Weigel (We78) except for those  of  n-Hexane,
 which were obtained by regression from the data of Loomis  (Lo28).

                                     Table 2

        Agreement Between Indoor Grab Samples and Integrated Alpha Track
                              Measurements in Maine
Sample Type
First Floor Grab
Alpha Track
First Floor



P-value of


0.74 *
0.71 *
0.73 *
0.46 *
0.15 *
*p less than 0.01

                                 FIGURE  CAPTIONS
Figure 1:  Extrapolated Radon  Solubilites  (L) in Various  Solvents  as a  Func-
           tion of Temperature.

Figure 2:  Predicted Radon Retention in 20 ml of  Solvent  in  Equilibrium with a
           Range  of  Air Volumes  as  a Function  of  Solubility (L).   Triangles
           Indicate Experimental Values.

Figure 3:  Spectra  of  Radon-222  in a  Hexane-Toluene Solution  at  Two Quench
           Levels.  At  the  Indicated  Window  Setting,  the Count Rate  is Nearly
           Independent of Quench Within the Range Shown.

Figure 4:  Diagram of Extraction Train (Not to Scale).


8  100
                        Rn BOILS
                                 TEMPERATURE, °C
                                   Figure  1




                                        VOLUME OF AIR (I)

                                            Figure 2




                                    ESR   CR

                                — 0.497  0.716

                                — 0.395  0.703
/ /
/ /
1 \ \
  400               600

                                                    Figure 3

                                                           Figure 4

                        RADON AND RADON DAUGHTERS IN AIR
                                 Andreas Wicke*
              Bundesgesundheitsamt, Institut fur Strahlenhygiene,
                    D-8042 Miinchen-Neuherberg,  West Germany
                              Justin Porstendorfer
               Zentrales Isotopenlabor der Universitat Gottingen,
                Burckhardtweg  2, D-3400 Gottingen, West Germany

    Surface-barrier detectors  have  been applied for  alpha-spectroscopic meas-
urements  of  environmental  levels  of  airborne  radon  (Rn-222)  and  thoron
 (Rn-220) as well as for their individual short-lived daughters.

    For  the purpose of  radon  and  thoron  detection,  the air  has  been  dried,
 filtered   and  sucked   into   an   evacuated  aluminum  sphere,  (14.2£).   The
positively  charged  Po-218 (l.aA)- and  Po-216  (ThA)-  atoms  are collected  on a
surface-barrier detector  by an applied  positive  voltage of  20kV between sphere
and detector surface.

    The  alpha  disintegrations  of  Po-218 (6.00MeV)  and  of Po-216  (6.77MeV) are
counted separately by  a  field  multichannel  analyzer.  Taking  a one hour count-
 ing  time at least 0.03  pCi/X,  radon can be measured  within 20%  of statistical

    The  same  sensitivity is  achieved  using a  flow-through  version of  that
 instrument  for continuous  operation.   For  thoron  the  sensitivity is  only
1  pCi/i corresponding  to  a flow rate of 1.6£/min.

    Using a separate  instrument radon  and thoron daughters are  collected  on a
membrane  filter  and   the decays  of  Po-218  (RaA),   Po-214  (RaC1) and  Po-212
 (The1) are  counted  by means of  alpha-spectroscopy during and after  sampling.
Good  results  have been obtained  for  a counting/sampling period  of 30 minutes
and  two  30  minutes  counting  intervals after  sampling.    At  a  flow rate  of
22.5«,/min  and  a   detector  efficiency  of  17%  the  sensitivity  for  each
individual  daughter  concentration  is  better  than 0.03 pCi/&  based  on  20%  of
statistical accuracy.
*Present  address:   New  Mexico Institute  of Mining  and  Technology,  Graduate
Office, Socorro, New Mexico 87801, U.S.A.

    This system is well suited for continuous operation, too.  However,- radon-
and thoron-daughters can only be determined in terms of potential alpha energy


    In order to obtain  detailed  information about the levels of radon, thoron
and  their  daughters  indoors,  sensitive equipment  is needed.   Since outdoor
concentrations play  an important  role in  well-ventilated  houses  the methods
should be capable of  detecting  variations  of atmospheric levels with satisfy-
ing accuracy.

    For this  purpose  existing methods,  e.g.  scintillation flasks,  ionization
chambers or two-filter devices for radon or thoron detection are not  sensitive
enough or the equipment is too bulky to be  used  in homes.

    Adequate methods to measure radon  daughters using the filter technique may
be satisfying for grab  sampling of radon daughters only.  Since a considerable
contribution of the  potential alpha energy concentration  derives  from thoron
daughters, too  (WICKE and PORSTENDORFER, 1980)  only  such  methods  can be used
which are able to discriminate between both radon  and  thoron daughters.

    Many  of  those naturally  occurring radionuclides are  alpha  emitters with
almost different alpha  energies, hence  alpha spectroscopic methods are advant-
ageous  to use.  The  type of  detector which can easily be  applied for this
purpose is the surface-barrier detector, a  semiconductor.

    In  our  studies  specific  instrumentation  has  been developed  which  uses
those detectors for measuring low  levels of radon, thoron and their  daughters
in air.

Radon and Thoron Measurement

    The method we applied for radon  and thoron detection is a specific version
of the two-filter technique.  Radon  and thoron  gas enter a decay volume while
their daughters are trapped  by  an entrance filter.  Instead of collecting the
daughters which are produced during  the passage of  the  gas on an exit filter,
initially positively charged Po-281  (RaA)- and Po-216 (ThA)- atoms are removed
by a  strong  electric  field.   The daughters, collected  directly  on the alumi-
nized surface of a  surface-barrier  detector can easily be  identified by alpha
spectroscopy.  Under  adequate conditions it can be shown  that  the decay rate
of Po-218 and Po-216  is proportional to the radon  concentration and  the thoron
concentration, respectively.

    Such type of instrumentation  was presented  by JACOBI  (1963)  with a volume
of 140A,  capable of  detecting  concentrations  of radon as low as  0.1 pCi/SL
within an hour.  Similar devices with  smaller volumes and  scintillation detec-
tors were  used  by  ROBERTS AND DAVIES  (1966) and  DALU and  DALU  (1971) for the
detection of atmospheric concentrations of  radon.

    Since  those  instruments, in  contrast to the  two-filter  technique, do  not
need  high  flowrates, diffusion  of  radon gas was  sometimes sufficient.   Those
"passive"  instruments  were  proposed  for  the  application  in  mines  (COSTA
RIBEIRO et. al., 1969) and in homes  (WRENN,  SPITZ  and  COHEN, 1975).

    The type  of  instrument we are using  consists of a 14.2£ aluminium sphere,
which  incorporates  a surface-barrier  detector*  isolated  in  a  PVC-mounting
(Fig.  1).   The  entering  air  is   dehumidified  by   Ca  C12,  and  radon-  and
thoron-daughters  are removed  by an  entrance  filter.   For grab sampling  the
sphere  can be evacuated  and the  air be  sucked in at  the place of  interest.
After applying the high-voltage counting  can start initially.
                     SURFACE BARRIER
                    (19.5mm active diameter)


    r +20 kV
         1-   Electrostatic  collection  chamber  system  for  radon  and  thoron
    For  field  purposes  we  combined  the  detector  with  a "state-of-the-art"
preamplifier,  shaping  amplifier and  a portable  multichannel  analyzer  with a
fast tape recorder** (Fig. 2).
* Ortec CR-24-300-100, 19.5mm active diameter.

**Silena NIM Standard Multichannel Analyser.



    Fig. 2:   Alpha particle signal analysis system

    Since  radon and  thoron daughters  are deposited  directly on the  sensitive
surface  of the detector,  a very  good  energy resolution  is  obtained  (~80KeV).
Fig. 3  shows a  typical alpha  spectrum of a radon-thoron gas analyzed  by  this
                     5           10          15
                      APPLIED VOLTAGE, kV
    Fig. 4:   Collection  efficiency for  Po-218  (RaA) and  Po-216  (ThA)  as  a
              function  of the applied voltage
    In order to obtain a high-efficient electrodeposition the strength  of  the
electric field is  very  important.   According  to Fig. 4  saturation  for  Po-218
collection is  achieved  above  16kV,  while  the  collection rate  for  Po-216  is
still increasing.

    For practical  purposes  we  applied  20kV  to  the  system and obtained  a  satis-
factory collection rate.

    A critical  factor  is  the influence of humidity.   Fig. 5  shows that between
0 and  50%  relative humidity,  small  changes have  a  considerable  influence  on
the collection efficiency of  Po-218.   Best results  are  obtained  when the  air
is carefully  dried.


    O  60


    2  40

    8 20
I     I    I     I    I     I    I    I    T
                    I     I    I     i    I     I    I    I    I
                   20       40      60      80       100

                        RELATIVE HUMIDITY, %
Fig. 5;   Collection  efficiency of Po-218 (RaA)  as a function of relative


    The calibration  of  the system  was  done by  introducing  a known amount  of
radon into the sphere emanating a standard radium  solution.   The  average value
of  several  calibrations  was  12.0  cpm(pCi/J!.) •   At  a  one-hour  counting  time
the sensitivity  for  radon is 0.03  pCi/fc  assuming  an error of 20%  to  counting

    The detection efficiency  for  thoron of  the  static version of  the instru-
ment  is  so  poor   (~4   counts/(pCi/2.)   for  counting   times  >   10  minutes)
that only the flow-through type of  this instrument is of  interest.   For moving
the  air  a small diaphragm pump*  and a flowmeter  are  added  to the instrument
(see Fig. 1).

    A  flowrate  of 1.68,/min results thoron detection efficiency  of about 0.4
cpm/(pCi/«,)    and  a  sensitivity   of  1   pCi/Jl  for   a  one-hour   counting
interval.  These values  improve for  higher flow  rates.   For radon,  however,
the  detection   efficiency  is  fairly   independent   of  the  flow  rate,  12
cpm/(pCi/Jl) .

    We  applied  this instrument  successfully  in  those  studies,   where  quick
changes  of  the  radon concentration should be monitored.   At a  flow  rate  of
1.6J./min  the  response time of  the  system was  only about  10  minutes depending
on the size of the drying tube.

Radon- and thoron-daughter measurement

     An  established  method for  the  determination of  radon-daughters in air  is
to draw  a certain volume  of air through a membrane or  a glass-fiber filter and
to count  the  alpha activity  on  the  filter.   The  Tsivoglou-Thomas  method
 (THOMAS  1972)  uses  gross  alpha  counting in three  time  intervals  after collec-
tion to calculate  the  RaA,  RaB and  RaC-concentration in air.   The accuracy,
especially  for  RaA,  was  improved   by CLIFF  (1978)  taking  the first  counting
period  during sampling.   Although more  sophisticated instrumentation is neces-
sary,  an alpha spectroscopic  analysis fo  the filter sample  can  improve accur-
acy.   Such  a  procedure  was  proposed by  MARTZ  and co-workers  (1969)  and pre-
sented   for  practical  applications  in  a   recent  paper  by  NAZAROFF  et.al.
 (1981).   Making a similar approach as  CLIFF,  DUGGAN AND HOWELL  showed how to
improve  the  spectroscopic technique by taking the first  counting while sampl-
ing was  in progress.

     Adopting  this  method for  our  purposes  radon-  and  thoron-daughters  were
collected on a  membrane  filter during  a  sampling period of  to,  while alpha-
spectroscopic   counting   of   Po-218  (RaA)  and  Po-214   (RaC1)   took  place
 (tl~t2)   using  a surface-barrier   detector.   During  a  second   time   interval
from t3-t4  only  those  counts from  Po-214  were  taken.    Once  those  three
alpha counts  are obtained it  is easy to calculate the  individual radon daugh-
ter concentrations by the method of "Simultaneous Equations"  (WICKE 1981).
 *Wisa type 203

    To optimize  the  precision of the  technique  one must carefully select the
sampling and counting  intervals.  The  calculations show that the precision of
Po-218 measurement  can be  improved  by longer sampling/counting intervals, if
the assumption  of a constant  concentration  in air would  be applicable.  The
precision of  Pb-214  and  Bi-214  (Po-214)* measurement  is  associated  with the
time  lag  between the  end of  sampling  and the  start of  the  second counting
interval.   Table  1 displays  the selected  sampling  and counting periods applied
for this  method  with  associated  sensitivities.   Here  the  sensitivity  is de-
fined as  the  concentration  in air of  a radon daughter  such that the standard
deviation in the  measurement,  associated only  with  the  random nature of  radio-
active decay,  is 20%  of  the  sampled  concentration.   For  comparison purposes
the product of  flow rate and detector efficiency  is assumed to  be  equal to
ll/min  and  the  activity   ratio  of   the  daughters  Po-218:Pb-214:Bi-214  =
1:0.6:0.4.  Comparing  the sensitivities of this alpha-spectroscopic technique
with  other methods  (Table 1)  it  is  evident that the procedure presented here
can be used to make radon daughter measurement more precisely.

    As an extension of  this  method counts  of  Po-212 at  8.77 MeV at  two differ-
ent counting intervals  after sampling  can successfully  be  used for the  deter-
mination of  thoron  daughter concentration in  air.  In  order to obtain a good
sensitivity for  both  Pb-212 and Bi-212 specific  sampling  and counting  inter-
vals  have been  selected  (Table 1).   The sensitivities, due  to 20%  statistical
error,  which  are presented in this table are  based on  an assumed activity
ratio Pb-212:Bi-212= 1:0.5.

    To obtain  a  good  energy resolution of the alpha-spectroscopic instrument
as  well  as  a good  detection efficiency  a  sampling unit  has  been designed
consisting  of  a  surface-barrier  detector^   in  a  distance  of  6mm  from  a
membrane  filter^  (Fig.  6)  .  The   air is  drawn  through   this  filter   by  a
diaphragm pump^  at a  relative constant  flow rate of  22.5H/min.   The  signal
analysis system  is similar  to  the  one  used for the radon/thoron monitor  (Fig.

    The detection efficiency  of the system can easily be calculated using a
Monte-Carlo method  (WILLIAMS 1966) and be checked  by a standard plane source.
For the  geometry of our  system a  detector  efficiency of  17% _+  0.5%  was ob-

    The energy  resolution  of  this  alpha-spectroscopic device  is good  enough
for a satisfactory separation  of Po-218 (6.00MeV),  Po-214  (7.69MeV) and  Po-212
(8.77MeV)  with an overlap significantly below 1% (Fig. 7).
*Po-214 is assumed to be an initial radioactive equilibrium with Bi-214.

1:  Ortec type CR-24-300-100, 19.5mm active diameter
2:  Sartorius  membrane  filter type  SM,  1.2mm pore  size,  25mm  diameter,  20mm
    effective diameter
3:  Neuberger type N035.1.2.AN18

                            TABLE 1:  COMPARISON  OF METHODS FOR MEASURING
                                       RADON DAUGHTER CONCENTRATIONS

THOMAS (1972)
CLIFF (1978)

KEEFE ET.AL (1978)




TI- T2

T3- T4 T5- T6 Po-218 PB-214 Bi-214
11-25 26-35 37,0 5,9 13,4
11-26 27-42
16-36 37-57

3,7 1,6 2,5
2,4 0,5 1,2
•2,9 4,2 2,6
1,4 1,7 1,0
.1,1 0,7 0,6
0,4 0,4 0,4
0,2 0,2 0,2
0,1 0,1 0,1
PB-212 Bi-212
0,35 0,18
0,13 0,2
0,27 0,20
0,13 0,10

   EQUILIBRIUM RATIO:  Po-218:RB-214:Bi-214 = 1:0,6:0,4, PB-2J2:Bi-212 = 1:0,5


                 SURFACE BARRIER
               (19.5mm active diameter)
                     FILTER -_
                                             SCALE: 10mm
Fig.  6;   Sampling  unit   for a  simultaneous  collection  and  detection  of
          alpha  emitters in  air
                 O  50
300 keV
                                  6.00     7.69  8.77
                                ALPHA ENERGY, MeV
Fig. 7:   Alpha   spectrum   obtained   by  collecting  airborne  radon-  and
          thoron-daughters on a membrane filter.

    At  a  flow  rate  of  22.5&/min  and  a  detector  efficiency  of  17%  the
sensitivity  for  each   individual   daughter   concentration  is   better   than
0.03 pCi/J.  based on  the  data  of  Table  1  for  30  minutes  of  sampling  and
counting periods of 3 x 30 minutes after sampling.

    Continuous  operation of  the  system  is  possible.  Since  sampling  takes
place  all the  time  information  is  available  only  about  the  current  alpha
activity of Po-218 (RaA), Po-214  (RaC1) and Po-212  (ThC1) on  the  filter.   This
is  not sufficient to  determine all  5 individual  radon- and  thoron-daughter
concentrations.  But  it is possible to calculate  the  potential alpha energy
concentration  for  each time interval  from the integrated  counts  (HAIDER  and
JACOBI 1974, KAWAJI 1981).  The advantage of this instrument  is  its capability
to  distinguish between  the  potential  energy  concentrations  originating  from
radon- and  thoron-daughters.    Work  on  this   subject  is  still  in  progress;
details will be presented later.


    Alpha-spectroscopic  methods  using  surface-barrier  detectors   are   well
suited for the  analysis  of  radon,  thoron  and  their daughters in  air.  It  has
been  demonstrated  that a specific design  of   instrumentation leads to a  high
sensitivity for grab-sampling application as well as  to good time-response at
continuous operation.


    This work was financially supported by the Federal Ministry of  the Interi-
or, Federal Republic  of Germany.

    Participation at  this conference was made possible by support of  the
New Mexico Institute  of Mining and Technology,  Socorro,  New Mexico.


CLIFF, K.D.  (1978):  The measurement of low concentrations of radon-222
    daughters  in air, with  emphasis on  RaA  assessment.  Phys. Med.  Biol-  23,

    A  radon  detector suitable for  personal  or area monitoring.   Health  Phys.
    17, 193-198.

DALU, G., DALU, G.A. (1971):  An automatic counter for direct measurement  of
    radon concentration.  J. Aerosol Sci. 2, 247-255.

DUGGAN, M.J., HOWELL, D.M. (1968):  A method for measuring the concentration
    of the short-lived daughter  products  of radon-222  in  the atmosphere.   Int.
    J. Appl. Radiat. Isot. 19, 865-870.

HAIDER, B., JACOBI, W. (1974):  A new monitor for long-term measurement of
    radon daughter activity  in mines.   P-  217-226  in;   Radiation  Protection  in
    Mining and Milling of Uranium and  Thorium.   Proceedings  of  a  conference  in
    Bordeaux, France, 9-11 September 1974.

JACOBI, W. (1963):  A new method to measure radon and thoron in flowing gases
    and its  use  to determine the thoron  content of  atmospheric air.   Colloque
    International  sur la  Pollution Radioactive  des  Milieux Gaseux,  Saclay,

KAWAJI, J., LIN PAI, H., PHILLIPS, C.R. (1981):  Use of gross filters
    activities  in  a  continuous working level  monitor.   Health  Phys. 40,

KEEFE, D.J., McDOWELL, W.P., GROER,  P.G. (1978):  The environmental working
    level monitor.   Report  No.  P7628C,  Argonne National Laboratory,  Argonne,

MARTZ, D.E., HOLLEMAN, D.F., McCURDY, D.E., SCHIAGER, K. (1969):  Analysis of
    atmospheric  concentrations   of  RaA,  RaB   and  RaC  by  alpha-spectroscopy.
    Health Phys. 17, 131-138.

NAZAROFF, W.W., NERO, A.V.,  REZRAN,  K.L., (1981):  Alpha-spectroscopic
    technique for  field  measurement of radon  daughters.   Paper  presented  at
    the Second Special Symposium on Natural Radiation Environment,  Bombay.

ROBERTS,  P.B., DAVIES, B.L.  (1966):   A transistorized radon  measuring
    equipment.  J. Sci. Instr. 43,  32-35.

THOMAS, J.W. (1972):  Measurement of radon daughters in air.   Health Phys.
    23, 783-789-

WICKE, A. (1981):  Comparison of different methods for radon daughter
     measurement.   Laboratory  Report  New  Mexico  Institute  of  Mining  and
     Technology, Socorro, New Mexico.

WICKE, A., PORSTENDORFER, J. (1980):   Exposure to radon daughters in
     dwellings.  Paper  presented at  the  Specialist  Meeting on  the Assessment
     of  Radon  and  Daughter Exposure  and  Related  Biological  Effects,  Rome,
     March 3-7, 1980.

WILLIAMS, I.R.  (1966):   Monte-Carlo calculation of source-to-detector
     geometry.   Nucl.  Instr. Meth.  44, 160-162.

WRENN, M.E., SPITZ,  H.,  COHEN,  N. (1975):   Design of  continuous  digital
     output  environmental  radon  monitor.    IEEE Trans.   Nucl.   Sci.  NS-22,

                           OF PASSIVE RADON MONITORS

                                  Lyle  Rathbun

                               AeroVironment Inc.
                                145 Vista Avenue
                          Pasadena, California  91107

    A  passive  environmental  radon monitor  (PRM)  based  on the  electrostatic
collection  of  218po  ions  on a  thermoluminescent  dosimeter  (TLD)  was  intro-
duced by Andreas George  (1977).   Since then, the device has undergone consid-
erable testing, modification and field use by several developers.

    The important changes to  the  passive  environmental radon monitor have  been
increased  volume,  a  more  sensitive  TLD, convenient  dessicant  cartridges, a
safe and reliable case, an  improved TLD holder,-  and  a  non-changeable dessicant

    This paper  describes AeroVironment's  experience  with the development  and
use of the  PRM and the present state of the art.


    In  1978,  AeroVironment  Inc.  performed  environmental  monitoring  in  the
vicinity of proposed  uranium mines in  north-central Wyoming.  One of our  tasks
was to collect ambient radon concentration data.

    Following accepted  procedures, we used  Tedlar  bags  and  small, regulated
air pumps  to collect representative  48-hour samples.  We  had  the usual  pro-
blems associated with this  method.  The  air  bags leaked.   The pumps quit.  On
occasion snow and mud prevented us from  reaching the  sampling locations  at the
appointed times.

    These problems  in monitoring  radon led  us  to  search  for a better method.
We  found the  answer in  a  passive  radon monitor,   described  by  George  and
Breslin  (1979)  and  Cowper  and Davenport  (1978).  Using their description  as a
starting point, an equipment manufacturer built  the FR-3 for us.

    We began field  testing  several of the passive radon  monitors  (PRMs)  early
in  1979, comparing  results with  the  Tedlar  bag method.  The  results  of  this
effort  were presented  to  the  IEEE  (Friedland  et  al. ,  1980)  and eventually
referenced  by the  revised  version of NRC  Ref. Guide  4.14.   As  Tables  1 and 2
show,  we  found our  PRMs  results  essentially   comparable  to the  Tedlar  bag
results, without the problems.

    TABLE 1.  Comparison results of Tedlar bag method vs. PRM method.
Tedlar Method
0.79 + 0.05
0.43 .+ 0.04
0.43 + 0.04
0.65 + 0.04
0.62 + 0.05
1.07 + 0.06
0.69 + 0.05
PRM Method
1.03 + 0.08
0.52 + 0.04
1.68 + 0.05
1.70 + 0.06
0.42 + 0.03
0.81 + 0.06
0.68 + 0.05
TABLE 2.   Averaged comparison  results of Tedlar bag method vs. PRM method,
                              Tedlar Method
                     PRM Method
     Combined  average

     over a season  at

     a given site
0.75 + 0.06

0.52 + 0.04

0.75 + 0.05

0.66 + 0.04
0.69 + 0.06

0.47 + 0.03

0.74 + 0.05

0.69 + 0.05


    The PRM is usually comprised of a  three-compartment cylinder.  The princi-
ple component  of  the unit  is  the ionization chamber,  with its thermolumine-
scent dosimeter  (TLD)-bearing  cathode.  Accessory to this  are  a small bottom
chamber, which contains a layer of dessicant sandwiched between two wire grids
with filters;  and a top chamber,  which  contains  the battery supply for main-
taining an electric field between  the  cathode and the walls of the ionization
chamber.  The  entire  PRM is encased in a cover of PVC or metal.  The battery
and ionization chambers  are separated  by a solid plate,  which incorporates a
central opening for the cathode.  The top  of  the battery compartment is pro-
tected by a lid.

    The primary TLD, which  is  covered  by an aluminized mylar window, is posi-
tioned on the  tip of a cathode, which protrudes  downward into the ionization
chamber.  Radon diffuses through the filters and  dessicant and  into the bottom
of this chamber.   Radon-daughters are removed by  the filters.  As the radon in
the chamber undergoes radioactive decay, each emitted alpha particle is likely
to induce  a loss of more  than two electrons.   Hence,  the  resultant radium-A
ions generally are positively  charged and  tend  to  be  deposited on the nega-
tively charged window of  the  cathode.   As  each  deposited radium-A ion decays,
the  energy  from  the  emitted  radiation  has a definable  probability  of being
trapped by  the TLD.  This also holds true  for  successive members of the decay
chain.  Most  of  the energy deposited  in the TLD  is from the alpha particles
which  have been  emitted  by   radium-A and  radium-C1.   However,  due  to  the
possibility of significant TLD exposures  by  cosmic  and ambient  terrestrial
radiations, background  TLDs  may prove  necessary.   These  are  placed on  the
cathode but are shielded from alpha radiation.

    Our first  PRM,  the FR-3, has  the  basic  construction we  have described but
it differs  from  later  models  in that it is all  metal.  Other distinguishing
features  include  short  metal  legs  and  a  battery  supply  consisting of  two
510-volt cells.   While we have  fairly  good  results with  this  unit, we were not
entirely happy with it.  The  battery  arrangement did not  allow good wiring.
Inserting  the cathode   between the  batteries  could  result  in  an  electric
shock.  Inconveniently, the dessicant had to be poured  out  a small window for
drying.  The  LiF TLDS  we  used at  first  were  not as  sensitive as we thought
they should be.


    Intending to correct the problems  noted above, we asked our supplier for a
new  model,  dubbed  the  FR-32.   It was  constructed mostly  of  PVC material.
Graphite powder  was used to  coat the inside of  the ionization chamber.   The
dessicant cartridge was easily  removable,  allowing  the  TLD  probe to  be  in-
serted  from the  bottom.   Four  batteries of 300 volts  each  were used in place
of the  previous  battery supply.  The  more  sensitive TLD-200's  (CaF2~Dy)  were
used as the dosimeters.   However,  we  still had a few problems.  The graphite
lining of the  ionization chambers  proved unsatisfactory and was replaced by a
metal liner.   The  battery  compartment  lids often leaked and allowed battery

 damage.   The dessicant cartridges were  made of PVC  and  hence could not with-
 stand  the high  temperatures  required for  speedy  dessicant  drying.   This was
 corrected  in the later model  of the  FR-32 substituting a metal dessicant cart-

    Our  latest PRM,  called  the LR-5  (diagrammed in Figure 1), has no dessicant
 cartridge  at all.   Instead we have  a thin  sheet of  material at  the bottom
 which  allows radon  to  diffuse in,  but keeps the moisture out.  The new  battery
 compartment  lid  is  a press-fit PVC cap.   This never  has  to  be removed, except
 for  the infrequent battery changes.   A small PVC pipe  passes through  the lid
 and battery  compartment.  The  TLD  probe  is  inserted  into this small pipe where
 it makes electrical  contact.   The  portion of the probe containing the TLD pro-
 trudes down  into the  ionization chamber.   The  new probe  also eliminates some
 of the TLD breakage we had  experienced earlier.

    Thus,  our present  PRM  is  the  result  of  considerable development  and re-
 finement.  However,  nearly  all the modifications were performed with an aim of
 greater safety,   convenience  and  dependability.  The  size   and  shape  of  the
 ionization chamber  and the positioning of  the  TLD have changed very little in
 the  development  of  our PRM.  The 8-liter, eight-inch diameter ionization cham-
 ber  has  been standard.  In  each  PRM, the  active TLD  is  covered by  a mylar
 window and  positioned  1-3/4  inches  from  the top of the ionization chamber.
 Following  early  testing with a 1/8" LiF  TLD, the  current 1/8" x 1/8" x 0.015"
 CaF2~Dy TLD  was chosen  and  has  been the  exclusive  TLD  in  our  monitoring.
 Since  the TLD and  ionization chamber constitute  the  heart  of the  system,  we
 have  had a  consistent  frame  of reference for  comparison and calibration pur-

    Performance  of  the LR-5  and its  predecessor,  the  FR-32,  were  compared in
 March  and April, 1981.   Table  3 shows the  results,  which indicate  that,  in-
 deed,  the  two are comparable.


    The  calibration  of the PRMs involves two separate processes.   First,  the
 TLDs  are calibrated using  gamma source  at a known  distance according  to  NRC
 and  ANSI standards.   A  secondary  standard is  used  to check TLD performance.
 Then,  semi-annually, individual PRMs are  exposed to  known radon concentrations
 and calibrated from  the TLD readouts  of  these PRMs.   Several of these calibra-
 tions  have been  performed,  including  both laboratory  and  field exposures.   The
 results are  fairly comparable, as evident in Table  4.

    For  TLD  readout, we have  used  both the  Harshaw  Model 2000  A & B  and  the
Teledyne  7300.   We have  been pleased with  the performance  of each of these
units.   It  should be  noted that each  TLD reader has its own  characteristic
response,  which  will be  different from that of other TLD readers,  even  with
identical settings on units of the same model number.   Hence, new calibrations
are required  with each  reader.

                                  Access Tube Lid
4 each, 300 V
Eveready No.
    Screen with
                                                              TLD Probe
                                                          Cathode Assembly
      FIGURE 1.  Schematic diagram of the LR-5 Passive Radon Monitor.

TABLE 3.   Field intercomparison of  different PRM models.

Site I March 81
April 81
Site 2 March 81
April 81
Site 3 March 81
April 81
           Table 4.   PRM calibration results.

Model No.





New Mexico Institute
of Mining & Technology
Denver Mining
Research Center
New Mexico
Field Calibration
New Mexico
Field Calibration
Field Calibration

Nov 79

Apr 80

Dec 80

Jul 81

Jul 81






pCi/1 hr







    In  conclusion,  it should  be  noted the LR-5  Passive  Radon Monitor  is  the
end product  of  considerable testing  and  modification.   It  is  safer and  more
reliable than earlier  models,  yet it maintains the desired  level  of durabili-
ty.  It also retains the capability to monitor normal ambient concentrations.

    We would like to thank the following individuals for their help and advice:

    Of the Denver Mining Research Center of  the U.S.  Bureau  of  Mines  — Robert
Droullard, Ted Davis, Emery Smith and Robert Holub.

    Of  the  Environmental  Measurements  Laboratory  of the U.S.  Department  of
Energy — Al Breslin, John Harley, and Andreas George.


George, A.C., and A.J.  Breslin (1979):   Measurements of environmental radon
    with integrating instruments.   Health and  Safety Laboratory,  U.S.  Energy
    Research and  Development Laboratory  (now EML,  DOE) New  York, NY  10014.
    Published in the Workshop on Methods for  Measuring Radiation in and Around
    Uranium  Mills,  by  the  Atomic  Industrial  Forum,   Inc.,  Washington,  D.C.

Friedland,  S.F.,  L.  Rathbun  and  H.M.  Goldstein (1980):   Radon monitoring:
    uranium mill field experience with a passive detector.   Paper presented at
    the  Nuclear   Science  Symposium  of  the   IEEE,  October  17-19,  1979,  San
    Francisco,  CA.

Friedland,  S.F.,  L.  Rathbun  and  J.  Smith  (1979):  Radon monitoring near a
    uranium mill with a passive  detector.  Paper presented at  the 2nd  Symposi-
    um on  Uranium Mill Tailings Management.   Colorado State University,  Fort
    Collins, CO.,  Nov.  19-20.

Cowper, G., and M.R. Davenport (1978):  An  instrument for the measurement of
    long-term  average   radon   levels.   Chalk  River  Nuclear   Laboratories,
    Ontario, Canada.  Published  in  International  Atomic Energy Agency,  publi-
    cation  number SM-229/100.

                                AND FIELD RESULTS
                                  H.  Ward Alter
                                James E.  Gingrich

                              Terradex Corporation
                                460 N. Wiget Lane
                         Walnut Creek, California 94598


    Initial  calibration results  with  the  Track Etch  radon  detection  system
have  been previously  reported.(1)   Additional calibration data  have  recently
been  obtained  in  a wide variety of  standard radon  test  chambers  and prelimin-
ary results  from  this  work  can be reported.  Seven  different radon test facil-
ities have been utilized for these tests in addition to  the  DOE  Environmental
Measurements Laboratory  which was  utilized  for the  initial  tests.   Six differ-
ent detector configurations  have  been calibrated including  two new miniaturiz-
ed  configurations which may  be suitable for  personnel  dosimeters as well  as
other applications.  Individual calibration tests have used  2 to  20 replicates
of each detector  configuration.   Radon  exposures have ranged  from 20 to 10,000
(pCi/l)-days at radon  concentrations from 5 to 1000 pCi/1.   Results give  mean
values for each  set of  replicates as a  function of the  total exposure.   Cali-
bration  factors   and  detector  sensitivities  derived from  the tests  will  be

    Large  numbers of  Track  Etch   detectors  have  been and  are being  used  for
radon monitoring  in homes,  buildings and outdoors.   The results   from  some  of
these measurements will  also be discussed.


    The  Track  Etch  *   radon measuring  technique  has been  widely  used in  a
passive manner to make  time integrated  radon concentration  measurements.^2)
Applications of  this  technique have  included  radon soil gas  measurements  for
uranium   exploration,   earthquake  prediction,  geothermal   exploration,   and
fundamental  geological  and  geophysical  studies.'3,4)    Track  Etch detectors
are  now   being extensively   used  for monitoring  radon  in  houses,  buildings,
uranium mines  and processing  facilities  and  in other areas  where there is  a
need to know exact radon concentrations. ^-* ,6.)

    With  the Track Etch  technique, an alpha sensitive plastic detector (usual-
ly mounted in  a  plastic container or on  a  card) is exposed  to the atmosphere
to be measured.   Alpha particles  from radon in  the air;  or  from  radon  daugh-
ters plated out on various adjacent  surfaces,  penetrate  the  detector and cause
radiation damage  tracks  that are  subsequently revealed by an etching  process.

The number  of  alpha tracks counted per  unit area (T/sq.mm) is therefore  pro-
portional  to  the average  exposure  (eg.  (pCi/1)-hours) .    Exposure  times  can
range up to a  year  or more, if desired, using improved detectors  which  retain
alpha tracks without fading for very long time periods.

    Early types of Track Etch detectors  suffered  from deficiencies in sensiti-
vity, track fading  and  optical quality,  although early  tests showed  the  pro-
mise  of  a passive  integrating  device.   In recent years,  an improved type of
detector has  been developed which  overcomes the deficiencies  of the earlier
type.  This detector  has  already  seen wide use in a number  of  radon measuring
applications during the last three years.

    Track  Etch detectors  have several  features  that make them  particularly
attractive  for environmental  radon monitoring  uses.   They  are simple,  small
and  lightweight,  and  they have  no moving  parts,  electronics or power  sup-
plies.  They are  also sensitive only  to  alpha particles  and are not sensitive
to  any  gamma  or  beta radiation that  may be present.   Since the detectors are
so small and have no moving parts, they can be easily used  for  making unobtru-
sive  radon measurements  in occupied houses and buildings.   The etched detect-
ors also provide  a permanent record for future reference.


    Several different Track Etch  detector  configurations have been developed
for  different applications.   Either  radon-only or  combined radon  and  radon
daughter  measurements can be  made.   The various  configurations are described

    Open  Cup   (Type C) .  The Type C Track Etch detector  system consists  of  a
small plastic  cup with  the detector element attached to  the inside  bottom of
the  cup.    This  open cup  system  is the one  usually  used  in  making  soil  gas
measurements  for  mineral exploration and  earthquake  prediction  applications.
In  the Type C  configuration, the  detector measures the primary  radon gas alpha
particles  originating within the  cup  as  well as  the  alpha particles  from the
radon daughters  plated-out on  the inside walls  of  the  cup.   The  Track  Etch
reading for an open cup is not only a function of the radon  and radon daughter
activities, but  it  is also a  function of the plate-out characteristics  of the
ambient atmosphere.   It  is therefore  not recommended for  other than  soil  gas
measurements.   (Figure 1)

    Membrane Covered  Cup  (Type M).  The Type M Track Etch  detector  system is
the same as the  Type C except that the  opening of  the cup  is  covered  with  a
semi-permeable  plastic   membrane.(7,8)   Tnis membrane   (sometimes   called  a
Thoron Filter)  slows the normal diffusion of noble gases into the  cup and  thus
discriminates  against radon-220 (sometimes called thoron  that has a  55  second
half-life)   while  permitting 60  to  70%  of  radon-222  (which  has  a  3.8  day
half-life)  to  enter  the  cup.   The membrane  covered cup  system is mainly  used
in  making   soil   gas  measurements  for  mineral  exploration applications  to
eliminate  thoron interference  and  water accumulation in the  cup.   it also  pre-
vents  the  entrance of radon daughters  and thus measures a  radon-only  signal.


Figure 1. Type-C Open Cup Track Etch Detector


    Filter Covered Cup (Type F).  The Type F Track  Etch  detector  system is  the
same as the Type C except that the opening of the cup is covered  with  a hydro-
phobic, microporous  filter which  permits rapid  and complete infiltration  of
the  gaseous  radon  isotopes  but  discriminates  against  the  nongaseous  radon
daughters.  Because of its higher  sensitivity  to radon, the  Type F configura-
tion is preferred  to  the Type M configuration  in all radon-only applications
where thoron is not an important component.

    Bare Detector Card (Type B) .   In  this configuration the  Track Etch detec-
tor  element  is mounted flat on  a  card facing  the  atmosphere to be measured.
The  effective  measured  volume  is a  hemisphere  with  a  radius  of  3.0 inches
 (7.68  centimeters),  the  range of. the polonium-212  alpha  particle  in  air,  or
2.5  inches  (6.30  centimeters),  the  range  of  polonium-214  alpha  particle  if
only  the  radon-222 series is present.  In  using the Type B  card detector,  it
is mounted in  such a  way that  it is  always  at  least 4 inches (10 centimeters)
.from  any  adjacent surface to eliminate any alpha contribution  from plated-out
radon  daughters on that  surface.   The bare  detector measures radon plus ambi-
ent  radon daughters and  is  sensitive  to  plate-out  only with respect  to radon
daughters  plated  out directly  on the  detector.   The  reading   on the  bare
detector  will  therefore  be a  function of both radon and  the radon daughters.
The  Type  B  is  mainly used for indoor house and building  monitoring.   It  has
been  found  that  in  most of  these  environments the  Working  Level  Ratio  is
normally  in  the  range of 0.3 to  0.7 with an  average  of about  0.5(9'10)   By
using  this  ratio,  the Type  B detector  reading  can be  reported in  radon  or
radon  daughter  concentration  units (pCi/1 or WL).  Approximately  26,000 indoor
measurements have  been made with the  Type  B detectors.   (Figure  2)

    New  Detector   Systems  (Type  SF and Type SM) .   Recently two  new  detector
configurations  have  been  developed  and preliminary calibration  results  have
been  obtained  on the two  systems.   In the two  new configurations  (identified
as Type SF and Type SM) ,  the Track Etch detector element is mounted in a much
smaller cup  which is 1 1/4 inches in diameter  and  3/4  inches high  (3  cm O.D.
by  2 cm) .   The opening  of  the cup  is  covered by the standard microporous
filter  (Type SF)  or  the  standard  membrane  filter  (Type SM) .   In both config-
urations,  the  detector  element  responds  only  to  the  radon  signal.   These
miniature  cups  are expected to  be ideal  for  use as personal radon dosimeters
or as  devices to  measure  radon in  homes or  in  other applications  where a very
small passive device is needed.  (Figure   3)


    In using any  radon detector system it  is  important that it  be fully  and
accurately calibrated at  known  radon  concentrations.  This can pose a problem
in calibrating all  types  of  radon detectors because of  the  difficulty of pro-
viding a radon  source of uniform concentration  for a  long  continuous  period.
In addition,  the radon provided for calibration  should have  low thoron content
and  be well characterized  with  regard  to  radon  daughter  concentrations,
aerosol size and concentration, and humidity.  The  calibration  facility should
also provide adequate  space  to permit simultaneous testing of many replicate
devices.    Only  a  few radon  test  chambers  around   the  world  can  meet  these
criteria.   The test chambers chosen for most of the Track Etch detector cali-
brations  reported  here were those operated by  the U.S. Department of _Energy


Figure 2.   Type-B Bare Detector Card  Track  Etch  Detector


Figure 3.   Type-SF Track Etch Personnel Radon Dosimeter


at  the  Environmental Measurements  Laboratory (EML), although  we  have subse-
quently calibrated  the  detectors in  four other test  chambers.   The earliest
EML calibration  tests were conducted in  a  small  (2  cubic meter)  test chamber
while a  larger  test  chamber  (20 cubic meters) was  extensively used  in the
later phases of  the calibrations.   Radon concentrations in both test chambers
were measured  using a continuous scintillation flask  monitoring  system.   The
Working Level Ratios approximated the ratios  found in most indoor environments
(0.3 to 0.7).

    The test exposure conditions  for  the calibrations  in the EML are shown in
Table 1.   Each of  four  different detector types  was  calibrated and  in  some
instances as  many as 20  replicate  samples were used  for  each detector type.
After the  detectors were exposed, they  were  returned  to Terradex Corporation
for processing and  reading.  At least 100 tracks were counted  for each detect-
or  to keep the relative  standard deviation to less  than  10%  due  to counting
statistics.  The  relative  standard  deviation  measured  from each set of repli-
cates agreed with counting statistics.

    Table 2 combines the  Track Etch detector  readings with the radon exposures
as  reported by the  Environmental Measurements Laboratory.  Each detector read-
ing is the mean  of  the  replicate  readings.  To obtain  the  appropriate calibra-
tion factors,  the track density  readings  were  divided by  the measured radon
exposure values and the resulting data are  presented in Table  3.  Table 3 also
shows,  for each  configuration,  the mean  value  and related  statistics.   The
mean and its standard deviation are calculated by weighting the Table 3 values
by  the number of  replicates at each exposure.  The error includes any error in
the radon  exposure  as measured by  the  chamber  operator.   The higher standard
deviations of  the Type  C and Type B configurations  reflect  the varying radon
daughter concentration at each exposure.

    An  example of  the  replicate detector  response  from a single  radon  test
exposure is shown in Table  4.  These data show that  there is excellent sample
to  sample  reproducibility of the four  different detector  types.   The percent
uncertainty (relative standard deviation) of  a single detector is in the range
of  7  to  13%. (1)   This  is  consistent  with  the  expected uncertainty  due  to
counting statistics for  a single  detector.

    Since this calibration  work was completed, several other calibration tests
have been  made,  both  at EML and  in  other chambers in 1980 and  1981,  for the
Type F  and Type B configurations.  Additional calibration and intercomparison
work is continuing at six different  radon test chambers on  a regular basis.

    Mean calibration factors  for  all  1979 and 1980 runs fall within 1% of the
1979 FILT value  reported  in Table 3 and  within  10% of  the  Type B value.  Mean
T/sq.mm values for  all runs are plotted  against  (pCi/1)- days  in Figs. 4 and 5
together with  the weighted mean lines  from Table  3.  These  figures  show the
data from all  of the EML tests as well as  that obtained from  three other test


Test Chamber
1 2m3
2 20m3
3 2m3
4 20m3
5 20m3
6 20m3
7 20m3
8 20m3
9 20m3


Average Radon
Exposure Rate




Radon Exposure

Detector Reading
(Mean Tracks/sq.mm)
Type C
Type M
Type F
Type B


Weighted Mean
Relative Standard
Deviation of the
Mean (%)

Type C

Type M

Type F

Type B

                              TABLE 4

Std. Dev.


Type C
Type M
Type F
Type B
Note:  Test conducted  in  20  cubic  meter  EML  test  chamber  at  radon
       exposures  of  996  (pCi/l)-days





1979  EML RUNS:

1980  EML RUNS:*,A,



                                                            1000   3000   10,000  30,000   100,000
                                       RADON  EXPOSURE   ((PCi/l)-DAYS)
                               Figure 4.  Type-F  (Filtered Cup) Detector  Calibration


         1979 EML RUNS!

         1980 EML RUNS!

         CHAMBER A!*

         CHAMBER B: +

         CHAMBER C:%
                         10     30       100    300    1000   3000   10,000  30,000   100,000

                                  RADON  EXPOSURE   ((PCi/l)-DAYS)

                          Figure 5.  Type-B  (Bare Card) Detector Calibration

       The  plots  show  the  somewhat  increased   (over  counting  statistics)
scatter of  the  data for the  Type  B configuration which is  responding  to the
variable daughter and plate-out environment in each calibration test.

       Using  the  statistical  information  from  the  calibration  tests,  the
sensitivity  of   the different  detector  configurations  was  calculated  as  a
function of the area  counted on each  detector. (5)   The results  show  that a
one to three month  integrating  period is  adequate for  a  variety  of outdoor or
indoor measurements at typical radon concentrations  (See  Table 5).
                                    TABLE 5



Type C
( (pCi/1) -Months^

Type M

Type F

Type B
*Sensitivity  is  that  measured  radon  exposure  whose  relative standard deviation
due to all sources of error is 50%.

    At  the  present time, Track Etch detectors are being used to make measure-
ments in  a  variety of  environmental monitoring situations.  The largest use  is
for house and building  monitoring, mostly  in selected areas where high  radon
concentrations are expected.   Other uses  include monitoring around operating
uranium mines,  mine tailings and  mill  sites, baseline studies around planned
uranium mill sites, monitoring around  phosphate  and radium processing sites,
monitoring  in radium  needle  storage facilities  and measuring radon  in  soil

    Outdoor  Surveys

    To  make outdoor radon measurements, the  Type F detector configuration  is
usually employed,  and  it is often  placed in a canister to protect it from the
weather elements.   The  Track Etch detectors are ideally  suited  for outdoor
measurements since  they  can  make  long   time-integrated  readings  and   they
require no  power supplies or  maintenance  of any kind.  These features are  of
particular  value  in the outdoor environment, such  as  near uranium mining and
milling facilities  where the  radon  levels vary  greatly  as  a  function  of
production  activity,  weather  and  season,  and where a  temperature-insensitive
passive device is essential.  (Figure 6)

Figure 6.   Type-F Track Etch Detector Being Placed in Its Canister

    Preliminary  results  from  the  radon  measurements  around  one  operating
uranium  mine  and  mill  complex have  been  previously  reported. (6)    in this
measurement program,  12  sampling  stations  were established around the facili-
ties and the detectors were  left  in place  for  a month at  a time.  This program
has  now been  continued  for a  full  year  and  the data from  three  typical
sampling stations  are shown in Fig. 7.  At  each station  it can be noted that
there is a  fairly  large month to month  variation depending on the activities
at the facility and the  changing  weather patterns.  The maximum  variations are
typically two  to  three  times  the  average  concentrations,  and they  vary  by
factors of  5 to 10 from the highest months to the lowest months at the same
locations.   Normally  the on-site measurements  were significantly  higher than
either  the  site boundary  or the off-site measurement.   The on-site measure-
ments for the  year averaged 2.62 pCi/1 while  the site  boundary averaged 1.18
pCi/1  and   the off-site averaged  0.89  pCi/1.   It is   interesting to  note,
however, that  during  the  month of May  the  off-site location  had  a slightly
higher  average radon  concentration than  the other two  sites.   The year-long
average  radon  concentrations show  a  typical fall-off with  distance  from the
center of the mine/mill complex.  (Figure 8)

    Indoor Surveys

    To  make indoor measurements,  either the  Type  B  detectors or  the  Type F
detectors are  usually employed.  Most indoor  measurements  made  to  date have
been made with the Type B detector cards, since  they are  easier to handle and
use in occupied dwellings.   (The  difference  in response  of  these two detector
types was  discussed previously.)   The detector  cards  are usually attached to
the wall or  hung from the ceiling of  the rooms being monitored.  Often two or
more detectors are placed in each home being monitored, and the  sampling loca-
tions are selected on the  basis of  expected  occupancy rate and  those locations
suspected of having high radon concentrations.

    The  statistical data from almost  30,000 indoor Track Etch radon measure-
ments in seven different major locations are shown in Table 6.  The locations
include:   Northern California;  Houston,  Texas;  Northeastern  U.S.;  Eastern
Pennsylvania;  the  State  of Maine; Saskatchewan,  Canada, and numerous locations
in  Sweden.    Many   of these  measurements were  made  in  ordinary homes  not
suspected of having high radon  concentrations,  but  many  (particularly those in
Sweden) were made  in  homes suspected of having  higher  than normal radon con-
centrations due to location or due to building  materials.   Most of the  houses
were built  with the  normal construction  practices for  that general location
and  few were  known to  be especially constructed  for  energy efficiency.  In
most instances the Track Etch  detectors  were  left in place for  3 months but
individual detectors  were left  in place  as  short as 1 month or as long as 10
months.  The summertime  periods were  usually avoided when making the measure-
ments  because  of  the tendency  to have  unusually high  ventilation  rates  at
these times due to warmer weather.

    The general statistical  results from the indoor surveys  are  shown in Table
7.   In  Table  7, radon  concentration data are given in  picoCuries  per liter
(pCi/1)   averaged  over  the exposure period.   Statistics   show  the  mean radon
concentration  (which is close to the median  value in all  surveys), the low and
high values,  and  the percent of the  data exceeding 20  pCi/1.   Table  7 also
shows these  results converted to Working  Level  (WL)  assuming a Working Level
Ratio (WLR) of 0.5.   The 20  pCi/1  level  (0.1 WL) would expose  an individual to
about 2 Working Level Months  (WLM) per year.

                         On Site Average 2.62
               —— — — Site Boundary Average 1.18
                         Of f Site Average 0.89
              Figure  7.   Atmospheric Radon Around a Uranium Mine-Mill  Complex

2   2.0
                                       ONE-YEAR-LONG READINGS
         Figure 8.
                        6000                12000

                    DISTANCE FROM  URANIUM  MILL  (FEET)

                Reductions in Radon Concentrations Away From a Uranium Mill

                                    TABLE 6
N. Calif, winter-summer
Houston fall-winter
No. East U.S. summer
E. Penna. winter-spring
Maine fall-spring
Canada winter-spring
Sweden winter-summer
Sweden spring-summer^
Sweden fall-winter^
Sweden winter
Sweden winter-spring
Time (Mos . )
    (1) Selected Homes
                                    TABLE 7

N. Calif, winter-summer
Houston Fall-winter
No. East U.S. summer
E. Penna. winter
Maine fall-spring
Canada winter-spring
Sweden winter-summer
Sweden spring-summer
Sweden fall-winter
Sweden winter
Sweden winter-spring

0.1- 91
0.1- 34
0.4- 22
Working Level Greater Than
0.002 -0.11
0.001 -0.53
0.001 -2.6
0.002 -2.3
0.002 -1.8
*Expressed  in  Average  Radon  or  Radon  Daughter  Concentrations  and  Assuming
Working Level Ratio of 0.5

    The radon  concentrations  in those homes from Northern California, Houston
and Northeastern U.S.  in  the  summertime were relatively  low and they clearly
show  the  influence of  warm  weather  and  increased  ventilation.    The  high
concentrations  from Eastern Pennsylvania and Maine  show the  dual  effects of
reduced  air change  rates  and  high  radon  production  rates  from  the uranium
containing soils and rocks under and around  these  homes.

    In Sweden,  the  radon concentrations in ordinary houses are higher than the
low U.S.  results,  undoubtedly  due  to  tight  Swedish  construction practice.
(Selected  Swedish  homes  are  from  communities  pinpointed  by  geology  and
construction materials which  are thought to lead to high radon values.)   The
Canadian  homes were  selected  from  a  group showing  0.01  WL or  greater  by
previous grab sampling.

    A  significant  number of  the homes  measured  in Eastern Pennsylvania and
Sweden  have radon  concentrations exceeding  20  pCi/1  (0.1   WL) .   This  is  a
concentration at which there may be  a significant concern about an  increase in
lifetime lung  cancer  risk.   The high concentrations  were somewhat expected in
the Swedish locations but were  unexpected in many of the  other locations.  The
highest concentrations measured in  a number of homes  are at  levels unaccept-
able for  workers  in uranium mines.   It  should also  be noted that present in-
door provisional standards  in  the 0.02  WL  (4  pCi/1)  range may be considerably
below the actual levels in many normal homes.

    Figure  9 shows  the statistical distribution of the indoor  radon concentra-
tions from  4709 measurements from Sweden and from all  of  the measurements made
in Maine  and Houston.(H)  The data from these  and all  other indoor surveys
have  a log-normal  distribution.  In  addition,  the data show  the  range  of
values  that can be expected as  they are  influenced  by  geology and building
ventilation.  Both  the Swedish  and  Maine  results  are high because of geology,
construction materials and  the  lower air change  rates in cold climates.   The
Houston data represent well-ventilated homes in a warm climate with apparently
little uranium mineralization in the underlying soil  and rocks.


    The results from calibration tests  and  field  measurements  demonstrate that
the Track  Etch technique for  radon  monitoring provides  adequate  sensitivity
and reproducibility for  both  indoor  and outdoor  measurements.  The technique
is also  practical  and easy to  use  when making  large-scale,  long-term radon
measurements.   The  completely  passive nature  of  the  detectors makes the Track
Etch technique  particularly attractive for  making  measurements  in any loca-
tions  where it is desirable  that  they  be left unattended for a period of
time.   The  use of  selected types of Track  Etch  detector  systems permits the
selective measurement of radon only and the  elimination of thoron  response.

                                                         4709 Detectors
                                                          427 Detectors
                                                          304 Detectors
 1.0             10.0
     pCi per Liter
   Figure 9.   Statistical Distribution  of Indoor Radon  Concentrations


    Field use of  two different detector types has been demonstrated around an
operating uranium  mine/mill  complex and in approximately  30,000  indoor loca-
tions.  The results  from the measurements around the uranium mine/mill complex
indicate that  the average monthly  radon concentrations  change by significant
amounts  (a  factor of  5 to  10)  and that the  average  concentration decreases
away  from  the  mine  site as  might be expected.  The highest  radon concentra-
tions near the mill  were below the  concentrations measured  in  many homes.  The
results  from  indoor measurements indicate  that  the radon  concentrations in
homes  may  be  significantly  influenced by  their  location with regard  to the
amount of uranium  in the soil  and rocks beneath and  around the house, as  well
as the  uranium content of the construction materials  and the construction of
"tightness".  In a large percentage of  homes already monitored, the radon  con-
centrations exceed provisional indoor standards and a  few have radon concen-
trations that  are unacceptable in uranium mines.   This work also demonstrates
that  the radon concentrations  in significant numbers of homes exceed 20 pCi/1
(0.1  WL).   This  concentration approaches  the  maximum permitted  exposure in
uranium mines.  Should current attempts to  reduce the 4 WLM mine limit to 0.7
WLM be  successful,  up  to  50  percent of  homes monitored  in  the  current  work
would exceed the  mine  standard.   A  significantly larger amount of indoor radon
data  will  be  required in the  U.S.  to establish a  reasonable  data  base  and to
determine  if  radon  is a major  public  health problem.  The  magnitude  of the
potential problem  will become  greater as there is  greater emphasis on energy
conservation in houses and they are sealed for greater  energy  efficiency.


    This paper could not have  been  written  without the  kind cooperation of the
following  organizations,  under  whose  sponsorship  the  work  was  done:    U.S.
Environmental  Protection Administration,  Las  Vegas,  Nevada; Statens Provnings-
anstalt, Boras, Sweden;  General Electric  Company, Fairfield,  Connecticut; and
the Pennsylvania Power and Light Company,  Allentown,  Pennsylvania.

1.  Alter,  H.W.  and  R.L.   Fleischer  1981,  "Passive  Integrating Monitor  for
    Environmental Monitoring," Health Physics Vol. 40, pp. 693-702.

2.  Alter,  H.W.  and  P.B. Price,  1972,  "Radon Detection  Using  Track Registra-
    tion Material," U.S. Patent 3,665,194.

3.  Gingrich, J.E. and  J.C. Fisher,  1976,  "Exploration  for  Uranium Utilizing
    the Track Etch Technique,"  25th International Geological Congress, Sydney,

4.  King, C.Y., 1978, "Radon Emanation  on the  San Andreas Fault", Nature, Vol.
    271, p. 516.

5.  Alter,.  H.W.,  1981, "Passive  Integrating  Radon Monitor  for Environmental
    Monitoring," 2nd  Special  Symposium on  the Natural Radiation Environment,
    Bombay, India.

6.  Gingrich,  J.E.,   1980,  "Field  Measurements  with  the Passive  Integrating
    Track  Etch  System," 3rd  Symposium on  Uranium  Mill  Tailings  Management,
    Colorado State University,  Fort Collins, Colorado, p.  505.

7.  Ward,  W.J.,  1977,  "A Convenient  Method for  Reducing the  Radon-220  Back-
    ground in Uranium Exploration," U.S. Patent 4,064,436.

8.  Ward, W.J., R.L.  Fleischer  and A. Mogro-Campero, 1977,  "Barrier Technique
    for Separate Measurements of  Radon  Isotopes," Rev.Sci. Instrum.  V.  48,  p.

9.  McGregor, R.G. et.al.,   1980,  "Background Concentrations  of  Radon and Radon
    Daughters in Canadian Homes,"  Health Physics,  V.  39,  p.  285-289.

10. Stranden, E.,  et.al.,  1979,   "A  Study on   Radon  in  Dwellings,"  Health
    Physics, V.  36, p. 413-421.

11. Prichard, H.M.,  Gesell,  T.F.,  Hess,  C.T.,  Weiffenbach, C.  and Nyberg,  P.,
    (1981)  "Integrating Radon  Detector  Data  from  Dwellings  in  Maine  and
    Texas," EPA International Meeting on  Radon and  Radon Progeny Measurement.
    Montgomery,  Alabama, August  1981.

                               LIST OF FIGURES

Figure 1.   Type-C Open Cup Track Etch Detector

Figure 2.   Type-B Bare Detector Card Track Etch Detector

Figure 3.   Type-SF Track Etch Personnel Radon Dosimeter

Figure 4.   Type-F (Filtered Cup) Detector Calibration

Figure 5.   Type-B (Bare Card) Detector Calibration

Figure 6.   Type-F Track Etch Detector Being Placed in its Canister

Figure 7.   Atmospheric Radon Around a Uranium Mine/Mill Complex

Figure 8.   Reductions in Radon Concentrations Away From a Uranium Mill

Figure 9.   Statistical Distribution of Indoor Radon Concentrations


                                 John C. Pacer
                      Bendix Field  Engineering Corporation
                            Grand Junction,  Colorado

    Bendix Field Engineering Corporation (BFEC),  the prime contractor for the
Department of Energy's (DOE) National Uranium Resource Evaluation (NURE)
program, has performed studies to characterize and evaluate radon measurement
techniques used for uranium exploration.  These studies have been performed in
the laboratory and in the field at known uranium deposits.  The BFEC radon
calibration unit has been the principal instrument in this work because it
provides a controlled radon atmosphere with which the calibration factor,
detection limit, sensitivity, and reproducibility can be determined for
various radon measurement devices.  The reliability of the measurement
techniques is also ascertained by field studies.   Radon measurement devices
studied have been field-portable zinc sulfide systems, alpha-track detectors,
thermoluminescence detectors, diffused-junction detectors, and activated
charcoal canisters.

    The calibration unit used for these studies includes a 700-liter chamber
through which radon in continuously pumped.  The radon concentration within
the chamber can be varied from 5 to 500 picoCuries per liter (pCi/1) and  is
continuously monitored.  Relative humidity in the chamber can be maintained at
a constant level with a microprocessor controller.  A radon measurement device
can either be placed directly in the chamber or a gas sample can be drawn from
the chamber into the device.

    Activated charcoal canisters have the lowest detection limit (0.3 pCi/1)
and the greatest sensitivity (0.1 pCi/1).  One type of activated charcoal
canister has a coefficient of variation of 5 percent for replicate samples
from the calibration unit.  The coefficient of variation for the thermo-
luminescence detectors is 42 percent.  The coefficient of variation measured
for all other devices ranges from 11 to  20 percent.   The thermoluminescence
detectors have the highest detection limit at greater than 30 pCi/1.  The
portable zinc sulfide detectors have the lowest sensitivity at 2 PCi/1.   Radon
measurements in the field show more variability than those obtained with  the
controlled atmosphere in the calibration unit.   Based on these studies, one
type of activated charcoal canister is superior to the other devices for  radon
measurements used in uranium exploration.


    As part of the NURE program, BFEC has conducted research and development
of new and improved uranium exploration techniques and systems.  One such
uranium exploration technique is the use of near-surface radon measurements to
locate subsurface uranium mineralization.  As part of the assessment program
to determine the usefulness of radon for uranium exploration,  a  number  of
radon measurement techniques were used.  In order to compare the performance
of the various techniques, it was necessary to calibrate (or normalize) them
to a common base and to determine their detection characteristics.   This was
done by studies performed in the laboratory and in the field at  known uranium

    It should be emphasized that the techniques discussed in this report were
used according to exploration procedures.  In this mode one is concerned with
relative values and the distinction of anomalous radon measurements from a
geochemical background value for an area.  Thus, accuracy and precision have
been balanced against time, ease of operation, field reliability, and cost in
formulating a measurement technique.  For many of these techniques  one  can
easily modify the measurement time, detector area or volume, or  handling
procedure to improve their performance for making environmental  measurements.
Hence, while the results presented in this report are a fair assessment of the
characteristics of a radon technique used in a uranium exploration  mode, one
needs to be careful in predicting the performance of a technique modified for
environmental monitoring.

    The BFEC radon calibration unit has been the principal instrument in the
assessment work.  The unit includes a 700-liter chamber through  which air
containing radon is continuously pumped.  A microprocessor controller is used
to maintain the radon concentration and relative humidity within the chamber
at constant levels.  The radon concentration within the chamber  can be  varied
from 5 to 500 picoCuries per liter (pCi/1) and the relative humidity from 5 to
95 percent.  The calibration unit thus provides a controlled radon  atmosphere
with which the calibration factor, detection limit, sensitivity, and
reproducibility can be determined for a radon measurement device.

    Field studies were also performed to assess the reliability  of  a radon
measurement technique for a natural situation.  Radon measurements  were taken
at a site of known subsurface uranium mineralization in the Red  Desert  area of
south-central Wyoming.  At a sampling location, radon measurements  were made
for duplicate samples taken 1-meter apart.  The simple layered geology  of the
site and sandy soil cover made this a reasonable approach for  determining the
influence of the natural environment on a radon measurement technique.

    The radon measurement devices studied were field-portable  zinc  sulfide
systems, alpha-track detectors, thermoluminescence detectors,  diffused-
junction detectors, and activated charcoal canisters.  The field-portable,
zinc sulfide systems were of the RD-200 type produced by EDA Electronics,
Ltd., of Ottawa, Canada, and of the RE-350 type produced by TSA, Inc.,  of
Boulder, Colorado.  The alpha-track detectors were Track Etch  *  cups
produced by Terradex Corporation^of Walnut Creek, California.  The  thermolumi-
nescence detectors were Alpha-2   cups manufactured by Westinghouse

Electric Corporation of  Pittsburgh,  Pennsylvania.   The  silicon  diffused-
junction detectors  were  alphaMETERS  produced by alphaNUCLEAR Company  of
Mississauga,  Ontario, Canada.   Two types  of activated charcoal  canisters  were
studied.  One was a modified U.S.  Army Mil gas canister obtained from Union
Carbide at Oak Ridge, Tennessee,  and the  other was  ROAC cups produced by
Inter-Science Research and Development Services (Pty) Ltd.,  of  Johannesburg,
South Africa.


    The BFEC radon calibration unit  is shown in Figure  1.   It has been the
principal instrument in  assessing the performance of radon measurement devices.
Radon is generated in four ion-exchange filters by  the  radioactive decay  of
radium deposited on the  filters.   The radon from one or more sources  is swept
along in airflow of 0.1  to 0.2 liters per minute and is continuously  delivered
into the radon chamber.   A larger airflow of 5 to  20 liters  per minute of
filtered air containing  variable  humidity is introduced into the chamber
simultaneously.  The air in the chamber is exhausted through a  vent outlet on
the top of the building  in which  the calibration unit is located.  Thus,  the
calibration  unit is a dynamic system and  radon-daughter buildup is minimized.

    The radon concentration and relative  humidity  in the chamber can  be
maintained at a constant level with  a microprocessor controller.  The radon
activity  in  the chamber  is monitored with a continuous  flow-through zinc
sulfide scintillation detector.  Air from the chamber is pumped at 1.5 liters
per minute through the detector and  returned to the chamber.  The output  of
the detector  is used by the microprocessor controller to adjust flowrates so
that the  radon concentration in the  chamber only varies by 2 percent.  The
radon concentration within the chamber can be varied from 5  to  500 picoCuries
per liter  (pCi/1) .  The relative  humidity in the chamber is  monitored with a
Hygrocon  II  relative humidity unit.   Using the output of this device, the
microprocessor controller has the dilution air for  the  chamber  going  through
either  a  high humidity or a heatless drying cycle  prior to entering the
chamber.  Under microprocessor control, the relative humidity in the  chamber
varies  by less than 1 percent.  The  temperature and pressure within the
chamber are  also monitored, but are  not controlled.

    The calibration chamber is a  large fiberglass  glove box  with an internal
volume of 700 liters.  It is equipped with an interchange box that allows the
insertion of radon devices to be  studied  into the  interior without
significantly disturbing the steady-state condition of  the chamber.  Figure 2
illustrates  the large number of devices that can be placed directly in the
chamber at a single time. The air in the  chamber is well-mixed  and the radon
is uniformly distributed even when a significant volume of the  chamber is
occupied by radon devices.  It is also possible to  draw a gas sample  from the
chamber as shown in Figure 3.

    The calibration of the unit is currently based  on calibration factors
obtained at the U.S Department of Energy's Environmental Measurements
Laboratory (EML) .   These factors  were obtained by calibrating a group of  zinc
sulfide scintillation flasks and  our counting system at EML. Figures 4 and 5
show this  calibration procedure.

    Figure 6 lists the characteristics of the radon  calibration  unit  used to
assess radon measurement techniques used for uranium exploration.   This list
summarizes the previous discussion of the unit.   In  using  the  unit  for the
assessment studies, the unit was always operated with the  relative  humidity
maintained at 30 percent.  For most studies the  radon concentration was
maintained at approximately 100 pCi/1.  One aspect of the  techniques  that
varies considerably is exposure or collection time.   While one could  treat
exposure time as a separate variable, the author has chosen  to define it as
fixed at the manufacturer's recommended exposure time and  as an  integral part
of a technique.

    The calibration factor for each device was determined  with the  calibration
unit.  This was done with each device operated in the manner used for uranium
exploration.  These calibration factors provide  a common basis for  determining
the detection limit, sensitivity, and reproducibility of a radon measurement
device.  The detection limit of a device is the  radon concentration equal to
the inherent background level of the device.  This was determined by  measuring
the signal present in devices that had not been  exposed to any radon.  The
sensitivity of a device is the minimum variation in  radon  concentrations that
is detectable by the device.  This was determined by the calibration  factor
for a device and a device's minimum unit of measure.  The  reproducibility was
determined for replicate radon measurements and  is indicated by  the
coefficient of variation which is the standard deviation of  the  replicate
measurements divided by the mean of the measurements and multiplied by 100


    In order to ascertain the reliability of a radon measurement technique in
the natural environment, radon measurements were made at sites of known
subsurface uranium mineralization.  The studies  for  only one site,  the Red
Desert area of south-central Wyoming, are presented  in this  paper.  The number
of techniques used at this site was more comprehensive than  the  other sites
and the geologic setting is fairly simple.  Descriptions of  the  site  can be
found in Bramlett et al., 1981 and Sherborne et  al., 1979.   The  site  is
situated on a gently rolling plain that generally dips to  the  south.  The soil
is essentially sandy and contains primarily quartz and feldspar  grains derived
from the underlying sandstone of the Battle Spring Formation.  The  site is
underlain by a sequence of interfingered sandstone,  shale, and coal layers.
The known deposit is a roll-front deposit that lies  within an  arkosic sand at
a depth of 140 meters.

    The reliability of a radon measurement technique was determined by making
duplicate radon measurements 1 meter apart at a  sampling location.  Thirty-
four locations on a grid with spacings of 300 meters were  designated  at the
site.  The radon measurements were made 0.7 meters below the surface. This is
a typical sampling depth for radon measurements  performed  for  uranium explora-
tion.  Soil gas at this depth was obtained with  a 6-millimeter-diameter probe
for the portable zinc sulfide systems.  The other radon measurement devices
were placed in 0.15-meter-diameter holes at a depth  of 0.7 meters.


    The radon measurement techniques  studied used field-portable  zinc  sulfide
systems, alpha-track detectors,  thermoluminescence detectors,   diffused-
junction detectors,  and  activated charcoal canisters.   Descriptions of these
techniques are found in Pacer and Czarnecki, 1980. The  field-portable  zinc
sulfide systems are  shown in  Figure 7.   They are the RD-200 produced by EDA
and the RE-350 produced by TSA.   The  RD-200 uses a bulb pump to acquire a gas
sample.  The alpha activity in the zinc sulfide chamber is counted for three
consecutive 1 minute intervals.   These  counts are used  to compute the  radon
count  rate according to a formula developed by Morse (1976) which accounts for
the presence of thoron and the ingrowth of radon daughters.  The RE-350 is
similar to the RD-200, but it has a microprocessor controller  which  controls a
diaphragm pump for consistent sample collection and the subsequent alpha acti-
vity counting.  The microprocessor can  then subtract background,  correct for
thoron decay and radon daughter ingrowth,  and use its calibration factor to
display the radon concentration of the  gas sample directly in pCi/1.

    The alpha-track detectors and thermoluminescence detectors are shown
in Figure 8.  The alpha-track detectors are Track Etch    cups  produced by
Terradex, that employed a polycarbonate track material.  The thermo-
luminescence detectors are Alpha-2 ™ cups  manufactured  by Westinghouse.
Both types of cups are typically exposed for 30 days in a uranium exploration
mode.   They are then returned to their  respective manufacturers for  processing.
The radon exposure for the Track Etch   cups is reported in tracks per
square millimeter for 30 days (T/mm2-30d).  The radon exposure for the
Alpha-2  cups is reported as net counts for 30 days.

    A  silicon diffused-junction detector is shown in Figure 9.  The  device is
an alphaMETER produced by alphaNUCLEAR.  This device detects and integrates
alpha  particles which impinge the surface  of the detector.  In a uranium
exploration mode, these devices are typically exposed for 1 to 3 days at a

    Two types of activated charcoal canisters are shown in Figure 10.   They
are the Mil canisters and the ROAC cups.  We modified the procedure  for using
the Mil canisters described by Countess, 1976, for uranium exploration.  The
canisters are exposed for 2 days, held  for at least 60  hours for thoron-
daughter decay, and then gross-gamma counted for 15 minutes.  This procedure
is also used for the ROAC cups,  though  the manufacturer recommends a 1-minute
gross-gamma measurement of the ROAC cartridge immediately on retrieval.

    Figure 11 summarizes the  radon measurement techniques.


    The results of the studies performed with the radon calibration  unit are
shown  in^Figure 12.   In terms of detection limits, all  the devices except the
Alpha-2   detectors  have reasonably low detection limits.  The charcoal
canisters have the lowest detection limit.   The thermoluminescence detector
material used in the Alpha-2  ™ cups is  sensitive to radiation other  than
alpha particles and  its background is continually increasing.   Earlier studies
by the author with Track Etch   cups  indicated that they have.a detection
limit of 100 pCi/1,  but improvements  by the manufacturer have  reduced  this to
1.7 pCi/1.


    All the measurement techniques are very sensitive  to measuring  radon.  The
activated charcoal canisters have the greatest sensitivity  and  the  portable
zinc sulfide systems have the lowest sensitivity.

    For replicate measurements in the controlled environment of the calibration
unit, most devices had a coefficient of variation from 11 to 20 percent.  The
exceptions were the ROACmcups with a coefficient of  variation of only  5
percent and the Alpha-2 ™ cups which were at 42 percent.  The results  of
the replicate measurements made at the field site are  shown in  Figure  13.
Unfortunately, the ROAC cups and the alphaMETERS were  not used  at the  field
site.  So, there is no certainty that the ROAC cups  would perform as well in
this environment.  The two portable zinc sulfide systems responded  differently
to the field environment.  While the TSA RE-350 emanometer  was  unaffected by
the field environment, the coefficient of variation  for the EDA RD-200 in-
creased from 11 percent in a laboratory environment  to 31 percent in a field
environment.  This increase in variability probably  reflects variability in
the operation of a device of this complexity when the  device is being  used in
the field.  The RE-350 emanometer did not demonstrate  this  variability since
it is microprocessor controlled.  The Track Etch * cups and Mil charcoal
canisters exhibit some decrease in reliability in going to  a field  environment.
The Alpha-2   cups show improved reliability for the field  environment.
This probably is due to the higher radon concentrations (mean radon
concentration of about 2000 pCi/1) encountered in the  soil  gas  of the  site
compared to the BFEC radon calibration unit.


The detection limit, sensitivity, and reproducibility  were  determined  for var-
ious radon measurement devices by studies performed  in the  laboratory  with the
BFEC radon calibration unit and in the field.  Based on these studies, one
type of activated charcoal canister, the ROAC cup, is  superior  to the  other
devices for radon measurement used in uranium exploration.

Bramlett,  L. ,  Pacer,  J.C.  and Moll,  S.,  1981,  Geochemical exploration for
    uranium in the  Red  Desert,  Wyoming:   U.S.  Department of Energy Open-File
    Report GJBX-125(81),  40  p.

Countess,  R.J., 1976, 222Rn  flux measurement with  a  charcoal canister:  Health
    Phys., v.  31, p.  456-457.

Morse, R.H., 1976,  Radon  counters  in uranium exploration:  Exploration for
    uranium ore deposits,  Vienna,  International  Atomic  Energy Agency,
    p. 229-239.

Pacer, J.C. and Czarnecki, R.F., 1980, Principles  and characteristics of
    surface radon and helium techniques  used in  uranium exploration:   U.S.
    Department of Energy  Open-File Report GJBX-177 (80) , 64 p.

Sherborne, J.E., Pavlak,  S.J.,  Peterson,  C.H., and Buckovic,  W.A. ,  1979,
    Uranium deposits  of the  Sweetwater Mine area,  Great Divide Basin,
    Wyoming:  Regional  AIME  meeting,  Casper, Wyoming, September,  1979.

1.  BFEC radon calibration unit:  the rack (left)  contains recorder  for

    various sensors, microprocessor controller,  flowmeters,  and radium sources

    and the glove box (right) is the radon chamber.

2.  View of inside of radon chamber containing many Track Etch cups  and  an

    operator holding an alphaMETER.

3.  RE-350 radon emanometer being tested with the BFEC radon calibration unit.

4.  Zinc sulfide cells being filled with the EML radon chamber.

5.  The zinc sulfide cells being counted at EML.

6.  List of radon calibration unit specifications.

7.  Portable zinc sulfide systems:  the RE-350  (left)  is produced by TSA, Inc.

    and the RD-200  (right) is produced by EDA Electronics.

8.  Alpha track detectors, Track Etch cups (bottom row), are produced by

    Terradex Corporation and thermoluminescence detectors, Alpha-2 cups    (top

    row), are produced by Westinghouse Electric Corporation.

9.   A silicon diffused  junction  device,  the  alphaMETER (left)  is produced by

    alphaNUCLEAR Company  and  a recorder  (right)  for  ten simultaneously used


10.  Activated charcoal  devices:  the ROAC cup  (left) and the ROAC cartridge

    (center)  are produced by  Inter-Science Research  and Development  and the

    Mil canister (right)  is obtained from Union  Carbide.

11.  List of radon measurement techniques used for uranium exploration.

12.  Characteristics of  radon  detectors determined with  the radon calibration


13.  Reliability of radon  measurement techniques  determined by  field  studies.

Figure 1.  BFEC radon calibration unit: the rack (left) contains  recorder for various sensors, microprocessor
controller, flowmeters, and radium sources and the glove box (right) is the radon chamber.

Figure 2. View of inside of radon chamber containing many Track Etch cups and an operator holding an
          Figure 3. RE-350 radon emanometer being tested with the BFEC radon calibration unit.

Figure 4. Zinc sulfide cells being filled with the EML radon chamber.
     Figure 5. The zinc sulfide cells being counted at EML.

   Radon Calibration Unit Specifications

  1.  Dynamic System
  2.  Chamber Volume 700 Liters
  3.  Microprocessor Controlled Humidity: 5-95%
  4.  Radon Concentration Range: 5-500 pCi/liter
  5.  Temporal Radon Variation:
       a. Hourly: ±2.6% at 60p Ci/liter mean
       b. Daily: ±2.2% at 60p Ci/liter mean
                Figure 6. List of radon calibration unit specifications.

Figure 7. Portable zinc sulfide systems: the RE-350 (left) is produced by TSA, Inc. and the RD-200 (right) is
produced by EDA Electronics.
Hgutt %. Alpfitt tncjc detectors, Track Etc* cups (bottom row), are produced by Terradex Corporation and
ttennohuninesctece detectors, Alpha-2 cups (top row), are produced by Westinghouse Electric Corporation.

 Figure 9. A silicon diffused junction device, the alphaMETER (left) is produced by alphaNUCLEAR Company
 and a recorder (right) for ten simultaneously used devices.
Figure 10. Activated charcoal devices: the ROAC cup (left) and the ROAC cartridge (center) are produced by
Inter-Science Research and Development and the Ml 1 canister (right) is obtained from Union Carbide.

              Radon Measurement Techniques
               Used for Uranium Exploration
Portable Zinc Sulfide System

Alpha Track Detector
Thermoluminescence Detector
Diffused Junction Detector
Activated Charcoal Canister
EDA RD-200
ISA RE-350
Track Etch Cup
Alpha - 2 Cup
Sample Collection
   or Exposure
   10 minutes
   15 minutes
    30 days
    30 days
    2 days
    2 days
    2 days
                   Figure 11. List of radon measurement techniques used for uranium exploration.

        Characteristics of Radon Detectors
  Determined With the Radon Calibration Unit
EDA RD-200
ISA RE-350
Track Etch
Alpha - 2
M11 Charcoal
  Detection Limit
30.4 & 2.4/month


of Variation

              Figure 12. Characteristics of radon detectors determined with the radon calibration unit.

Reliability of Radon Measurement Techniques
          Determined by Field Studies

                               Coefficient of
          Detector                  (%)
  EDA RD-200                      31
  ISA RE-350                       13
  Track Etch                        22
  Alpha - 2                         15
  M11 Charcoal Canister             23
            Figure 13, Reliability of radon measurement techniques determined by field studies.

                        A REVIEW OF SOME EXACT METHODS
                        FOR 222RN-DAUGHTER MEASUREMENT*

                                 Peter  G.  Groer

                         Institute for Energy Analysis
                       Oak Ridge Associated Universities
                           Oak Ridge,  Tennessee  37830

    Some exact methods for the  measurement of  the  short-lived  222Rn progeny
are discussed.  All methods use three  separate measurements of alpha activity
or a combination of alpha activity and total beta  activity.  The latter
methods permit a rapid determination of individual Rn-daughter concentrations
with several Instant Working Level Meters  (IWLM) and Environmental Working
Level Monitors.  The former methods, based only on the measurement of alpha
activity, were originally used  for the calibration of these devices but can,
of course, be used separately.   They differ from the usual methods in their
counting philosophy.  Alpha counts accumulated during the first counting
interval are recorded, and counts  from subsequent  intervals are added.  The
resulting larger total counts yield greater statistical precision than similar
methods that reset the counter  at  the  end of each  counting interval.  A brief
discussion of the probabilistic treatment of Rn-daughter build-up and decay on
a filter paper and some comments on counting statistics conclude the paper.
*This paper is based on  work performed under Contract No. DE-AC05-760R00033
between the Department of Energy, Office of Health and Environmental Research,
and Oak Ridge Associated Universities.

                            TRACK ETCH CALIBRATION:
                      DESIGN AND RESULTS OF A PILOT  STUDY

                               Susan M. Hinkins

    The Office of Radiation Programs will be conducting a field study of radon
and radon progeny measurement techniques.  One of the measurement devices used
will  be  the  Track Etch  Device'®  (TED).   Prior to  their deployment  in the
field we will  run  a calibration study of the TED's in  a  radon chamber.  This
paper briefly describes a pilot study done in anticipation of  this calibration

    The purpose of  the pilot study is to gather  information which will help us
anticipate and avoid  problems  which  might otherwise occur in  the  main study.
A successful pilot  study  will  identify  problems in data quality control which
can then be  corrected before  the main study begins.  This  study  identified 3
data problems:

         the data contain at least one outlier.

         the number of tracks  at the highest  exposure  is substantially lower
         than the number of tracks at the next  highest exposure.

         there appears to  be  a "family resemblance" in data  from  devices ex-
         posed and read at the  same time.


    The independent variable is the total exposure to radon, which is measured
as  the average radon concentration  (p  Ci/1)  times the exposure  time  (days).
Because the  radon chamber  was  being held at  an approximately constant radon
level, we could  vary  the  exposure only by varying  the  exposure time.  There-
fore total exposure is confounded with exposure  time; in the calibration study
we would avoid this by varying  both  radon  level and exposure  time.   The aver-
age radon concentration in  the  chamber was calculated from hourly measurements
from  flow-through  radon gas  monitors.   Scintillation  cell  measurements  were
taken several times each day,  also.

    The  response  measure on  the TED's  is the  total number  of  tracks  in  100
microscope  fields  (approximately  5.75  mm2).   For  a given  exposure,  X,   we
expect the  total  counts,  T,  to be distributed as  a Poisson variable with mean
proportional to exposure.   Therefore we expect  the relationship to be  T =  bX,
a straight  line through  the  origin.  The Poisson distribution has  the  property
that the variance is equal to the mean.

    The  chamber was  run for 38  days  and six exposures were planned:   3,  4, 7,
14,  21  and 35-day  exposures.   We  later included  38-day measurements  because
they were available.

    Unfortunately, we  failed  to  specify the  placement of  the  devices  in  the
chamber, so exposure is  confounded  with location in  the  chamber.  We will  not
make the same mistake in the calibration study.

    All  TED's were filter cups from batch 8.


    The  largest  value  in the 3-day  data is  an outlier.   Assuming all  3-day
values  come from  the  same  Poisson distribution,  the probability of  getting
such a large value is less than  .0001.

    We found no reason  to believe that this extreme value was due  to mislabel-
ing or other clerical errors.  The plot of the 3-day  data is given  in Figure 1.
                                    Figure 1
          Total Tracks

















    The data at exposures less than 35 days  can  be  fit  very well by a straight
line through the origin.  However,  the number of tracks at the 38-day exposure
is significantly smaller than the number of tracks at the 35-day exposure.

    Again, we  could find no  evidence of clerical  error.   After investigating
several possible  explanations we were left  with the following  two unresolved

    i)   While we do not believe  that the  location of the devices in the cham-
ber could cause such an effect, we  cannot  discard this  possibility because the
exposure groups were confounded with  location.   A study currently underway may
resolve this question.

    ii)  Since the  35  and 38-day TED's  were sent in at  different  times, they
were presumably  etched at different  times.   Possibly the  etching  process can
cause this much variation.

    What originally  interested  us in the possibility of  an etching or reading
effect was  that  the  35-day  values,  which were higher than  expected,  were read
at  the  same time  as the highest 7-day readings  and  the  highest  4-day read-
ings.  For  a given  exposure,  3,  4, or  7-day,  we  looked  at blocks  of  TED's,
where a  block  is  defined  as  the  set  of  TED's  which  were  exposed  together,
mailed together,  and presumably  etched and  read  together.  When  we examined
the data  by blocks, we  found  that  the sample variance within  blocks is often
much  smaller than  expected for  a  Poisson  model.   Also  one-way  analyses  of
variance show significant differences between blocks.

    As an example, the summary of the 7-day data  is given in Figure 2.  Recall
that under  the assumption of a  Poisson  distribution, the  variance should equal
the mean.

    A smaller variance than expected  indicates  an error in the assumptions,  or
an  unexplained  and uncontrolled  source  of  correlation  and  variability.   This
source could be a reader effect, smoothing the data toward the mean.

                                   Figure 2
7-day Series
All data
1 way Analysis
C. Total
Sample Mean
of Variance
d.f. S.S.
4 190.6
20 186.4
24 377.0
Sample Variance


    The pilot  study  identified  three  data  problems.

    1)    The  3-day data  contain  one  extreme  value.   While  this  is not  cause
         for  concern,  it is important  to  assess  the  likelihood  of  outliers
         before  deploying the devices.

    2)    Although  at the five lower  exposures  (21  days, or less) the number  of
         tracks  is proportional  to  exposure, the number of tracks at the  high-
         est  exposure  (38-days) is  substantially  lower  than the  number  of
         tracks  at the  next  highest exposure (35 days).   Could this be due  to
         an etching  or  other procedural effect?

    3)    Measurements from TED's exposed and  read  at the  same time vary  sub-
         stantially  less  than  would  be   expected  under   a  Poisson model.
         Measurements from TED's with similar exposures  but  read at different
         times  vary  more  than  would be  expected,  but  this  may  be caused  in
         part  by the small variance  within  blocks.   This "family resemblance"
         may be  due  to  the  etching process or  to a  reader error, a smoothing


                   W.M. Nazaroff, M.L. Boegel, and A.V. Nero

              Building Ventilation and Indoor Air Quality Program
                        Energy and Environment Division
                          Lawrence Berkeley Laboratory
                            University of California
                          Berkeley, California  94720

                                  August 1981

    We describe the procedures we use in residences  for rapid "grab-sample"
and time-dependent measurements of the air-exchange  rate and radon
concentration; the radon source magnitude is calculated from the results of
simultaneous measurements of these parameters.   Grab-sample measurements in
three survey groups comprising 101 U.S.  houses  showed  the radon source
magnitude to vary approximately log-normally with  a  geometric mean of 0.37 and
a range of 0.01 to 6.0 pCi 1~^ h~l.  Successive measurements in six houses
in the northeastern United States showed considerable  variability in source
magnitude within a given house; in two of these houses the source magnitude
showed a strong correlation with the air-exchange  rate, suggesting that soil
gas influx can be an important transport process for indoor radon.

    Keywords:  radon, houses, air-exchange rate, pollutant sources.


    Characterizing radon sources and the processes by  which radon is
transported into buildings is an important component in developing a
comprehensive understanding of radon as a contaminant  of indoor air.  For
sources such as building materials, domestic water,  and natural gas,
characterization is relatively straightforward; however, these sources do not
appear to contribute to indoor radon concentrations  in amounts sufficient to
account for the levels observed in a large number  of houses surveyed in the
United States and Canada.  Rather, evidence is  mounting that local rock and
soil (referred to simply as "soil" throughout this paper) are the dominant
sources of radon in these houses.  The characterization of soil as a source of
indoor radon is difficult, because, for example, the permeability and radon
emanating power of soil vary with changes in moisture  content, radon can
migrate over long distances in soil; and radon  can enter a house by either
bulk flow or diffusion.

    A potentially important approach in investigating the origin of radon
indoors is to calculate the effective radon source magnitude by simultaneously
measuring the radon concentration and the air-exchange rate in a house.   It is
the initial results obtained from such measurements that strongly suggest soil
to be the dominant source of radon in a significant proportion of U.S.
houses.  Presuming this to be the case, we can use measurements of radon
source magnitude to ascertain the potential for high radon levels in a
geographic area.  Furthermore, by measuring radon source magnitude
continuously in conjunction with meteorological parameters such as wind  speed,
indoor-outdoor temperature difference, and barometric pressure, we can  gain a
better understanding of processes which transport radon from its source
material to indoor air.

    To date, we have measured radon source magnitudes in about 100 U.S.
houses.  In most of these homes we used a single simultaneous measurement of
radon concentration and air-exchange rate to determine the radon source
magnitude at one point in time; however, in several we made successive
measurements of the radon source magnitude over periods of time ranging  from 3
to 30 days.  Our most significant findings are:  (1) that the radon source
magnitude varies over a wide range from one house to another; (2) that  the
radon source magnitude in a house can change dramatically even over a period
as short as several hours; and (3) that frequently the radon source magnitude
in a single house shows a positive correlation with the infiltration rate —
consistent with the hypothesis that soil is important as a source of indoor
radon and that bulk flow is important as a transport mechanism, as we shall

    In this paper we describe the instrumentation and techniques used in our
studies to calculate the radon source magnitudes in residences.  We present
the results of successive measurements of source magnitudes taken in six
houses in the eastern pact of the United States, and use these along with the
results of a grab-sample survey of 101 houses variously located in the United
States as a basis for tiiscussing some aspects of the origin and transport of
radon in U.S. houses.



    The mass-balance equation for indoor radon can be written as

           = SR(t) + Xv(t)R0(t) - ARnRi(t) - Av(t)Bi(t)                  (1)
where R^(t) is the indoor radon concentration,  Ro(t)  is the outdoor radon
concentration, Sg(t)  is the indoor radon source magnitude per unit volume,
^v^t' is the air-exchange rate, and \Rn is  the  time constant for
radioactive decay of  radon (0.00756 h~l).   Since Xv (almost always greater

than 0.1 h 1) is much larger than XRU, we ignore the third term on the
right hand side of Eq.(l).  We define the "effective" indoor radon source
magnitude, Qg, as the sum of the first two terms on the right hand side of
Eq.(l) so that for steady-state conditions we have

    Qs = Ss + Xs Rs = Xs Rs ,
     R    R    v  o    v  i
where the superscript s denotes the steady-state value.  In this way a
simultaneous measurement of the air-exchange rate and the indoor radon
concentration can be used to compute the effective source magnitude, assuming
that steady-state conditions prevail.  The difference between the effective
source magnitude, Qg, and the value for indoor sources, Sg, is small if
the indoor radon concentration is much greater than the outdoor concentration,
as is often the case.

    By continuously measuring the radon concentration and the air-exchange
rate, we can calculate the source magnitude even under changing conditions.
Equation (1)  is solved to obtain

                     Ri(t') - Ri(O)         + o,t',

         0,t'            t1
where <>Q (-•  indicates a time-weighted average over the interval 0 to t'
of the contents of the brackets.  Because we measure radon averaged over
finite time intervals, we do not know R^(t') or R^(0).  We approximate
these two quantities as

    Ri(t') =  1/2 «Ri>0,t' + t',2t') •


    Ri(0) = 1/2 (-t',0 + 0,t1^ •

We further approximate
    <8i(t) Xv(t)>0,t' by 
    Grab-Sampling:  Technique and Instrumentation


    Grab-sample measurements of radon are made using scintillation cells,
either cells fabricated in our laboratory after the design of Lucas (Lu57) or
commercially-available cells (EDA,  Model RDX 388).  The procedure used for
taking the sample varies: (1) The sample is taken directly with the
scintillation cell which is then either counted in the field or sent back to
our laboratory for analysis.  (2) The sample is taken in a radon-impermeable
sampling bag (Environmental Measurements, Inc., air-sampling bag, made of
Tedlar) which is sent back to our laboratory and analyzed by transferring a
portion of the sample to a scintillation cell.  (3) A metal vessel of known
volume (typically 1 liter) is used to collect the sample, and then returned to
our laboratory where the radon is extracted by passing the sample through a
glass-wool trap cooled to -196° C in a liquid N£ bath; the extracted radon
is then transferred to a scintillation cell.  This procedure is described by
Lucas (Lu77), and our implementation of it is reported by Ingersoll (In80).
The procedure for concentrating radon is important for measurement precision
and improvement of our sample throughput rate when measuring radon
concentrations below about IpCi/l.

    The scintillation cells are calibrated in batches.  A calibration factor
is determined for the batch by filling several of the cells with a known
amount of radon, derived from a standard-reference-method solution of
(National Bureau of Standards).   The response of individual cells within the
batch is checked by filling several cells with a constant,  unknown
concentration of radon,  and comparing their count rates.   The responses of
individual cells fabricated at Lawrence Berkeley laboratory (LBL) have been
found to be within a range of 5%.  We occasionally check our calibration
factor against those independently  determined by other laboratories.   (In a
recent check we were among six of the nine laboratories participating whose
calibration factors agreed within a range of 10% (Ge81).)

    Air-Exchange Rate

    We measure air-exchange rate by tracer gas decay,  most  commonly using
sulfur hexafluoride (SF6), but occasionally ethane,  as the  tracer.   The
tracer gas is injected into the house and mixed to a uniform concentration by
means of portable fans or the furnace fan.   The concentration is monitored
over time and the resulting data are fit to an exponential  decay curve of the

    C(t) = C(0) exp (-Xvt) ,                                             (6)

where the time constant,  Xv,  is the air-exchange rate.
        concentrations  are  measured  with  a  commercially-available,  non-
dispersive infrared (NDIR)  analyzer  (Foxboro-Wilks ,  Model  Miran 101).  The
analyzer is calibrated  by measuring  its response  to  gases  derived  from
compressed-air tanks containing known  concentrations of SFg.

    Continuous Measurement:  Techniques and Instruments


    The radon concentration is measured over time with a Continuous Radon
Monitor (CRM) developed by Thomas (Th79) ,  and designed and fabricated at LBL.
Figure 1 presents a schematic diagram of our most recent version of the CRM,
whose unique feature is that the output is provided as an analog rather than a
digital signal, simplifying the interface of the CRM to recording devices
(strip-chart recorder or a data logger).  The output voltage has a range of 0
to 1023 mV, corresponding to 0 to 1023 counts in the counter.   After reaching
1023 the counter automatically resets.  Converting the count total to an
analog voltage introduces an uncertainty on the order of one count, which is
insignificant compared to the statistical uncertainty in measuring radioactive
decay.  We use three-hour integration intervals for analyzing the CKM data.

    The CRM is calibrated following the procedure of Busigin et al.  (Bu79);
the radon concentration is calculated as follows:

    t',t'+3h = 2.19t')t.+3h - 0.13 t'-3hjt. ,                 (7)

where n is the net count rate in min~l.  Aside from calibration uncertain-
ties, the relative standard deviation in the measurement of a steady-state
radon concentration is typically 6% at 5 pCi/1, 21X at 1 pCi/1, and 48% at  0.4
pCi/1 (Na81).

    Air-Exchange Rate

    We have developed an automated system, the Aardvark, for continuously
measuring the air-exchange rate and the radon concentration in an occupied
residence.  It is described in detail in another report (NaSlJ.  Briefly, the
Aardvark uses the tracer gas decay technique with SF^ to measure the
air-exchange rate.  The operation of the system is controlled by a
microcomputer system, which also performs preliminary data analysis.   The data
are recorded on magnetic tape and printed on a terminal.

    The mechanical system of the Aardvark is presented in Figure 2.  As
illustrated, a blower is used to draw and return air at a total rate of 80  1/m
through as many as four sampling lines and up to four delivery lines.   A
bypass loop provides 20 1/m to the SFf, analyzer and 2 1/m to a CbiM which is
interfaced to the system.  The analyzer is automatically calibrated at
user-specified intervals by measuring its response to three concentrations  of
    drawn from compressed-air tanks.
    To ensure good distribution and thorough mixing of the tracer gas in a
house with a forced-air furnace, we use only one delivery line,  installing its
end in the return-air duct of the furnace system.  A relay is used to bypass
the thermostat and turn on the furnace fan during injection of the tracer
gas.  In a house that does not have air-distribution ductwork, mixing is
accomplished by using three or four delivery lines, attaching the end of each
to a portable fan which is turned on during injection.   In either case,  three
or four sampling lines are used to reduce the dependence of the air-exchange
rate measurement on the location of the sampling point(s).

    Air-exchange rates are measured by the Aardvark over  90-minute  intervals.
The average of two consecutive measurements is used in equation  (5)  to
calculate the radon source magnitude.


    Survey of Radon Source Magnitude in U.S.  Housing

    Radon source magnitudes were determined in a survey comprising  three
housing groups:  "energy-efficient" houses in the U.S.  (one  in Canada)  (16
houses, Ho80); conventional houses in  the San Francisco Bay  Area (29 houses,
Be79); and conventional houses in rural Maryland (56 houses, Mo81).   Although
this  survey was not intended to be a random sampling,  we  believe the results
provide an indication of what we might expect generally in the U.S.  housing

    On the night prior to measurements occupants were  asked  to close windows
and doors to establish a correspondence between the radon concentration
measured and the ratio of source magnitude to air-exchange rate.  Upon first
entering the house, we collected one to three grab-samples of  air for radon
analysis; the samples were taken from  the main living  space, well away from
doors and windows.  A tracer-gas decay measurement of  the air-exchange rate
was then made over a one to two-hour interval.   A summary of these measuremens
is presented in Table 1.

    In all three housing groups the air-exchange rates were  found to be  fairly
low — an arithmetic mean of 0.41 h~l.  In as much as  all doors  and  windows
were  closed during the measurements, these values represent  infiltration only
 (i.e., uncontrolled leakage through the building envelope).  Furthermore,
because most of the measurements were  made in the spring  and summer  when the
weather is mild, we can assume that the average annual infiltration  rate for
these houses is higher than the values we report.

    A histogram (see Figure 3)  of radon source magnitudes measured  in all 101
houses shows a broad distribution of values,  ranging from 0.01 to 6  pCi  1~1
h"1.  Individual measurements are distributed in an approximately log-normal
fashion, with a geometric mean of 0.37 pCi 1~1 h~l,  and a geometric
standard deviation of 4.0.  The geometric means of the  radon concentration and
radon source magnitude measured in the San Francisco Bay  Area  houses were
considerably lower than the corresponding values for the  other two housing

    Time-Dependent Measurements of Radon Source Magnitude in Selected U.S.

    In the six houses located in the northeastern United  States,  we  measured
radon source magnitudes over periods of 3 to  30 days.   Five of these houses,
located in New York and Maryland,  were built  within the last ten years;  all
were tract houses,  built with some attention  to energy  efficiency.   The  sixth
house, located in  New Jersey, is over  100 years old but has been  retrofitted
to reduce energy use.   In three of the houses radon measurements  were made

continuously and air-exchange rate measurements were made intermittently,
typically once per day.  In the other three houses, both air-exchange rate and
radon concentration were monitored continuously over three-to five-day

    Four of the houses, those in New York and New Jersey, were occupied during
the measurements.  Because of the cold weather, the windows and doors in all
of the houses were closed almost all of the time.  A summary of the results is
presented in Table 2, which gives both the arithmetic mean and the range of
the radon concentration, the air-exchange rate, and the radon source magnitude
for each of the six houses.  With the exception of Roch 49, which exhibits a
much lower radon concentration and source magnitude, the mean values of each
parameter obtained for these houses are narrowly distributed.   The radon
source magnitudes, especially in Cb, AM-1 and SG-1, were found to vary
significantly over the monitoring period, and the air-exchange rates were
found to vary over a comparable or narrower range.

    We observed a positive correlation between the measured air-exchange rate
and radon source magnitude for five of the houses, as indicated in Figures 4
through 8.  (In the case of Roch 49, the concentrations of radon were too low
to determine whether or not such a correlation existed.)  Figures 4 and 5 are
scatter plots of the measurements made in CB and Roch 6, respectively, and the
line plotted in each figure represents the least-squares fit to the data.   For
CB this line has a slope of 2.57 pCi 1~1, and a y-intercept of 0.23 pCi
I"1 h~l; the respective values for Roch 6 are 1.09 pCi 1~1 and 0.21 pCi
1~1 h~l.  Figures 6, 7, and 8 are plots of the air-exchange rate and radon
source magnitude vs. time measured in Roch 60, AM-1 and SG-1,  respectively.
The average wind speed, derived from measurements made on-site at five-minute
intervals, is also plotted for AM-1 (Ke81).  In each of the three houses,  the
highest peaks in the radon source magnitude correspond in time with the
highest peaks in the air-exchange rate.  This correspondence is most evident
in Figure 7, the plot for Am-1, showing that on 4/18 when the radon source
magnitude increased from 0.7 to 2.6 and then decreased to 0.35 pCi 1~1
h~l, the air-exchange rate increased from 0.26 to 1.22 and then decreased to
0.72 hr1-


    Source Magnitude Distribution

    Knowledge of the distribution of radon source magnitudes in U.S. housing
would contribute greatly to efforts towards characterizing public exposure to
indoor radon, as well as towards designing programs to reduce that exposure.
Because our survey did not represent a random sampling of U.S. houses, the
distribution of values we have reported may well differ from the actual
distribution of radon source magnitudes.  For example, half of the 16 energy-
efficient houses surveyed were solar homes, generally relying on rock-bed heat
storage, a potentially significant source of radon.  The other two housing
groups,  although they reflect more typical design and construction practices,
were selected from only two geographic areas of the country.  Our data may
also show a seasonal bias since most measurements were made in the spring and
summer months;  if,  as we have observed, large variations in radon source

 magnitude occur over time,  then it  is  very  possible  that  systematic  differences
 also occur from one season  to  another-  Another  item affecting  the
 interpretation of these  measurements is the contribution  of  radon in the
 outside air to the effective radon  source magnitude.  When trying to ascertain
 the effect of  changes  in the air-exchange rate on  indoor  radon  concentrations,
 particularly for houses  at  the  lower end of the  distribution, this factor
 assumes importance.

     We mention two other factors which, although potentially important, are
 unlikely to have much  influence on  the measured  distribution:   (1) the
 uncertainties  associated wih the imprecision in measuring radon concentrations
 and the air-exchange rates  as well  as in approximating steady-state
 conditions,  and (2)  the  apparent variability in  radon source magnitude over
 short time periods within a given house.  These  two  factors could cause a
 dispersion of  the values  reported here even if all of the houses had the same
 mean value;  however, even this dispersion would not  account for the  wide range
 of source magnitudes observed in our survey.  In other words, in spite of
 these considerations we  conclude that the radon  source magnitudes in U.S.
 houses are indeed widely distributed, and a substantial fraction are on the
 order of several pCi 1~1  h~l.

     In most  of the houses with source magnitudes on  the order of 1 pCI 1~1
 h~l or more, it is unlikely that building materials or domestic water are
 the dominant sources.  Ingersoll measured radon emanation rates from 100
 samples of concrete  from across the United  States and found values ranging
 from 0.2 to  2.0 pCi  kg"1  h~l (In81).  If we take the case of a one-story
 house with 2.4 m ceilings,  then a 0.2 m-thick concrete slab with a typical
 density of 2000 kg m~3 would contribute only 0.02 to 0.2 pCi I"1 h"1  to
 the source magnitude.  We took a few measurements of radon in tap water in
 Maryland and found concentrations on the order of 1000 pCi/1; using  Hess'
 observation  that 10,000  pCi/1 of radon in water typically results in  0.65
 pCi/1 of radon in air in  a house with an air-exchange rate of 1 h~l,   the
 resulting source magnitude would be 0.07 pCi 1~1 h~l.  On the other hand,
 Wilkening reports  that data from roughly 1000 measurements of radon flux from
 soil  range up  to 1.4 pCi m~2 s~2 (mean value of 0.4 pCi m~2 s~2),  a
 value which  would  contribute 2.1 pCi I"1 h~l (mean value of 0.7 pCi  I"1
 h~l)  to  the  source magnitude in the house  postulated above,  assuming the
 flux  from the  soil into the house is the same as  the flux from uncovered soil
 (Wi72).   The fact  that even the maximum flux from uncovered soil is not large
 enough  to  account  for the highest indoor source magnitudes observed suggests
 the possibility  that in some cases  radon may be more efficiently transported
 from  soil  into a house than into the atmosphere.

    In addition to the variations observed  from one house to  another  our
 survey revealed a significant difference in  radon source magnitude  between  the
 San Francisco Bay Area and the  communities  studied  in rural Maryland.  This
difference suggests that  radon  source magnitude may depend on geological or
structural factors.  If so,  it  may  be useful to conduct  limited  surveys of
radon source magnitudes in a region to  determine  whether radon  levels are high
enough to justify further measurements  in  that  area.

    That radon source magnitudes show a broad log-normal distribution has
important implications for any effort to control public exposure to radon in
houses.  For such a distribution, the best cost-benefit ratio can be obtained
by addressing control efforts toward those houses with the highest source
magnitudes.  Since the source magnitude in closed houses appears to be more
broadly distributed than the air-exchange rate, specifying a minimum
air-exchange rate for all houses does not appear to be a cost-effective
control strategy.

    Correlation Between Air-Exchange .Rates and Radon Source Magnitude

    The correlation between air-exchange rate and radon source magnitude is
evident in Figures 4-8.  The correlation is particularly clear for the two
houses with the highest (mean) source magnitudes, CB and AM-1.  It is our
hypothesis that in these two houses, and perhaps in Roch 6, Roch 60 and SG-1,
as well, surrounding soil is the dominant source of radon; the radon enters
the house in high concentration with soil gas that is driven by the same
forces that cause infiltration — wind speed and indoor-outdoor temperature
difference.  A simple model consistent with this correlation is that on
average a constant fraction of the air that infiltrates the house comes from
the soil and carries with it a high concentration of radon; as the wind speed
and temperature difference increase, the flow rate of soil gas into the house
also increases, therefore, the radon source magnitude increases.   For example,
if the radon concentration in the soil gas is 500 pCi/1, a commonly reported
value, then the soil gas portion of infiltration air need be only 0.5X of the
total (corresponding to 16 1/m at an air-exchange rate of 0.4 ti~l) to
account for the slope of the line in Figure 4.  The y-intercept of 0.2 pCi
1~1 h~l could represent the contribution to the source magnitude of radon
diffusing from and through the building materials.  Much of the scatter in
these data could reflect changes in the concentration of radon in the soil gas
at the soil/house interface, or changes in the fraction of infiltrating air
coming from the soil.

    The data for AM-1 also show a strong correspondence between radon source
magnitude and air-exchange rate, particularly on 4/18 when the air-exchange
rate steadily rose over a period of 20 hours, increasing greatly during the
middle of this period concomitantly with an increase in wind speed.  During
the first 15 hours of this period, the radon source magnitude also increased
strongly; however at noon, when the last significant increase in air-exchange
rate (to 1.2 h~l) occurred, the radon source magnitude declined by nearly
25%.  In fact, throughout this period it appears that the increases in radon
source magnitude become smaller relative to the increases in the air-exchange
rate.  It is possible that the concentration of radon in the soil gas drops as
the soil gas is drawn into the house at a higher rate.  That the source
magnitude dropped to a low value of 0.35 pCi 1~1 h~l that afternoon, even
though the air-exchange rate was still fairly high at 0.7 h~l, is consistent
with this hypothesis.

    The correlations we observed between air-exchange rate and radon source
magnitude would not be expected to apply if the changes in air-exchange rate
were a result of opening windows and doors or changing the rate of mechanical
ventilation.   As pointed out earlier, measurements in these houses were made

during winter and spring when the outside temperatures  are  relatively  low,  and
therefore, we assume that the doors and windows were closed as much  as
possible.  Given these conditions the changes in air-exchange rates  can be
assumed to reflect changes in weather conditions which  appear also to  effect
the radon source magnitude in some of these houses.

    We have considered various explanations for the  observed correlation
between radon source magnitude and air-exchange rate, but in the case  of two
houses, CB and AM-1, the data seem to point quite clearly to the
aforementioned soil-gas theory.  For example, since  we  include  in our
calculation of source magnitude the contribution from outdoor radon, we expect
a positive correlation between the source magnitude  and the air-exchange rate;
however the observed slopes are much steeper than the slope of  the regression
line resulting from this effect alone, which would equal the outdoor radon
concentration, typically 0.1 pCi/1.  Errors in measuring the air-exchange rate
could also lead to a false correlation.   For example, if the radon
concentration and air-exchange rate were constant, then a scatter plot of a
series of measurements of radon source magnitude and air-exchange rate would
show a positive correlation with a slope equal to the radon concentration.
The range of air-exchange rates in such a plot would reflect the precision  of
the measurement.  How precise our air-exchange rate  measurements are is
difficult to estimate and probably varies from one house to another, depending
on the mixing rates within the house and on the number  and  location  of
sampling  lines.  The first eight measurements in Roch 60 indicate that it is
possible  to achieve a standard deviation of less than 0.05  h~l  in the
measurement of air-exchange rate.  Such a level of precision is  sufficient  to
preclude  the possibility that the correlation between air-exchange rate and
radon source magnitude for CB and AM-1 is due to measurement uncertainty.  In
the case  of Roch 6, Roch 60 and SG-1 however, the apparent  correlations could
reflect measurement uncertainty in as much as the measurements of air-exchange
rate in each of these houses are distributed over fairly narrow  ranges.


    We have measured radon source magnitudes in over 100 houses  by
simultaneously measuring the radon concentration and the air-exchange  rate; in
six of these houses source magnitude measurements were  made over time, in some
cases continuously.  The data from these measurements corroborate the
hypothesis that soil is an important source of indoor radon in U.S.  houses,
and, furthermore, support the theory that soil gas influx,  rather than
molecular diffusion, is the dominant transport process  by which  radon  from
soil enters houses, at least in those with large source magnitudes.

    Such studies could be fruitfully extended in at  least two directions:  A
well-designed survey project could be undertaken to  determine the actual
distribution of radon source magnitudes  in the U.S.  housing stock to provide a
basis for regulatory action aimed at controlling public exposures to radon.
Continuous measurement of the source magnitude over  a period of  three  months
to a year in individual houses where source magnitudes  are  high  would  also  be
useful,  especially if measurements of the radon concentration in soil  gas,  the
radon flux through an opening between the house and  the soil, and weather
parameters were made simultaneously.


    We thank the following people for their assistance in collecting and
analyzing the data on radon concentrations and air-exchange rates  for the  six
houses monitored over time:  M. Lints of the Rochester Institute of
Technology, and R. Young, S. Brown, J. Dillworth, J.  Pepper,  F.  Offermann,
S. Doyle, and B. Moed, all of Lawrence Berkeley Laboratory.

    This work was supported by the Assistant Secretary for Conservation and
Solar Energy, Office of Building and Community Systems,  Buildings  Division,
and the Assistant Secretary for Environment, Office of Health and
Environmental Research, Human Health and Assessments Division of the U.S.
Department of Energy under Contract No. W-7405-ENG-48.

Be79   Berk J.V.,  Boegel M.L.,  Ingersoll J.B., Nazaroff W.W.,  Stitt B.D.,  and
       Zapalac G.H.,  1979,  "Radon Measurements and Emanation Studies," in
       Energy Efficient Buildings Program:   Chapter from the Energy and
       Environment Division Annual Report 1979, Lawrence Berkeley Laboratory
       report, LBL-10704, Berkeley, CA.

Ge81   George A.C. and Fisenne  I.M., 1981,  private communication,
       Environmental  Measurements Laboratory,  New York, May 12.

He81   Hess C.T.,  Weiffenbach C.V., Norton S.A.,  Brutsaert W.F. ,  and Hess
       A,L. , 1981, "Radon-222 in Potable Water Supplies in Maine:  the
       Geology, Hydrology,  Physics, and  Health Effects," presented at the
       Second Special Symposium on Natural  Radiation Environment, Bhabha
       Atomic Research Centre,  Bombay,  India,  January 19-23,  1981.

Ho80   Hollowell C.D., Berk J.V., Boegel M.L., Ingersoll J.G., Krinkel D.L.,
       and Nazaroff W.W., 1980, Radon in Energy-Efficient Residences, Lawrence
       Berkeley Laboratory report, LBL-9560, Berkeley,  CA.

In80   Ingersoll J.G., 1980, Operating Instructions for LBL Radon- Measurement
       Facilities, Lawrence Berkeley Laboratory report, LBL-11097, Berkeley CA.

In81   Ingersoll J.G., 1981, "A Survey of Radionuclide  Contents and Radon
       Emanation Rates in Building Materials Used in the United States,"
       Lawrence Berkeley Laboratory report, LBL-11771,  Berkeley,  CA.
       Submitted to Health Phys.

Ke81   Kelly C.J., 1981, private communication, Automation Industries, Inc.,
       Vitro Laboratories Division, Silver  Spring, MD,  May 12.

Lu57   Lucas H.F., 1957, "Improved Low-Level Alpha Scintillation  Counter for
       Radon," Rev.Sci Instrum. 28, 680-683.

Lu77   Lucas H.F., 1977, "Alpha Scintillation Kadon Counting," in Workshop on
       Methods for Measuring Radiation in and around Uranium Mills Albuquerque
       (E.D. Harwood, ed.),  Vol. 3, No.  9,  Atomic Industrial Forum,
       Washington, B.C.

Mo81   Moschandreas D.J.,  Rector H.E., and  Tierney P.O., 1981, A  Survey Study
       of Residential Radon Levels, Geomet  Technologies report, ES-877,
       Rockville,  MD.

Na81   Nazaroff W.W., Offermann F.J.,  and Robb A.W., 1981, "Automated System
       for Measuring  Air^xchange Rate and  Radon Concentration in Houses,"
       Lawrence Berkeley Laboratory report, LBL-12945,  Berkeley,  CA.
       Submitted to Health  Phys.

Wi72   Wilkening M.H.,  Clements W.E . ,  and Stanley D.,  1972,  "Radon-222 Flux
       Measurements in Widely Separated Areas," Natural  Radiation Environment
       II (J.A.S. Adams et al, eds.),  U.S.  Energy Research and Development
       Administration report, CONF 26-720805,  Washington,  D.C.

                                Figure Captions
1.  Schematic diagram of the Continuous  Radon Monitor  (CRM).

2.  Schematic diagram of the mechanical  system of  the  Aardvark,  an  automated
    system for continuously  measuring  air-exchange rate and radon
    concentration in occupied houses.

3.  Frequency distribution of radon  source magnitude.  For each  house,  the
    radon entry rate per unit house  volume is calculated as the  product of  the
    radon concentration and  the  air-exchange  rate,  measured after the house
    had been closed for several  hours.

4.  Scatter plot of radon source magnitude versus  air-exchange rate  for house,
    CB.  The data were collected between November  16 and December 17, 1980.
    The line represents a least-squares  fit to the data.

5.  Scatter plot of radon source magnitude versus  air-exchange rate  for house,
    Roch 6.   The data were collected between  January 8 and January 22,  1981.
    The line represents a least-squares  fit to the  data.

6.  Plot of air-exchange rate and radon  source magnitude versus  time for
    house,  Roch 60.

7.  Plot of wind speed,  air-exchange rate, and radon source magnitude versus
    time for house,  AM-1.

8.  Plot of air-exchange rate and radon  source magnitude versus  time for
    house,  SG-1.

         2 L/m
(EDA, Model RDX 388)

        Model R269)
     HV power supply
    Model PMT-20A-P)
                                                   Continuous Radon Monitor
  DAC      Buffer
(10 bits)   amplifier
          PMT base
        Model E990)
                                                                                                             Analog panel meter
                                                                                                             Model ICL7106EV)
     Figure  1

                        AARDVARK MECHANICAL SYSTEM

         10ppm 25 ppm 50 ppm

                                                                  PURE SFg SUPPLY
Figure 2
                                                               XBL 817-1036

§   20



1   2.0



                                   •         I
                                                                               -i	r
                                                                QR = 0.23+2.57 X
                                                                 r\               »
                                                           i	i
0.4        0.6         0.8

  Air-exchange rate  (h  )
                                                                                    XBL 818-1148
                 Figure  4


1   2.0


                 Roch    6
                                             •QR =0.21+1.09 X
          Figure 5
                                Air-exchange rate (h )
                                                                           XBL 818-1147

  _cu  1.0






  ® 0.6





-^ 2.5





 §   ,.o


-S  0.5
                                                            Air-exchange  rate

                                                            Wind speed
                                                    J  L
                                                                                 10    Q.





        Figure  7



 o>   1.0







^  2.5

 &  1.5

                                                                         XBL 818-1144
       Figure 8

    Table  1.   Data  from 101-House  Survey of  Radon Concentrat.on and Air-Exchange Rate
U.S. Energy-0
San Francisco
Bay Area
Maryland '
Monitoring Period
May- August


No. of

Radon (pCi/1)
2.6 2.2

0.4 2.2
1.8 4.0
1.2 4.0
Rate (IT1)
0.23 2.2

0.28 2.5
0.35 2.2
0.31 2.3
Source Magnitude
(pCi I'1 h'1)
0.61 2.5

0.10 2.5
0.62 3.4
0.37 4.0
 GSD = exp
 I  [In X  - ln(GM)]'
i=l      1	
 Source:  Hollowell, C.D., Berk, J.V., Boegel, M.L.,  Ingersoll, J.G., Krinkel, D.L.,
          and Nazaroff, W.W., 1980, Radon in Energy-Efficient Residences. Lawrence
          Berkeley Laboratory Report LBL-9560, Berkeley, CA.

dSource:  Berk, J.V., Boegel, M.L., Ingersoll, J.G.,  Nazaroff, W.W.,  Stitt, B.D. ,
          and Zapalac, G.H., 1979, "Radon Measurements and Emanation  Studies," in
          Energy Efficient Buildings Program:  chapter from the Energy and Environ-
          ment Division Annual Report 1979, Lawrence  Berkeley Laboratory Report
          LBL-10704, Berkeley, CA.

6Source:  Moschandreas, D.J., Rector, H.E., and Tierney, P.O., 1981,  A Survey Study
          of Residential Radon Levels, Geomet Technologies Report ES-877, Rockville,

 In six of these 56 houses, the radon concentration was found to be less than the
 detection limit of the measurement procedure  (0.4 pCi/1 with 50% relative standard
 deviation).  For these houses the radon concentration was assumed to be 0.1 pCi/1
 for purposes of calculating the geometric mean radon concentration and source

 Table 2.  Summary of Radon Source Magnitude Measurements
           for Six Houses Monitored Over Time
House Monitoring No. of Radon
ID Location Period Meas. (pCi/1)


ROCH 49 N.Y.

ROCH 60°' N.Y.

AM-1 MD.

SG-1 MD.















Air- Exchange
Rate (h~l)




Radon Source
Magnitude3 '
(pCi I'1 h'1)




  Arithmetic mean values for three-hour intervals; range of values given
  in parentheses.

  Occupied house, radon measured by CRM, air-exchange rates measured by
  tracer gas decay one or two times per day.
  Occupied house, radon measured by CRM, air-exchange rates measured
  continuously by tracer gas decay.

  Unoccupied house, radon measured by CRM, air-exchange rate measured
  continuously by tracer decay.
e                                                                     _i   _ I
  The mean radon source magnitude is rounded  to the nearest 0.05 pCi In.

                                IN MAINE HOUSES

                          C.T. Hess, C.V. Weiffenbach
                       University of Maine, Orono, Maine
                         H.R. Prichard and T.F. Gesell
                      University of Texas,  Houston,  Texas

    Radon measurements using  track  etch  cups  and diffusion alpha detectors in
the field are  compared  with air grab samples and  water  grab samples measured
in the laboratory for 100 houses in Maine.  These  measurements have been taken
for the  winter  months  from  October 1980 to April  1981 as  part  of an on-going
year  round  study.   Airborne radon  concentrations ranging  from O.05  to  21O
pCi/1 have  been measured.   Radon  in  water  concentrations from  20  to  180,000
pCi/1  have  been  measured.    A  linear regression  radon  concentration  in  air
versus radon concentration in water is significant at the r = 0.5 significance
level for 40 laboratory samples of  each.  The concentration of radon in air is
1.3 pCi/1 for 10,000 pCi/1 in water.  Analysis of constant sources of radon in
air such as soil, bedrock,  or building materials and variable sources of radon
such as radon rich water shows a strong  inverse effect on concentration due to
building ventilation.

    There has been great interest  in  indoor radon  in houses in recent publica-
tions by Barnaby on radon  in Swedish houses,  (Ba  79),  by Cliff on  radon in
dwellings in Great  Britain  (Cl  78, Cl 79) , by Fleisher in Rochester, New York
(Fl 81), and by the  U.S.E.P.A.  in  houses built  on phosphate tailings piles in
Florida  (US 75).  Recommendations  for changes have been  made for radon levels
above 3  pCi/1  in homes.  In Canada recommendations for changes  have been made
for radon greater  than 5 pCi/1, in places  such as Uranium City, Saskatchewan
(Ke 78).

    This paper  presents  the initial  results  of  a  cooperative project  between
the University  of  Maine Department of  Physics, and  the  University  of Texas
School of Public Health to  measure indoor  air  and water  radon  levels  in 10O
houses in Maine.  The  radon in  air was  measured in  the  laboratory  with using
96 grab samples and  with 514  track  etch  cups.   The radon  in water was measured
in the  laboratory  using 100 grab  samples  from  the same  houses.   The  houses
were also sampled  for  uranium  in  water  and were characterized  for  fuel used,
estimate of  ventilation, presence of  granite   building  materials,  and  with
types of construction,  number of residents,  and  volume  of house.  The survey
took place in  October,  1980, through May,  1981,  and is  part of a  year round
study of radon  level in houses.  The  initial grab samples,  of  air  and water,
and the initial place of Terradex - Fl track etch cups with paper filters took
place in October, 198O,  and the cups were collected in April - May, 1981, and
then analyzed by Terradex Corp.  The  data thus  represents radon levels during

fall and  winter  months,  when the ventilation  is  at a minimum  in Maine.  The
track etch  cups  were  placed in  bedroom,  bathroom,  living  room  or  kitchen,
basement,  and  outside on  a  porch.   The  grab  sample of  5 liters of  air was
collected in Tedlar or heavy plastic bags, in the  kitchen  or  living  room of a
house,  while  students were collecting  the  information about the house.  The
water sample was then taken after the water was allowed to flow in a sink for
at least  10  minutes to permit fresh water  to be  brought  up  out of  the well.
The  10  ml  water  sample  was  taken  with  a  syringe  and placed in   liquid
scintillation vials with  10  ml mineral oil fluor  and analyzed  for 40 minutes
with a  commercial  liquid scintillation  counter.   Samples were  standardized
with radium  solutions;  for  further details  see  the  method  of  Prichard and
Gesell (Pr 80) .  A 10 liter water  sample was  taken and acidified with 10 ml
1 normal HC1 for radium or uranium analysis.  Track etch cups were provided by
the  U.S.E.P.A.  Laboratory  in  Las  Vegas,  Nevada  and  analyzed  with  low
sensitivity and  high  sensitivity depending on  the density of  tracks.   After
the cups were exposed for 6 months - 7 months,  the cups were collected,  sealed
together  with  plastic inserts  and stored in Tedlar  bags  until  returning them
to be measured at Terradex Corp.  in California.  Cups were identified by track
etch cup number and house code  number.

    The houses measured  were taken from a  group  of houses previously studied
for radon in water.   The houses were selected for  a  high  radon  in water group
and a low radon  in water group.   Three fourths of the houses were measured by
student teams  and  one fourth by  the State  of  Maine Department  of Health En-
gineering.  All  the cups  were  handled  in the  same way and the survey is being
continued this summer using multiple cups at the same  point  in  the house.  Of
the original 53O cups, 513 have  been  used and measured,  a return of more than
95% of cups placed.

    Radon in  air was  measured in the grab  samples by extracting the radon in
-70°C Hexane,  and  counting it with commercial  liquid scintillation  counters.
Additional data  were  collected  in  several  houses  using  electronic  diffusion
alpha detectors  of  the Wrenn type (Wr  75) .   The diffusion alpha detector was
calibrated using National  Bureau  of Standards  radium sources  and  recalibrated
at  the  Eastern  Environmental  Measurements  Laboratory,  and   other university
laboratories,  at  the University  of   Texas  and  University  of  Maine,  and
U.S.E.P.A. lab at  Montgomery,  Alabama.   The alpha measurements  of radon were
taken over  a  period  of  two  days to  7 days at  10 minute  intervals.   These
measurements were  used to  determine  the variation  of radon with time  in  a
variety of houses with various sources of radon in air.  The data  was analyzed
for peaks  and compared to  records  of  water use  and changes in  ventilation.
Analysis of pulse shapes permitted determination of ventilation times for some
of the houses.


    The radon  in air values  for  grab  samples  of  air ranged from .03  to 201
pCi/1 while the  track etch cups  ranged from 0.06  to  132  pCi/1.  These  values
are illustrated  in Figure 1 which  compares the grab  sample  taken in October

with the average  value  for  the track etch cup taken from October through May.
The data are correlated significantly with r = 0.991,  with a slope of 1.49 and
an  intercept  of -0.48  pCi/1.   This result may  show  a systematic measurement
difference or may  show  a difference in average  ventilation.   To test average
radon levels  due  to radon from water, which  ranged from  20 to 115,000 pCi/1,
plots were made of bathroom air measured  with  track etch  cups versus radon in
water concentration.   These  results are  shown in Figure  2.   This  data  has a
slope of 0.107 and an intercept of 1.59 with r = .307, n = 85  significance.  A
similar curve  was made for radon  in the  basement air measured  by  track etch
cup and water radon concentration in Figure 3.   The data  has  a slope of O.127
and an  intercept  of 4.49 with r =  O.151  significance.  A summary of all data
for track etch cups versus water radon is shown in Table I.  Surprisingly, the
highest value  for slope is for  radon  in  basement air versus  radon in water.
Significant lower slopes are shown for  all bathroom and kitchen-living room.

    Histograms  number  of occurrences  versus  radon  levels by room, measured
using track etch cups are shown in  Figure  4.   The average  values for radon in
air are shown  with arrows and are as follows; bathrooms 3.51 _+ 7.12, bedrooms
3.10 +  6.98,  kitchen-living  rooms  3.21 + 8.27,  basement  6.79  +  16.8, outside
0.83 _+  .79.   Since the radon is highest in basements, this shows that a major
source  is  associated with soil gas, or water  radon in floor  drains, or  base-
ment building materials.  Reductions on the second  floor  are by 2,  and reduc-
tions on  third floors  are  also observed  to  be factors of  2 from  two  story
house  data.   Error  estimates are  taken  for  the data to simulate  a  normal

    By  using calibrated  Spitz-Wrenn meters  (Wr 75) , measurements  have  been
made of levels of radon in air during water uses such as showering,  dish wash-
ing, and  clothes  washing.   The fraction of radon lost ranges  from  50%  to 95%
for these  activities  (Pa  79).  Weekly averages of the radon  in  the 20 houses
showed  radon levels from  0.5 pCi/1  to  87 pCi/1  during  the  heating season.
Occasional radon daughter measurements by the method of Kusnetz  (Ku 56)  showed
equilibrium of  daughters  from 0.3  to  0.5.  Portions  of records  of  radon con-
centration in air  of  three  houses A, B and 18 are  shown  in Figures  5  and 6.
House A has  the highest levels we have found.  This house, with an average of
87  pCi/1  is  a solar heated  house  with connected greenhouse  built  on granite
bedrock with  2  feet of granite  gravel fill.   House B shows  no  222Rn contri-
bution other than water supply.  These results show that the radon in water is
not always the  most  significant  source of radon in the air; but that radon in
water usually high when radon from all sources is high.  Other sources include
granite blocks  in  basement   walls  or  granite  gravels in concrete  or  direct
emanation into houses from granite through soil and foundations.

    Results exhibiting  the  variation of radon  in air  in  homes with the radon
in  the  water  are  summarized  in Figure 7.    The open  circles represent average
radon in  the air  due  to water supply only obtained  by  considering only the
radon appearing during  times of major water uses and which decreased with time
according  to apparent ventilation rates of the houses studied - see Figure 5.
For the component of radon due to water,  the radon levels  in the air have been
normalized to these  levels  expected for  a  ventilation rate of one  air change
per hour.   The  unadjusted average  radon levels  are also  plotted as asterisks
for all the  houses studied.   All data presented are  for  winter, with minimum
ventilation  in  the  houses.    Since summer  ventilation  decreases   levels  in
houses,  the expected exposures from radon  in  these  sixteen houses range from
0.01 to 14 working level months per  year.


Signif .
No. Pts.
No .2.3 pCi/1

Ba79   Barnaby,  W.  1979,   "Very   high   radiation  levels  found  in  Swedish
       houses."  Nature 281, 6.

Cl 78  Cliff, K.D.  1978,  "Assessment  of  airborne radon daughter concentrations
       in dwellings in Great Britain."  Phys . Med . Biol. 23, 696.

Cl 79  Cliff, K.D., Davies,  B.C. and  Reissland,  J.A.  1979, "Little danger from
       radon."  Nature 279, 12.

Fl 81  Fleischer,  R.L.,   Mongro-Campero ,   A.  and  Turner,  C.G.,   1981,  "Radon
       levels  in homes  in the  Northeastern  United  States."    Second  Special
       Symposium  on  Natural  Radiation  Environment  Bhabha  Atomic  Research
       Center, Bombay, India.

US 75  United  States  Environmental  Protection  Agency   (1975).   Preliminary
       findings:  radon daughter levels in structures  constructed on reclaimed
       Florida phosphate land.  Technical note,  EPA ORP/CSD 75-4.

Ge 79  George,  A.C.  and Breslin,  A.J.,  1979,  "Distribution  of  ambient  radon
       and  radon daughters  in New York  - New  Jersey residences."  Paper  in
       Proc. Nat. Rad.  Environ III  Edited by  Gesell, T.F., Lowder, W.M.  and
       Mclaughlin, J.E. Con. 780422 (Dept. of Energy, Washington).

Ke 78  Keith Consulting,  1978,  Report on  investigative and  remedial measures,
       radiation  reduction  and  radioactive  decontamination  in  Uranium  City,
       Saskatchewan  for  the  Atomic  Energy  Control  Board  of  Canada.   Keith
       Consulting of Regina, Prince Albert, Lethbridge, Edmonton, Canada.

Pa 79  Partridge, J.E.,  Horton,  T.R., and  Sensintaffar ,  E.L.,  1979, "A  study
       of radon -222  released from water  during  typical  household activities"
       U.S.E.P.A. Technical  note ORP/EERF-79-1  Eastern Environmental Radiation
       Facility, Montgomery, Alabama  36109.
Pr 77  Prichard,  H.M.  and  Gesell,  T.F.,  1977,  "Rapid  measurements  of
       concentrations  in   water  with   a  commercial   liquid   scintillation
       counter-"  Health Physics 33 No. 6, 577-581.

Pr 81  Prichard,  H.M.,  1981  "The  measurement  of radon-222  in  air  by  direct
       extraction  into  a  liquid  scintillator -"   International  Meeting  on
       radon-radon daughter measurements, Montgomery, Alabama.

Wr 75  Wrenn,  M.D.,  Spitz, H.  and  Cohen,  N. ,  1975,  "Design of  a continuous
       digital  output  environmental  radon  monitor-"    IEEE  Transactions  of
       Nuclear Science N.E. -33, 645.

Ku 56  Kusnets,  H.L.,  1956,  "Radon daughters  in mine  atmospheres:  A  field
       method  for  determining  concentrations."   AM .  IND .   HYG .   Q.   17(1):
       85-88 .

                                FIGURE CAPTIONS
Figure 1. Radon concentration  in air pCi/1  measured by  track etch cups  from
          October, 1980 to May; 1981 and grab  samples of  air taken in October,

Figure 2. Radon concentration  in  bathroom  air  measured  by  track etch  cups
          versus radon in water measured by liquid scintillation.

Figure 3. Radon concentration  in  basement  air  measured  by  track etch  cups
          versus radon in water measured by liquid scintillation.

Figure 4. Histograms of number of occurrences of  radon levels  by  room  measured
          with track etch cups versus concentration  of  levels in  the  room for
          outside  air, kitchen/living room air,  bedroom air,  bathroom  air and
          basement air.  Average values  indicated by  arrows.

Figure 5. Radon concentration  in  air   for two  different houses  measured  by
          diffusion  alpha  detectors at 10  minute  intervals versus  time  of
          day:  a) open window  to  outside; b) close window  to outside;  c) run
          clothes  washer; d)  shower; e) partially close off greenhouse  from
          house;  f  and g) showers;  h)  dishwasher; i)  bath  and shower.   Note
          for  houses  the effect  of individual  water  uses   in  comparison  to
          background 222Rn-

Figure 6. Radon concentration in House  18  versus time  of day  for  6  days,  noon
          marked with n.   Showers are marked  with letter  s,   baths with  letter
          b, laundry with letter  1  and dishwasher marked with letter  d.

Figure 7. Mean values over two to  seven days at  radon in air of  homes  during
          the heating  season.   Asterisks are  for uncorrected for  ventilation
          while circles are corrected for ventilation.

                                        133.0,201.0 24.5,23.3
                                          RADON IN AIR (pCi/L)

                                          TRACK ETCH CUPS

                                          VERSUS GRAB SAMPLES

                               TRACK ETCH CUPS

                       RADON IN  AIR  VERSUS
                       RADON IN  WATER  FOR
                       THE BATH
   24         40         56



O  8
                                133.0     54.6
                                         RADON IN AIR VERSUS

                                         RADON IN WATER FOR

                                         THE  BASEMENT
                     24          40

                   RADON IN WATER (nCi/L)

                               8  0
                          80     4     8
                      RADON IN AIR(pCi/L)
                                                      Figure 4

                       Figure 5



*   0
-  15-1
               HOUSE A  (28nCi Rn/l  WATER)
HOUSE B (27nCi Rn/l WATER)

                     24            12

                       TIME OF  DAY

            Figure 6



.•.'•'•• '
*i* * *S* »*..**
18 ( 18 nCi Rn/ 1 water) /':•'••
• "'••••..
"' '"Vl"
'•""fit ' 't
III 1 s




• * »
•••."••,.. 	 "''., ,...•„ •'.,./, ••"..,-•
'"./••" '... ,.•''
d ^ •'••''



' i " "" "" 	 ~>~r-
h d

•-'"'""'""t'H t!

             TIME   (n= noon)

                         Figure 7
                                ORADON  FROM WATER

                                * RADON FROM ALL  SOURCES
               20         4O          60
              RADON IN WATER SUPPLY, nCi/l

                         SURVEYS AND DECISION MAKING IN
                           A REMEDIAL ACTION PROGRAM

                                   A.G.  Scott

                                   DSMA Acres
                             4195 Dundas Street, W.
                           Toronto, Ontario, Canada

    As the observed frequency distribution of grab WL measurements in houses
 was found to be lognormal with a GSD of 2.0, lognormal sampling theory was
 used to guide the choice of sampling frequency for an extensive survey program
 in over 500 houses.  Analysis of lognormal data is illustrated by a Monte
 Carlo sampling simulation.

    The tasks of the survey portion of the Remedial Action Program at Elliot
 Lake are:

    1.   identify those houses with annual average WL's in excess of the
         action criterion of 0.02 WL,

    2.   demonstrate that the annual average WL is below 0.02 WL in those
         houses where remedial work is not carried out,

    3.   demonstrate that the annual average WL following remedial work is
         less than the action criterion of 0.02 WL.

    The program started in July 1977, and the belief at that time was that the
 elevated WL's in houses were due to the presence of radioactive contamination
 adjacent to the houses.  As a result, relatively few WL measurements were
 made, the major effort being in gamma-ray survey.  By late 1977 it was clear
 that radioactive contamination was not the major cause of elevated WL, and a
 systematic WL survey would be required to identify problem houses.  The survey
 would have to be based on repeated grab-samples, as that was the only equip-
 ment available at the time.

    By then sufficient WL measurements had been made to show that the distri-
 bution of WL measurements over a period of time in individual houses was near-
 ly lognormal with a geometric standard deviation (GSD) of approximately 2 (see
 Figure 1).   As a result of this, sampling theory for the lognormal distribu-
 tion was reviewed to determine sampling strategy as a basis for survey organi-
 zation.   Fortunately many pollutants also have a lognormal distribution, and
 so our task was greatly simplified by the use of a NIOSH publication (NI 75)
on this  subject.


in -
8 -
c _
D —i
A -
0 _





X = 39
GSD = 2.03





D 5


D ;


rO 9

1 HhUKh 1 ILAL
J = 2 0
GSD = 2.0

D 95

? 9!

                                                    CUMULATIVE FREQUENCY (%)
                                                           FIGURE  1

    Figure 2 is copied from (NI  75)  and shows  the variation of the 2 sigma
confidence limits on the mean of a lognormal distribution with the number of
samples taken and the GSD of the distribution.   For  a  GSD of  2.0, at least  9
measurements are required to reduce the confidence limits to  below 50%,  and a
large number is required to reduce the  limits  to less  than 40%.  From a  prac-
tical point of view, there was little gain  in  accuracy in taking more than  10
independent measurements, and so that number was chosen as the minimum number
of samples in a survey.  For convenience, the  "annual  average survey" was set
at 13 measurements, as this fitted with an  initial measurement followed  by  a
sample a month over a 12 month period.

    As readings tend to be higher in the summer, it  is necessary to take read-
ings over a full year to avoid biasing  the  results according  to season.  There
will still be a slight theoretical bias in  the  estimate, for  the samples are
not truly distributed at random, since  they are not  taken at  night nor at
weekends which together comprise nearly 3/4 of  the year.  However, this  is
unavoidable in practice.

    To avoid waiting to start remedial  work until a  year's survey was complet-
ed, a set of rules based on "distribution free" statistics was generated to
identify those houses where the  annual  average  was clearly in excess of  20  mWL
so that they could receive prompt remedial  work.

    When enough readings have been made to  end  the survey, the average WL in
the house is calculated.  For those houses  where the average  is less than or
equal to 20 mWL, the data are sent to the RAP manager  at the  AECB* for
review.  If he agrees that the average  is probably less than  20 mWL, he  issues
the homeowner a letter to that effect.   If  he  feels  that the  calculated  aver-
age is not fully representative  of the  average  in the  house,  he may request
extra survey readings, or that remedial action  be carried out in the house.

    For those houses where the average  is greater than 20 mWL, the data  are
sent to the RAP manager for review.  The AECB  then contact the homeowners to
obtain their agreement to remedial work.

    After the remedial work is completed, a Post Remedial Survey is carried
out to show that the WL in the house has been successfully reduced to less
than 20 mWL.

    Major restoration of finish  and appearance  is not  started until this sur-
vey is completed.  As the homeowner  is  naturally interested in completion of
remedial work as soon as possible,  a rapid  answer is required.

    The procedure adopted is to  take 1  WL measurement  a day for 10 consecutive
working days.   If the best estimate  of  the  mean is less than  0.02 WL, the
house is forwarded to the RAP manager as probably complying with the primary
criterion.   The Compliance Division  of  the  AECB subsequently  check these
houses to verify that the WL has been reduced.   The  interrelationship between
the survey,  the remedial program,  and the AECB  is shown in Figure 3.
    AECB-Atomic Energy Control Board.  The AECB directs the Remedial Action
    Programs carried out and financed by the Joint Federal Provincial Task
    Force on Radioactivity.

                     NUMBER OF GRAB SAMPLES
Figure 2 - Effect of small  grab  sample sizes on requirements for
           demonstration of noncompliance.  Three different data
           geometric standard  deviations  (GSD) are shown which
           reflect the amount  of variability in the environment.


       STAGE  1
                    STAGE 2
         FIX      \N_
     PASSES POST    \N_
                                          _V  AECB ACCEPTS
                                            \ SURVEY RESULTS
          AECB ACCEPTS  \N
                               FIGURE 3


    As a large number of houses (500+)  were  under  survey,  it was not practical
to keep an overview of the program by manual summaries.  Therefore, a computer
accessible data base was developed which lists  the survey  readings made in a
house, produces estimates of the house average  WL  and  also keeps track of the
progress of the house through the remedial action  program.

    A demonstration of the analysis problems posed by  the  lognormal distribu-
tion, a Monte Carlo sampling simulation was  carried out.   Figure 4 shows a
histogram of 5500 numbers generated randomly from  a lognormal distribution
with mean 20 and GSD of 2.0.  If these 5500  lognormal  numbers are analyzed in
the usual manner (which assumes a gaussian distribution) the mean is 19.7 and
the standard deviation is 15.6.  The gaussian with these parameters is shown
in Figure 4, and can be seen to be a rather  inadequate description of the
actual distribution of numbers.  Approximately  16% of  the  numbers would have
to be negative for the gaussian to apply.

    However, we do not make decisions based  on  a single reading, but only on
the average of groups of readings.  It is our practice to  take at least 10
readings in a house before we make any statistical decision, and so it is the
frequency distribution of samples of 10 that is of interest.  This distribu-
tion was obtained by taking 550 groups of 10 numbers from  the 5500 lognormal
numbers generated earlier.  The arithmetic mean and the best-estimate mean
were calculated for each group of 10.  The best-estimate mean is based on the
logarithmic mean of the sample numbers, and  is  the theoretically correct
method to estimate the average value of a lognormal distribution from a
sample.  Figure 5 shows the CDF of the 550 best-estimate means.  They are
lognormally distributed as expected, with variability  reduced close to the
theoretical value of GSD = 1.25.  The distribution of  arithmetic means was
virtually identical to this, showing that the uncertainty  in the estimate of
the mean was not a function of the method of analysis, but of the variability
of the measurements themselves.  The slight  asymmetry  in the distribution does
require the confidence limits to be asymmetric, and so confidence limits based
on gaussian analysis will be incorrect.  For example,  gaussian analysis of the
5500 numbers gives 95% confidence limits on  the mean groups of 10 samples as
11.6 and 27.8.  Figure 5 enables the 95% confidence limits to be read direct-
ly.  They are approximately 13 and 28.5, which  are close to the gaussian
values, but slightly higher.

    In practice the use of confidence limits creates yet another problem -
what to do in the case of those measurements that  fall between the confidence
limits.  For example, if we wish to work to  90% two tailed confidence limits,
and we make 10 measurements, Figure 5 shows  us  that we can make no decision if
the mean of the measurements is more than 13 mWL and less than 28.5 mWL.  Now,
the one thing we cannot do in the real world is to take measurements and then
fail to make a decision, and although we can narrow the limits by taking more
readings, the zone of indecision cannot be removed completely except by an
arbitrary decision.   We prefer to make our arbitrary decision first, and so
for our surveys we work at 50% confidence.   At  this level  there is no zone of
indecision,  and it has the intuitively appealing feature that we apparently
believe the average of our measurements when it comes  to deciding if the
annual average WL in a house is, or is not,  greater than 20 mWL.

                                                               GAUSSIAN ESTIMATED
                                                               FROM MEAN AND VARIANCE
                                                               OF GENERATED DISTRIBUTION
                                                                                 70      80
                                                 FIGURE 4

                             BEST ESTIMATE MEAN AND
                              ARITHMETIC MEAN
           30    50     70      90
           Cumulative Frequency (%)

             FIGURE 5


Ni 75  "Statistical  Methods  for  the  Determination  of  Non-compliance  with
       Occupational Health  Standards."   HEW  Publication  No.  (NIOSH)  75-159,
       April,  1975.

                  Bernd Kahn,  Marcia Wilson  and  John T. Gasper

                School of Nuclear  Engineering  and  Health Physics
                        Georgia Institute of Technology
                            Atlanta, Georgia  30332

    The relation of Rn-222 in building  air  to Ra-226  in structural materials
was examined by measuring the Rn-222 flux density  at  surfaces and
concentration in air in two office and  laboratory  buildings.  The buildings
have walls constructed partially with concrete blocks that contain Ra-226 at
relatively high levels.  The Rn-222 flux density at walls and floors was
measured by gamma-ray spectral analysis of  charcoal collectors for radon that
had been sealed to surfaces for 3-day periods.   The Rn-222 concentration in
air was measured by collecting air in evacuated 125-cc radon scintillation
(Lucas) cells and counting alpha particles  in the  cells.  Results were
compared with samples collected outside and in a building that did not have
elevated Ra-226 levels in structural materials.  To determine source terms,
Ra-226 concentrations were measured in  building materials and soil.
Contributions to the Rn-222 concentration in building air from the ground
beneath the building, outside air, and  construction materials were estimated.
The identified sources contributed between  67 and  100 percent of the average
of measured concentrations; in every instance,  most of the Rn-222 was from the
ground beneath the building.


    Some concrete made from phosphate slag  contains  elevated Ra-226 levels.
Buildings constructed with this concrete were  identified  in surveys through
increased external radiation exposure rates due  to the gamma radiations  that
Ra-226 and its Pb-214 and Bi-214 progeny emit.   Exhalation of  the gaseous
Rn-222 daughter from concrete walls,  floors, or  pillars into building  air will
also increase the radiation dose to the  lungs.   This dose is due to alpha
particles emitted by the Po-218 and Po-214  progeny,  at concentrations  usually
specified in terms of the working level  (WL).  An  increased WL value is
generally considered to indicate a greater  hazard  than the accompanying
external exposure rate in a building, but is not as  easily quantified  because
radon and its progeny in building air fluctuate  widely.   Complicating  factors
in determining Rn-222 levels in air due  to  building  materials  are the
dependence of these levels on the ventilation  rate and the fraction of Rn-222
emanating from the material, as well as  relatively high levels from other
sources.  This report describes measurements of  Rn-222 concentrations  in air
and exhalation rates from surfaces to distinguish  among the sources of Rn-222
in air in three buildings, two of which  were built with concrete blocks  that
contain high-radium phosphate slag.

    The slag from the thermal process for phosphorus production is a calcium
silicate that has been used variously for the  fine and coarse  aggregate  in
concrete (Bo77, He79).  Phosphorus plants in Idaho (Bo77, Pe78), Montana
(L178) and Alabama (Ma78, Ka79) have produced  slags  that  contain Ra-226  at
concentrations between 30 and 60 picocurie  per gram  (pCi/g).   The Ra-226
concentration in concrete depends on the Ra-226  level in  the slag derived from
phosphate rock and on the fraction of slag  used  in the concrete.  Radiation
exposure rates of approximately 40 microroentgen per hour (uR/hr) have been
measured in structures built with this material  (Bo77, Ka79).  In comparison,
exposure rates in buildings due to cosmic rays and Ra-226, Th-232, and K-40 in
building material and the ground usually range from  6 to  18 uR/hr  (Ka79).

    The EPA had proposed rules under the Resources Conservation and Recovery
Act of 1976 to prevent use of byproduct  materials  such as phosphate slag if
they increase gamma ray exposure rates by 5 uR/hr  or WL values by 0.03.  At
equilibrium, 0.03 WL results from a Rn-222  concentration  of 3  picocurie  per
liter (pCi/L).  In buildings, the ratio  of  WL  to Rn-222 concentrations
normally is well below this value, but differs according  to various
conditions.  The short-lived progeny do  not attain equilibrium due to  air
turnover and deposit on surfaces to some extent.   Mixtures of  Rn-222 and
progeny in building air typically result in approximately 0.004 WL per pCi/L
(UN77).   Increases of 5 uR/hr and 0.03 WL or the corresponding 7.5 pCi/1 are
readily detected.

    A number of reviews have considered  Rn-222 in  building air and its sources
(Co81, Ea75, Ha78, Ka79, Le81, Mo76,  NC75,  OE79, Ta80, UN77).  Elevated  Rn-222
concentrations in building air can be due to high  concentrations in outside
air, the ground beneath, or the water or gas supply.  Initially low levels can
rise steeply due to accumulation of radon and  its  daughters within a tightly
closed room or structure (Sp80).   The Rn-222 concentration in  outside  air is
typically a few tenths of a pCi/L, but may  fluctuate in the course of  the year

by almost two orders of magnitude at a single location (Ha78a).   Radon inflow
from beneath the building is particularly apparent in houses built on uranium
tailings and at locations of former phosphate mines (Gu80, Ea75).

    Average WL values in buildings not known to be associated with any
elevated Ra-226 levels appear to be between 0.003 and 0.007 WL (Ka79, Ge80)-
Because of the above-cited factors, the individual values on which these
averages are based may vary by almost two orders of magnitude among the
buildings included in a particular survey.

    Radon-222 progeny levels in buildings constructed with high-radium
concrete were between 0.0005 and 0.05 WL in 107 structures in Idaho (Pe78);
between 0.001 and 0.1 WL (except for one higher value) in 69 Montana houses
(L178); and averaged 0.018 WL in 17 Alabama houses (US80).  These measurements
were performed for extended periods to determine average radiation doses.   In
Montana, averages in two towns were 0.02 and 0.013 WL (US80). The log-mean
value of 0.006 WL in the Idaho town is reported to be the same as the normal
background, and the corresponding log-mean Rn-222 concentration  of 1.4 pCi/L
is also approximately the same as in normal structures (Pe78, US80).   The
Rn-222 concentration and WL value in outside air and the Rn-222  flux density
from the ground beneath buildings were not reported.

    In this study, air samples for Rn-222 analysis were collected in brief
(1-minute) periods at 3 multi-story office/laboratory buildings  located near
each other.  In building NE, interior walls on the second and third,  but not
the first floor, are of concrete blocks with a Ra-226 concentration of 19
pCi/g; in building ER, most walls, interior and exterior, on all three floors
were constructed with these blocks; while in building CE the building
materials are not elevated in Ra-226.  The Ra-226 in all concrete except the
blocks was estimated to average 1.5 pCi/g on the basis of external radiation
exposure rates (Ka79).  Air samples were collected simultaneously to compare
Rn-222 levels inside and outside, on the various floors of individual
buildings, and on corresponding floors in the three buildings.   Samples were
collected in the mornings, when the highest levels for daytime occupancy were
expected due to the early-morning peak in Rn-222 concentrations  outside and
the overnight accumulation inside while the buildings were closed.  The
turnover rate of air within the three buildings was measured by  releasing and
collecting a tracer gas.

    The exhalation rate of Rn-222 from the concrete blocks was measured for
3-day periods in the NE and ER buildings and compared with that  from poured
concrete walls in the NE and CE buildings.  The Ra-226 content and Rn-222
emanation fraction (or emanation power) in crushed blocks was also measured.
The exhalation rates of Rn-222 from soil and through the floors  of the three
buildings were measured to determine the relative contribution by the ground
to the total Rn-222 concentrations in air.  The external radiation exposure
rate and Ra-226 concentration in surface soil were measured as possible
indicators of the magnitude of Rn-222 exhalation rates from the  ground. The
Rn-222 concentration was measured in the public water supply to  determine the
contribution of this potential source to radon levels in building air.


    The 3 measurement sites are combined office and laboratory buildings at
the Georgia Institute of Technology.  The NE building is built on a slab and
is three stories high.  Its outside walls are brick with some concrete and
metal panels; the inside walls are mostly of concrete block.  The second and
third stories were added in 1965.  Only the blocks on these floors contain
elevated amounts of Ra-226.  The ER building was also built in 1965, except
for a small west wing added later.  It has a basement that is above ground on
two sides, 2 floors above basement level, and a small partial third floor.
The exterior walls are concrete block faced with brick and some concrete and
metal panels above the basement; the interior walls are of concrete block.
Except in the newer wing, the blocks have the same elevated Ra-226 levels as
those in the NE building.  The CE building, used as control, is much older.
It has a basement that is above ground on two sides, and three stories above
the basement.  The walls are of poured concrete, faced with brick.  All
buildings have structural supports and floors of poured concrete.  The two
newer buildings are centrally air conditioned.  The older building has window
or area air conditioners in most offices and laboratories, but the hallways
and a laboratory extending throughout the basement are not air conditioned.

    Cylindrical cells (6.0-cm dia. x 9.4 cm outside) coated on the inside with
ZnS(Ag) scintillator were used to measure alpha particles from radon and its
progeny (Gu80).  The cells have a quartz window at one end and a glass
stopcock at the other.  The cells were evacuated to approximately 0.04
atmosphere with a vacuum pump before sample collection.   They were counted
within a shield to keep out light by placing the window on a photomultiplier
tube.  The cells were calibrated with standard Ra-226 solutions at 5 pCi/L
levels.  The counting efficiency for combined Rn-222, Po-218, and Po-214 alpha
particles was 2.5 +_ 0.1 counts/disintegration of Rn-222.   Before and after
each measurement of Rn-222, the detector background was  determined.   Each
sample was counted twice in succession.   The presence of Rn-222 and the
absence of interference was confirmed by observing the 3.82-day half life of
Rn-222.  The combined counting periods of the background measured before and
after counters were typically 1,200 minutes, and the two sample counts were
for a similar period.  Intervals of 150  minutes between sample collection and
measurement and between cell flushing and background measurement were
maintained to assure decay of any unsupported short-lived radon progeny.

    Air samples were collected by opening the stopcock of an evacuated cell
for approximately 1 minute.  The cell was usually held 1 m above the floor in
the middle of corridors.   Outdoors, air  was collected 1  m above open ground
and several meters distant from the building.  Weather conditions were

    The standard deviation of counting these samples, which is believed to be
the major  source of random error, was typically 0.04 pCi/L, and the minimum
detectable level,  defined as the 3-sigma counting error  of the background, was
0.08 pCi/L.   Three samples collected side by side on 9-6-81 (see Table 1)
measured 0.85,  0.81,  and  0.86 pCi/L.   Hence, differences by more than 0.1
pCi/L are  considered significant.

               Table 1
Radon-222 Concentrations  in Air,  pCi/L
1981 Time Conditions
3-6 0910 S, NC, BW*
3-12 0845 S, NC, BW
3-17 0900 S, NC, LW
3-19 0900 S, NC, BW
3-23 0820 OC, LW,
rained earlier
7-8 1040 S, NC, LW
7-10 1040 S, NC, C
7-15 1030 S, NC, LW

7-21 1050 S, NC, LW
CE Building
out- base-
side ment 1st 2nd 3rd

<0.1 0.8 1.0

0.1 0.7 0.4 0.5
<0.1 0.4 0.1 <0.1

0.1 0.2
NE Building
side 1st 2nd 3rd
<0.1 1.0 0.6
<0.1 1.0 0.5

0.2 1.5 0.9 0.6

0.2 1.1 0.6
1.5 0.7

ER Building
out- base-
side ment 1st 2nd

0.1 0.5 0.1 0.9
0.4 1.1

<0.1 3.2 2.9 2.2
0.2 2.9

                                                      Table 1 (cont'd)
1981 Time Conditions
7-25 1040 S, NC, LW
7-28 0730 S, NC, LW
7-30 0730 OC, LW
0930 light rain
1130 at 11:30
8-4 0810 S, NC, C
8-6 0750 S, SC, LW
8-13 0830 S, SC, LW
CE Building
out- base-
side ment 1st 2nd 3rd
<0.1 0.6

0.3 1.9 0.5 0.4 0.6

NE Building
side 1st 2nd 3rd
0.3 1.8
0.4 1.6
0.2 1.3
<0.1 0.9
0.2 0.8
0.3 0.5

0.2 1.3 0.9 0.8t

ER Building
out- base-
side ment 1st 2nd
0.2 3.5

0.3 2.6 2.9 2.1
3.2 2.2
       *  S: sunny; NC: no clouds; SC: scattered clouds; OC:  overcast; C:  calm;  LW:  light  wind;  BW:  brisk  wind;
       t  triplicate samples

    The Rn-222 content in tap water collected in the CE building on 11-17-81
was measured in duplicate samples.  Aliquots of 18 ml water were placed in
emanation tubes and radon was flushed from them into Lucas cells,  which were
counted twice in succession for a total period of 1,300 minutes.  The
background in these cells was measured before and after the samples for 500
minutes each. The lower limit of detection (3-sigma value of the background
count) was 0.6 pCi/L.  The same water—Atlanta city water obtained from the
Chattahoochee River—is used in all three buildings.

    The rate of air turnover was determined by releasing SF^ gas on separate
occasions on the lowest and second lowest floor of each of the  three
buildings.  Approximately 1 cc of the gas was released.  Air samples of 1  cc
each were then collected in 2-cc plastic graduated syringes and injected into
a gas chromatograph (Varian model 3700).  The SF£  peak was observed just
before the oxygen peak.  The peak areas were measured for samples  collected
generally at 10-minute intervals for one to two hours after gas release.
These areas were plotted on semilogarithmic paper as a function of time after
gas release to determine the air exchange (ventilation) rate,Xa, in hr~ -

    The Rn-222 flux density was determined by collecting radon  on  activated
cocoanut charcoal (6 - 14 mesh) in cylindrical plastic containers  and counting
the charcoal after ingrowth of the Pb-214 and Bi-214 progeny with  a Ge(Li)
detector plus multichannel analyzer.  The container has a 500-cc volume and  is
filled with 200 cc charcoal (122 g) retained by a paper filter  and metal mesh
at the bottom of the container.  A plastic cover seals the container before
and after collection.  The container was pressed tightly against the surface
to be monitored and sealed to it with putty; for soil measurements,  the
container was pressed into the soil.  Samples were collected for 3-day periods
and counted for 10,000 to 50,000 seconds, depending on the count rate due  to
Rn-222 progeny.  The count rate was corrected for background radiation due to
Rn-222 progeny in air and decay of Rn-222 during collection,  between
collection and counting, and during counting.  The measurement  was converted
to units of picocurie per square meter per hour (pCi/m^hr) on the  basis of
the counting efficiency of the detector for a 200-cc sample volume,  the period
of collection, and the 80-cm^ surface area of the collector.

    Measurements of Rn-222 flux density were performed on walls constructed  of
high-radium concrete blocks and of poured concrete, on poured concrete floors,
and on soil near the buildings.  Two of the samplers were left  open to air in
the CE building and two were sealed with the cover for 3-day periods to test,
respectively, the extent to which still air collects on the charcoal and  the
extent of leakage into the container.  The open charcoal sampler collected the
equivalent of 500 pCi/nrhr.  The closed container collected no  Rn-222,  at
the detection limit of 10 pCi/m2hr.

    Concrete blocks from the NE, ER, and EA buildings (the latter  a warehouse
constructed with high-Ra-226 blocks) were broken into small pieces,  ranging
from powder to approximately 5 mm in diameter, and weighed 500-cc  aliquots
were analyzed for Ra-226 concentration by gamma-ray spectrometer.   To
determine the emanation fraction, a 10-gm aliquot was placed in a  50-cc glass
tube with stopcocks at each end, connected to a second 50-cc tube.  The sample
was stored for 2 to 6 days at accumulate Rn-222 in both tubes.   The connection

between the two tubes was then closed and the air in the second tube
(containing no concrete) was flushed into an evacuated radon scintillation
flask.  The measured volume was 40 percent of the air over the concrete
sample.  The ratio of the Rn-222 emanation rate to the generation rate was
determined from the Rn-222 in the sample and the Ra-226 in the material,
corrected for decay of Rn-222 during collection, between collection and
measurement, and during measurement.

    Soil samples were collected from open ground at the two locations where
the Rn-222 flux density had been measured outside each building.  Volumes of
500 cc were scraped from the top 2 cm of ground.  The samples were placed in
plastic 500 cc containers, weighed, and measured by gamma-ray spectrometry.

    Radium-226 was determined in soil and pieces of concrete block in closed
plastic containers by measuring the 352- and 609-keV gamma rays of Pb-214 and
Bi-214, respectively.  To check these values, the 186-keV gamma ray of Ra-226
and U-235 were also measured.  It was assumed that 58 percent was due to
Ra-226 and the rest due to U-235, and that the Ra-226 gamma ray represented
3.5 percent of Ra-226 decays.  The 583- and 911-keV gamma rays of Ra-228
progeny were used to determine Ra-228,  and the 1,461-keV gamma ray was used
for determining K-40.  The radium results are uncertain to approximately 5
percent because of uncertainties in the applied gamma-ray fractions.  In
addition, the standard deviation of the count rate—5 to 10 percent—depends
on radionuclide concentrations and counting periods.

    A Reuter-Stokes pressurized ionization chamber held 1 m above ground was
used to measure external radiation exposure rates at the same locations.  The
detector was calibrated with a 10 mg Ra-226 standardized source.


    The measured Rn-222 concentrations  in air listed in Table 1 average 0.6,
1.0 and 2.2 pCi/L respectively in the CE, NE, and ER buildings compared to 0.2
pCi/L outside.  Each of these averages  represents between 16 and 25
measurements.  The averages are consistent with previous observations of much
higher Rn-222 concentrations that have  been encountered inside buildings than
outside.  Outside levels were near 0.2  pCi/L during extensive surveys in
Sweden (Hu56) and Austria (St80).   Indoor concentrations in Austria averaged
0.6 pCi/L and ranged from <0.05 to 5.2  pCi/L; those in Sweden generally were
in the same range.   Some much higher levels in Swedish houses were found
associated with low ventilation rates and concrete that contained alum shale

    Radon-222 concentrations measured in July-August were only slightly above
the March measurements  in the CE  and NE  buildings,  but much higher in the ER
building.   Averages  were as  follows  outside the building and inside on the
floor  nearest the  ground:

         Location            March          July-August

         outside             0.1 pCi/L      0.2 pCi/L

         CE basement         0.6            0.8

         NE ground floor     1.2            1.4

         ER basement         0.5            3.2

The increases in the CE and NE buildings may be associated with the  increased
levels in outside air, but the large increase in the ER building must  be due
to another cause.

    Within each building, the highest concentration of Rn-222 was consistently
in the story nearest the ground, as summarized in Table 2.   Levels in  upper
floors of the same building were similar.  The highest concentrations  in
buildings are commonly found nearest the ground, particularly in basements
(Ge80).  This observation gives support to the hypothesis  that the main  source
of RN-222 in building air is the flux from the ground beneath the building.
In the NE building, the elevated Rn-222 concentration on the ground  floor,
which does not have the high-radium concrete blocks used on the upper  floors,
suggests that these blocks are not the major source of RN-222.

    The measurements on 7-28 (see Table 1) show the decrease of Rn-222 levels
outside between 7:30 and 11:30 A.M.  This is consistent with previous
observations of diurnal fluctuations that result in a gradual increase at
night to a maximum near sunrise associated with stable atmospheric conditions
and possibly inversions during the night, and rapid decrease thereafter
(NC75)-  The changing indoor Rn-222 concentrations at higher levels  paralleled
those outside.  Some of the reduction of Rn-222 concentrations  indoors may be
due to greater air turnover as more persons used the building in the course of
the morning.  The measurements on the morning of 7-30,  undertaken on an
overcast day with relatively stable atmospheric conditions  and  light rain
beginning toward the end of the study period, showed relatively uniform  Rn-222
levels outdoors and only a slight decrease in the building.

    The highest exhalation rates were found at cracks and  seams in the
concrete floors of the buildings, and at one location in the ground  just
outside the ER building, as shown in Table 3.  In general,  the  exhalation
rates were several thousand pCi/m^hr from soil and several  hundred
pCi/m^hr from the concrete blocks.   The flux densities at  the floors nearest
the ground were much less where the floor was continuous than at cracks  or
seams.  The exhalation rates from poured concrete walls in  the  CE and  NE
buildings, which have only ordinary Ra-226 contents,  were  between 10 and 100
    A survey of exhalation rates  from soil  reported an  average value of  1,500
     ^hr (0.75 atoms/cm^sec)  and  values  that  ranged approximately
three-fold above and below this average  in  the  continental  United  States
(Wi72).   Considerable fluctuation can be caused by  changes  in air  pressure and
soil moisture content (Co81).  The values in  Table  3 for  the soil  near the CE

                             Table  2

                 Average Rn-222  Concentrations  In
                     Building and Outside  Air



• 7 (8)
.4 (7)


Number of measurements  in  parentheses

                                         Table  3

                             Radon-222  Flux  Density,  pCi/m2hr








CE Building
ground floor wall

2,100(SE) 4,630(C) 10
2,050(SW) 8,380(C) 14

NE Building
ground floor wall
3,430(E) 22,000(C) 580(W)
4,360(N) 48,300(5)
10,500(N) 270(E)



ER Building
ground floor wall

18,300(NE) 8,270(5) 75(W)
5,000(SW) 6,170(5) 220(E)


Notes:   1.   Letters  in parentheses  for ground  and wall  samples  show direction
            relative to building; for  floor  samples,  (C) or  (S)  indicates
            measurement over crack  or  seam,  respectively.
        2.   Samples  were collected  for 3-day periods, beginning  on indicated
        3.   The wall samples labelled  (E)  and  (W) were  for high-radium concrete
            blocks;  all  others  were for poured  concrete.

 building are near the cited average and were reproducible.  Flux densities at
 some of the other soil locations, however, were relatively high and differed
 for successive 3-day sampling periods by factors between 2 and 3.

    Flux densities between 30 and 650 pCi/m2hr (0.8 to 18 attocurie/cm2sec)
 had been measured on cellar floors in New Jersey, where the Ra-226
 concentration in soil averaged approximately 1 pCi/g (Ge80).  The flux
 densities measured on unbroken surfaces at the ground floors in the NE and ER
 buildings were toward the upper end of this range.  The high values above
 cracks or seams undoubtedly are higher still relative to the area of the
 breaks because most of the area of measurement was unbroken floor. As a
 result, the flux density for the floor as a whole is higher than on unbroken
 floor, and may reach the exhalation rate from the soil beneath the building

    Application of the reported ratio of 20 pCi/m2hr (0.005 pCi/m2sec) per
 Ra-226 concentration of 1 pCi/g in 10-cm-thick concrete (UN77) indicates that
 most of the exhaled Rn-222 is from beneath the floor.  This ratio also gives
 the order of magnitude of the observed values for the flux density at poured
 concrete and concrete block walls if their Ra-226 contents are 1.5 and 19
 pCi/g, respectively.

    The emanation fraction based on the flux density measured for four
 concrete blocks in the walls averaged 3.0 percent (see Table 4) and ranged
 from 0.5 to 5.7 percent.  The calculation is based on an average Ra-226
 daughter concentration of 19;+ 1 pCi/g measured in four blocks that had been
 removed from walls in these buildings (see Table 5), and the average weight
 and dimensions determined with these blocks.  Note that all of these blocks
 are hollow.  The loss of 3 percent of the Rn-222 by exhalation would increase
 the Ra-226 concentration to 19.6 pCi/g, consistent with the direct Ra-226
 measurement (Table 5) of 20 pCi/g.

    For comparison, the emanation fraction for two blocks from the EA building
 that had been crushed was 3.9 percent, as shown in Table 4.  These average
 values of 3.0 and 3.9 percent are above the range of 0.2 to 1.5 percent
 observed for 11 samples of Polish concrete that contain slag (Pe80).  The
 emanation fraction measured in 116 samples of ordinary concrete in Hungary,
 however, ranged from 15 to 60 percent, with a log-mean value of 30 percent
 (To80).  Agreement between the Ra-226 and radium-progeny concentrations in 4
 concrete blocks measured here supports a smaller emanation fraction.

    The radioactivity levels in surface soil just outside the three buildings,
 shown in Table 6, were relatively uniform.  The overall Ra-226 average of 1.5
 pCi/g is above the local average of 1 pCi/g, and the range of 1.0 to 2.6 pCi/g
 is at the upper end of values between 0.4 and 2.6 pCi/g found in the Atlanta
 area (Ka79).   The measured terrestrial exposure rates listed in Table 6 are
 consistent with the exposure rates calculated for the radionuclide
 concentrations.   These exposure rates were measured near the soil sampling
 locations but represent a much larger area, thus providing a measure of
confidence that the few samples represent radionuclide levels near the
buildings.   The terrestrial exposure rate was highest at the location of the
highest Rn-222 flux from soil,  but the much higher flux density was not

                           Table 4

      Emanation Fraction of Rn-222 from Concrete Blocks
NE (W)
NE (E)
ER (W)
ER (E)
Exhalation rate,
546 (2)a
316 (2)
45 (2)
217 (2)
Emanation fraction, %
3.9C (3)

a  Number of replicate measurements is indicated  in  parentheses.

b  Complete exhalation from one side of a block would  be  19
   pCi/g x 0.95 g/cm3 x 7.0 cm x 104cm2/m2 x 0.00755
   hr-1 = 9,450 pCi/m2hr.

c  Amount of Rn-222 in air per weight of block, relative  to
   Ra-226 concentrations of 19 pCi/g.

                               Table 5

                Radionuclide Content of Concrete Blocks
Sampling date

Ra-226 progeny
ty, pCi/g
Ra-228 progeny


Note:   The last  two  sets  of values are from Ka79, Table 4; block EA
       was reanalyzed  in  this study as indicated.

                                 Table 6

                Radium-226 Concentrations  in  Soil  Outside
          Buildings and Confirmatory Exposure Rate Measurements

Radionuclide concentration, pCi/g
Ra-226 progeny
Ra-228 progeny
Terrestrial exposure rate, yR/hr



5 ±
0 ±
3 ±
2 ±

7 ±
4 ±







.0 ±
.3 ±
.4 ±
.8 ±

.7 ±
.5 ±







.4 ±
.0 ±
.3 ±
.0 ±

.9 ±
.8 ±


*  _+ values are deviations from mean of duplicate measurements

t  Calculated on basis of

   I = 1.90 CRa_226 progeny + 2-82 CRa-228 progeny + 0.179CK_40

   where C is concentration in pCi/g

** Measured with pressurized ionization chamber;  3.8 yR/hr cosmic-
   ray exposure rate has been subtracted

proportional to the incremental exposure rate  above  the average.  The reported
ratio of 2,000 pCi/m2hr (0.5  pCi/m2sec)  per  soil  Ra-226 concentration of  1
pCi/g (UN77) is consistent with observed values except for  the highest
exhalation rate near the ER building.  The Ra-226 level of  only  1.8 pCi/g at
the surface suggests that the much higher Rn-222  flux density must be
attributed to elevated Ra-226 concentrations well below the surface or  freer
movement of radon through the soil at  that location.

    Five of the six ventilation measurements obtained on  each of the two  lower
floors in the three buildings yielded  rates  near  1 per hour, as  shown in
Figure 1.  These rates of exchange of  inside with outside air are consistent
with the previously reported  range in  buildings (Ha73).   The very high
turnover rate for the first floor  of the CE  building was  not used in
calculations because the nearby open windows and  doors during the ventilation
measurement were closed when  Rn-222 concentrations had been measured.

    The Rn-222 radioactivity  level in  a  structure can be  computed from  the
following balancing of Rn-222 sources  from the ground, building materials, and
outside air against radioactive decay  and ventilation:

    d GRn   Sm1m   Sgzgfw
    	 = 	 + 	 -  *Rn CRn - Xa (CRn -  L0)                    (1)

      dt      V       V

    d °Rn
if  	 = 0 (equilibrium) and \a »\ Rn (normal ventilation)

                m-^-m + Sglgfyr
    GRn = C0 + 	                                            (2)
                    Xa V
    Im = 10^ CRa XRnfepw                                               (3)

In the above equations,  the symbols represent  the following:

    CRn :  inside Rn-222 concentration,  pCi/L

    Co :   outside Rn-222 concentration,  pCi/L

    t   :  time,  hr

    Sm  :  material surface,  m^

    Sg  :  ground surface,  m^

    Im  :   inward Rn-222 flux density from material  surfaces, pCi ra

    Ig  :   inward Rn-222 flux density from ground surface,  pCi m~2hr~1

    V   :   building air  volume,  L

                                    Figure 1
   5,000 -
   2,000 -
   1,000 -
     500 -
r.    200 -
     100 -
                                            1st FLOOR (A	A) 1.38/hr
 (0	0) 1.24/hr.
                               2nd FLOOR
                               (V	V) 1.34/hr.
                                           1st FLOOR
                                            (D	D) 0.70/hr.
                                          (O	0)l.
                                         1st FLOOR
                                         (A—A) 4.5/hr.
                                           60    70    80

                                           TIME (MINUTES)
100   110   120    130    140

        :   air turnover rate,  hr"-"-

        :   Rn-222 decay rate,  0.00755  hr~l

    fw :   fractional Rn-222 diffusion  through  floor

        :   Ra-226 in material,  pCi  g~l
    fe  :   exhalation fraction from material

    P   :   density of material,  g  cm" 3

    w   :   one-half of wall  thickness,  cm

In the presence of two types of  wall materials,  one containing more Ra-226
than the other ; the terms  Smlm can be replaced by  Smlm + Sm*Im*,
where the unstarred and starred  terms refer to the low-radium and high-radium
materials, respectively.

    The building dimensions, Rn-222 flux densities and other data in Table 7
were used to compute Rn-222  concentrations in building air.  The value of V
applies to the entire building.  The values of Sm  refer to concrete and
brick surfaces, including  outside  walls, inside  walls, and floors that are not
unusually elevated in Ra-226 concentrations, whereas Sm* refers to the
surfaces of concrete blocks  with elevated Ra-226 content.  The area of
basement walls facing soil was added to the floor  area for Sg.  The averages
of the pertinent measured  flux densities in Table  3 were used for Im* at the
NE and ER buildings and for  Im at  all three buildings.  The average flux
measured in soil near the  building was  used for  !„ in preference to the
values measured on the lowest building  floor.  The differences between
measurements on unbroken floor and over cracks or  seams were so large that it
was assumed that all of the  Rn-222 from the ground entered the building
through passages through or  around the  floor (i.e., fw = 1.0).  The air
turnover rate for a building is  the average of the slopes of the pairs of
lines drawn through the data in Figure  1.

    The Rn-222 concentration in  tap water was 1.0  +_ 0.6 pCi/L.  This value
indicates that water use is  not  a  significant source of Rn-222 in building
air.  At an estimated hourly water use  of 1,000  litres, building volumes in
excess of 1 x 10? liters (see Table 7), and an air turnover rate of
approximately 1 per hour,  the concentration of Rn-222 in building air from
water would be only 1 x 10~^ pCi/L.  This estimate is consistent with the
expectation that water from  a surface supply would contribute only minor
amounts of Rn-222 to building air  (E079).

    The results of these calculations,  shown in  Table 8, suggest that Rn-222
emanating from the concrete  blocks high in Ra-226  has a concentration of 0.1
and 0.2 pCi/L in the NE and  ER buildings, respectively.  This contribution
would approximately double the concentration due to outside air.  The concrete
materials at the usual lower Ra-226 level in the three buildings contribute an
extremely small fraction of  the  total Rn-222 in  building air.  The major
source of Rn-222 in building air appears to be Ra-226 in soil beneath the

                                Table  7

     Factors for Calculating Rn-222  Concentrations  in  Building Air
dimensions CE Bui
plan, m2 40 x 16
14 x 11
height, m 5.5+2
V, m3 12,
Sm, m2 16,
Sm*, m
Sg, m2 1,
wp, g cm-2
Radon exhalation
and concentration
I , pCi m"2hr"1 2,
I , pCi m hr
Im*, pCi m-Zhr-1
CRa, pCi g-1
C*n3, pCi g"-1-
Air turnover
V hr-l
+ 27 x 4 +
x 3.6 + 3.0
16.1 inside
39. outside
NE Building
69 x 27
3 x 3.0
ER Building
65 x 29
3.7 + 2 x 3.0
*  material  with elevated Ra-226 content
t  does not  include outside panel  walls
() assumed value

                                Table 8
                  Calculated Rn-222 Concentrations
                       in Building Air, pCi/L
Source of Rn-222        CE Building    NE Building     ER Building

Outside air                 0.1            0.2             0.2

Ra-226 in ground            0.2            0.6             1.1

Ra-226 in building
 material -  1.5 pCi/g      0.06           0.05            0.03

          - 19   pCi/g       —            0.1             0.2

Total                        0.4            1.0             1.5

Measured average            0.6            1.0             2.2

building.  The large increase in the Rn-222 concentration in ER building air
from March to July may be related to an increased exhalation rate  from the

    Differences between the measured and computed Rn-222 concentrations shown
in Table 8 may be attributable to the following sources of uncertainty:  (1)
assuming that the Rn-222 flux density at the lowest floor equals that  at the
ground outside (not considering partial retention of radon by the  floor or
possibly higher flux rates from soil beneath the building than outside);  (2)
averaging Rn-222 concentrations for various times and locations; (3) combining
concentration and flux density measurements not collected simultaneously or at
the same locations; (4) measuring flux density and ventilation rates on only a
few occasions; (5) using approximate values for building volumes and surface
areas; and (6) obtaining only a few measurements of Ra-226 concentrations and
Rn-222 emanation fractions.


    In two office/laboratory buildings constructed in part with concrete
blocks that contain Ra-226 at the elevated level of 19'pCi/g,  average  Rn-222
concentrations in air were 1.0 and 2.2 pCi/L,  respectively.  Radon-222
concentrations averaged 0.6 pCi/L in a control building,  and 0.2 pCi/L in
outside air.  Samples collected simultaneously at comparison locations for
1-min periods showed that Rn-222 concentrations were highest nearest the
ground and similar to each other on upper floors.

    Radon-222 flux densities averaged 53 pCi/m^hr from poured concrete with
normal Ra-226 levels and 280 pCi/m^hr from the concrete blocks with a  Ra-226
concentration of 19 pCi/g.  The emanation fraction of Rn-222 from  the
high-radium concrete blocks averaged 3.0 percent on the basis  of flux  density
measurements on site and 3.9 percent for crushed block.  The flux  density at
the floor nearest the ground was several hundred pCi/m^hr over unbroken
concrete but one to two orders of magnitude higher over cracks and seams.  The
flux density over soil near these buildings ranged from 2,000  to 18,000
    According to calculations based on flux density and ventilation  rate
measurements, the concrete blocks with elevated radium content  contributed 0.1
to 0.2 pCi/L to the Rn-222 concentration in building air,  equal to the Rn-222
concentration due to outside air during this period.   Most of the Rn-222  in
air in the two study buildings and the control building is attributed to
Rn-222 exhalation from the ground, moving with little or no retention at
cracks, seams, and other passages through the concrete. The main differences
in building air Rn-222 concentration among the three buildings  appear to  be
due to differences in exhalation rates from the ground.

    The calculated Rn-222 concentrations in air in two of  the three  buildings
were one-third below the measured values.  Unless a source of Rn-222 has  been
overlooked, these differences may arise from basing comparisons on averages of
relatively few measurements with large differences in flux densities and
concentrations.  Assuring better comparability would require many more
measurements, obtained simultaneously.

    Despite the relatively elevated Ra-226  concentrations  in the concrete
blocks and the extensive use of these  blocks  in  the  two buildings, the Rn-222
and inferred WL values in building air were low  and  the blocks apparently
contributed only a small fraction of these.   At  a  typical  ratio of 0.004 WL
per Rn-222 concentration of 1 pCi/L in building  air,  the highest average value
was 0.009 WL, and the value due to the high-radium concrete blocks was 0.0008
WL.  Unless the ventilation rate is drastically  reduced and, at the same time,
the emanation fraction of the blocks increases greatly, Rn-222 levels from the
blocks would not reach limiting levels in the two  buildings.  The magnitude of
the emanation fraction and its dependence on  factors  such  as pressure,
temperature, and humidity changes is probably the  crucial  point in predicting
Rn-222 exhalation from building material, and requires considerable additional

Acknowledgments:   We thank Cornelia Jackson for  collecting and measuring the
July samples of Rn-222 in air,  Jacqualyn Gasper  and John Oliver for measuring
ventilation rates, and Dr.  John Garden for  use of  the gas  chromatograph.


Bo77  Soothe G. F.,  1977, "The Need for Radiation Controls in the Phosphate
    and Related Industries," Health Phys.  32, 285.

Ca75  Caruthers L. T. and Waltner A. W.,  1975, "Need for Standards for Natural
    Airborne Radioactivity (Radon and Daughters)  Concentration in Modern
    Buildings," Health Phys. 2J), 814.

Co81  Colle R., Rubin R. J., Knab L.T., and Hutchinson J.M.R.,  1981,  Radon
    Transport Through and Exhalation from Building  Materials:   A Review and
    Assessment, NBS  Technical Note 1139.

Ea75  Eadie G. G., 1975, "Radioactivity in Construction Materials, A  Literature
    Review and Bibliography," US EPA Rept. ORP/LV-75-1.

Ge80  George A.C.  and Breslin A.J., 1980,  "Distribution of Ambient Radon and
    Radon Daughters  in New York-New Jersey Residences," The Natural Radiation
    Environment III, T.F. Gesell and W.M.  Lowder,  eds., US DOE Rept.
    CONF-780422, p.  1237.

Gu80  Guimond R.J. and Windham S.T., 1980, "Radiological Evaluation of
    Structures Constructed on Phosphate-Related Land," The Natural Radiation
    Environment III, T.F. Gesell and W.M.  Lowder,  eds., US DOE Rept.
    CONF-780422, p.  1457.

Ha73  Handley T.H. and Barton C.J. , 1973,  "Home Ventilation Rates:  A
    Literature Survey," US DOE Rept. ORNL-TM-4318.

Ha78  Harley J.H., 1978, "Radioactivity in Building Materials," Radioactivity
    in Consumer Products, A.A. Moghissi,  et al.,  eds., Pergamon Press,  New
    York, 336.

Ha78a  Harley J. H., 1978, "Radon-222 Measurement," Regional  Baseline Station,
    Chester, NJ, US  DOE Rept. EML-347.

He79  Hendricks, D.W., 1979, "Elemental Phosphorus  Plants," US EPA Office  of
    Radiation Programs, Las  Vegas Facility, Las Vegas, NV,  unpublished

Hu56  Hultqvist B.,  1956, "Studies on Naturally Occurring Ionizing Radiations,
    with Special Reference to Radiation Doses in  Swedish Houses of Various
    Types," Kungliga Svenska vetenskapsakademiens hadlingar,  fjarde serien,
    Band 6, nr 3,  Stockholm.

Ka79  Kahn B., Eicholz G.G., and Clarke F.J., 1979,  "Assessment of the
    Critical Populations at  Risk Due to Radiation Exposure in  Structures,"
    Report to Office of Radiation Programs, US EPA,  NTIS No.  PB 81-103764.

Le81  Lepman S.R. , Boegel M.L.,  and Hallowell C.D.,  1981,  Radon:   A
    Bibliography,  USDOE Rept. LBL-12200.

L178  Lloyd L., 1978, Montana Dept. of Health and Environmental Sciences,
    Helena, Montana, personal communication.

Ma78  Maxwell R., Belvin E.,  and Reed R., 1978, TVA, Muscle Shoals,  Alabama,
    personal communication.

Mo76  Moeller D. W. and Underbill D.W., 1976, "Final Report on Study of the
    Effects of Building Materials on Population Dose Equivalents,"  Harvard
    School of Public Health,  Boston, Mass.

NC75  National Council on Radiation Protection and Measurements, 1975,
    "Natural Background Radiation in the United States," NCRP Rept.  No. 45,
    Washington, D.C.

Pe80  Pensko J. and Stpiczynska Z., 1980, "Emanating Power of Rn-222 Measured
    in Building Materials," The Natural Radiation Environment III, T.F. Gesell
    and W.M. Lowder, eds., US DOE Rept. CONF-780422, p.  1407.

Pe78  Peterson B.H., 1978, Idaho Dept. of Health and Welfare, Boise,  Idaho,
    personal communication.

OE79  OECD Nuclear Energy Agency, 1979, "Exposure to Radiation from the
    Natural Radioactivity in  Building Materials," Organization for Economic
    Cooperation and Development, Paris, France.

Ta80  Task Force on Radon in  Structures,  1980,  Report of the Task Force on
    Radon in Structures,  U.S. Radiation Policy Council RPC-80-002.

To80  Toth A., Feher I.,  Lakatos S.N., Koszarus L.,  and  Kezthelyi B.,  1980,
    "Distribution of Natural  Radioactive Isotope Concentrations and  Exhalation
    Factors Measured on Concrete and Brick Samples Produced in Hungary," The
    Natural Radiation Environment III, T.F.  Gesell and W.M. Lowder,  eds.,  US
    DOE Rept. CONF-780422, p. 1396.

Tr79  Travis C.C., Watson A.P. ,  McDowell-Boyer  L.M., Cotter S.J. , Randolph
    M.L. ,  and Fields D.E., 1979, "Natural and Technologically Enhanced  Sources
    of Radon-222,  Nucl. Safety 20,  722.

UN77  United Nations Scientific  Committee on the Effects of Atomic Radiation,
    1977,  "Sources and Effects of Ionizing Radiation," United Nations,  New

US80  U.S. Radiation Policy Council,  1980, "Report of the Task Force on Radon
    in Structures," Rept. RPC-80-002.

Wi72  Wilkening M.H.,  Clements W.E.,  and Stanley D., 1972,  "Radon-222 Flux
    Measurements in Widely Separated Regions,"  The Natural Radiation
    Environment II, J.A.S. Adams et al.,  eds.,  AEG Rept. CONF-720805.


                                Gustav Akerblom

                          Geological Survey of Sweden
                                    Box 801
                             S-951  28 Lulea, SWEDEN

    In Sweden, radon daughter concentrations of  200-800 Bq m~3  (0.05-0.2 WL)
are known to occur in some buildings  constructed of areated concrete which
have been manufactured from alum shale,  a  Cambrium, uranium-rich black shale.
As a result of these alarmingly high  levels, the Swedish government set up, in
1979, a Commission whose task is to initiate research around the problem of
radon, and to recommend remedial measures  against natural radiation in

    Based on the recommendations of the  Commission, the Government proposed
the following measures:

         the introduction of provisional limits  for permitted radon daughter
         concentrations in dwellings. These limits are 400 Bq m~3  (0.11 WL)
         for existing buildings and 70 Bq  m~3  (0.02 WL) for new development.

         the use of specific building techniques when developing areas with
         high soil gas radon contents.

         an immediate search for all  buildings constructed of alum shale based
         on aerated concrete.

    About 300,000 houses in Sweden are constructed entirely or partially of
aerated concrete containing alum shale.  During  1980, radon daughter levels
have been measured in 20,000 of these houses by  Local Health Authorities.  The
first results show that about 14% of  the measured houses have radon daughter
levels exceeding 400 Bq m~3.  In about 2%  of the investigated houses the
value exceeded 1,000 Bq m~3 (0.27 WL), and maximum values of 4,000 to 9,000
Bq m~3 (1.1 -2.4 WL) have been measured  in some  houses.

    Radon daughter levels exceeding 800  Bq m~3 are not caused entirely by
radon emanating from building materials.  A contribution of radon from the
ground is necessary.  Most of the houses with values exceeding 800 Bq m~3
have proved to be sited on ground containing alum shale, uranium-rich granite
or uranium-rich pegmatite or on eskers.  Many thousands of houses in Sweden
are built in such situations.

    The measurements so far obtained indicate that the problem of  radon
penetrating into buildings from the ground is a far greater  problem than  that
of radon emanating from building materials.   Geological and  geophysical
investigations of risk areas are of the utmost importance if all new
development is to comply with the provisional limits for radon daughter levels
proposed by the Radon Commission.


    Approximately 300,000 dwellings in Sweden are constructed, either
partially or entirely, of radioactive aerated concrete.  This concrete is
manufactured from alum shale, a uranium-rich (U 50-350 ppm), Upper Cambrian,
black shale which is sometimes sufficiently rich in organic  matter that it may
be used as fuel.  Alum shale occurs as a horizontal layer which varies in
thickness from 2-75 metres.  The distribution of the alum  shale  occurrences
in Sweden is shown in Fig. 1.

    The manufacture of alum shale base aerated concrete dates back to 1928.
It was known at the time, that the concrete  and the alum shale from which it
was produced, had high uranium contents, but it was not until during the
1950's that the suitability of alum shale as a building material was
questioned (Hultqvist, 1956).

    During the 1970's, high indoor radon levels were measured in houses built
of alum shale based aerated concrete, but manufacture of this material
continued until 1975. Indoor radon measurements made by the  National Institute
of Radiation Protection during the 1970's revealed that radon levels in houses
increased in connection with energy saving measures.   Houses were  being made
more air-tight and rates of ventilation were being lowered.   In some of the
houses investigated by the Institute, radon  levels of 200 -  800 Bq m~3
(Becquerel per cubic metre) were encountered (Swedjemark, 1977, 1978 a and b,
1980) .

    A new dimension to the radon problem became apparent in  1978 when the
National Institute of Radiation Protection measured radon levels in houses
built of normal materials on a foundation of alum shale tailings.   The
measured levels were found to be between 400 and 1,600 Bq m~3 (Swedjemark,
et al 1979).


    In February 1979, the Swedish Government appointed a Commission to
investigate measures against radiation hazards in buildings.  The  Commission
was also instructed to initiate a search for all such buildings, and to
promote research around the radon problem.  The Commission is assisted by
various experts, for example, from the National Institute of Radiation
Protection,  the National Board of Health and Welfare, the National Board  of
Urban Planning, the Geological Survey of Sweden, the building industry and the
Swedish Association of Local Authorities. A preliminary report from the
Commission was presented in May 1979 (Radonutredningen 1979).   The report
contains proposals on:  provisional limits for the maximum radon daughter
levels in dwellings;  maximum permitted radium content in building  materials
and gamma radiation from these materials; searches for houses with high radon
daughter levels; and research programs.

                       Figure 1
                                             24° E
                                      GULF OF BOTHNIA
                                         ALUM SHALE
                                       a KNOWN AREAS
                                       I OF RADIOACTIVE
                                         GRANITE >25,uR/h
                                        SGU 1981

    A further proposal made  by  the  Commission  in  its report was that maps,
termed GEO-radiation maps, should be  prepared  to  show the distribution in
Sweden of all areas with particularly radioactive rock and soils  (alum shale
and uranium-rich granites and pegmatites).   A  full description of these maps
is given in Wilson (1981).


    Acting on instructions from the Government, the National Board of Health
and Welfare and the National Board  of Urban  Planning in consultation with  the
National Institute of Radiation Protection have issued provisional regulations
based on the proposals presented by the  Radon  Commission  (Socialstyrelsen
1980, Statens planverk 1980).  In short,  the regulations are as follows
(complete text is given in Appendix 1.):

    Existing houses

    400 Bq m~3  (0.11 WL) is  a provisional action  limit for the annual
average of the equilibrium concentration of  radon.  Houses with levels
exceeding 400 Bq m~3 are declared to  be  insanitary.

    After remedial action or building alterations, the annual average of the
equilibrium equivalent concentration  of  radon  shall not exceed 200 Bq m~3.

    New development

    70 Bq m~3  (0.02 WL) is the  maximum permitted  limit for the annual
average of the equilibrium equivalent concentration of radon in rooms which
are in continual use.

    50 yR/h  (micro Roentgen  per hour) is the maximum permitted  gamma radiation
for rooms in continual use.

    Building materials

    Building materials for  use  in continuously inhabited constructions shall
not have a gamma index or radium index exceeding  1.0.  It  is also recommended
that gamma radiation shall  not  exceed 100 pR/h for outdoor areas  which are
in regular use, for example, playgrounds.
The gamma and radium indexes are defined as follows:

               CK           CRa
Gamma index = 10,000  +    1,000  +

Radium index =  cRa

where CR, CRS and Cjh are the concentrations of potassium-40,  radium-226
and thorium-232 expressed in terms of Bq m~3 of building material.

    70 Bq m-3 (0.02 WL),  the maximum permitted  level  for new development, is
a functional requirement  implying  that  technical measures may be necessary
against all sources of radon gas,  for example,  radon  in the ground, in
building materials and in household water  supplies.   This regulation
supercedes the proposal from the Radon  Commission  that, in areas where gamma
radiation in the ground at foundation level  exceeds 30 uR/h, houses shall
either be constructed "radon safe" or the  site  investigated to determine
whether a "radon safe" construction is  necessary.


    Searches for houses with high  radon daughter levels have been in progress
since the spring of 1979.  Initially the search was aimed at finding houses
constructed with alum shale based  aerated  concrete, but more recently
attention has been focused on the  search for houses where high indoor radon
levels are related to high soil radon concentrations.

    The search for houses constructed with radioactive concrete faced certain
problems.  As house deeds generally contain  no  information as to the type of
concrete used, most houses in the  country  need  to  be  visited in order to
determine whether or not  they are  constructed of radioactive concrete.

    The searches have been the responsibility of the  Local Health
Authorities.  These authorities have relied  upon information from householders
and measurements of gamma radiation using  handborne scintillometers or
carborne scintillometer measurements (Wahren et al 1979).  The latter method
has proved to be very effective permitting control of radiation from the outer
walls of 600 to 1,000 houses per day.

    The carborne measurements carried out  by the Geological Survey of Sweden
use a scintillometer with a sodium iodine  (Nal) crystal which has a volume of
6,130 cubic centimetres.   The instrument is  mounted in a car with a shield
against radiation from the ground  in order to reduce  the background radiation
contribution (Fig. 2). A  follow-up program of measurements of indoor radon
daughter levels has commenced for  those houses  which  are judged to have a

    To date (Spring 1981), radon daughter  measurements have been performed in
approximately 20,000 houses.  Most of the  measurements have been made using
alpha-sensitive film.  On a more limited scale, measurements have been made
using the filter method or thermo-luminiscence  detectors  (TLD).

    For most of the film  measurements,  Track Etch  films have been used.  The
films (2 per house) have  been used without radon daughter filters.  The
measuring period has generally been 3 months, summers preferably being
avoided.  The results of  the two measurements for  each individual house have
been converted to give an average  annual radon  daughter concentration
(Samuelson, 1980).


    The first compilation of measurements  of radon daughter concentrations
from 5,600 houses built of radioactive  aerated  concrete gives the following
results:  46% of the houses have levels exceeding  200 Bq m~3  (0.05 WL), 14%

Figure 2

levels exceeding 400 Bq m~3 (0.11 WL)  and 2%  levels  exceeding 1,000 Bq m 3
(0.27 WL) (Hildingsson, 1981).   A more recent compilation  for 12,000 houses
gives very similar results:  48% with  levels  exceeding  200 Bq m~3, 13%
levels exceeding 400 Bq m~3 and 1% levels exceeding  1,000  Bq m~3
(Hildingsson, pers. comm.).

    Clear geographical variations in the radon daughter levels are apparent
from the results so far obtained.  These variations  are caused, either by
variations in the radium content of the aerated concrete used, or by
variations in the soil gas radon concentrations, the latter being closely
related to the geology.  Variations in the radium content  of the aerated
concrete are due to the fact that this type of concrete was manufactured in
different parts of Sweden using alum shale with widely  varying uranium

    There is a marked increase  in the  number  of houses  with high indoor radon
levels in areas where the bedrock consists of alum shale or uranium-rich
granites and pegmatites (Table  1).  It has been found,  however, that sporadic
occurrences of high indoor radon levels also  occur in areas with normal
uranium contents in the bedrock or soils.  Enchanced soil  gas radon levels
have been noted in glacial eskers with normal uranium contents of 3-8 ppm U in
the soil and in areas with scattered occurrences in  uranium-rich pegmatites.

    The initial results of the  program of indoor measurements show clearly
that the problem of radon emanation from the  ground  is  greater than that from
building materials.  At a rough estimate, 3,000 - 15,000 houses in Sweden are
in need of remedial measures against radon penetration  from the ground.
Tracing these houses will be difficult.  One  way would  be  to carry out indoor
measurements in all houses built in areas known to have high uranium contents
in the bedrock or soil cover.  Tracing of the sporadic  ocurrences is however a
problem.  It should also be borne in mind that exceptionally high soil gas
levels exist in houses which are ventilated in such  a way  that large
quantities of air are drawn into the building through cracks, etc. in the
foundation or cellar walls.


    A research program, initiated by the Radon Commission, is underway to
investigate the relationship between radon concentrations  and uranium content
in the ground and indoor radon  and gamma ray  levels.  100  houses with high
indoor radon daughter levels from areas of alum shale or uranium-rich granite
have been selected for detailed studies of the bedrock,  soils, groundwater and
building techniques.  These houses are not constructed  from radioactive
aerated concrete.  The investigations  will include in situ determinations of
the U-, Th- and K-content of the rocks and soils using  gamma spectrometers,
gamma logging, measurements of  the soil gas radon content  and/or radon
emanation from the surface, measurement of radon content in the groundwater
and in the drainage layer under the houses, and measurements of radon daughter
levels in the houses.  The results of  the investigations should be available
in 1982.

    Other research programs are also in progress where  various types of
remedial measures are being studied, including means of ventilation and the
laying of aluminum foil on the  floor to prevent radon penetration from the


(houses and
Building Radon daughter levels, Bq m~3, in the investigated Geology
material houses and apartments
% > 200 % > 400 % > 100 no. houses investigated
aerated 3.5 1.0 516 Till, sand, gravel and clay overlying
concrete granodiorite and acid to basic
volcanics (U in bedrock 1-8 ppm)
"- 79 42 9* 380 Sand, clay and alum shale overlying
sandstone, alum shale and limestone
(U in alum shale 50-300 ppm)
"- 65 42 62 278 Till, sand and clay overlying
gneisses younger U-rich granites and
pegmatites (U in gneiss 2-b ppm, in
granites 10-20 ppm). Well exposed.
other type 90 48 22 31 Alum shale rich till, sand and
of b. material gravel overlying sandstone and alum
shale (U in alum shale 50-300 ppm)
"- 20 10 2 63 Uranium-rich granite, exposed (U
content 12-37 ppm)
1    6 houses 4,000 Bq m~3
     1 house  8,000 Bq m~3

2    1 house  6,000 Bq m~3

Table 1.  Variations in indoor radon daughter levels reflecting the geology.


    In order to comply with the regulations  concerning radon daughter levels
in the new buildings (70 Bq m~3),  one  needs  either  to build houses "radon
safe", or to know that the soil gas  radon  content is so low that no indoor
problem will arise.  In the case of  building on ground with enhanced radon
concentrations, the type of "radon safe" construction used may be varied
according to the actual soil gas radon levels.

    The purpose of the site investigations is to determine whether or not the
soil gas radon concentration within  a  planned area  will give rise to high
indoor levels once the houses are  built.

    Account must be taken of the changes in  the soil gas radon concentrations
which will be caused by the actual process of building, for example, lowering
of the water table or removal of some  of the soil cover so that the
foundations lie at a deeper level.

    For the site investigations carried out  by SGU, soil gas radon
measurements are made at a depth of  one metre in order to minimize temporary
fluctuations caused by variations  in wind, temperature and precipitation
without increasing the measuring costs to  unacceptable levels.  In conjunction
with the radon measurements, observations  are made  on soil types, bedrock
geology, ground water and gamma radiation  both above the ground and in the

    The methods used by SGU are ROAC-cups  which are filled with activated
charcoal (Hambleton-Jones and Smit,  1980).  Track Etch alpha sensitive film
(Gingrich and Fisher, 1976) and emanometers.   The measuring time for ROAC-cups
is 5-7 days, and for Track Etch film 3 weeks. These methods were tested
during 1980 in a research project  initiated  by the  Radon Commission (Hesselbom
et al, 1981).  Measurements were performed in both  granite and alum shale
environments.  The ROAC and Track  Etch cups  gave comparable results, but the
emanometer readings varied, compared with  ROAC and  Track Etch, from area to
area probably because varying amounts  of atmospheric air were pumped into the
instrument.  The amount of interference from atmospheric air will depend on
the porosity and permeability of the soil  in which  the measurements are made.
The advantage of the emanometer method is  that it is quick, and it gives an
idea as to whether radon levels in the soil  are enhanced or not.  Fig. 3 shows
the results of test measurements carried out by SGU in an area of alum shale.

    The bedrock along the profile  consists of alum  shale  (U 150 - 300 ppm)
overlain to the west by limestone  (U less  than 2 ppm).  The soil cover is 5-10
metres thick and consists of clay  to sandy till, and in the central part of
the profile of clay.  The water table  lies 2-3 metres above the bedrock
surface.  In the western part of the profile the till is comprised of
limestone fragments, alum shale fragments  are absent.  This also applies to
the upper 4 metres of the till in  the  eastern part  of the profile.  Here the
content of alum shale fragments increases  successively with depth.

 ROAC units


                                     Figure  3
                      ROAC  (activated charcoal)
                                                              50 E
                                                                                                  150 E




                                                                                100 E
                                                                                                   150 E
                      EMANOMETER  Re 279
                        »  day 1
                        •  day 3
                                                                                                  150 E


                        • uranium
        100 W
                                                                               100 E
                                                                                                   150 E
        •* _»-9if>»i:PfiQ-* _ • /^•^»-o->__QVr;-?.-Jf(--i-i>j!Tx?

       )  «tj U • 'r^C OQ, •  • •M«*5W;-X»'O*
                                                                                                   150 E

    The radon measurements show that the  radon  concentration in the soil for
the western part of the profile is low, approximately  3,000 to 30,000 Bq
m~3.  This can be related to the low uranium content of the till combined
with the shielding effect of the limestone  horizon, the till cover and the
high water table all of which prevent radon from  the alum shale horizon from
reaching the surface.  In the eastern part  of the profile the shielding effect
is no longer present and the soil gas radon content is approximately 120,000 -
160,000 Bq m~3.

    The results show a good correlation between the measurements performed
with ROAC and Track Etch cups and the emanometer.

    None of the methods used by SGU are calibrated to  give absolute
concentrations of radon in the soil in Becquerels per  cubic metre.  Track
Etch, however, quote an average value of  approximately 6,000 Bq m~3 for all
soil gas radon measurements performed by  their  method.  Of the measurements
carried out by SGU at a depth of one metre  in till or  gravel containing alum
shale, values some 30 to 300 times larger than  the average value quoted by
Track Etch have been obtained.  For till  or gravel containing fragments of
uranium-rich granite, values 25 to 75 times the average Track Etch value have
been measured.  In other words, soil gas  radon  levels  in areas of alum shale
can be as high as approximately 500,000 - 1,800,000 Bq m~3, and in areas of
uranium-rich granites 200,000 - 500,000 Bq  m~3.   With  such high soil gas
radon levels, intake of more than 70 litres per hour of soil air  (radon
content 500,000 Bq m~3) into a house with a ventilation rate of 0.5 air
exchanges per hour will be sufficient to  produce  indoor radon daughter levels
exceeding 70 Bq m~3.

    Fig. 4 illustrates an area investigated by  SGU in  the county of Narke,
central Sweden.  The purpose of the investigation was  to determine the radon
content in the soil air in an area planned  for  development.  The investigation
included radon measurements using both an emanometer and Track-Etch film,
gamma-radiation measurements with a handborne scintillometer, and
gamma-logging of drill holes.

    The bedrock of the area consists of alum shale which is overlain to the
west and north by limestone.  The bedrock is covered by till and sand.  In the
eastern part of the area, layers of clay  occur  interbedded in the sand.  The
surface of the till lies at 1-3.4 metres  depth.   In the northern and western
parts of the investigated area, the till  is dominated  by limestone fragments,
whereas in the southern and eastern part  alum shale fragments dominate.  The
water table lies at a depth of 0.4-1.1 metres.  Tailings of burnt alum shale
(rodfyr) have been tipped in the northern part  of the  area.

    The gamma radiation measurements made at ground level gave low readings of
6-11 yR/H, except for the area with rodfyr  where  the radioactivity is 90-115
yR/h.  The gamma-log results also gave low  readings of 3-10 yR/h with a
slight increase to 14-19 yR/h for soils containing a few fragments of alum
shale.  It should be noted, however, that the drill holes only penetrated the
upper soil layers down to the upper surface of  the till.  The readings
therefore do not apply to the radiation levels  of the  till.

< 15 000 Bq m3
> 15 000 Bq m3
>100000 Bq m3
High  water table
Drill holes
Emanometer  measurements

Track-Etch measurements
                                                                                           Investigated area

                                                                                           Planned  development
                                                                                                    150 M

    The radon measurements were made at the  greatest possible depth  (0.9
metres for Track Etch and 0.7 metres for the emanometer) but the high level of
the water table often prohibited this.

    On the basis of the results of the  radon measurements, the investigated
area can be subdivided into three (Fig.  4).   In  the north and northwest where
the till is dominated by limestone fragments, the  radon concentration in the
soil gas is normal to slightly raised.

    In the southern and eastern parts of the area, where the till is dominated
by alum shale fragments, the soil gas radon  level  is higher than normal,
30,000 - 100,000 Bq m~3.  Within the latter  area however, lower values were
obtained where the water table was very near the surface.  The high ground
water level hinders the radon gas from  the alum  shale till from reaching
ground level.  If, however, the water table  is lowered, either naturally or as
a result of construction work, the content of radon in the soil gas will
increase in these areas.  Soil gas radon levels  in the areas of tipped rSdfyr
exceed 100,000 Bq m~3.

    In the areas with high levels of soil gas radon, the planned dwelling
should be so constructed as to prevent  infiltration of radon from the ground
into the dwellings.

Gingrich,  J.E.  and Risher,  J.C.,  1976.   Uranium  exploration  using  the  Track-
    Etch method.   IAEA-SM-208/19,  213-225.

Rambleton-Jones,  B.B.  and Smit, M.C.B.,  1980.  ROAC  -  a  new  dimension  in radon
    prospecting.   Atomic Energy Board, Pretoria,  South Africa.   ISBN 0 86960
    7081,  24 pp.

Hesselbom, A.,  Israelsson,  S.  and Tovedal,  H., 1981.   Radon  in  the ground.   A
    study of methods and instruments  for determining radon concentrations in
    the ground.  Byggforskningsradet,  Report  R47:1981  (in Swedish), 75 pp.

Hildingsson, 0.,  1971.  Radon  daughter levels in 5,600 houses.   Measurements
    made using  alpha sensitive film and  filter method.  Technical  Report,
    Statens Provningsanstalt,  SP  Rapp. 1981:27,  (in  Swedish).

Hultqvist, B. 1956.  Studies on naturally occurring  ionizing radiations with
    special reference to radiation doses in Swedish  houses of various  types.
    Svenska vetenskapsakademins handlingar, fjarde serien, Band 6, nr  3,
    Stockholm,  Sweden.

Radonutredningen, 1979.   Preliminary  proposal for measures against radiation
    hazards in  buildings.  Report from the  Radon Commission, Jordbruksdeparte-
    mentet, Ds  Jo 1979:9 (in Swedish), 114  pp.

Samuelson, I. 1980.  Determination of the annual average radon  daughter con-
    centration  in a house measured with  alpha sensitive  films.   Statens
    Provningsanstalt, SP A2 601,  8 pp.

Socialstyrelsen,  1980.  Statement from the  National  Board of Health and Welfare
    with advise and regulations concerning  remedial  measures against radon in
    dwellings,  September 1, 1980.   Socalstyrelsens forfattningssamling, SOSFS
    (M) 1980:71,  4 pp, (in Swedish).

Statens Planverk, 1980.   Svensk Byggnorm 1980 (Swedish Building Code 1980).
    Statens planverks forfattningssamling 1980:1 (in Swedish).

Swedjemark, G.A., 1977.   The ionizing radiation  in dwellings related to the
    building materials.   Statens  stralskyddsinstitut.  Report 381:1977-004, 1-9.

    1978a.  Radon in dwellings in Sweden.  In Proceedings from  the Symposium
    on Natural  Radiation Environment  III, Houston, Texas, April 23-28, 1978.

    1978b.  The effects  of energy saving reduced ventilation on radiation dose
    in dwellings.  Statens stralskyddsinstitut,  Report SSI:1978-014, 1-13 (in

    1980.   Radioactivity in houses built of aerated  concrete based on  alum
    shale.  Statens stralskyddsinstitut, Report  SSI:1980-14, 1-6.

Swedjemark, G.A., Hakansson,  B.  and Hagberg,  N.,  1979.   Radiation levels  in
    houses built on wastes from  processing  of alum shale.   Statens
    stralskyddsinstitut, Report  531:1979-006, 1-43 (in  Swedish).

Wahren, H., Makitalo, A., Persson.  T.  and Svensson,  C.-E,  1979.  An attempt to
    trace 'radon houses' in the  municipality  of Uppsala.   Delrapport 1979,
    Lanslakarorganisationen i Uppsala  Ian,  62 p,  (in Swedish).

Wilson, C., in press.  Regional  environmental documentation of natural radia-
    tion in Sweden.

               Appendix:  Summary of provisional regulations and
       recommenda.tions concerning radiation hazards in buildings in Sweden
1.  Directions issued by the National Board of Health  and  Welfare,  in  force
    from September, 1980.

    A building is regarded as insanitary when  the  radon  daughter concentration
    in rooms in continual use exceeds an annual average  of 400 Bq m~3.
    SOSFS(M) 1980:71 (Socialstyrelsen 1980).

2.  Mandatory regulations issued by the  National Board of  Urban Planning,
    SWEDISH BUILDING CODE (SEN)  1980, in force from  January 1, 1980.   National
    Board of Urban Planning Statute Books 1980:1 (Statens  planverk  1980).

New buildings

    Buildings shall be so constructed that the annual  average concentration of
radon daughters in rooms in continual use shall not  exceed 70 Bq m~3.

    The annual average concentration of  radon  daughters  is determined
according to methods laid down by the National Board of  Urban Planning in
consultation with the National Institute of Radiation  Protection.

                                  SEN 36:41

    The regulation relating to a maximum permitted indoor  radon daughter
concentration of 70 Bq m~3, according to 36:41,  can  require  the application
of remedial measures against sources of  radon  in the ground, in household
water supplies and in building material.

                                  SEN 31:142

    A building shall be so constructed that gamma  radiation  in rooms in
continual use shall not exceed 50 yR/h.

                                  SEN 31:141

    A satisfactory level for gamma radiation is  achieved if  the regulation
relating to building materials,  31:143,  is applied,  and  if gamma radiation
from the ground is screened by a concrete structure  or filling of low
radioactive material.

                                  SEN 31:1411


    Buildings shall be so constructed that  the  annual  average concentration of
radon daughters in rooms in continual use shall not exceed  200 Bq m~3-
Exception is made where the radon daughter  concentration cannot be brought
below this level by means of increased ventilation and remedial measures
according to 31 rebuilding: 14.

                                  SEN 36: Rebuilding:   41

    Exception to the regulation  quoted in SEN 31:141 may be granted.   (That is
to say, there are no regulations which apply to lowering of gamma radiation
levels in cases of rebuilding.)

                                  SEN 31: Rebuilding:   141

    Exception to the regulation  quoted in SEN 31:143 regarding existing
building materials, may be granted.   However, an accessible, strongly radon
emanating filling material shall be  removed where required  in order to conform
with the regulation concerning radon daughter concentrations quoted in SEN 36
Rebuilding:  41.

    Easily accessible filling may be found  against cellar walls, and in the
framing of joists and the crawl  space.

                                  SEN 31: Rebuilding:   143

Building materials

    Building materials used in buildings in continual  use shall not have gamma
or radium indexes exceeding 1.0.

                                  SEN 31: 143

    Filling materials and material used in  the  drainage layer under the
foundations also come under the  heading building materials.

                                  SEN 31: 1431

    Gamma- and radium indexes are defined as follows:

               CK           CRa
Gamma index = 10,000  +    1,000

Radium index =
where CR, CRa and C-pn are the concentrations  of  potassium-40, radium-226
and thorium-232 expressed as Bq kg"1 of bulding  material.

Out of doors

    There are no regulations which apply to gamma radiation  levels out of
doors.  However, it is recommended that in areas which  are in regular use,  for
example playgrounds, gamma radiation should not  exceed  100 yR/h.

                                Figure Captions
1.  Distribution of the  alum  shale formation and of areas known to contain
    radioactive granites.

2.  Car and equipment  used  in the search for houses with aerated concrete.

3.  Results of  test measurements made in an area of alum shale (after
    Hesselbom et al 1981).

4.  Map showing the results of radon measurements in an area planned for

                           EFFECT OF AIR CIRCULATION
                          AND DUST REMOVAL ON INDOOR
                     S.N. Rudnick, W.C. Hinds, E.F. Maher,
                          J.M. Price, and M.W. First

                  Department of Environmental Health Sciences
                               Harvard University
                            School of Public Health
                             665 Huntington Avenue
                         Boston, Massachusetts  02115

    The  effectiveness  of  increased  room  air  motion  and  dust   removal  in
reducing working levels in  residences subject to radon intrusion was evaluated
in  a  90-m3 chamber  under steady-state  conditions for  air infiltration rates
between 0.2 and  0.8/hr.   Room-sized electrostatic precipitators and high-effi-
ciency  fibrous  filters  were  tested  as  typical   residential  air  cleaning
devices; a portable  box fan and a ceiling fan were  employed  as typical resi-
dential air  movers.  Reductions  in  working levels  from 40 to  90% were mea-
sured.  The  fate of  radon  decay products  was  determined  by  direct measure-
ment.  When mixing fans were  used, most of  the potential alpha energy deposit-
ed  was plated out on  the  room surfaces; less than 5% was deposited on the fan
blades.  Results were  compared  to a mathematical  model  based  on well-mixed
room and good agreement was obtained.


    Strategies to  control radon decay  products  in  residences  can be divided
into three types:   (1)  prevention of radon entry into homes  (e.g., caulking,
sealing, barrier  paints,  and  judicious selection of building materials);  (2)
dilution of outside  air directly or  through  heat exchangers;  (3)  removal of
radon  or  radon  decay products from  indoor atmosphere  by  various air  treat-
ments.   Historically,  strategy  (2)   has  been  most   widely  used,  perhaps
inadvertently.  In recent years,  it has become  less acceptable because  of the
high cost of energy for space heating  and  cooling.

    The present  study was  directed  to  strategy  (3) ,  although  no  attempt was
made  to cleanse the  air  of  radon  gas.   The  treatment methods evaluated
included the  use  of room-sized  electrostatic  precipitators, high-efficiency

air filters,  and  air mixing by  fans.   Because much of the  data  collected has
not been analyzed, this paper should be  considered  a  first reporting of on-go-
ing research.  More complete results will be available in three months.(1)

    Very little has  been published  on  the use of  methods to  remove airborne
radon decay products.  The  experimental  work  that has been reported, concerned
primarily  with  the  reduction   in  radon decay   product  concentrations  in
mines, (2~8)  has  only  limited   application  to control  in  residences  because
of  the  higher dust  concentration,  higher air  infiltration rates,  and higher
humidity found in mines.


    Experimental  studies  were conducted in  a  90-m-^  chamber,  about  the  size
of  a  large  room  in  a house.   Special  attention  was  given  to making  this
chamber as much like a residence as  possible, although certain differences are
noteworthy.   The  chamber's  4-m  high ceiling  is not  characteristic  of  houses,
nor are its  painted  metal walls.   We  do  not  believe,  though,   that  these
differences are  important.   Transient  effects  in  houses,  such as  changes  in
air infiltration or  radon  intrusion rates with time,  however,  are  undoubtedly
important,   but  we made  no attempt  to  simulate  this behavior.   In fact,  to
facilitate  interpretation of  our experimental results, we  held all parameters
within  our control  constant,  and  no  data were  retained  for analysis  until
steady-state conditions had been reached.

    Air Infiltration

    The natural air  infiltration rate into our experimental chamber was  very
low;  from  the decay rate of  a  CO  tracer  gas,(l)  we measured  about 0.015/hr.
Because we wanted to hold the air infiltration rate  constant  at  various  sett-
ings and still  simulate air infiltration into a room  of a  house, a blower was
installed  outside  the  chamber to  exhaust air  from it.   (See  Figure  1.)   No
inlet air  port  was provided and,  thus,  incoming  air was  forced  to infiltrate
through the  cracks  and pores in the walls or  the  seals  around  the door  and
windows.  Pressure in the chamber was never more than 3-mm Hg below barometric
pressure and  was  usually  less.   The exhaust  airflow  rate  was  measured with  a
calibrated  venturi  flowmeter,   and  the  rate  controlled  with  a  butterfly
damper.  The  air  infiltration  rate was equal to  the measured  airflow  rate
divided by the chamber volume.   The  particulate level and  size distribution of
the infiltrating air was  about  the  same as  laboratory air,  although somewhat
variable,  and is reported  elsewhere.(1)

    Radon Generation

    Radon-222 gas  was  generated  by de-emanation  of  a   100-|iCi  solution  of
Ra-226.  About  0.2  L/min of  prehumidified air flowed continuously  through  a
bubbler containing  the  radium solution,  through  a  liquid carryover trap  and
membrane filter, and into a distribution manifold  placed  on the floor of  the
chamber to simulate radon emanating from the  subsoil  and  entering  through the
floor-  This is  shown schematically in Figure 1.  The manifold was  constructed
from 3.5-cm I.D. plastic pipe in which  0.6-mm holes were  drilled.  Inasmuch as
the holes  are much  smaller than  the pipe diameter,  gas  is evenly  discharged

from  all  holes.  These  holes face downward  and  are located on  60-cm  centers
throughout  the  chamber.   About  5  L/min  of  dilution air  was also  introduced
into  the distribution manifold.

    Sampling and Measurement of Radon Decay Products

    As shown in Figure 1,  sampling ports  were provided  in the exhaust duct and
on  one wall  of  the  chamber.   Both  ports  accept  50-mm  diameter  open-faced
filter holders, which  minimize aerosol sampling  losses.  To  determine  whether
these samples would be representative  of  the contents of the chamber,  we  made
tests  in which  a  tracer  gas  (CO)  was  introduced  continuously through  the  dis-
tribution manifold.  After steady state  had  been reached,  samples were  taken
from  the  two sampling ports  as  well  as from eight  other locations at  various
heights in  the  chamber.   All  samples  were  in  reasonably  good agreement  over
the  range  of   air  infiltration  rates we  studied.CD   Further  verification
that  the chamber was well mixed can be  made  from  a  comparison of  samples  taken
from  the  exhaust  duct and wall port.  For  25 pairs  of samples, the  exhaust
duct  gave an average  of  0.0907 working level  (WL)  compared  to  0.0890  WL  from
the wall port, about a 27, difference.

    Filter  samples  were  counted  by  alpha  scintillating   using disposable
Ag-activated  ZnS  phosphors  coated on  one  side  of  a Mylar  film (William B.
Johnson  &  Associates,  Inc.,  Research Park,  Montville, New  Jersey   07045),
which  were   placed  in direct  contact  with  a  bare photomultiplier  tube  and
enclosed  in a  light-tight box.  Working  level  was  determined by  the modified
Tsivoglou method.C9)


    Because  samples  taken from various locations in the experimental  chamber
showed good agreement, we  believe that  a material  balance for  radon decay
products using a steady-state, well-mixed model should give  reasonable  predic-
tions of working level.  When incoming  air is essentially free of  radon or  its
decay  products, as  was  found for our  experiments,  the  following equation is
obtained: CD
         WL =
              (Xr + I)(xa + I + R)
where   WL  =  potential  a-energy  concentration  of  short-lived  radon decay
               products in air in units of working level

        S   =  radon exhalation rate per room volume  in Bq/(hr*nP-'

        I   =  air infiltration rate in 1/hr

        X   =  radioactive decay constant (Xr =  0.00758/hr,
               Xa = 13.7/hr,  Xb = 1.55/hr,  and Xc  = 2.11/hr)
=  dimensional conversion factors (3.79 X 10~
   2.13 X 10~4» respectively)

=  removal rate for particulate matter; 1/hr.

Approx. Scale:
1 cm = 0.36 m l"
, )
^ N
I* '
{ /
y S
s y
^ N
V /
/ \*
(^ X1
X \
< /
C >
X. I—I


X— V
Qll -5c
Vf K '


- Denotes location of 0.024-in diameter holes in
4-in diameter pipe used for ditribution of radon.

< >
f N,
S /
t/ N
A x
X \
S '
» »
^ X

^ s

' N
: ;
X \
\ x3
' N
^ s
^ y
< \
^ s

s N
•* /
<: >
C >
; >
< >
' x
\ s
f \
c >

< )
c >
: >
: •>
^ )
/ N
s /
< N
^ /

c ;
( >
r >
V ^
^\ /^
^ >
•s r

(. ^

<}' >
' \ '
< 5
c ;
< >
^ N
S /

c >
C3+- Radon Generator


^ Exhaust Ventilation ^ ^ s.mp|,nfl Port (2*
(2.6 ft. above floor) ft abov. f|oof)
              •Sampling  Port

       Venturl Flowmeter

R can be subdivided into various particulate  removal mechanisms,  i.e.,

                                R = Pn + Pf + A                             (2)

where    Pn=   rate  of  plateout  due  to  normal  air  motion  (including that
               caused by air infiltration)

         Pf=   rate  of  plateout due  to forced convection  (e.g.,  from  mixing

         A  =  air  cleaning  rate   (e.g.,   from   filters  and  electrostatic

These terms are not  necessarily independent.  For example, a filter may  remove
particles, while its fan enhances plateout.


    Experimental  protocol  was  the  same for  all  runs.   The  air infiltration
rate was  set,  and,  after  the contents of the chamber  were allowed  to come to
steady-state conditions, various measurements were made, some of which will be
reported  at  a later  time.(1)    Whenever  possible,  the  data  obtained were
compared to predicted concentrations from Equations  (1)  and (2) .

    Effect of Air Infiltration Rate

    A  log-log  plot  of  the experimentally measured  working  level  versus air
infiltration rate is shown in Figure 2.  (The solid line is the  linear regres-
sion  line.)    Also  shown  are predicted values  from  Equation  (1)   when only
plateout due to normal air  motion  is  considered.   The  data follow the general
slope of  these lines,  and a plateout  rate of between  0.25 and 0.50/hr can be
inferred.   (A  background  plateout rate of  0.4/hr  was   assumed  when  the model
was  used  for  prediction  in other  sections  of this paper.)   Verification of
these results by direct measurements of potential alpha energy on the  surfaces
of the chamber were made and will be discussed later.

    Effect of Forced Convection

    The effect of  various  mixing  fans is shown in Figure 3,  a plot of working
level  fraction remaining  after  a  fan  was  started  versus   air infiltration
rate.  Reductions  in WL varied from about 40 to  80%.   Based on Equations  (1)
and  (2) ,  predicted  values  of  WL  fraction remaining  versus  air infiltration
rate, with rate of  plateout due to forced convection as a parameter,  are also
shown in  Figure 3.  From  this  plot,  a rate of plateout due to the fans of
between about  2 and  5/hr  can be inferred.    (This plateout rate  is in  addition
to that obtained from normal air motion.)  The data also show  a  greater  reduc-
tion in WL at  lower  air infiltration rate, although the reason for this  is not

    Manufacturer's specifications for  these fans are given in Table I.

         0.5  r
0.2 -
         0.1  -
       0.05  -
            Regression line for
            experimental data
                                                               RATE, hr-1
                      Well-mixed  model:
                      Eqs.  (1)  and (2).
                      AIR INFILTRATION RATE, hr-1
                       RATE  WITH NO AIR TREATMENT.

z   1.0
                          Regression  line for
                          experimental data
                           RATE, hr-1
                                                               	 Well-mixed model: Eqs.
                                                                    (1)  and  (2).  (Adjusted
                                                                    for  normal plateout rate
                                                                    of 0.4/hr.)
                                                                                       20-in. BOX FAN

O 52-in. CEILING FAN (high speed)
                                            AIR INFILTRATION RATE,  hr-1
                                    FIGURE 3:   EFFECT OF MIXING FANS ON WORKING LEVEL

                                    TABLE I:

Hunter 20-in box fan
(Model 11077)
Hunter 52-in ceiling fan
(Model 22306-7J)
(NEMA Method)
	 • —
Although the  box  and ceiling fans gave comparable  reductions  in WL,  the ceil-
ing fan has significant advantages:  it consumes  less  energy than the box fan,
is  considerably quieter,  is designed  to operate  continuously,  and produces
less  noticeable air motion.  It  also  can be  used all  year round;  the manu-
facturer(lO)  claims  it   "reduces  air  conditioning  cost  during the  cooling
season and reduces wasteful heat stratification during the heating season."

    Effect of Particulate Air Cleaners

    Figure  4,  a plot  of  WL fraction  remaining versus air  infiltration rate,
shows  the  effect  of various particulate-removing  air  cleaners.   Predicted  WL
fractions  remaining  versus  air  infiltration rate  with air  cleaning  rate  as a
parameter are also shown.  (A background plateout rate of 0.4/hr was assumed.)

    The  electrostatic precipitators  (Sears  Model  No.  156.73300 Console  Air
Cleaner),  which we  measured to  give an  air  throughput  rate  of 5/hr  for  our
chamber,  yielded  WL reductions  that  were  in  excellent  agreement  with  the
predicted values, although they assume 100% collection efficiency.  About  half
of this reduction was obtained with  the high voltage  turned off.  Whether this
latter reduction was due  to removal by the unit's  fan and  charcoal  filter  or
by enhanced plateout on the walls is not known.

    A high-efficiency particulate air (HEPA) filter, which  removes essentially
all particulate matter, was also tested at an  air throughput rate of  5/hr.   It
yielded an effect equivalent to  between  10 and 20/hr-  This difference can  be
attributed  to increased  plateout.   Removal  of  particles  from  the  air  would
increase the fraction of  radon decay products  that  are unattached to  particles
and  thus   significantly  enhance  plateout;  air  motion  caused  by  the  HEPA's
blower would also promote plateout.  We noted  other indications  of this pheno-
menon:  limited  data on  the  unattached  fraction(l)  indicated that  about  80%
of the working level was unattached, whereas only about  20% was  unattached for
most other runs.  Extreme disequilibrium, another  indication of  high  unattach-
ed fraction, was also observed.

I  1.0
            Regression line
            for experimental data
           AIR CLEANING
            RATE, hr-1
                            	  Well-mixed model:
                                 Eqs. (1) and (2)
                                 (adjusted for normal
                                 plateout rate of 0.4/hr)
                                        D NORMAL OPERATION
                                        • NO HIGH VOLTAGE
                                   PRECIPITATOR (6.0/hr)
         A  1.1/hr
         O  6.0/hr
                                                     HEPA FILTER



    Paper or aluminum-foil disks, affixed to various  locations on  the  surfaces
of  the  chamber, were  removed from  the chamber  and counted  sequentially  by
alpha  spectrometry.   The  fan  blades were  also removed  from  the chamber  and
selected  areas  counted  simultaneously  by  direct  contact with  six  ZnS  alpha
scintillation counters, which were protected from  room  light by  thin aluminiz-
ed mylar windows.  Counting was  done  over  two  time periods for  alpha  spectro-
metry and three time periods for  scintillation, and  the potential  alpha-energy
areal density  (MeV/m2) was back-calculated  to the  time at which the  samples
were removed from the chamber.^'

    The results of a material  balance on potential alpha-energy  for five runs
are given in Table II.   Contributions  from  the air,  chamber surfaces,  and  fan
blades  and  housing  were  summed and  compared to  the   predicted  values  from
Equation  (1);   between 69% and   97%  of the   predicted  potential  energy  was
accounted for.

    Greater than 95%  of the plateout  was  found to  take place on the  chamber
surfaces.  Holub et  al.^°' found deposition on fan  blades, but none on  walls
for similar experiments.  The reasons for these conflicting results may be  due
to  differences  in  measuring technique.  In our experiments,  less than  5%  of
the  potential  alpha-energy lost  from  the  air was  found on  the  fan  blades,
which had only  0.4%  of  the surface  area of the chamber.  Thus, the potential
alpha-energy areal  density on  the fan blades was greater  than  10 times  that  on
the walls.  This might explain why  the relative  external activity on  the  fan
blades measured by  Holub et al.'s beta-gamma detection  system was  in a  measur-
able range, whereas  the wall  activity  was  not detectable by  their equipment.
Other differences between  Holub  et  al.'s and  our  experiment,  such as  aerosol
size and concentration,  chamber  surface to  volume  ratio, and  infiltration  air
sources, make comparison difficult.


    Various methods  may be used to  reduce working  levels  in residences.  A
ceiling fan, for example,  was shown to give  up  to  75% reduction in  working
level.   Because  it was  most  effective at  low air infiltration  rates,   it  is
probably a good choice for an energy-efficient  house.  It consumes about  150 W
of power and may also  reduce cooling and heating requirements.   Greater reduc-
tions are possible  using HEPA filters,  but higher  energy  consumption and  noise
may  prove  to be a deterrent  to their  use.   Higher fractions  of unattached
radon decay products found when  HEPA filters  were used  may be  less desirable
from a health-effects  standpoint, as the increased hazard commonly believed  to
be  associated  with unattached  radon  decay product  may offset  some  of  the
benefits of lower working levels.


    This work was  supported  by the  U.S. Environmental  Protection Agency  Con-
tract No.  68-01-6250.

                                                TABLE  II
Air Treatment


132-cm ceiling
fan (high speed)
132-cm ceiling
fan (high speed)
Two 51-cm box
fans (high speed)
Rate, Hr"1



4.095xl07 2.808x10?
(2.7%; (2%)


89 . 9%
90 . 2%

Predicted  potential alpha  energy  based  on  a well-mixed  model,  de-emanation  of lOO-yCi  Ra-226  into  a
90-nP chamber, no plateout,  and radon-free infiltration  air.

% of the predicted total potential alpha energy is  shown in  parentheses.

On still blades of 132-cm ceiling fan.

1.   Rudnick, S.N., Hinds, W.C., Leith, D., First, M.W., Maher,  E.,  and Price,
     J.,  "Effect  of  Indoor  Air  Circulation Systems  on  Radon  Decay  Product
     Concentration," Final Report  under EPA  Contract  68-01-6250  (in  prepara-
     tion) .

2.   Goodwin, A.,  "Review of  Problems  and Techniques for  Removal  of Radon and
     Radon Daughter  Products  from  Mine Atmospheres,"  in  Proceedings  of  the
     12th AEG  Air  Cleaning   Conference,  CONF-720823,   NTIS,  Springfield,  VA,
     1973, pp.  378-398.

3.   Coleman, R.D.,  Kuznetz,  H.L., Woolrich, P.F.,  and Holaday,  D.A.,  "Radon
     and Radon Daughter Hazards  in Mine Atmospheres,"  Ind. Hyg. Quart.  17;405

4.   Shreve,  J.D.,  and  Cleveland,  J.E.,  "Effects  of  Depressing  Attachment
     Ration  of  Radon Daughters  in Uranium  Mine Atomosphere,"  Am.  Ind.  Hyg.
     Assoc.  J.  33:304 (1972).

5.   Rock, R.L.,   "Control  of  Radon  Daughters   in  U.S.  Underground  Uranium
     Mines,"   in  Proceedings  of   the  12th  AEC  Air  Cleaning   Conference,
     CONF-720823,  NTIS,  Springfield,  VA, 1973, pp. 348-354.

6.   Washington,   R.A.,  Chi,  W.,  and Regan,  R.,  "The  Use of  Vermiculite  to
     Control  Dust  and Radon  Daughters in  Underground  Uranium  Mine Air,"  in
     Proceedings  of  the 12th  AEC Air  Cleaning  Conference, NTIS,  Springfield,
     VA, 1973,  pp. 355-376.

7.   Wrenn,   M.E.,  Eisenbud,   M.,   Costa-Ribeiro,  C.,  Hazle,  A.J.,  and  Siek,
     R.D., "Reduction  of  Radon  Daughter  Concentrations  in  Mines  by  Rapid
     Mixing Without Makeup Air,"  Health Phys.  r7:405  (1969).

8.   Holub,   R.F.,  Droullard,  R.F.,  Ho,   W.,  Hopke,  P.K.,  Parsley  R.,  and
     Stukel,  J.J., "The Reduction  of  Airborne Radon Daughter Concentration  by
     Plateout on  an Air  Mixing Fan,"  Health Phys.  _36:497 (1979).

9.   Thomas,  J.W.,   "Measurement  of  Radon  Daughters   in  Air,"  Health  Phys.
     22_:783  (1972).

10.  Robins & Myers,  Inc., manufacterer's  literature  provided  to  retailers.


               R.E. Toohey, M.A. Essling, H. Wang+ and J.  Rundo

               Radiological and Environmental Research Division
                          Argonne  National  Laboratory
                           Argonne, Illinois  60439

    The problem  of  relating  radon  levels  in houses  to the  radiation doses
received  by  their  residents  would  be  greatly  simplified  if  the parameters
which  determine  the  equilibrium  factor,  or  working  level  ratio,  could  be
identified and their  effects  predicted.   The  first step in such  a program is
to measure  the equilibrium factor  in a  number  of houses under  a variety of
conditions, by simultaneously measuring  both  the  radon  concentration and the
working level.

    Our preliminary measurements in a few houses indicate that the equilibrium
factor  normally  varies over  a  range of 0.05 to  0.50,  but  on  occasions  it
approaches 1.0.  The mean of eight sets  of observations  was  0.2O +_ O.14 (one
std.  dev.).   Some  of  the parameters which have  been  observed to  affect the
equilibrium factor  include the relative  areas of room  surfaces  and airborne
particles, the ventilation rate,  and the air  circulation.   Human activities,
such  as cooking  and smoking, also  affect the equilibrium factor, by directly
affecting the aforementioned variables.


    Because the  radiation dose  to the  lung  from environmental  radon  is due
almost  entirely   to  the   inhalation  of  the  short-lived daughter   products,
measurements of  radon levels alone are  not  sufficient  to determine  the dose.
Typically the activities of the daughter products in air are not  in equilibri-
um with the  parent radon, nor  is  there  equilibrium among the daughters them-
selves.  The working  level (WL)  was introduced as a convenient  unit to over-
come this problem in uranium mines  and it is  also  applicable to other environ-
ments where radiation dosimetry is the  principal  concern.   It  is defined as
that concentration of short-lived radon daughters in any combination  which has
a potential alpha-particle energy of 1.3  x 10^ MeV  per liter of air.   Since
*Work supported by U.S.  Department of Energy under contract  IW-31-109-ENG-38.

+Visiting Scientist  from the  North China  Institute  of Radiation Protection,
Taiyuan, Shanxi, PRC.

 this  is equal  to  the potential  alpha-particle energy  of  100 pCi/1  of radon
 with  all  its  short-lived daughters in  equilibrium,  we  can  define  the equili-
 brium  factor,  F,  as one  hundred  times  the working level divided  by the radon
 concentration in pCi/1.   Since  it is easier to  determine  the radon concentra-
 tion  than  the  working  level,  the  task of  determining lung  doses  would be
 simplified  if  those factors which  determine  F  could  be identified  and their
 effects quantified.   In  order to  do  this,  simultaneous measurements  of radon
 concentrations  and  working  levels must  be  made in houses  under a  variety of
 conditions.  In addition, parameters which may  affect  the  equilibrium factor,
 such  as aerosol concentration  and ventilation  rate,  must  also  be identified
 and monitored.  Although  the effects  of  such  parameters may be best determined
 under  controlled conditions in  a  laboratory,  the results must  be  validated by
 measurements in houses.


    Radon  levels  were determined  by either  collecting air samples  in Lucas
 flasks  (Lu57),  or  by continuous monitoring with a  Spitz-Wrenn  chamber (Wr75).
 Concentrations  of  radon  daughters were determined and  working  levels computed
 with  an Environmental Working Level Monitor (Ke78).  Measurements  were made in
 three  houses  owned  and  occupied  by  ANL employees;  two of the  three  (H-01,
 H-23)  have radon  levels which  routinely exceed  370  Bq/m3 (10  pCi/1).   The
 radon  levels  in  the  third (H-02)  ranged from 11  to  300 Bq/m3  (0.3 - 8.0
 pCi/1).  The results are shown in Table 1.

    Values  for  F  reported  by other  workers  have  averaged around  0.5  (Ge78,
 McGSO).  Only one  of our series  of  observations approached  this  value,  while
 the mean  of all our observations  is  0.20 •+  0.14  (one  std.  dev.).   The radon
 concentration  (Spitz-Wrenn  chamber),  working  level   (EWLM),  and  equilibrium
 factor  in  the  basement of  house  23  during the  period  27-30 Sep  are  shown in
 Fig.  1.   Note   that  while the radon  concentration and working level undergo
 diurnal variations  by  a  factor  of four,  88% (71/81)  of  the values  of  F  lie
 within the range 0.15 - 0.30.   However,  some of  the variability in  F may be an
 artifact of the measurement techniques.   The  EWLM determines working  level on
 the basis  of  a three-minute sample  of  air,  whereas  the values obtained  with
 the Spitz-Wrenn chamber  were  averages of 60-minute counts;  furthermore,  there
 is a delay between a change in the radon  concentration  outside  the chamber and
 equilibration inside  it,  and another delay  before the  electrostatically  col-
 lected daughter products  are at  equilibrium with the radon producing them.

    Although the variations in  values of F have not as yet been  firmly linked
 to  such parameters  as  air  exchange  rate or airborne  particle  concentrations,
we have observed the lowest values of F  when  the fan  of the heating or cooling
 system is  in operation.   Presumably  this removes dust  particles from the  air
 by filtration and impaction, giving unattached daughter products  the chance to
 be removed from the air by deposition on surfaces within the house.


    Laboratory   experiments  have shown that  radon daughters  "plate  out," i.e.,
 deposit on  various  surfaces exposed  to  a  radon-containing  atmosphere  (Ho79,
Ge81).  We have recently  performed some direct measurements  of radon daughter
plate  out  on  surfaces  in a house.   The  data are  shown in  Fig.  2.   When  the
cover  was  removed   from  the  window  (316  cm2) of  a  gas-flow  proportional

counter, the counting rate immediately increased, presumably due to the detec-
tion of  the decay of radon  and daughters in the  air  within a-particle range
of the  counter  (about 4 cm).   The  counting  rate  then  continued  to increase,
showing  the characteristic build-up  of  activity  due  to deposition  of radon
daughters on  the surface  of the  window,  until  an equilibrium  was  reached.
When  the  cover  was  replaced,  the   counting   rate  doubled,  because  radon
daughters were  also deposited  on  the surface of  the  cover itself,  and then
decreased at a rate characteristic of  a mixture of  radon daughters.


    Simultaneous  measurements  of  radon  levels  and working levels  in  a  few
houses  have resulted in a mean value for  the  equilibrium  factor of  0.20 _+
0.14.  The  lowest values of F were  observed during  the  operation of the fan of
a house's heating or cooling system,  which  presumably removed radon daughters
from  the air by  both filtration and impaction  on walls and  other surfaces.
Direct observations of the plate out of radon daughters on surfaces in a house
were  performed,  confirming  the  results  of  laboratory experiments  by other

Ge80  A.C. George  and A.J.  Breslin,  1980, The  distribution of ambient  radon
      and radon daughters  in residential  buildings in the New  Jersey-New York
      area, in Natural Radiation Environment III,  T.F. Gesell and  W.M.  Lowder,
      Eds., U.S.  Department  of  Energy Symposium  Series  No. 51,  CONF-780422,
      Vol. 2,  p. 1272.

Ge81  A.C.  George  and  E.G.  Knutson,  1981,   Measurements   of  radon  daughter
      plateout, presented  at the  26th Annual  Meeting of  the  Health  Physics
      Society, Louisville,  KY, June 21-25, 1981,  Abstract  No. P/189.

Ho79  R.F.  Holub,  R.F.  Droullard, W.  Ho,  P.K.  Hopke,  R.  Parde,  and  J.J.
      Stukel,   1979, The  reduction  of  airborne radon daughter concentration by
      plateout on an air mixing fan, Health Phys. 36,  497.

Ke78  D.J. Kefe, W.P. McDowell, an  P.G. Groer, 1978, The  environmental  working
      level monitor,  Final  Report to  U.S.  Environmental  Protection  Agency,
      Argonne  National Laboratory Report #P7628C.

Lu57  H.F. Lucas,  Jr.,   1957,  Improved low-level  alpha-scintillation  counter
      for radon, Rev.  Sci.  Inst.  28, 680.

McG80 R.G.  McGregor,   P.  Vasudew, E.G.   Letourneau,  R.S.   McCullough,   F.A.
      Prantl,   and  H.  Taniguchi,  1980,  Background  concentrations  of radon  and
      radon daughters in Canadian homes, Health Phys.  39,  285.

Wr75  M.E.  Wrenn,  H.  Spitz,  and  N.  Cohen,  1975,  Design  of  a  continuous
      digital-output   environmental radon  monitor,   IEEE  Trans.   Nucl.   Sci.
      NS-22,  645.
              The submitted manuscript  has  been authored  by
              a contractor of the U.S.  Government under  con-
              tract  No.  W-31-109-ENG-38.   Accordingly,   the
              U.S.  Government retains a  nonexclusive,  royal-
              ty-free  license  to publish  or  reproduce  the
              published  form of  this contribution,  or allow
              others to  do so,  for U.S.  Government purposes.

Table I.  Observed values  for  the equilibrium factor, F.
Period No. Observations
Range of F
- 0.83
- 0.35
- 0.20
- 0.23
- 0.44
- 0.16
- 0.21


Fig. 1.  The radon concentration, working level, and equilibrium factor  in  the
         basement of house no.  H-23  for  a period of  four  days.

Fig. 2.  The build-up and  decay  of  radon daughters  on  the window of a  large-
         area proportional  counter  placed  in  the  utility room of  house  no.

                              ^        t

                                    « •




-£ 0.3

$ 0.2


                                •      «S
                              •         •••*•••

                               •  •  .•          •
           V. .
           ^b« •


  Figure 1

                      Figure 2

                          NATURAL RADIATION IN SWEDEN

                                 Carole Wilson

                          Geological Survey of Sweden
                                    Box 801
                             S-951  28  Lulea,  SWEDEN

    In 1979, when the  problem  of high radon daughter levels in Swedish houses
became widely publicized, the need for information on variations in the natur-
al radiation environment became very apparent.  The radon problem was at first
attributed to radon emanation  from alum shale based  aerated  concrete,  but it
was soon  obvious  that  ground with an  abnormally  high uranium content consti-
tutes  an  even  greater  risk for  high  radon and  radon  daughter  levels  in
houses.  The  Geological  Survey of Sweden was  commissioned  to produce a docu-
mentation in  map  form of  all  areas  and  rock  types with  gamma  ray  levels
exceeding 30  yR/h,  with the intention of delimiting  risk  areas for high soil
gas radon contents.

    The maps, known  as GEO-radiation maps,  are produced at  the scale of 1:5O
OOO.   They  are  based  primarily  upon  airborne  radiometric   surveys,  ground
measurements of gamma  radiation  and  geological mapping.  To date some 450 map
sheets  have  been published  covering approximately 55%  of the  country.   The
maps  provide primary   information  to  local  planning,  health  and  building
authorities as to variations in  the  natural radiation environment.  Within the
so-called risk areas marked  on the maps, local authorities are recommended to
investigate the  soil gas radon  content prior to any new development.

    Geological environments  known,  in  Sweden,  to  be  associated  with radon
daughter problems in dwellings are alum shale, a Cambrian, uranium-rich black
shale, uranium-rich granites and uranium-rich pegmatites.  Both alum shale and
uranium-rich granites constitute  extensive areas of bedrock.

    In  1978,  the National  Institute for Radiation  Protection measured radon
daughter levels  in  a number  of houses  built on  uranium-rich shale tailings
(Swedjemark et al.  1979).   The measured  levels  exceeded the maximum level of
1110  Bq m~3  (Becquerel per  cubic metre) of radon  in  equilibrium with radon
daughters permitted  in Swedish mines.  This  was the first proof in Sweden that
radon emanating  from  the  ground can give rise to such high concentrations of
radon  daughters   in  dwellings.   Earlier,  the radon  problem  in houses  was

attributed  to  radon emanation  from building  materials  containing  unusually
large  amounts  of  radium,  for  example,  alum  shale  based  aerated  concrete
(Swedjemark 1978a, 1979b, 1980),

    Early  in  1979,  a Government  Commission,  termed  the  Radon  Commission, was
set  up to  investigate  the problem of  radiation risks  in  dwellings  (Radon-
utredningen,  1979).   At  an  early  stage  in  the  Commissions work,  it  became
apparent  that  radon emanating from  the ground  could constitute as  great, if
not greater problem  than  radon emanation from building materials.  Measurement
programs  carried  out by  Local Health  Authorities during  the  last  year   have
proved  this  to be true»  In  Sweden,  areas with particular risk  for  high  soil
gas radon  contents  are  considered to be  those  with  a.  bedrock  of uranium-rich
alum  shale (a  Cambrian,  black  shale),  uranium-rich granite  or uranium-rich
pegmatites and  areas of  glacial  drift  cover comprised largely  of  these   rock
types.  Knowledge of the  geographical distribution of  these  areas  became  thus
necessary  in  order,  at least,  to  prevent future  building without the  use of
necessary technical measures against radon penetration from the ground.


    The source  of natural radiation  from the  ground is almost entirely related
to  the presence  in the  bedrock or  drift cover  of  the  radioactive elements
uranium and  thorium, and their  daughter products including  radon,  and  potas-
sium-40.   Their concentrations in various rock  types in  Sweden  is  very vari-
able.   (Table 1).

    The  geographical distribution  of  the uranium-rich  alum  shale  formation
(Fig.  1)  is  fairly  well-known due  to mapping and prospecting  activities, the
latter  owing  to  the  rocks'   unusually high  contents  of uranium,  vanadium,
molybdenum and  organic  matter.  The  uranium  content  can  range  from 50 - 350
ppm (Armands 1972:  Andersson et al. in press).

    Uranium- and  thorium-rich granites  and pegmatites  are known  from a  number
of areas  in Sweden  (Fig,  2).   The  granites (and pegmatites) occur in a variety
of geological  settings  in the Swedish Precambrian, and  represent a variety of
granite types  and ages, 1750-890 Ma  (Wilson and  Akerblom 1980).   Knowledge of
the  occurrence  and  extent of  the  granites and  pegmatities has  been obtained
mainly  from  airborne radiometric  surveys  and uranium  prospecting  carried out
principally by  the Geological Survey of Sweden (SGU).

    For  the  purpose of  documenting  the natural   radiation  environment in
Sweden, and  for delimiting risk areas  for high  soil gas radon  contents, the
Geological Survey of Sweden was  commissioned  to  produce a documentation in map
form of all  areas and  rock  types  known to  be  particularly  radioactive-   The
maps,  known  as GEO-radiation maps,  are produced at the  scale  of  1:50   000.
They are  based primarily upon airborne  radiometric  surveys,  ground measure-
ments of gamma radiation and geological mapping <>

    The choice  of  this method  for  delimiting  risk  areas for  high  soil gas
radon contents was determined  by:


Granite, normal
Granite U- and Th-
Alum Shale
U ppm
0.1- 2
0.5- 5
0.5- 2
10 -350
1- 2
% mRa
.1 -
.5 -
.1 -
.1 -21
5 -
12 -
5 -
2 -
2 -
5 -

10 -230
Table 1; Uranium, thorium and potassium contents, radium index and gamma radi-
         ation for different rock types  in Sweden.

Figure 1
     3 NARKE
     5 OLAND
     6 SKANE
     SGU 1977

         4	J 1969-1979
         25-60 EXPOSURE RATE (pR/h) MEASURED  1m
               ABOVE GROUND LEVEL
               SGU 1980

     1.    the  assumption that the greatest  concentrations  of radon in  the  soil
          are  directly related to  rocks  and  soils  rich  in uranium.

     2.    availability  of  regional  information  on gamma  radiation  from  the
          airborne  and  ground radiometric surveys  carried  out  by SGU.   These
          surveys cover to date approximately 45%  of the country.


     Airborne  gamma-ray spectrometry carried out  by SGU is  used  primarily  for
 regional prospecting  for  uranium, and  as a complementary geophysical method  to
 geological  mapping.   The instrumentation has   been designed to  cope  with  the
 special  problems  of prospecting  in  areas of glaciated terraine.   Most of  the
 bedrock  in  Sweden is  covered by  glacial till  which can vary in thickness  from
 4  to 10  or  more metres.  Much of this  material is  of  relatively  local deriva-
 tion,  but it has  always  been transported  to  some  extent.   This  latter factor
 is of  special interest with regard to the radon problem as will  be shown later.

     The  SGU  technique (Linden and Akerblom, 1976) is to use four  sodium iodide
 crystals giving a total volume of 17  litres.   Gamma  spectra  between  0.45 and
 2.85 MeV are recorded  on 258 channels.  The plane  flies at  an altitude of  30
 metres along  profiles  200 metres  apart,  and at a  speed of  70 m/sec.  Registra-
 tion takes  place digitally with  a measuring time  of  400  msec, readings being
 taken  every  40 metres.  After  computer  processing of the  data  from  all the
 channels,  the results  of  the measurements  are plotted automatically  as maps
 using  an ink-jet  color plotter.   Gamma radiation  from  uranium,  thorium and
 potassium for each measuring  station is marked on the maps as colored  lines  (U
 red, Th  blue and K  yellow),  the  lengths  of  which  are  proportional  to the
 registered  equivalent  contents  of  these elements  (U  and  Th  as  ppm and  K  as
 %).   The  three-component,   colored,  gamma radiation maps  give a   detailed
 picture  of  the  distribution of  the  radioactive  elements over  large  areas,  as
 well as  locating point  anomalies  (Fig.  3a and b).

     Follow-up of the  anomalous points  or areas for each gamma radiation map  is
 carried  out on  the ground using  handborne scintillometers.   Observations as  to
 terraine, bedrock  geology,  Quaternary  geology and  gamma-ray levels  are noted
 for  areas  where  gamma-spectrometry  measurements   on the  ground  have  been
 carried  out.  In  the  compilation of the  GEO-radiation maps,  all  this material
 is  used   together  with published geological  and  Quaternary  geological  maps
 (where  available), and detailed  local  knowledge   of  the  geologists  mapping
 specific areas.


    Maps showing radon contents  in the ground  would be the  ideal for environ-
mental purposes.   Such maps  cannot  be  produced  owing to  the  enormous  costs
 involved and  the  length of  time  necessary to  carry out measurement  programs.
An alternative would  be to  produce maps  showing  the uranium content  (in ppm)
of the ground, but again, this is not  yet 'feasible for more than  limited areas
in Sweden.




                    BEDROCK GEOLOGICAL MAP

                         COARSE PORPHYRITIC GRANITE

                         FINE GRAINED GRANITE
                    |   |  GRANITE  GNEISS

                    I  \\  METABASITE
                    01234   5km
                            SGU 1981

    The  GEO-radiation maps in  their  present form are  intended  as key maps  to
show  areas  where the level of  gamma  radiation from uranium  and/or thorium  in
the bedrock or  drift  cover  is particularly high.   For this  purpose the limit
of  30 pR/h  has  been arbitrarily  selected.   To  date  some 450  maps have been
produced  covering  approximately 55% of the  country.   The maps cover initially
areas  of alum  shale  and known  uranium-rich  granites.   The  latter  category
coincides in general with areas covered  by  airborne radiometric surveys.  The
maps  are  accompanied by short commentaries on the geology and radioactivity  of
the areas concerned.  Examples  of  the maps will now be  discussed.

    Fig.  4  shows the geological map over part of the  alum shale occurrence  in
Ostergotland,  southeastern Sweden, and  Fig.  5  is  the  GEO-radiation  map over
the same area.  An  enlarged  detail from the  gamma  radiation map over Fornasa
is  shown  in  Fig. 6.  The  lines  on  this map are related to gamma radiation from
uranium  in the near  surface alum shale.

    The  profile  in  Fig.  7 shows  how  Quaternary ice  transported fragments   of
alum  shale  from  the bedrock  exposure up  towards  the surface of  the till  in a
southerly  direction.   The  alum shale fragments   are   successively  mixed with
other material during transport.

    One  can see  how individual "nappes"  of the  shale  material  develop  in the
drift cover, the latter  being  about  15  metres thick.   Comparison  between the
bedrock  geological  map and  the  gamma  radiation map  reveals that  alum  shale
fragments  are present  in  the drift  cover several  kilometres  south  of  the
bedrock  source.  In  consequence, higher  than normal  levels of gamma radiation,
and probably enhanced  radon  emanation,  can occur beyond the area  of  bedrock
exposure  of  a radioactive rock.   Consideration must therefore  always  be  given
to  the  direction and  length  of transport of  rock  fragments  when  judging  the
risks for high soil  gas radon contents.

    A further illustration of  this point  is  made  in  Fig.  8 which  shows  the
gamma radiation  map  from an area  of  granite  gneiss  in central  Sweden.   The
bedrock  is   largely  concealed  by  a drift  cover  of glacial  till.  A  clearly
uranium  anomalous   area  is  distinguished  south  of  Svennevad.   This  area
coincides with a glacial esker and post glacial washed  sands.   The esker has
its source  some  10  kilometres  to  the north  of Svennevad in an area of  alum
shale bedrock.   The  gamma radiation map  therefore  indicates spread of  alum -
shale fragments  of the  esker  and  related  sands.  Enhanced gamma radiation from
uranium concentrations  in the granite gneiss bedrock is otherwise only a local
feature in connection with pegmatites and zones of migmatisation.

    The gamma radiation,  geological and GEO-radiation  maps over a uranium- and
thorium-rich granite in Molndal, south of Gothenburg,  are shown in Figs.  9,  10
and 11.   The suburb  of  Balltorp is planned to house 15,000 people.

    The granite  in  question  is a  red  to  grey alkali   granite  which partially
weathers to  a  coarse sand.   The granite  has a uranium content  of 10-30  ppm, a
thorium  content   of  30-70  ppm and radioactivity  of 30-60 uR/h.   The  granite
is  bounded   to   the  east  by   gneissose   granodiorite  and  to  the west  by  a
porphyritic  granite  with  normal uranium  and thorium contents.   The whole area
is  well  exposed  and the  till  cover,  where present,   is only  a half to  two
metres thick.  The  valleys contain sediuents,  mainly  clay.   Glacial transport
of till  material  was from east  to west.

^^_, ' i i ^j-^j^^-j ' |-;[ undlffer«ntlatad

                                                            2O  23 km
                                                   10km   SGU 1977




 EXCEED  30 pR/h

 BELOW 30pR/h




 SCALE  1:50000

0             5km
                                                                                               SGU 1980

                                                                                                                                                                                   I	-llMI-iiillllllll.l	ll-i	II	>llll|IMlli|lll.iliH	-ilil	|.
                                               I lllllllll.lllll

||ll|l'|..l||ll|l  'IMHI|.||l-l.l. ||ll|ll||l|||l||||lHl|.ili
  •I	..-••ill.-.-	-H.-I.-	H.-MIII	iiii||i"'ii||i||l|ll"il|iiii|||ii|i>i.iii||i||||||||||iiiii	I|I.....IH,.I|	i-.in  iiiniin	i	|n	i	...|.  .i..iii.i|i.i..iii..|i-i	I..IIH	HII... .|i--i	-m
 » -I	M'  	I-I'-M	IM	II.M—HI	-.lll..|||ll-l|l—I....I	." .H"llh.|.ll.|ll|||.I..IH||l|l|ij|ll|ll	ll-ll»|i|||ll|||||||.H..|l	l|l».|||l..ll||.||lll|.||-|l-l-.l-	I	|..||.||".|l.	IIIIMll..  ll-ll	||l|'"l.	  l||l..|. .1. ||.
  -l. .II.H-.H. ••.-••I. .1	».i-n|ii	iii|—.|ii..>i. i.	|i	i	•iiii|i-||iiiiHiii.||iiii|iili|||||j|j|ij|i..i|i-.i	ii>i||||||||j|j|ij||||iiiii||iiiiiiiii—i|iii.i||.i.ii"iii.i>i..|—  .I-......I...II	|i.t|||iiii.ii"i|iii	  -|||ii...||..|... .I...II|.IHI.I... HI	.I'.iin	

  •I'l'-ii".	li'lui "ii ij  i..IMI...I. .iiiiii"...  -i. >|i	HIII..IIII.-.I|.-I "•  •	>...|... ..n i|i||i.iiii...    ...H	i.-iiiiij|li||j|jij||||||i|i'i||||M..ii|||||r.i|iiiii|ii	i|»—	->i.i».    ..i- .||i.|..i.l- >i...	.••!	n iin.iii
  	  .....mil  ....  ...H	I.	iiiiMI'-*-'-	iiii||iii.||n  uii-i.ii.ii.ii n	.I..I....M	i.	i|. i.   n	I'llllllllilllllli'llilllill'llllllll-'l	'"'I! •	!• H-..|.	-| -MM- .|i..—n ..in  i.	>	ii".it. .(..in-..
  ...l.ll ..... I......I"	HI ...HI..I	||	i..i	III).   |i.| .||.	H..I	...I		..|i—i|	.l.-l	  ,	..I	|. I... 	....... .......MII.	I,  .,i.||.  ,|i. ,.||. .illlllMlli.il
 •.lull...  .i-iii|.iin	 H-..I."	..I.....	i	..i... ...i	  -	I..IHII.  -ii -mi	i|iii|.|||.i|||iiiiiii|||il||iHi.i||l|in||||"i|liiiM|..<|iii. •(	n-i	!• in—- M • ••- -I	n  •	n-	MI  -i-  -	>|i"H

MI. .In   nun  i   i     4	n "in *	i	i	1*	i	iiiiu||i.t|ii.ii.  11  inii|iiiK|i...	i-.iliii	IIIHI	iiiii.ili.-.|ii|i.||i-i|.||.i	HI
   	nii.»h-  -i	........n	ii... ..*i*...»it .i...ii.i..i	i. .i*.i|ni.i	iii|i-i>»ii..i>i-i|i«»i'-»niiiiii.i"'i'.iiMi|iii"ii".i|!..iiiiiiii.iiiiii>M|>|i III.'|....I.>II.III||I|IHI..II||I.I	in >III'...HI«I|I<
   	||	|l.	|«.*...|.*|l|......"|l  ,|..l	|1||.	|....... ||l.|l..|*l||||	l|||||.> -l..|.ll'.|lll"l|.|..'>||lll. l.'lll'.|.  'IIHlll- 1|.	I.IIMII.II|.||||....-.'.||.||.t|..|...||.	*|l.l*..	 . .|..||  ..II..  ......Illll,

            ICE TRANSPORT
                                                                  20-40 jiR/h

         ALUM  SHALE








               Figure 7
                                                                        SGU 1979

                              Figure 8
§ = i :- v
f i =«SVEN&E'VAD


                                                              GAMMA  RADIATION  MAP

               1	-Ill	
                V<- --/• .H||-i4|fJ|»A-M||MmHP--u
                                   BEDROCK  GEOLOGICAL MAP MOLNDAL
                                                01234     5km
Porphyritic granite
Uranium and thorium rich
Gneissose granodiorite
                                      Figure 10



          •t*-» » + + * + + + ** + * + *+ +


                                         **'™v  ***«*-»>
     r  ir 20- »•  36-
                                                               GEO - RADIATION   MAP


                            GAMMA  RADIATION CAN
                            EXCEED  30pR/h

                            GAMMA  RADIATION BELOW

                            ALKALI  GRANITE
                                                                     SCALE 1:50000

                                                                           SGU 1979
                                            Figure II

    A program  of  radon measurements was  carried out  in the  area  during  the
 summer  of 1980 (Hesselbom 1981).   The  results indicate  soil  gas radon levels
 ranging  from  30,000 to  500,000 Bq m~3 (relative values).


    During  the last  twelve  months, a  research program  has  been initiated  to
 investigate  the  relationship  between  radon  concentrations   in  soil  gas  and
 radon/radon  daughter levels  in  dwellings,  and  to  test  the  validity  of  the
 GEO-radiation maps.   Test areas  were  selected  in areas  of  both uranium-rich
 granite  (Lysekil) and alum shale (Fjugesta).


    In  the  town of Lysekil,  radon daughter concentrations have  been measured
 in  60 selected houses.   Seventy  percent of  these houses have radon daughter
 levels  exceeding  the permitted  limit  for  new development (70 Bq m~3, Swedish
 Building Code, 1981),  and 10%  have levels  exceeding  the permitted  limit  for
 existing dwellings (400 Bq m~3).

    The  town  of Lysekil lies  on the west coast of Sweden in an area of granite
 known as the  Bohus granite complex  (Fig. 12).   The granite occupies an area  of
 about 20 km by 90 km in Sweden and extends into SE Norway.  It has the form  of
 a  large  flat-lying,  sheet-like intrusion  consisting  of  a  series  of granite
 types ranging in color  from  red  to grey, and  in  texture from coarse-grained,
 sometimes  porphyritic  to  extremely fine-grained  (Asklund 1947).   Chemically
 and mineralogically,  however,  the granite is  fairly homogeneous  and the whole
 of  the   granite belt  has  enhanced uranium  and thorium  contents  (Wilson and
 Akerblom 1980).  The GEO-radiation maps covering  the  granite  mark  the entire
 complex  as having gamma-ray  levels  exceeding 30  pR/h.   In reality,  gamma
 radiation from the ground varies  between  10 and 50 pR/h, the lower values  in
 areas where the granite is covered by  loose sediments,  sands  and  clays.   The
 higher  values  refer  to  the  exposed  bedrock.   Numerous ground spectrometer
 measurements   over  the  granite  give   the   following  uranium-,  thorium-  and
 potassium contents:  U  9-37 ppm, Th 25-90 ppm and K 3.5-5.0%.

    Airborne  radiometric  surveying  has  been carried  out over  the southern part
 of  the  granite.  Fig. 13, showing  the  U-component  of  the gamma radiation map,
 illustrates clearly the intensity  of  gamma  radiation from the granite compared
 with  that from the adjacent gneisses with normal U and Th contents.

    Measurements of radon  concentrations in well waters from the Bohus granite
 give, in general,  high levels.   In the Lysekil area,  50% of the investigated
 wells (ca 200) have  levels exceeding  1000  Bq/1.  The  levels  recorded compare
 with  those  obtained  from  uranium-rich  granites in  Maine (Hess  et  al. 1980).
 In the gneisses adjacent  to  the Bohus  granite, radon levels  in well water are


    The   indoor measurements  of  radon  daughter concentrations carried  out  in
Fjugesta  (45  houses)  show  that  80% of the houses  have  levels  exceeding  the
permitted  limit  for  new  development  and  12%  have  levels  exceeding   the


                        + + + + + + t + + + -f+ + + + + + + + +
                        «• STR0M6TAD+ + + + + + + +
        SGU 19S1

Figure   13
                                                                                   Illllllllll"	I
                                                                                       II	I
                                                                      Illlllllllllllllllll	•    ..-I'	Il-
                                                minim	iiiiii'iiiu
                                                illllllllllliMliiiiiiniiii	mi	n
                                                            	                lillnili	inliiiiiiii

permitted limit  for  existing dwellings.   The bedrock  of  the area, consisting
of  flat-lying  Cambro-Ordovician  sediments  including  alum  shale,  is largely
covered by  up  to 1O metres  of  glacial  till and post glacial sediments.  Fig.
14, showing the U-component of the gamma radiation map and the bedrock outcrop
of  the alum  shale  formation,   indicates  clearly  the spread  of  alum  shale
fragments in the till.   The geological map over  the  area is shown  in Fig. 15.


    The problem  of high  soil gas  radon  levels in  Sweden  is seen now  to be
acute.  Rocks and soils with enhanced uranium contents occur  fairly extensive-
ly  in the country.  The  normal background gamma radiation  in  Sweden is 6-1O
yR/h,  and vast  areas  have  gamma-ray levels  of 12-20  pR/h.  Investigations
so  far  carried  out indicate  that  the GEO-radiation maps, based  on  gamma-ray
measurements  and  with  an  arbitrary  limit  of   30  yR/h   for   risk  areas,
adequately  define  the  areas with  particular  risk  for  high soil gas  radon
contents.   However,  extensive  programs  of indoor measurements  carried  out
during the past year by Local Health Authorities and by the National  Institute
for  Radiation  Protection,  reveal  that  levels exceeding  the permitted limits
occur  sporadically even  in areas where  gamma-ray  levels  for  the bedrock  and
drift  cover are  considerably lower than 30  uR/h.   A number  of soil gas radon
measurements carried out by SGU in such an area  show that radon concentrations
in  the ground can  vary  enormously  from one measuring  point to another.   A
lower  gamma-ray  level,  for  example  20   yR/h, could   be   adopted  for  the
GEO-radiation maps, but the consequences of this would be that  vast  areas of
the  country  would be  placed under  stringent building restrictions  which  are
probably not warranted.

            Figure 14

            Figure 15
0   1   2   3km
I	i	i	i
   SGU 1977

Andersson, A., Dahlman, B. and Gee, D.G. in press.  Kerogen and uranium re-
    sources  in  the  Cambrian  alum shales  of  the  Billingen-Falbyggden areas,
    Sweden.  Geologiska Foreningens i Stockholm Forhandlingar.

Armands, G. 1972.  Geochemical studies of uranium, molybdenum and vanadium in
    a  Swedish   alum  shale.   Acta  Universitatis  Stockholmiensis,  Stockholm
    contributions in geology, XXVII (1), 1-48.

Asklund, B. 1947.  "Gatsten och kantsten", Sveriges Geologiska undersokning,
    C 479 (in Swedish), 187 pp.

Hess, C.T., Norton, S.A., Brutsaert, W.F., Casparius, R.E., Coombs, E.G. and
    Hess, A.L. 1980.  Radon-222 in  potable  water  supplies  of  New England.   New
    England Water Works Association, 2,  113-128.

Hesselbom, A., Israelsson, S. and Tovedal, H. 1981.  Radon in the ground.  A
    study  of  methods and  instruments for determining radon  concentrations  in
    the ground.  Byggforskningsradet, Report R47:1981 (in Swedish), 75 pp.

Linden, A.H. and Akerblom, G.V. 1976. Method of  detecting small or indistinct
    radioactive  sources   by  airborne gamma-ray   spectrometry.   In:   Geology,
    mining and  extractive processing of  uranium.  Ed. M.J.  Jones.   Institute
    of Mining and Metallurgy, London, 113-120.

Radonutredningen, 1979.  Preliminary proposals  for measures against radiation
    hazards  in   buildings.   Report from  the Radon  Commission,  Jordbruks-de-
    partementet, Ds Jo 1979:9,  (in Swedish)  114 pp.

Statens Planverk, 1980.  Svensk Byggnorm (Swedish Building Code).  Statens
    Planverks forfattningssamling 1980:1 (in Swedish).

Swedjemark, G.A. 1978a.  Radon in dwellings  in  Sweden.   Statens Stralskydds-
    institut, Report SSI:  1978-013, 1-24.

Swedjemark, G.A. 1979b.  Indoor measurements of natural radioactivity in
    Sweden.  Statens stralskyddsinstitut,  Report  SSI:  1979-026, 1-11.

Swedjemark, G.A. 1980.  Radioactivity in houses built of  aerated concrete
    based on alum shale.   Statens stralskyddsinstitut,  Report  SSI:  1980-14  1-6.

Swedjemark, G.A., Hakansson,  B. and Hagberg, N. 1979.   Radiation levels in
    houses built on wastes from processing of alum shale.   Statens  stralskydds-
    institut, Report SSI:  1979-006, 1-43,  (in Swedish).

Wilson,  M.R.  and Akerblom, G.V- 1980.  Uranium-enriched granites in Sweden.
    Sveriges   geologiska   undersokning,   Rapporter  och  meddelanden,   nr.   19,

                                 FIGURE  CAPTIONS

Fig. 1.  Distribution of the alum shale formation.

Fig. 2.  Distribution of known radioactive granites.

Fig. 3a. Uranium-component  gamma radiation map  over a  uranium- and  thorium-
         rich granite near Vasteras, central Sweden.  Original scale 1:50,000.

Fig. 3b. Geological map over the same granite.  Original scale 1:50,000.

Fig. 4.  Geological map  for the. map  sheet Linkoping NW,  southeastern  Sweden.
         Original scale 1:50,000.

Fig. 5.  GEO-radiation  map  for  the  map  sheet  Linkoping  NW,  southeastern
         Sweden.  Original scale 1:50,000.

Fig. 6.  Uranium-component  gamma  radiation  map  over  Fornasa,  Ostergotland.
         Original scale 1:50,000.

Fig. 7.  Profile illustrating  the  distribution of alum shale fragments  in  the
         Quaternary drift cover.

Fig. 8.  Uranium-component  gamma radiation  map  over  Svennevad,  southeastern
         Sweden.  Original scale 1:50,000.

Fig. 9.  Uranium-component  gamma   radiation   map   over   Molndal,   Gothenburg.
         Original scale 1:50,000.

Fig. 10. Geological map over Molndal,  Gothenburg.   Original scale 1:50,000.

Fig. 11. GEO-radiation map over Molndal, Gothenburg.  Original scale 1:50,000.

Fig. 12. Geological map over the Bohus granite complex,  western Sweden.

Fig. 13. Uranium-component gamma radiation map over  the  Bohus  granite  around
         Lysekil.  Original scale 1:50,000.

Fig. 14. Uranium-component gamma radiation map over Fjugesta,  central  Sweden.
         Original scale 1:50,000.

Fig. 15. Geological map over Fjugetsa.  Original scale 1:50,000.

                       INTEGRATING RADON MONITORS (PRIMs)

                              Robert A.  Washington

                          Atomic Energy Control Board
                              Post Office Box 1046
                                 Ottawa,  Canada

    A description is given of a chamber used at  the  Atomic Energy  Control
Board laboratory for the calibration of passive  radon  integrating  monitors
(PRIMs).  It has been proven capable of providing  stable and  reproducible  (to
_+ 5% or better)  concentrations of Rn^22 in air frOm  about 20  pCi/1 (740
Bq/m3) to more than 3,000 pCi/1 (111,000 Bq/m3).   The  temperature  range
employed was 25° to 30°, and the relative humidity was in the range from
58% to 62%.  The results indicate that  a calibration curve of the  form y = mx
+ b (where y = cumulative radon exposure in picocurie-hours/litre, and x is
the TLD reading in nanocoulombs)  provides better information  about PRIM
performance than a simple calibration factor k = Y/x.

                      OPEN DISCUSSION  I:  AUGUST 27, 1981
Open Discussion;  Instrumentation and Measurement Methods

Moderator;        Peter G. Groer

Panel;            Speakers for Sessions I and II

                   Ron Colie
                   Andreas Wicke
                   James E. Gingrich
                   Howard H. Prichard
                   Andreas George
                   Donald J. Keefe
                   Susan Hinkins
                   Lyle H. Rathbun
                   John G. Pacer


    MR. PHILLIPS:  I know it's not convenient  to go to a microphone every time
you have a question, and in a discussion  session like this it will be even
more difficult.  But I think it's important  to try to get this portion of the
program in the proceedings.   We're not  only  taping these sessions, but we also
have a recorder—not a mechanical recorder obviously—with us for that purpose.
So it's important, again, that you identify  yourself into the microphone.
And, for the microphone on my left,  again, be  sure to speak directly into it.

    I made one mistake, one bad mistake,  this  morning.  The paper by Melvin
First listed in the proceedings i^s to be  presented by Stephen Rudnick, and it
will be presented.  I apologize for  that  confusion.

    The moderator for this session is Peter  Groer, who you know by now, if you
didn't know him beforehand.   So, Peter, I'll turn it over to you.

    MR. GROER:  Well, thanks.  For the  first few minutes, I'd like to see how
things are going.  So I'll open the  session  to the floor, and I think we'll
just accept questions.

    Please identify yourself clearly.  When  you use technical terms, try to
pronounce them as clearly as possible.  It's quite a job to record all of
these accurately.  I have experience from other meetings; it's very hard to do.

    We'll start just with an open question session and we'll see which way the
discussion turns.

    MR. HOLUB:  Question to you.  You mentioned the statistics.  I remember
that Lucas said that if you count the same nucleus twice, once as radium A and
once as radium C prime, you have greater  error than just Poisson statistics,
square root of the count, indicate.   Do you  have any feeling of how much that

    MR. GROER:  You are correct in pointing  to Henry Lucas.  He has done very
interesting work in this area—together with an applied mathematician at
Argonne.  The results he has are not applicable to radon daughter counting.  I
have looked at the problem,  and, as  I said in  my talk, I cannot calculate the
variance for the daughter and the granddaughter.  For radium A it's very easy.
Okay.  So, you're invited to try to  solve this.  I don't know the answer.

    MR. HOLUB:  In which sense did Lucas  make  a mistake?  Why was he wrong?

    MR. GROER:  I did not say that he made a mistake.  He looked at a
different problem.  He was not concerned  with  radon daughter counting.  It's
the paper—I hope we're talking about the same paper—it's the paper in the
Journal of Applied Physics,  I think.

    MR. HOLUB:  Yes.

    MR. GROER:  He is not concerned  with  the problem of radon daughter

    MR. SCOTT:   I was just struck by the  fact  that literally everyone seems to
be concerned with radon daughter  counting, with  the possible exception of the
Germans.  What about thoron?   I mean,  how come we're discriminating against
thoron so effectively?  Is thoron a  problem or do we just ignore it?

    MR. GROER:   I think that's a  very good question.  If there's anybody here
on the panel who would like to answer this, be my guest.  All I can say is
that I made many measurements in  the U.S. in uranium mines and in other
locales—then spectroscopically—and I've never  seen thoron daughters.  But I
know that it's a problem in Canada.   And  I am  sure there are people here who
know what the thoron daughter concentrations are in Canada.  David, to this

    MR. ROSENBAUM:  I think the basic policy reason and the reason in the ORP
why we haven't done anything  on thoron is that the health effects data are all
in terms of radon daughters.   And, therefore,  if you want to know the effect
on somebody you look at the radon daughter concentrations, the working levels,
and from that you induce the  danger,  the  risk.

    Now, the implicit assumption  here is—and  that's because the original
measurements in the mines threw away the  thoron—that either thoron is so
small that it isn't a problem, and therefore doesn't make any difference, or
the ratio of thoron daughters to  radon daughters is the same as it was in the
mines.  And, therefore, when you  throw it away the health effect still comes
out right.  Undoubtedly, there are situations  in which both those assumptions
are completely wrong.

    MR. SCOTT:   If I could just carry on  with  this.  One place where it does
strike me as being particularly important (I may point out, I am from Canada
and there isn't all that much thoron around, but it is present.) is in the
kind of instrument that does  estimates of radon  daughters by, say, alpha decay
or possibly even alpha spectroscopy  plus  beta  counting.  The resulting
estimates will be severely distorted with even the presence of a small amount
of thoron, particularly in the beta  plus  alpha ones.

    MR. GROER:   No doubt about it, if you have thoron daughter concentrations
that are high enough, they will,  since they fall into the two different alpha
channels, the upper alpha channel, there  is no doubt about it that this will
distort measurements.  So the instrument  you heard about today cannot be used
in its present form in atmospheres where  there are thoron daughters.  I agree
with you.

    MR. PERDUE:  If I might make  a comment on  the thoron.  Of course, there is
another radon we haven't talked about, which is  219.  But anyway, I think
we'll all agree that in the world in general,  thorium, in background levels,
is larger than the amount of  radium  there is in  the soil, quantitatively, or
by weight.  So if one takes measurements  approaching background levels, there
has to be thoron in it.  And  you  can measure this.

    MR. GROER:   Well, there's one big difference in the two decay chains.
That's the half life of the noble gas, right?  And that makes a difference.
So it depends very much on your situation.  I  don't dispute the existence of
it, but radon-220 does have a short  half-life.

    MR. PERDUE:   That is easily explained.  Radon-222 may decay in the next
state, but radon-220 decays right  where you are.

    MR. ROESSLER:  Just one small  comment on  the thoron question.  In the
heavy mineral industry, in particular  I'm looking at zircon mill and products,
using two total alpha counts at different times—I guess, a rock—we did find
both radon daughters and thoron daughters.  In fact, if you calculate in terms
of working levels,  the thoron daughter working level of that particular case
is slightly higher  than the radon  daughter working level.  But the summation
of the two is still several orders of  magnitude below occupational standards,
so we never really  followed through  on that in any great detail.  But there
are places where we're going to find thoron daughters, yes.

    MR. GROER:  Okay.  Any more comments on the thoron and thoron daughter

    MR. ROSENBAUM:   The National Radiation Protection Board estimates that
there is about ten  percent as much thoron as  radon in a typical house.  I
don't know if that  is peculiar to  Britain or  the building materials that are
used or what it's from, but that's the kind of numbers that they use.  And
they might be enough.  I'm no kind of  instrumentalist, but that might be
enough to interfere with some of the measuring techniques.

    MR. COHEN:  One last comment on  thoron.   My lab is in the United States,
and there's plenty  of thoron in it.  In fact, it puts a fair correction on the
radium A peak, enough to really make a difference in our lab at least.

    MR. GROER:  What causes the correction to your radium A peak?  Thoron?

    MR. COHEN:  There's a thoron alpha that is the same energy as the radium A.

    MR. GROER:  Yes, but why do you  think that you have thoron in your filter,
if you use a filter method?

    MR. COHEN:  Because there's also a thoron peak at 8 point—8 MEV.  But
there's also a thoron daughter at  6  MEV.

    MR. GROER:  I agree, but not that  it's thoron.  I mean, it was technical.
It's just a little  difference.  You  said thoron and 1 said no, but that it's a
thoron daughter I agree with you.  In  the houses where we have used this
instrument and in the uranium mines  with the  channel analyzer underground, we
have never seen a thoron daughter  peak at higher energies than, well, above
7.68 MEV.  What is  the alpha energy?  I think it is 8.7 or something like
that.  Our highest  flow rate, as I mentioned  to you, was 40 liters per minute.
But we are sampling only for two minutes.

    So, if there are no more questions on the thoron daughter, I'd like to
steer the discussion a little bit.  I  want to bring up the question of
availability of commercial instruments—how satisfied people are.  I was asked,
you know, to bring  this up.  How happy are investigators with commercial
instrumentation?  What are the problems?  What instruments should be produced
commercially?  Is there a need for a certain  type of instrument?  Is there any
interest in this topic?

    MR. HOLUB:  I'd like to mention the problem of acquiring a constant radon
source one can rely on for these calibrations.  And we have some commercially
available dry sources too—EPA brought some.  We have had bad experiences;
some people have had better experiences.   Now the onus is really on NBS to
produce something really good.   The idea  is to provide a known, reproducible
radon concentration at any time.  So,  I wonder how interested others are.  We
certainly are.  I think it would be useful to have them.

    MR. GROER:  If I understand you correctly, you're calling for a radon
meter that would measure—

    MR. HOLUB:  Radon source.   Always produce a given amount, hopefully a
solid source, so that you don't spill  it.  The more I question, the more I
think NBS is going to do it.

    MR. COLLE":  Well, yes—

    UNIDENTIFIED SPEAKER:  Stop right  there.  Good concise answer.

    MR. COLLE:  The onus may indeed be on us.  What he was referring to, of
course, is some type of solid source to replace our solution standards.
Although there are laboratories that can  use solution standards—use them very
well, perhaps better than we can—there is a real need out there for a solid
source, something like the commercially available source, which would be very
easy to use.  We have a very limited effort underway looking for a replacement
in a solid source.  And, if the onus is on us, the last comment I'd like to
make is, if you see me, I'll tell you  who to make the check out to.

    MR. KAHN:  What about a liquid source with two valves on it so you can
bubble?  What we need is some source that can be used over and over again.  Is
there any problem with that?

    MR. COLLE:  Well, we had a small meeting, with a few people, some EPA
people, and for such a type of source, we discussed several types of
configurations.  One of them was a solid  source like ion exchange resin
impregnated with radium, in which case somehow magically either the radon
would get out of this material into a  solution and then you would take it out
of the solution, or perhaps take an aliquot of that solution for making radon
in water analyses.

    We have considered about three or  four different designs, none of which
have been excluded yet, and your type of  proposal is being considered.  I must
say that this effort is relatively low key, but they are being looked at.
There are several possibilities and that  which you've mentioned hasn't been

    MR. PHILLIPS:  Since I suggested, Peter, the question about the commercial
instruments, I feel compelled now to try  to continue comment on that.  I have
not used the newer commercial radon daughter instruments, but I have looked at
a few of them and, just from initial observations, it appears to me that there
may be some problems in the intake portion in terms of plate out, prior to
deposition of the daughters on the filter.  I know in particular you were
concerned about this for the instrument that you and Don worked on.  I wonder
if there is any other comment on that, particularly from you.


    MR.  GROER:   As  you probably remember from the figures you saw on the
slides this morning,  we  tried to keep the air inlet completely unobstructed
because  we were worried  about the so-called umbrella effect, that we are
losing the aerosol  particles on possibly the detector that shields the air.
Now, I do know that there are commercial units around that measure radon
daughter concentration,  or  I should say total alpha counts, rather than
individual radon daughter concentrations.  Very little care has been taken to
avoid plate out problems.   I have seen an instrument where the air inlet is
quite a  small opening, and  then a few inches away from that is the filter.
There is just no doubt in my mind that you will lose some of the radon
daughters on the way  to  the filter.

    So,  I think this  is  an  important problem, and I think manufacturers should
be made  aware of that.   Unfortunately, as long as there are customers who buy
these units, the manufacturers will not be forced to change their assembly

    MR.  KEEFE:   I'd like to make one comment with regard to the problem of
plating  out on the  detector.  From an engineering standpoint, one of the
problems that you have is that it's easy to design an unobstructed air intake
port.  The basic problem is to get the filter membrane from point A to point
B, and this is the  area  that costs a lot of money.  This is also where your
mechanical problems lie, I'd say 30 percent of the cost of our instrument
lies in that particular  problem.  Industry is going to have the same problem,
except they're looking for  a product that is lower in price, that has a good
market and saleability,  and they're trying to cut in that area, I feel.

    But, if the scientific  community and the people who are interested can
bring out the point that plate, out is a serious problem and they should
consider these things even  though the costs of their instrumentation may go
up, I think they may  make some changes in their instrumentation and they could
solve that problem.

    MR.  NYBERG:  We have looked at—in our lab—quite a few instruments, not
as many as Andy George,  but then I don't think anyone has.  I think as an
electronics engineer  I can  say, speaking, I think, for some of us, that the
instruments are getting  better.  Just in the last few years we have seen an
awful lot of newer, better, smaller, lighter, more versatile instruments
coming out.  One of the  problems that we have that maybe could be addressed by
this group is how do  we  evaluate them?  Now we send them to Andy or some such
thing.  This business about plate out on the detectors is a very common
problem.  There's a much too tortuous path for these daughter instruments.
But there is not even a  semi-standard way of evaluating this at the present
time.  I'd like to  see if someone has some suggestions for looking at these
things in a little  more  organized fashion than we're doing right now.

    MR.  GROER:   Thank you.  I think that's a very good point.  It also leads
to the question of  intercalibration and things like that.  I'm sure that some
of our panelists here will  have some ideas on that.

    MR.  COLLE:   No, I don't have ideas on it.  I just want to reinforce the
point and perhaps give you  something else to think about.  What we have in
this room is a collection of probably the premier laboratories in this area.

What you have to keep in mind as well is  that most of  the radiation protection
being done in this country is being done  by state laboratories, and the fact
of the matter is that their instrumentation needs are  considerably different
than research laboratories' needs.   They  need instruments that are very
reliable and relatively inexpensive, particularly if they have many sites to
cover.  They cannot spend 10 or 20  thousand dollars for an instrument.

    Another matter related to calibration and the reliability of instruments
is that there really should be some type  of evaluation of instruments to guide
some of these other laboratories.   These  laboratories  are not equipped to do
the type of evaluations that your laboratory may be able to do, and it would
be a great service to them to have  this evaluation.  Also, there is a great
need, once an instrument and/or method is evaluated, to assure that it's
compatible, on a continuing basis,  with other things.

    These are just additional thoughts that I think we should keep in mind.
We shouldn't overlook that there is a bigger user community out there than
just the people in this room.  Now, I will leave the floor open for the ways
to do that.

    MR. KEEFE:  Speaking of calibration,  again, and the problems, and state
agencies.  I have received calls from several state agencies with regard to
calibration of the EWLM, which belongs to the EPA, which they have loaned out
to some state agencies.  They are presently involved in one specifib case in
the state of Montana in lawsuits versus various cities and mining companies
and so forth.  Immediately they get involved in having to prove the
calibration.  The measurements that are made with this instrument must be
documented and proved.  Again, you're back into the calibration problem.

    One of the problems I'd like to mention about calibration, and it was
mentioned before, is that people have been using these radon chambers, I hope
basically for calibration of radon  instruments, not necessarily radon daughter
instruments.  One of the problems that you have in these large bags is that
the volume of gas that's contained  inside the bag is not sufficient for our
calibration.  What we have to do in all our calibrations is go to the uranium
mines.  We have to go to the Bureau of Mines research  mines and use that large
volume of air and try to keep the air as  static as possible to make our

    Certainly the scientific community could use some  sort of a calibration
laboratory where it wouldn't have to go that far.  Try that in February
sometime—in the snow.  It's an interesting experience.

    MR. GROER:  February in Colorado is okay if you ski down.  With this,
maybe we can leave this topic, or is there further comment?  I don't think we
really have solved anything.  What  I personally don't  see is how we could get
a national quality standard for instrumentation like that.  Okay, why not?
Let's hear your ideas.

    UNIDENTIFIED SPEAKER:  I don't  have anything more  to say.  I just wondered
why not.

    MR. GROER:   I  just wanted  to hear some suggestions.  I mean, who would you
suggest to be the  supervisory  agency?  The National Bureau of Standards?  Some
other agency?

    MR. PERDUE:  I'd just  like to say that we've been here most of the week
making intercomparison studies of existing instrumentation, and I feel that
unless a lot more  of these studies are made we're never going to really solve
what's going on.   There's  a number of problems that haven't even been
approached.  What  happens  with humidity?  Does this change the radon daughter
attachment?  What  happens  to the other isotopes?  We know sometimes they are
indeed present.  So, unless we can address first the things we know, we can
get Bureau of Standards radium that can be calibrated and, of course, can go
back to what they've got.   Can we really know what's going on unless we keep
this thing up?  We are, each one of us, working in a little pocket and we're
getting data and we don't  know whether it matches somebody else's data or not.

    MR. COLLE:   I  think the important point is that it's very admirable that
these things keep  up.  But the problem is this.  You have an informal group of
six or seven laboratories—and perhaps these six or seven laboratories can
answer some of these questions for the other laboratories—but what do the
other 30 or 40 laboratories who are also making these measurements do?

    MR. RATHBUN:   It occurs to me that a lot of the other work like this
that's being done  has been farmed out to independent companies.  And it seems
like at this point there is no national agency that wants to be the arbiter
for such a project.  Why not have an independent group bid on a contract to do
this type of work  and let  each person that wants something calibrated pay a
certain fee to bring it in?

    MS. FISENNE:   I'll help you out on that one.

    MR. GROER:  We should  have comments on this one.

    MS. FISENNE:   All I can say to that is that's why we have the National
Bureau of Standards.  So,  if somebody is willing to come up with contract
money, they will obviously address this problem.  I don't think it has to be
done as a private  industry project.  Obviously, the National Bureau of
Standards is exactly that, at  least in this country.  They provide us with our
radium standards right now. So if everybody is that much interested, then we
should be able to  get the  money together to allow them to remain again on
national standards in this area as well.

    MR. COLLE:   Thank you.  That's indeed the point.  We have had this type of
initiative literally proposed  for perhaps a little longer, but definitely for
three years.  It has ranged from a couple of hundred thousand dollars to
multimillion dollar efforts proposed from every possible point of view that
you could perhaps  think of. You know, we've tried to beat the bushes in all
sorts of ways.   We thought we  hit all the right things and everything else.
We have not been successful yet.

    We recognize there is  a need out there for basic national standards for
these measurements.  We would  like to make an effort to try to provide this.
There are probably other laboratories which right at the moment could do a
better job of it,  but by congressional fiat we have the national


responsibility to maintain these things.   We  have  both the long-term interests
and commitment to maintaining these things.   And I can't agree more with Sue
that it should be done on a national basis.   That's why we were established as
this type of laboratory.   And it just takes more of a push than we've been
able to provide ourselves.

    MR. GROER:  Great.  What could be the first step?

    MR. HOLUB:  To make a list of problems and agree on that.

    MR. GROER:  Can you be a little more  specific?

    MR. HOLUB:  For instance, as I said before, the plate out, the charge, the
trace gases, and the humidity—these I consider a  beginning—and then the
techniques of spectroscopy versus total alpha and  error propagation, perhaps
in more detail.  What happens if the concentration changed during the sampling
and it's simply error propagation?

    MR. GROER:  I agree with you.  It is  an important research problem.  But I
think we just addressed slightly different questions.  The question was about
intercalibration, standardization, possibly of instruments.  My concern is how
do we take the first step as a group if the National Bureau of Standards is
the agency that would be entrusted with such  a venture?

    MR. COLLE:  I'm not sure.  I'm not sure of the mechanism.  I just know
that it requires money and personnel, both of which we are drastically short
of.  How the money comes is not our most  important concern.  We would
certainly not turn down a multimillion dollar contract or interagency
agreement.  We would prefer direct funding to our  own appropriations since
that's more stable and one can build continuity in programs.  But we have some
plans.  We have had these plans for a number  of years, and it's just a
question of putting them into our program and funding them.  The mechanism to
do that is to have Dr. Rosenbaum write us out a check.

    MR. GROER:  Well, that was certainly  a clear answer.  I think Bob Holub
addressed some interesting questions this morning, and he just reiterated some
of them.  I'd like maybe to take up the first one. What do we know about
radon daughters immediately after?  By immediately after I mean, maybe a
picosecond or maybe a nanosecond after they were formed in vacuum.  What do we
know about that?  What charges are there  on these, and then what happens from
there on?  Would anybody like to tackle this  question?

    Well, there are papers in the literature  that  tell you exactly what the
charge states are for a radium A or for a thorium  A.  But this is all measured
in vacuum—not the typical condition.  And then, of course, the neutralization
of these daughter products after they are formed is an area that's very poorly

    I think there's a research project going  on at the University of
Illinois.  Does anybody knows about this  project?  Is there somebody here that
works in this area?  Bob?

    MR. HOLUB:   That's  the project that was funded by the Bureau of Mines, and
it has been stopped,  not  for  lack of funds.  But the basic facts are that
usually people  assume that 90 percent of all daughters in the air are
positively charged.   Now, in  1914, as I mentioned before, Welish found-that
the few have ether, lots  of ether; he obviously didn't have a hundred percent,
but lots of ether in  the  atmosphere,  (and) none of the daughters was charged,
none.  In the arrangement similar to what Andreas Wicke has shown there simply
was no correction due to  charge.

    MR. GROER:   Excuse  me.  Can  I interrupt you?  This reference will have to
go into the record.   Could you possibly share it with the group?  I think it's
an interesting  reference, which  I personally didn't know.

    MR. HOLUB:   Yes.  Most people reference it, but no one reads it.
Unfortunately,  I don't  remember  the full reference, but I can give it to you

    In the experiment done in Illinois, published in January in Science, they
did exactly what Jess Thomas  did, except instead of air passing through they
had nitrogen with various trace  gases.  And when they had 10-ppm and N02,
the diffusion coefficient was such that it must have been neutral.  It was, by
the way, about  twice  as fast  as  when it was charged like impure nitrogen.
It's a very simple and  easily understood experiment.

    The Canadians did a similar  experiment (published in Physics in March) ,
and they had pure gases.  When their ionization potential was lower than
argon's their results basically  confirmed what Hopky found at the University
of Illinois.  Instead of  trace gases, they simply used other gases, like
acetylene and formaldehydes and  C02 or something—I forgot, there are quite
a few gases.  They all  have lower ionization potential than argon; however,
they still have higher  ionization potentials than polonium.

    In Bombay there is  an Indian group, Cotropi.  They looked at ether.
Apparently, besides the ionization potential, the ion molecule reaction also
plays a role.  All these  ion  molecule reactions are taking place within a
period of microseconds.   And, by the way, some people use the term "born
neutral," even  though,  of course, it's not very likely that the daughter is
neutral and is  slowed down.

    So, that is a very  difficult problem.  And it depends on trace gases.
Now, the number of hits per molecule per second is ten to the ninth, so it's
basically, if you are ten to  the minus ninth of something in the air, still
it's going to be a hit  once a second.  So you can imagine the terrible problem
this might be.   Even  though,  as  you all say, for practical purposes maybe it's
not necessary,  but it should  be  looked into.  That's the review of what
happened in this in the last  half a year or year.

    MR. GROER:   Okay.   So, if I  interpret the situation correctly, there is no
systematic research effort in this area—neutralization of, let's say, newly
formed radon daughter products in the atmosphere, whatever this means.

    MR. HOLUB:  That's it exactly.   We  are  trying to do something, but it is
very difficult.

    MR. GROER:  Is David Rosenbaum  here?  No?  Well, this is certainly, I
think, an interesting question.   Humidity was brought up.  Now, there are
many, many different questions related  to humidity.  Calibration, filters,
TLD, track etch, plate out,  I don't really  know where to start.  If anybody
here, the panelists, the audience,  has  a particular humidity problem he would
like to bring up, please do  so.

    MR. COHEN:  This question is related to what you just said.  If I
understood Jim Gingrich this morning, he said that what a track etch detector
measures is what is plated out on it.   Now,  if that's true, then it must be
sensitive to what the plate  out conditions  are.  And, if plate out conditions
are sensitive to humidity and all these other things, how can they just be a
measure of radon, the radon  daughters?

    MR. GROER:  Jim?

    MR. GINGRICH:  The current type of  detector we have now will measure
whatever is plated out on the surface.  But for precise radon measurement,
Bernie, we use that type F cup.   All right,  what it does is filter out any
radon daughters that are outside the cup.   So the only thing that gets through
the filter is radon itself.   Radon  itself will then go into the cup, plate out
on the walls and sides of the cup,  and  plate out on the detector's surface
itself.  You'll get some signal from the radon that's still in the cup and
some signal from the plated  out materials on the side of the cup.  Does that
answer your question?

    MR. COHEN:  It's a clear answer, except it doesn't agree with things I've
heard.  I've heard that, for example, if you put a solid state detector where
the detector is, you get a clean spectrum.   It's as though everything that is
being detected is the stuff  that is actually plated out on the surface.  And
if that's true,  then even in the cup you're measuring stuff which plates out
on the surface of the detector.   Am I wrong somewhere?

    MR. GINGRICH:  Oh, yes,  you are measuring stuff that plates out on the
surface of the detector in the cup, but you've got to remember that the only
thing that will get through  the filter  is radon.  So the signal that the
detector sees is directly proportional  to the total amount of radon that has
come through the filter.

    MR. COHEN:  The radon comes through the filter and is in the cup.  All

    MR. GINGRICH:  Okay.

    MR. COHEN:  Then it decays into daughters and, as I understand it,
essentially what happens then is only the radon daughters which plate out on
the detector are detected.

    MR. GINGRICH:   No,  that's not correct.  It is seeing those radon daughters
as well as the alpha particles  from other radon daughters that would have been
plated out on the  walls of  the  cup—and also a signal that is directly from
the radon itself because the detector does see a certain volume of radon gas.

    MR. COHEN:  But what's  the  support for this?  I mean, is it true that if
you put a solid state detector  there you get a clean spectrum with only
peaks?  That's what I've heard.

    MR. GINGRICH:   I'm  not  an expert on solid state detectors.

    MR. COHEN:  Well, how about with your tracks?  I mean, are all your tracks
the same?  Do they have only two different lengths?

    MR. GINGRICH:   No.   You'll  see a full spectrum of tracks.  And the reason
you will is because if  you  look at the volume, first, the important thing is
where do the alpha particles start within the volume?  All right.  If it
starts some distance out, then  by the time it gets to the detector the track
that you see in the detector will be fairly small compared to a track that
would have started from an  alpha particle fairly near.

    MR. COHEN:  That's  what I was asking.  You actually have data looking at
the length of tracks, and you found that the tracks do have all different

    MR. GINGRICH:   Oh,  yes.  It's a full spectrum.

    MR. PHILLIPS:   What you said is true; at least I have done that.  If you
put a solid state  detector  i-n those cups, then what you see is a clean A and C
prime spectrum. But I  think you have to consider the conditions and the solid
state detector construction; it has a bias on it, too.  So that may not be
directly relatable to the track etch situation.

    MR. HOLUB:  We have done this, too, and the detector on the surface is
neutral.  The bias probably would not play a role if the daughters were not
charged.  But there is  no bias  on the surface of the detector.

    MR. PHILLIPS:   That's true, if it's operated at a positive potential with
a negative potential or ground  on the outer surface of the detector.

    MR. KAHN:  Well, what does  that say about the radon?  You also see the
radon alphas.

    MR. PHILLIPS:   No,  I didn't.  In the counts that I did, and these were
integrated counts  over  fairly long periods of time—about an hour at
concentrations, let's say,  of 50 picocuries per liter, equilibrium around 25
percent—I saw essentially  no radon in the spectrum.

    MR. KAHN:  Can you  explain  that?

    MR. PHILLIPS:   My explanation—

    MR. GINGRICH:   Well,  I would  agree with you.  There's actually a paper
that's published on this,  and I could give you a reference to it.  It's by Dr.
Bob Fleisher and tells the contribution of radon and radon daughters, the
alpha particles, where they plate out, and so forth.  But, as I remember,
something like 30  percent  of our  signal in the cups, in the normal geometry,
is due to radon, and the other  65 percent is due to plate out on the side and
on the detector.  I can check those  numbers for you, but I can certainly give
you a reference to it.

    MR. KAHN:  But that doesn't seem consistent with his observation.

    MR. GINGRICH:   I don't know how  to explain that inconsistency, I'm sorry.

    MR. PHILLIPS:   There's another interesting situation that I haven't been
able to explain.  (I might add that  all of this is preliminary work; I have
not followed up on it because I got  busy with other things.)  If you put it in
a filtered cup, you see a  clean spectrum, and the heights of the A and C prime
peaks are about the same.   If you remove the filter at the same concentration,
the C prime peak is roughly twice the height, or as many counts, as the A.
Somebody tell me why.

    MR. GROER:  Bob?

    MR. HOLUB:  In answer  to the  first question, it depends simply on how big
the cup is.  If you see the radium A and C prime from the walls, then, of
course, you get contributions;  it's  not a clean spectrum.  But, you remember,
for instance, in the case  of Dr.  Vickers' experiment, he had a clean spectrum,
too.  But, of course, he  had a charge there.  And then the answer to your
other question is a matter of equilibrium.  If you don't put the filter there,
then you have a different  case than  if it's coming from the outside.

    MR. PHILLIPS:   Right.   But why would you expect to see twice the amount of
C prime as you did A, irrespective of the equilibrium?  I can't explain that.

    MR. GEORGE:  You said you have no filter in the cup?

    MR. PHILLIPS:   Correct.

    MR. GEORGE:  So you're getting the radium C prime from the decay of radium
A that's in the clean cup, and if you remove the filter you get an additional
radium C that's in the air.

    UNIDENTIFIED SPEAKER:   But you always have more A than you have C prime in
the air.

    MR. GEORGE:  I don't  know much about spectrometers, but I think that's
probably one of the reasons.

    UNIDENTIFIED SPEAKER:   How do you get more C prime than A in the air?

    MR. PHILLIPS:   I would believe,  though, that if you had the filter on the
cup and you left it for a  sufficient time, wouldn't you expect a hundred
percent equilibrium in the cup?   That's what I would expect.

    MR.  GEORGE:   I  don't  know.

    MR.  PHILLIPS:   So I would think that the highest count would be when the
filter was on the cup.  I think that you'd have an equilibrium situation.

    MR.  GINGRICH:   Excuse me, Chick.  But I'm pretty sure you wouldn't expect
a hundred percent equilibrium in that cup.  The reason for it is that you've
got a lot of plate  out surfaces.

    MR.  PHILLIPS:   Yes, but what I'm looking at is all the plate out.  Because
I'm seeing a clean  spectrum.  I'm seeing nothing from the walls.  Or,
essentially nothing.

    MR.  GROER:  Bob?

    MR.  HOLUB:  One more.  We can discuss this later in detail, but in
experiments done by Eugene Benson at the University of San Francisco they had
a cup, a collimator,  and  it was supposed to measure radon daughters in the air
together with radon by means of collimating in such a way that from one volume
you would get only  radium A and from the other radium C prime and radon.  In
other words, a perfect system.  However, the collimator—it's really a sink
for radon daughters for plate out—is just a few centimeters long in diameter,
quite small.  And still the daughters practically never made it, the alpha
particles from those  appropriate volumes never made it to the surface, to the
track etch.  If you look  at the depletion of concentration of the-daughters
close to the wall,  it simply is that there are no radon daughters.   It happens
very quickly and it must  be very complicated inside the cup.  You simply can
have practically no daughters there.  It can go anywhere almost.  I'm not
surprised that you  see all kinds of things.

    MR.  PHILLIPS:   Well,  I would agree wigh you, except the only plate out I'm
looking at is that  which  is on the detector.  I'm not seeing the plate out
from the walls.

    MR.  HOLUB:  That's right.  You are not taking care of the whole thing.
Maybe it's escaping here  and there.  Almost anything can happen.

    MR.  PHILLIPS:   Well,  you might be right, but I don't know why taking the
filter off of the cup makes it behave that strangely.

    MR.  HOLUB:  These cups, when you touch them they're very easily affected—
four thousand volts,  and  then inside it's plus and outside it's minus and
there are spots that  are  even higher.  So it can happen just by touching it.

    MR.  PHILLIPS:   Yes, but it's the same every time, and I don't think I
touched it.

    MR.  HOLUB:  No, no.   You have ten cups altogether and you take the one
apart from the other  and  they are all completely different.

    MR.  PHILLIPS:   I  realize that.  But every time I performed the experiment
I just described, the same thing happened.

    UNIDENTIFIED SPEAKER:   Got the  same  results regardless of the variance.

    MR. HOLUB:  Then I don't know.


    MR. GROER:  The record should show that.   I think without a blackboard,
it's sort of hard to follow.  I've  sort  of  lost track visually at some point
in this discussion.  It's  a very interesting question, but I don't think we
can settle it.  At least,  my visual imagination is not good enough.  Maybe
with a pad of papers it can be settled at dinner this evening, I don't know.

    Time passes  quickly.  I've asked in  the initial phase for input from the
audience.  I would now turn the discussion  around so the speakers can ask our
panelists to make comments, suggestions, raise questions.  I will start on the
right end of the table here.  Ron?

    MR. COLLE:  I think I've said enough.   I have nothing else.

    MR. GROER:  Okay.

    MR. WICKE:  I want to  ask Jim about  his experience with charging his cups,
if he did some experiments on this.  It  seems  to me an important factor
because when you make calibrations  you assume  that the distribution of the
daughters which are plated out in these  cups is reproducible.  I think it
might be due to charges that it is  not reproducible, and that will produce
some errors.

    MR. GINGRICH:  I think the whole question  of charge on the cups has been
overplayed a little bit and I say that because of our experience of making a
lot of measurements, not that we've made a  lot of precise charge measurements,
but a lot of radon measurements. We've  calibrated these detectors in
something like six or seven different chambers over various periods of time
under quite different conditions and so  forth, and yet the data all fall
pretty good on that calibration curve line. If there were charges that are
maintained on the cups for long periods  of  time, and other considerations like
that, I'm sure that we would see a  much  bigger scatter in our data.  And we
just do not see that scatter.  So that's one empirical reason we think that
the charge thing is maybe  overplayed a little  bit.

    MR. PRICHARD:  I'd like to direct some  questions to anyone who may have
been doing some work with  activated carbon  or  other sorbents in concentrating
or making integrated radon measurements.  In particular, I'm interested in
relative humidity and volume dependences.   Any comments?  Andy?

    MR. GEORGE:   Now, as I said earlier  this morning, we just really began to
look into this M-ll activator charcoal canister, and we haven't done too much
in terms of humidity effects and temperature.  Although I looked at the
temperature range that you find normally in the indoor environment—and I
assume for this kind of application for  getting data like screen data, I
thought that was sufficient.  As I  said, this  area needs to be investigated.
So I don't want to say this technique is ideal until you check it further.
I'm only suggesting that it may be  a convenient technique, and it's very easy
to test this technique.


    Now Jess Thomas back many years  ago did  some of these experiments, and I
know there is some application,  and  I  thought  I sent you a copy.  See, I don't
forget.  I sent you a copy,  and  I  thought maybe you found the information that
you needed.  If you are not  happy, I guess you have to contact Jess Thomas,
because he spent a lot of time on  that.

    MR. PRICHARD:  I was referring less to the cases in which you actually
pull air through the canister.   Where  you just use the canister, rather, a
volume of charcoal as a passive  absorber, such as the previously mentioned
unit and the M-ll.

    MR. PACER:  We did some  time tests using charcoal as a flow-through device.
At room temperature and humidity, which in Grand Junction is probably about
10, 15 percent (we just wanted to get  some initial numbers) one M-ll had
absorbed something like 90 percent of  the radon that it was exposed to within
the first five minutes, and  then it  absorbed up to 97 percent after the first
hour.  The other units were  a lot slower.  It  took about 20 minutes to absorb
the first 85, 90 percent, and it took  on the order of four hours to get to 97,
98 percent.  So, there is a  time dependency on that absorption, and I think it
has to do with the difference in the volumes of the charcoal.

    MS. HINKINS:  I just have a  couple of comments.  Since I'm a statistician,
my comments are mostly based on  the  data.  This is probably not a discussion
topic and maybe it's something you all do already, but it seems important to
document your data in the sense  that if  you want to go back and look at the
calibration results you can  find not just the  amount of time that the devices
were in the chamber, but the actual  dates they were in the chamber—not just
the total exposure, but the  radon concentration and the exposure time.  And as
much data as you can think of to put in.  That's very helpful when it comes
down to analyzing the data.

    And one other comment on calibration.  I find the method of displaying the
calibration data a little confusing.   From my  background I look at it as an
independent variable and a dependent variable  and you're trying to fit a
slope.  And, for a Poisson distribution, a better estimate than the sum of the
ratios is to take the ratio  of the sums.  Those are the only comments I have.

    MR. PACER:  I just have  two  comments.  One is, I agree that we need some
radon standards and I think  we'd support NBS doing that.  The other thing is,
one of the things I was interested in  coming to the meeting for was to learn
something about radon flux measurements  and their calibration, especially
since EPA does have preliminary  standards on radon flux.  And I really haven't
heard anything at all about  that.

    MR. RATHBUN:   I would just say that  when I measure radon flux, I do it
with an M-ll, and when I want them calibrated  I just send them to Andy George.

    (Laughter. )

    MR. GEORGE:  I guess you want to put me on the spot again.  Well, we have
been doing radon flux on and off, but  we really don't do it on a regular
basis.  We tried both the accumulation method and the radon flux canister:
Both methods work equally well.   If  you want data in a hurry, you do the
accumulation method because you  can  do it on the spot.  All you need is a can
you turn upside down.  You invert it over the soil with a couple of vials and
let it emanate for maybe a half  hour,  no longer than two hours or you get into
problems.  You start building the radon in the can and then start pushing that
into the soil, but you have to be careful.  In our experiments with this, we
found that for different size cans the time that you can allow for this
build-up varies.  So, we usually limit it to two hours.  From soils, two hours
is sufficient to accumulate enough radon, then you can transfer it to a Lucas-
type flask and count in the laboratory, and then after that release' the flux.

    Now, on the other hand, the  canister is a very convenient tool, but you
need a much longer time for accumulation, usually, on the soils, something on
the order of 24 hours.  From the surfaces and the concrete basements or the
walls, you need approximately 48 to  60 hours.  In both cases it works fairly
well, and canisters can be shipped back and forth very easily.  In our
experiments, we found that if you seal them after you take them apart from an
illumination area you do not lose any  radon.  We counted them immediately
after we moved them and then we  sealed them and counted them over several days
after we closed them up.  We found no  loss of radon.  The reduction you see is
sort of a decay of radon.  So, they  both work well, and you can measure almost
any surface with these two techniques. There is no limit with this.
Sensitivity, as I said, with the charcoal canister is about two-tenths to the
minus eighteen curies per centimeter squared per second.  I don't know what
EPA is using as a standard, but  I think one of the numbers they mentioned was
like two hundred times the minus eighteen curies per centimeter square per
second, after you clean up the tailings.  Now, we found numbers higher than
that in soils in the United States.  That's background soils.  They can go up
as far as 500, with averages about 200 to 300 over an entire year.  So, I
don't know whether that is above background or is just 200; I don't know what
the standard is.  Does anybody know  the answer to that?

    MR. BERNHARDT:  In response  to your question, the proposed number right
now is two picocuries per meter  squared per second.  It is within the observed
background values, yes.  I think it's  intended to be excess above that.

    Could I just make a couple of comments on what we have done on flux
measurements?  I don't have a lot to add to what Andy George says.  The main
thing we have used in Las Vegas  is the accumulation technique.  I guess I'll
note that when you're applying it more to a research thing or trying to put
more effort in I like the concept of taking several accumulation points so you
plot a slope of the accumulation in  the vessel and you get a slope-type thing
versus just a single accumulation point.

    We've done a little bit with the charcoal cartridges to get comparison.
The comparisons have been reasonable.  We really haven't pursued it as far as
we would like.  I guess agreeing with  your concept of a problem, the
accumulation vessel is only good for periods of like an hour to two hours at
the most.  It is hard even, therefore, to compare that to a charcoal

cartridge.  I've compared them over  the  same periods of time, but you're
pushing the sensitivity of the charcoal  cartridge by trying to only do the
charcoal for several hours.   It's  difficult to do an accurate accumulation
over the one week period that you  need for the charcoal.  So, to get a direct
comparison is a bit difficult, but it's  not impossible.

    MR. GEORGE:  Now, as far  as comparing these two methods, we do have a
radioactive concrete, with which you can do this kind of an experiment.  It's
a very homogeneous mixture, and it's big enough to use flux cans and
canisters, and then you can remove the canisters.  You can take your samples
with the can and then remove  the canisters at different times, and you can
determine this kind of a relationship  between the two.  I agree with you; it's
very difficult.  In the field one  time we had to stay overnight to remove
canisters and take the flux can measurements, taking a measurement every two
hours.  You have to identify  a whole area and place different cans every so
many hours.  That's how we found out that you cannot exceed, let's say, in
this area, five hours.  You start  leveling off and you're getting out of

    MR. COLLE:  An alternative to  sending it to Andy that seems to us to be
quite feasible would be to actually  develop the relatively small artifact
standard for radon flux, which would be  a physical surface that would not be
too large, that literally could be sent  to various laboratories for their own
standard.  We've looked into  this  in a preliminary way, and, although a lot of
development work has to be done, it  seems that it is quite feasible and that
it would have a dynamic range over which it would work and that one could
literally dope this.  It would be  a  ceramic material of perhaps a meter by a
meter.  Thickness is somewhat indeterminable, but it would be impermeable on
five sides.  The exhaling surface  is only on one side and you would only be
allowed to use an inner portion of the surface area.  It would be suitable for
charcoal canisters or for small accumulators; and it would have a dynamic
range under which it would work.

    We have considered this quite  extensively.  If someone would like to talk
to us sometime about perhaps  picking it  up, developing it on your own, we'd be
quite willing to talk to you  about it.   It does seem like it's feasible.

    MR. ROESSLER:  I'm surprised we  haven't heard from either Chick or Sam
Windham about this whole question  of flux.  It's true that both the
accumulators and the charcoal cartridges do a pretty good job of measuring the
flux that occurred during the time interval for which the device was deployed,
or if you're doing successive sampling you can sample a week and get a good
measurement for a week.   But  the real  question is whether this is a useful
number for us because of the  temporal  variation that by hour, by day, by week,
whatever, seasonal variations.   Usually, the meaningful parameter for
evaluating, in our case, a piece of  reclaimed phosphate land or some other
situation is the long-term average.  And we can't think of anything short of a
long-term sampling experiment to tell  us what the long-term average is.  So
that's the real problem as far as  I  can  see.  What is a measurement that tells
us what the long-term average flux is?  These sorts of measurements don't seem
to be that unless you repeat  them  over long time intervals.  The question
probably is, then, is there some other parameter that's more stable with time?

    MR. KEEFE:  One of the problems  that has perplexed me for quite a while is
the problem of plate out,  which  I've mentioned before.  The most disturbing
part of it is the inconsistency  of the  various different answers I get from
different laboratories and different people.  It seems as though there are a
thousand different answers as to what effect humidity, temperature, charge and
so on have on plate out.   I think that's an area that needs to be tied down
before we can develop good instruments  that will accurately measure working
levels of radon daughters.

    MR. GEORGE:  I wanted  to say something about plate out in the design of
instrumentation.  Our experience so  far shows that plate out is only related
to the number of particles in the air.  Our experiment (we have a paper on it
in Health Physics coming out soon) shows that if you can measure the particle
concentration in the room, you can almost predict plate out exactly from
theory.  All you have to do to test  if  there is any plate out effect with your
instrumentation is to take your  instrument, put it in a calibration chamber,
generate a known particle  concentration and, if possible, a known particle
size, and then repeat this experiment with your instrument at different
concentrations and go from high  to low.  If you see any reduction in your
measurement, let's say of  working level, then you know you have plate out.  If
you have no reduction, then you  can  assume you have no plate out.  I suggest
that it is one way of testing for plate out effects.  As a matter of fact,
every time we design an instrument we go down to five thousand particles per
CC because I don't think you're  going to find many atmospheres with particle
concentration less than five thousand particles per CC.  So I consider that a
lower limit of particle concentration to which you should subject your
instrument.  I don't know  whether you can do that in the mine because you have
to have a filter to make sure that if you want to test for low particles you
have to filter the air and perhaps inject some known aerosol concentration so
you can maintain your chamber at the same level for some time until you get
reasonable results.

    In a chamber this can  be done very  easily.  You can put in an absolute
filter and then you can inject your  aerosol, and you can control it very
easily.  Particle concentrations from five thousand to a hundred thousand are
very typical.  And I think most  of our  data show that you begin to get some
plate out in some instruments that we've seen around between five and ten
thousand particles per CC.  About ten thousand we couldn't see plate out in
any instrument.  So if you can do that, I think it's a beginning.  If you can
check it around five and ten thousand and you don't get a plate out effect,
you can almost forget about it.   But if you want to be really scientific, go
down to five thousand or maybe three thousand.  Then you can argue with

    MR. GROER:  Jim Gingrich.

    MR. GINGRICH:  I thought it  might be interesting to set something in
perspective.  You all have read  some of the Three Mile Island accident
analyses and so forth that indicate  a very low potential number of deaths from
the Three Mile Island accident.   Right  now, based on very limited data—and
it's obviously very limited—on  eastern Pennsylvania, we have about 15 percent
of the homes giving radon  values above  20 picocuries per liter.  If you

convert that to working  levels  (again, above .1 working level) and use the
same criteria that were  used for  the number of people in the area around Three
Mile Island and assume that  15  percent of the homes have radon levels of this
magnitude, you can see that  the potential lung cancer deaths are rather large.

    MR. GROER:  Thank you very  much, Jim.  It is a very interesting slide, but
I feel very strongly we  should  stay out of the area of epidemiology.  There
are very special meetings for this sort of thing.  Please don't take me;wrong,
but I think we should stay out  of epidemiology.  There's a meeting, as a
matter of fact, coming up in October in Golden.  Many of you know that.  So
there will be some discussions  on this very topic.

    Every one of our panelists  has had a chance to comment.  I'm coming back
again to the audience.  Any  hot items you would like to see discussed,

    MR. TOOHEY:  This isn't  a very hot item or question, but it's something
that you might be interested in.  And that is, if you want to take grab
samples of radon in houses,  how do you do it?  Normally what we've done in the
past is we take a Lucas  cell out  and have somebody bring it back to the lab
and count it.  Well, that's  nice  if you're doing a hundred or so houses as
we've done before.  But  in the  next step in our study of ambient radon levels
in Chicago area houses,  we want to do between three and five thousand; namely,
by giving everybody who  works at  Argonne a sampling device to take home, take
an air sample and bring  it back in.  So we need something that's fairly
rugged, will hold a vacuum,  and is also cheap.  And what we have come up with
are these things, which  you  can pass around if you want to look at them.

    They're basically aerosol paint cans, and they're 80 cents apiece.  I
don't know exactly what  their volume is.  My guess is it's about three-tenths
of a liter or something.  We get  them from a factory; they say they are
evacuated.  We've measured that they only get them down to about 40
millibars.  So we pump them  down  to almost zero.  We take them into a house,
give them to somebody, and just tell them, "Hold it in the middle of the
living room, press the button."  Instead of something coming out like it does
in your normal aerosol can,  something goes in.  We then take them back to the
lab, hook it up to an evacuated Lucas flask, let the two equilibrate, and then
count the Lucas flask as normally.

    This week is our first real test of these things under more or less field
conditions.  We filled four  of  them yesterday over at Ed's chamber at EERF.
He's going to fill another four of them tomorrow for me.  I'll take them back
with me and we'll count  them next week and I'll let you know how well they

    UNIDENTIFIED SPEAKER: How  do you hook them up?

    MR. TOOHEY:  To what, the Lucas?  The cylinder?  Okay.  Well, I'm glad you
asked that question, Senator.  We have another transfer device.  This end will
simply connect to the Lucas  chamber somehow, with a shorter piece of pipe, of
course.  I brought this  one  down  to hook up to the chamber.  On the other end
there's a piece of brass tubing inside a little Tygon sleeve.  We take the
button off of them.  This just  fits over the needle valve on here, you depress
that, and it will hold it down  and they equilibrate that way.  We did these on


a sampling port at the chamber,  just by hooking  the tube up and then slipping
them over and pushing it down.   We are making  the  assumption, which we'll test
when we count them, that the valve wasn't depressed until it was actually in
contact with the radon atmosphere.

    But, as you probably know,  Lucas cells are kind of expensive.  They're
fairly sturdy but not all that  sturdy, especially  around the windows; and they
don't hold a vacuum for more than a couple of  days.  So we hope this will
solve some of our problem with  it.  Maybe by the next time any of us get
together I'll have some data on just how well  they work.  And maybe we'll put
a note on something fairly brief on it in Health Physics or something.  But we
think this is one way we're going to have of doing a lot of grab sampling
fairly conveniently.

    MR. COLLE:  I just have a question, and it's something that we have been
burned on ourselves relatively  recently.   I seriously question when you hook
up this can to the cell whether you're indeed  going to get first mixing and
then any type of equilibration.   I really wonder if that's going to work.
Before you go out making five thousand, I suggest  that you check that
transfer, because those transfers are more difficult than they appear on first

    MR. TOOHEY:  On, I agree completely.   That's one of the reasons I brought
them down this week.  A twofold thing.  First  of all, see how well they hold
up under field conditions.  We  presumably will know the concentration in the
chamber.  Sue Fisenne also took some samples in  the bomb-type devices and some
other samples have been taken.  So we'll have a pretty good handle on what the
concentration was and then we'll see what we get.  So, maybe they will work;
maybe they won't.

    Certainly we will validate  the technique before we go into the sampling.
But that's another year down the pike probably.

    MR. PRICHARD:  Is there some particular reason you chose to go this route
rather than, say, traditional plastic bag of some  size which could then be
blown up and returned to the laboratory for, well, possibly even concentrated
techniques?  Here it looks like your effective sample size is going to be
somewhat less than 100 ML, accounting for the  partial pressure.

    MR. TOOHEY:  That's probably true.  The primary reason is to do it
extremely simply.  We don't have to send somebody  out with these things to
take the sample.  It's a difficult question of what it is you want to do.
Granted, grab sampling has its  limitations.  We  all know about diurnal
variations, this, that, and the other thing.   Arthur Scott will tell us
tomorrow maybe how to use grab  sampling to decide  what to do.

    But we thought the first thing we should do  is just get at least, one,
preferably several air samples  out of a lot of homes and see what we've got
out there—because we've been surprised by what  we've seen so far in a few
homes.  The highest we've seen  is about 60 picocuries per liter.  But, 10, 20,
30, is becoming kind of common.

    This is a first stage of  the project.  Then we want to look at this and
decide how to do a more  accurate measurement if need be, or more houses if
need be.  And look at some other things.  What really determined the level of
the house?  We're tending right now to think it's the geological
characteristics of the soil underneath it.

    MR. PRICHARD:  Well,  I agree with your goal of doing some preliminary grab
samples.  I may be missing something, but I don't see what's so difficult
about sending people home with a bag of some sort to blow up, especially when
you consider your potential problems with loss of vacuum or transfer problems.

    MR. TOOHEY  Well, I'm not that familiar with the bag technique, so I can't
really comment on it. But it seems to me that there would be a lot more
problems there than handing somebody an aerosol can, which he's already
familiar with, whether it was paint or bug spray or whatever, and saying,
"Just hold it up and push the button."

    MR. COHEN:  You have  a volume 95 aquarium pump and a $12 bag—have them
taped together—and tell  them to put the plug in the wall and leave it there
for five minutes and pull the plug out of the wall and that's the end.

    MR. AKERBLOM:  I would go back to a topic that we discussed earlier, the
reliability of instruments.   When you buy different field instruments—track
etch, meters for measuring, you are rarely told how they react to humidity,
temperature, and so on.   I think that we should go together and actually tell
the producers of the equipment that if they sell us the equipment they have to
tell us much more.  We have found it very difficult to get such information
from producers.  We tried in  South Africa to get some figures on the Rowak
caps.  They haven't given them yet.  They say they work.  Okay.  They work.
But how?

    MR. GROER:  Okay. Any manufacturers here who want to comment on this?


    MR. COLLE:  I think  the presumption was made that the manufacturer knows
and is not telling you.   In fact, the manufacturer does not know.  I mean,
most of the instruments  that  are manufactured were usually developed by
someone like EML, in which case they got the specifications, mass produced
them, wrote up a sheet for it and a price list, and distributed them.  And
that's the extent of what they know about them.

    MR. AKERBLOM:  Yes,  that  is perfectly true.  They don't know.  But they
don't seem to test it either. They should have their instruments tested
before they sell them.  I found this with track etch.  We had to start in
Sweden to test the track  etch.  The manufacturing company should know if it
was good enough.  If we  do not put pressure on the manufacturers, we never
have any good instruments.

    MR. GINGRICH:  I guess one of the comments from our end of the line is
that we welcome questions of  any kind.  Quite often people will ask us
questions after they've  been  using our services for a while, not knowing that

we may have already solved that problem for  somebody else.  So if you've got
questions about our devices or how to use them,  we would more than welcome any
kind of communication and questions that may come up.  There are still a lot
of things we don't know and would be happy to admit it.

    MR. GROER:  Silence?  I'd like to bring  the  discussion then to one
question that has arisen for me.  All the methods assume, among other things,
that once a radon daughter atom or aerosol particle attaches to the filter
that it really stays there.  Is that really  so?  What do you people here think?

    I'm referring especially to a paper by Johanson.  Some people have
probably read this paper.  Arid I was really  soliciting comments on this paper.

    MR. HOLUB:  I think he concludes that nothing really gets out.  If it
does, it's in a vacuum.

    MR. GROER:  Yes, that's one point.  This paper refers to a study in
vacuum.  Now, my experience has been that in air I have no evidence that
daughters are leaving without being asked.  So,  once they're in the filter
they stay there.  I have a very poor understanding of the mechanism by which
filters work.  I'm thinking of membrane filters. I thought that it might be
worthwhile to get into this if there's interest.

    MR. HOLUB:  The University of Illinois,  similar to Dr. Postadorf"s group
in Germany, is now examining the possibility of  reentry from rough surfaces,
whether it's grass or rocks sticking from the wall in a mine, into the
airstream after recall.  So it's definitely  not  very serious, and it's related
to this problem.  But the assumption for these studies is that there must be
an airflow because other people already found out that if you have just recall
then it simply returns to the surface where  it was before.

    MR. GROER:  Okay.  Well, there seems to  be then a universal agreement, in
the absence of any other comments, that daughters do not leave the filter.
Okay.  Then, maybe I can go on to the next assumption that concerns me—
constancy.  Now, I'm turning directly to Bob. I know that some years ago,
three or four years ago, you started an investigation into time dependent
concentration during the time of sampling.  In one of the papers I wrote it
said, "Bob Holub, private communication." Have  you followed this subject?
Where does the matter stand?

    MR. HOLUB:  I'm very sorry. It's still lying on the desk and nothing

    MR. GROER:  Okay.

    MR. SCOTT:  Some years ago I did a study on  atmospheric radioactivity, and
this was piggybacked onto an atmospheric dust collection thing.  The results
then were somewhat disappointing because these measurements, because they're
collecting dust, have to take place over about an hour at least.  There were
evidently changes in concentration there, and, despite analyzing everything
with a multi-Z-squares fit program, we got nonsense answers out of it.  This
was particularly true if the radium A concentration increased near the end of

sampling, because you then got radium A being one quarter of the concentration
of atmospheric radium C or B.   So,  it's hard to say how large an effect it may
have, but certainly if it does vary,  it does make a large effect.

    MR. GROER:  Okay.  Thank you very much.  It is obvious that to evaluate
this really quantitatively, by measurement, is impossible, because all of the
methods you have assume constancy,  and, if you want to check constancy, which
methods are you going to use to do  that? So, it has to be done indirectly.
And I have seen instances where this  assumption was certainly not true.  One
way of checking is to count while you are sampling.  In this case we put the
detector—it was a surface detector—underneath the filter paper.  If the
concentration is constant, then the radium A count you observe, say, during
the first minute of a two-minute sampling to the count to observe in the
second minute should be a certain ratio, within, of course, statistical
fluctuations.  If that ratio is not what it should be, assuming you take
proper care of the statistics, then I think you have an indication that this
assumption is not true.  So that's  what we have done in this area.

    So what this means to me,  and I think you will agree with that, is that
there should be a trend to go  to smaller instead of larger sampling times when
you want to measure low concentrations. If you agree with that, I'll go on to
something else.  Okay.

    MR. COHEN:  This is a practical question.  If the detector is underneath
the filter paper, what supports the filter paper?  The filter, I mean.

    MR. GROER:  If the flow rate is not too high, sometimes you don't need a
support for this.  Filter paper is  really a plastic material.  Now, sometimes
you can support it with a grade of  a  certain reasonable mesh size.  So, you
lose a little bit of detection efficiency, but it doesn't really make much

    MR. KAHN:  You could certainly  test this, if you had two collectors, by
offsetting the collection by a minute or two.  See what happens then.

    MR. GROER:  Good point. I think  maybe I don't quite understand you, but
the problem there would be, can the two collectors be exactly at the same
place in space?

    MR. KAHN:  Obviously not.

    MR. GROER:  Right.

    MR. KAHN:  As a matter of  fact, if they were in the same place or pretty
close to the same place, they  may tend to interfere with each other regarding
the air pull.  But, if they did seem  to be consistently similar, even though
they were a couple of meters apart, that would reassure you.

    MR. GROER:  I agree with that.  We've tried to do it in one place, and
that's the only way I could think of  looking at the ratio.  Any more comments
on that?

    MR. NAZAROFF:  It seems to me that there  are  two possible time trends that
you would be interested in.  In one case,  if  the  fluctuations are of a fairly
high frequency compared to the period in which you're measuring and there is
no long frequency bias (I can't say this with total assurance, but it's my
impression), it won't make very much difference in your analysis procedure.
That is, if you collect a lot of radium A  now and don't collect any for a
second interval, then a lot and then a little and then a lot, and there are
many of these intervals during your collection period, you are getting to the
point of talking about discrete atoms—if  you take it down to fine enough time

    The second sort of case where you would get into trouble is when the
period is of the same order or greater than your  sampling period.  Some
evidence can be gained in that question by making rapid successive
measurements of radon daughters at one place. If you do that and see big
differences over a fairly short period of  time, then you begin to suspect that
maybe this is going to be a significant problem.

    I've done this sort of thing in one house, but the measurements were only
made at 40 minute intervals.  I could see  trends  of longer frequency than
that, and they were not sufficiently rapid to indicate that there would be a
problem with it.  I guess that doesn't conclude that there couldn't have been
things going on in shorter time intervals  that could have caused the problem.

    MR. SCOTT:  Maybe I could comment on this question of fluctuation in time
periods.  The important question really is, "Under what circumstances are you
measuring?"  If you're measuring outside,  then we would know what the
frequency distribution of fluctuations in  the atmosphere is.  If you're
measuring inside, we have a totally different spectrum of frequency—a totally
different spectrum of frequency variation.

    Now, given that our main interests, and I'll  make an assumption here that
we are measuring inside, we know that we have an  effective filtering mechanism
which cuts out the shortest periods of all.  So the real question is, "Are our
sample times sufficiently short that we do not expect there to be a
significant variation?"  We have done measurements with the Nork gas tracers,
with spot samples as short as 30 second intervals; and you can show that in a
house there are indeed fluctuations of the order  of 30 seconds.  But, in fact,
the major fluctuations are of a considerably  longer period than that.

    A continuous radon monitor run so as to give  repeated spot samples
demonstrates that in a house there can be  significant changes of concentration
in a period of a half hour.  In a room with a radon source in it the radon can
change by a considerable number of percent in as  short a period as five
minutes.  Now, we do have an advantage by  measuring daughters in that things
are averaged to some extent because the daughters we measure now are not due
to the radon that is in the room now but due  to the radon that was in the room
ten minutes ago, 20 minutes ago, 30 minutes ago.

    So, I suspect that if you don't put your  sampling period much over about
five minutes—and, in fact, I would raise  another question, Why would anyone
measure for more than five minutes?—that  is  quite adequate sensitivity.  The
problem is not likely to be of great significance indoors.


    MR.  GROER:   Did you say  the problem will not be of great significance

    MR.  SCOTT:   I  think so.

    MR.  KEEFE:   In some of the measurements that we've made inside the home at
Argonne, you can definitely  see the effects from an air conditioner fan
turning  on.   (We take  measurements every ten minutes or so.  We take grab
samples  with a  nine minute measuring cycle, with a three minute pumping time
or sampling  time every ten minutes.)  Up goes the radium A concentration and
the equilibrium shifts.  I suspect you can see as much just by somebody
opening  the  door,  disturbing the equilibrium in a home.  If you have these
instruments  in  a home  and you are making measurements, I suspect that you're
going to see the effects of  a change in the equilibrium concentration of the
radon daughters during that  three minute time.

    MR.  HOLUB:   What we propose is simply to simulate it—make a step function
first, of the sampling and then nothing, and do it numerically.  The Canadians,
the group in Toronto,  did something like that last year.  Of course, it makes
a difference, increases the  error of all these methods.  I forgot exactly how

    MR.  GROER:   Thank  you.   I think those of us who are inclined to modeling
and playing  around with the  numbers should really try to get at least a
feeling  for  what happens if  the concentration changes during the time of
sampling.  There were  some experiments, but, as far as I know, maybe with the
exception of this  Canadian attempt, nobody has tried to get a quantitative
feeling  via  modeling this effect.  Maybe I'm wrong.

    MR.  SCOTT:   As a matter  of fact, I'm working on that now.

    MR.  GROER:   Thank  you.   Any more comments on this matter?

    MR.  PRICHARD:   While certainly not ignoring some of the interesting
computational problems that  would be introduced by changes in the daughter
concentrations  or  equilibriums while you are sampling, I don't think I'd be in
any danger of violating your suggestion that we not get into epidemiology by
pointing out that  the  real bottom line in all the measurements we're taking
does relate  to  long-term averages.  You know, lifetime exposures to people.
The end thrust  of  all  our computations should be those measurements which help
us determine those life-long averages.  While there are many interesting and,
in fact, fascinating technical problems, I would not want to see us lose sight
of the forest on account of  the trees, let alone the branches.

    MR.  GROER:   Thank  you, Howard, for this comment.  Although I don't want to
get into epidemiology, it is the working level months that are of concern.  In
the same vein,  let me  point  out that there's another problem.  The question
is:  Over what  period  of time—and after this I would like to leave this
topic—should you  integrate  the exposure a miner gets?  Should you integrate
until the cancer was diagnosed, five years before the cancer was diagnosed,
ten years before that?  That's really a problem that has to do with
carcinogenesis, and that's not solved at this moment.  But I agree that
epidemiology is based  on accumulative measures.  So there's no doubt about it.


    Okay.  Now, on epidemiology,  any  comments?  If not, time is getting
short.  I would like to give the  panelists one last opportunity to express
their views, if they wish to do so.   I will  start again on this side of the
table.  Thank you very much.

    (Whereupon, the meeting was adjourned at 5:20 p.m.)

                      OPEN DISCUSSION II:   AUGUST 28,  1981
Open Discussion;  Measurements and Related Topics

Moderator;  Arthur Scott

Panel;  Speakers for Measurements and Related Topics

                   Richard E. Toohey
                   William W. Nazaroff
                   R. A. Washington
                   Charles T. Hess
                   Bernd Kahn
                   Carole Wilson
                   Gustav Akerblom
                   Andreas George
                   Stephen Rudnick


    MR. PHILLIPS:  The moderator for  the  discussion  this afternoon is Arthur
Scott.  I'm sure Arthur by now needs  no introduction, but I do have to tell
you my favorite Arthur Scottism I heard him say once and I've thought a lot
about during the last couple of days.   He said, "The subtle things we see
right away; it's the obvious that takes us so much time."

    MR. SCOTT:  Well, good afternoon  and  welcome.  I see there are almost more
people up here than there are in the  audience.  Nonetheless, I'm sure the
panelists will do their best.

    Now, the purpose of a moderator in nuclear physics is to slow down the
neutrons so that they can interact with uranium, and I suspect the purpose of
a moderator in a circumstance like this is to slow down ideas so that they can
plummet into the fertile minds and imaginations of the members of the panel.
There are some questions left over from this morning.  So, it would seem to be
a good idea if we started with questions.  As moderator, I will take advantage
of my position by starting off with a question statement, and I'd like to
direct this to our Swedish colleagues who were speaking of introducing
regulations for building.

    Now, I happen to have had some experience with this, and I would be
interested to hear what answers they  have on two points.  One of them is:
What mechanism is to be set up to verify  that the buildings indeed do meet a
particular standard?  Second:  What is to happen if  the building does not meet
it?  I'm dealing with new construction here. What mechanism is to be set up
to check the houses?

    MR. AKERBLOM:  Well, there has been a long discussion within the Radon
Commission in Sweden on how you should go on with this.  What should be the
standard and how should you actually  tell people that they are below the
standard limit or over the standard limit.  We found out that there was no
easy way to do measurements in houses and get the correct answer.  You have
also shown that today.  You can't go  in and take grab samples, and you can't
actually rely on track etch films to  give you a good answer, especially if you
are measuring both radon daughters and radon at the  same time.

    We know enough to say, "This is regarded to be dangerous to people living
in houses like this; something should be  done about  it."  Should we be
concerned about the exact number? Or could we say that in an area which
exceeds approximately such-and-such amount of radon  daughter concentrations,
we would not allow people to live? In Sweden the amount considered
uninhabitable is 400 Becquerels per cubic meter.

    But whose definition of danger do we  accept?  It's not we who are
measuring—we can only give our figures to the people who are working with
it.  We measure as well as we can. Should we be measuring more?  It costs a
lot.  You do seven measurements the whole time, and  still you wouldn't be
sure, although much more sure than when you have done it.  We have seen
Scott's excellent work, yet he doesn't know everything.  It's not low, it's
not high.

    No, you leave it to the health authorities.  And tell them, "This is in a
level where we think it's  high;  it's proved it's higher than it should be
anyhow."  And, if you measured over that level, you'd declare that place
unsanitary.  Once it is declared unsanitary, they have the means to go to the
house owner and say, "This place is unsanitary."  If he owns the house
himself, he can do something  about it.  Perhaps he can just let it be.  The
laws are not so hard on that.  But, if he's got children there, you can always
start arguing with him about  the risks for the children, and, in Sweden, where
child advocacy is strong,  it  is  not just his problem.  The country also has to
take into account the risks for  the children.  If he's renting the house to
somebody else, of course,  it's an easy thing.  Because then he has to do
something about it or he can't rent it out.

    But this has been a discussion of people not willing to do something.  And
this will be difficult for a  long time.  Until now we have not met up with
anyone who says, "I don't  want to do this."  "I don't want to follow your
instructions."  They are happy to know that you can do something to the
houses.  And you could tell them to enforce the ventilation in the house where
concrete is used.  But that would be a simple solution.  But, if you were in a
house where it comes from  the ground, it's more difficult; it costs you quite
a lot.  So to give them a  start, the State has provided loans for people who
want to do something to their houses.  The problem is, of course, that you
don't actually know what to do.   This is going to take some time.  But we are
coming out with some advisements which, I think, will be successful.  They
have been working in Canada,  and I think they will also be working in Sweden.

    Did I answer that question?

    MR. SCOTT:  Yes.  Thank you. For the people who have just joined us, we
are still beating the questions  on this morning's session.  If anyone does
have any questions, would  you please quest.  Yes.  The first one to the
microphone gets it.

    MR. RATHBUN:  I just have a  comment on the last presentation of the
morning session, Mr. Washington's presentation.  He was talking about the
possibility of going to lithium  fluoride dosimeters as opposed to the calcium
flouride-dysprosium which  he's now using.  I would suggest that's not a good
step since I had to go the other direction.  You'll find that the calcium
flouride-dysprosium has at least 30 times better sensitivity for gamma, which
you're not really interested  in, but it also has considerably better
sensitivity for alpha, which  you are interested in.

    I used a .015 thickness,  which is thinner than most TLD's that you get.
Also, concerning the nonlinearity of the calibration factors, I also found
this.  But I found it to be much more pronounced at high levels than I did at
low levels.

    MR. KAHN:  In connection  with this, have you checked on the intrinsic beta
background of the cans, just  to  see whether there would be a zero, non-zero

    MR. RATHBUN:  Yes.  Of course, I used a background dosimeter and always
subtracted the difference  out.

    MR. KAHN:   But they're not outside,  are  they?  They wouldn't be exposed to
the surface of the can.

    MR. RATHBUN:   They're not, of course,  in the ionization chamber with the
other chips, although I  have run some  right  in  the ionization chamber—but not
on the probe.   I  found that the background measurement isn't much different
above the case of the perm or below it or  beside it.  The place I had the
biggest problem was unfortunately on one of  our probes.  It's designed to hold
two background dosimeters on the side, and they are  in a different plane, of
course, 90 degrees out from the active dosimeters, so they don't get as much
background if you stand them on edge like  that. But, as long as you keep them
flat, you're okay.

    MR. KAHN:   Well, I was just thinking that if there was a base background
due to anything within the ionization  chamber itself, that obviously would
cause the curve to go, not to zero, but  to the  positive number.

    MR. RATHBUN:   Yes, I think if you  use  the background chip, you'll subtract
out not only the gamma,  but also the beta.

    MR. GINGRICH:  Back again to Mr. Washington's paper of this morning.  I
spent some time with the Australian Radiation Lab a  couple months ago, and
they had done a rather thorough electrostatic field  evaluation of that
particular device and came to the conclusion that most of the problem of the
huge variability of the device was due to  the placement of the TLD in the
electrostatic field.  They did not feel  that that device would give them the
kind of confidence that they wanted in making the measurements there.  They
had done quite a bit of testing on it.

    MR. SCOTT:  First one to the microphone  gets it.

    MR. PHILLIPS:  This is to Richard  Toohey.  I've  seen those slides before
on the equilibrium values.  Have you verified those  values, using any
different instrumentation than the two you described this morning?  I think,
namely, you are looking at radon with  the  Wrenn chamber and the working level
with the Environmental Working Level Meter.

    MR. TOOHEY:  Yes and no.  The Wrenn  chambers are.  We always take a few
Lucas flask grab samples, and they seem  to work pretty well.  But we have not
used another device to cross calibrate or  check the  working level monitor.

    MR. PHILLIPS:  Thank you.

    MR. SCOTT:  I'd just like to add a comment  to this.  First of all, the
working level—equilibrium fraction ratio, rather—is defined as 100 times
working level over a radon for a co-existing concentration.  So, in fact,
under normal circumstances, no parent  measurements  actually measure
equilibrium fraction—because it is extremely rare  for people to take air
through a filter and put it into a plastic bag  and  then take the radon sample
from the plastic bag.

    So, we in fact are bedeviled in  this question by a further source of
variability.  First of all,  there's  the spatial variability because you're not
measuring the same space;  and,  secondly, the sample times are so far
different.  I mean, take  a short sample of two minutes.  It only takes a
matter of a second or so  to  fill a Lucas flask.  And, your sampling theory
says that, your estimate  must be different.

    MR. TOOHEY:  I agree  completely.   In fact, it's worse than that.  We found
that our particular Wrenn chamber takes about an hour to come to equilibrium
with an external radon concentration.  Most of the measurements I showed were
hour-long counts.  So, obviously, we're seeing an hour-long average, whereas
the working level monitor takes a two  or three minute sample and counts for
ten minutes or whatever it is.   This is certainly a source of some of the
variability we see.  But  I think on  a  long-term average, hopefully, that will
even out.

    MR. PHILLIPS:  I'm aware of all  that.  I guess the thing that disturbs me
about some of that data is the  extremely low radium A values when compared to
the radon.

    MR. TOOHEY:  Well, most  of  it is on the surfaces.  We have made the
measurements, and we can  show that the missing radium A has, in fact, plated

    MR. PHILLIPS:  I've just never observed that, I guess.  If I make
simultaneous radon and working  level measurements in a room, I see a lot
closer equilibrium of radium A  to radon—in any situation.

    MR. SCOTT:  We had a  number of papers this morning that touched on the
topic of plate out generally.   Is there anybody else who is interested in
plate out or would like to say  something about it?  Mr. Holub?

    MR. HOLUB:  I have to say something about the fan, obviously.  There was
an additional up-date paper  on  plate out published in a proceedings from the
University of Illinois in 1979.   I can send a copy of the proceedings to
anyone who's interested.   The reason we did this was simply to try to
understand better what is happening.   The first thing one has to look at is
the condensation nuclei concentration—its size, of course, its charge,
humidity, and perhaps even the  chemical character, of the condensation
nuclei—and then try to guess the mechanism.  So, for instance, it's much more
likely that small condensation  nuclei  would stick to the fan rather than the
big ones, which might bounce off because of inertia effects.

    Then also there is the charge or even image charge on the blades.  We have
explained it in our paper—we had put  charge on the blades and it affected the
sticking.  Then there is  the big question of what are the sticking
coefficients, because the typical aerosol phycisist always insists it's one.
But there is now growing  evidence from other sources—the Indians measured it,
and certainly Hopkie measured it, and  Krugar in South Africa—that the
sticking coefficient is much less than one.

    All this says that one should be careful about assuming what's on the fan
and what's on the wall.  It  depends  on many things.  So I would somewhat take
exception to calling it an artifact.   It's the most natural thing.


    MR. SCOTT:  And what is probably a concluding  comment on this, I've been
looking at this question of deposition, too,  by measuring deposited activity
on the walls of buildings.   This is a field measurement, not a laboratory
measurement.  It is difficult because if one  uses  an ordinary alpha
scintillation counter your  sensitivity isn't  that  great.  But, nonetheless, we
do have a few places where  the radon concentration is high, and we're able to
demonstrate from this a little bit of theory  that,  in fact, deposition
velocities for unattached daughters are in the region of .1 to .5 centimeters
per second, which is what we measured, and will, in fact, account for most of
the variation of working level ratios that we actually  see in houses.  So the
question of squaring the material balance does not seem to be beyond the wit
of man at the moment.  The  difficulty seems indeed that deposition velocity is
a function of air turbulence—of air velocity over the  surfaces.  So, as
usual, we were able to replace one fundamental question by an even more
fundamental one.

    UNIDENTIFIED SPEAKER:  I'd just like to make a comment.  I have read your
paper.  You were nice enough to send it to me, and, as  I recollect, I didn't
refer to it in my write-up, but your deposition rates were about the same as
what I was getting.  So, I  think there's some agreement.

    MR. SCOTT:  Are there any more questions  arising out of this morning's

    MR. TOOHEY:  I'll follow in our esteemed  chairman's footsteps by asking a
question of a co-panelist,  and that is Bill Nazaroff.   In your paper this
morning you talked about correlating the radon source term with wind speed
measurements.  I wanted to  know what happens  to the radon level.  We've seen
essentially no correlation  between wind speed and  the radon level in a house.

    MR. NAZAROFF:  For any  of you who have been to a meeting with Arthur Scott
before, if he was in this situation, he would pull out  a view graph and show
you exactly what had happened, but I'm unfortunately not that well prepared.

    We made wind speed measurements in only one house.  When you look at the
radon data from that house, you see that when the  wind  speed increases, the
radon concentration increases slightly and then after a few hours begins to
drop in response to the increased air exchange rate.  You could, in principle,
take the figure that I presented and work backwards, but I have the original
data which I would be happy to send you.

    MR. GINGRICH:  I just want to make a comment about  the wind speed thing as
long as we're on that subject again.  We made a lot of  soil gas measurements
with our track etch cup along with comparing  our comparison measurements with
some other techniques.  One of the most interesting things we found, and it
may have an application to  Bill Nazaroff's problem, is  that if you set out a
grid of samples or sample sites in a very hilly area and make soil gas
measurements, if you use a  radon emanometer where  you're taking grab samples,
you'll find that when the wind is blowing onto the surface of this side of the
hill the radon soil gas levels will be fairly low, in fact, quite low if the

wind is high.  If you then measure  over  the  top of the hill on the opposite
side, the radon levels go up again.   This  is due to either actual forcing of
the air through the soil on the hill  or  possibly Bernoulli effects bringing a
lower pressure zone on the other side.

    The same thing may be true in these  houses.  When the wind is blowing on
the house, it could be Bernoulli effects that are helping to bring radon out
of the soil gas and into the house.   Also, with regard to other measurements
we've made in soil gas (we've made  something like 300,000 soil gas
measurements in different parts of  the world), the average we're finding is
about 100 picocuries per liter in the soil gas.  But it's not unusual to have
soil gas concentrations ten times this.  And that's only 15 to 18 inches below
the surface.

    MR. SCOTT:  I'd just like to reinforce this.  If, in fact, the causation
of high radon levels in a house is  due to  the pressure in the house being
lower than outside, it's clear that Bernoulli's theorem requires that when the
air blows around the houses,  it's got further to go; the pressure inside the
house must be lower.  Therefore,  the  harder  the wind blows, the more negative
the house will become relative to some distant point on an infinite plane.

    In addition to that,  the upwind portion  of the house is a positive
pressure on it and the downwind portion  of a house has a negative pressure on
it; and these pressure zones do not terminate at the soil.  They in fact
extend into the earth to some distance.  We  have made measurements using an
emanometer around houses of the radon concentration and soil gas and find that
if you can make measurements on the upwind side, at times at a depth of two or
three feet, you get essentially outside  air.  On the downwind side, you can
get concentrations of perhaps a thousand picocuries per liter, which is
normally the kind of concentration you expect at several feet depth.

    The whole situation is incredibly dynamic.  Many of the things are not.
Many of the problems are partially aerodynamics, not those of the radium in
soil, per se.

    MR. COLLE:  One thing puzzles me, I've seen an awful lot of results lately
where people have looked at various correlations to meteorological parameters,
but I don't see how you can find a correlation between radium content, or
radon concentration, and a single variable unless you have some model you're
starting with, since all of the meteorological variables are
cross-correlated.  How can you separate  out  that cross-correlation?  I mean,
temperature and pressure are certainly cross-correlated, you know, and wind
speed.  In other words, since seldom  are all the other variables constant so
that you can really see a correlation between radon and wind speed, I just
don't see what these results mean.

    MR. SCOTT:  I have a particular bias on  this subject.  When we started on
our remedial projects, everybody knew certain things were true.  After a short
period, I decided that the simplest way  of evaluating them was to assume that
everyone was wrong.  And we made  measurements in adjacent houses for a period
of time and found that readings made  within  a matter of 15 minutes in adjacent
houses were not correlated.  Now, it  is  conceivable that for a house you could

develop a set of correlations with wind speed  and  so on.  If, for example,
there is only one hole in the house for radon  to come in out of the soil, when
the pressure is high on that side then indeed  radon may come in.  But that
particular correlation may only apply for  that particular house.  I would
require a lot of convincing to believe that  there  is any general correlation
with anything except perhaps the most physical of  variables, such as perhaps
wind speed, in the sense that may increase the Bernoulli forces.  Though,
unfortunately, it almost invariably leads  to an increase in ventilation rates,
so it's hard to see the difference, and perhaps the external temperature

    Presumably we've beaten this one more  or less  to death.  Would anyone like
to change the subject with questions?

    MR. BAILEY:  I have one general statement  to make.  I've seen a lot of
times when people amass paired data and then start to do regression analyses,
line fitting least square analyses.  I think we need to go further than just
the "R" factor.  I think whoever is massaging  the  data needs to go look at the
standard error of Y on X.  It's very easy  to take  a shotgun blast of data and
find a straight line that fits it very well  and assume that there's a good
regression or a good correlation, when it's  not necessarily so because the
error of that analysis is very, very broad.

    My second comment is directed to either  Mr. Akerblom or Ms. Wilson
concerning their limits for new structures of  two  picocuries per liter.  In
our study in the State of Florida, we found  houses that are outside of our
district of concern that we used for control structures to be in excess of two
picocuries per liter.  From that we assumed  that this is sort of like
background for maybe six picocuries per liter  radon indoor, being background,
or typical of anywhere.  Have studies been done in that area on houses not
presumed to be in the area of concern or houses built with crawl space or
mobile homes or whatever, and, if so, what kinds of indoor radon levels are

    MR. SCOTT:  On the two picocuries per  liter, I think that's two picocuries
per liter equilibrium equivalent radon, or .02 working levels.

    MR. BAILEY:  Okay.  I misunderstood the  equilibrium.

    MR. SCOTT:  A good SI unit there, I think.

    MR. AKERBLOM:  It is really a problem.   This is a very low limit, as I see
it, for radon daughter concentrations inside your  houses.  When we had the
problem in Sweden, it was decided, based perhaps upon too limited a number of
measurements, that, for health reasons, you  couldn't go with a higher dose and
allow it in your houses.  I think that is  something one of you has to come up
with—what dose will we permit for people  living in houses?  And, if you have
not decided that, well, you have to do it  sometime.

    The Canadians have done it from what I understood from you, Mr. Scott, for
houses on tailings.  We have done it for houses on ground anywhere.  But still

you might say these are  provisional limits, as I see it at least, which may
not be generally applicable.   We,  in Sweden, had simply come to a point where
we had to impose a limit.   It  was  done before we had full knowledge.

    MR. BERNHARDT:  Just for clarity, is that net above background or is that
a total gross, two picocuries  per  liter in equilibrium?

    MR. AKERBLOM:  It is a total measurement in the house.  I think I
misunderstood Mr. Scott's question earlier in the panel.  He asked me how we
measured this, and this, of course, is the question.  Until now I don't think
anybody has tried to measure this.  There has also been discussion about how
it should be measured.   They say it's not possible with the equipment that we
now have.  So it is possible that  one might permit a radon dose of double that
amount before doing anything about it.  However, in my opinion, anything over
that amount should be considered an emergency situation requiring immediate

    MR. SCOTT:  So, it looks as if we seem to have finished with questions of
this morning's presentation.   So I will, again, take advantage of my position
and propose a few questions for the panel.  It's always a good idea to start
on a fundamental question, and it  strikes me that in the group of people
gathered here we have had people who are worried about making measurements in
mines, people who are making measurements in the atmosphere, people who are
making measurements in houses, and, indeed, there are people who are concerned
about making measurements in laboratories.  Many of these people, in fact,
don't have much in common. For example, the question of plate out in sampling
heads is a non-question for probably about 95 percent of all people who take
samples because they use an open-faced filter holder and transfer the sample
to a separate sealer.  It's only the sophisticated guys who give themselves
sophisticated problems.

    But let's see if I can raise any interest on a fundamental question.  We
have an international meeting  here on radon-radon progeny measurements.
What's the purpose of the measurements?  This is a fundamental question
because it, in fact, raises a  lot  of questions and also simplifies things a
lot.  What purpose are the measurements that we're talking about?  Is this a
very general meeting or is there an underlying theme or interest that we
should be pursuing?

    MR. HESS:  I think that people who measure radon should have a
responsibility for measuring radon in their own area.  I think fulfilling this
responsibility can provide information from a wide variety of different
sources of radon, and our picture  then of radon levels in the United States
or, for that matter, throughout the world will gradually become clearer.

    I must say, I have a lot more  enjoyment from measuring specific physical
effects, but I think that if each  of us could take it upon ourselves to
measure some local buildings—either private dwellings or factories—and try
to collect more information about  what the levels are, we would then have a
more complete picture that would help us determine some of these questions
about levels.

    In my case, I've spent most of  my time measuring in Maine, and I have
found that doing one State is more  than  enough  trouble—more than enough
difficulty to keep somebody busy for  a long  time.  I think that as scientists
in this area we have a certain moral  responsibility to do some work which will
be of significance, perhaps either  for health or  for understanding or to
alleviate fears or misconceptions in  our own areas.

    MR. KAHN:  In connection with that,  there seems to be now three possible
sources of elevated levels.  One is in Maine—and some other places, too, I
guess—where there seems to be a lot  of  water bringing in high levels of
radon.  Certainly in many parts of  Sweden there seems to be the underlying
geology of the shell bringing in radon.   We've  worried about building
materials, but I haven't seen any that are as high as that.  Even in Sweden it
seems as if it's underlying material  more than  the building material.
Probably a systematic way of getting  at  this is to say, "Well, okay, let's
look at these two areas."  Of course, there's also man-made materials being
placed underneath houses.  Let's look at these  areas from the point of view of
geology, for instance, for high radon water  and high radon emanating
materials.  Then let's do extensive measurements.  How about these places in
Pennsylvania?  What's the source there?

    MR. GINGRICH:  The area in eastern Pennsylvania we've been surveying is a
known uraniferous area.  .There are  some  old  uranium mines in the area and some
underlying rocks that are known to  contain uranium.  It extends not only to
eastern Pennsylvania, but down to New Jersey and  a few other places in that

    MR. KAHN:  I'm sure in Maine, for example,  you went at it because you knew
there was a lot of radon coming out of water supplies.  So, I think that if we
wanted to go at this from a public  health point of view, we'd need first the
kind of thing that's being done in  Sweden.   I guess we can get at that one of
these days in the United States. The EPA laboratory here has been looking at
radon—how far are we with radon, Chick, in  water? Will we within a few years
know where all the high radon levels  are in  the water?

    MR. PHILLIPS:  I think we'll have a  lot  better idea.  I think certainly
within a couple of years we'll be able to identify most of the problem areas.

    MR. KAHN:  Okay, so we may have these four  areas.  I think a way to go at
it is to not all go at it and measure but to define where these problems are
and then look at the high level areas and the marginal areas.

    MS. WILSON:  I think, in fact,  in the United  States here you've got an
awful lot of information that you haven't used  yet.  I've talked to John Pacer
from Bendix and he says that large  areas of  the States have been known for
prospecting purposes, and this information has  not been used by the
environmentalists.  All our radiometric  measurements were made for prospecting
purposes.  This was material that we  had at  hand  and so we just decided to use

    MR. BAILEY:  We have had a commercial group do a radiometric survey of the
north end of Florida for mineral mining  purposes.  In the southern end, the
phosphate region of Florida, we did our  own  aerial survey for gamma only, just

gross gamma exposure.   There  has  been a lot of controversy going on between
various groups as to whether  or not gamma is a useful indicator of the radon
problem or potential radon  problem.  We happen to think that gamma is actually
of very little value whatsoever as far as a radon problem, unless the exposure
is well over 25 or 30  microroentgens per hour.  Otherwise, it's not much of an
indicator at all as a potential radon exposure problem.

    Part of the thing that  we're  stuck in the middle of—that probably many
other groups are going to be  in—is modeling.  As in the case of the people
from Sweden, they talk about  two  picocuries per liter being the maximum
allowed in new residences.  One of the things that we are looking for, that
we're trying to do something  about, is how can you go to a bare piece of
property that you haven't seen before, except maybe driven past on the
highway, and take a measurement,  do a little handwaving and say, "Yes, we'll
have a problem," or, "No, no  problem here."?  That's one of the problems that
we're really stuck with in  the States.  It's not measuring it so much or what
to do about it when you find  it,  because there's nothing you can do about it.
You can't force anybody to  spend  money on their house.  But what do you do
with a piece of property with cows on it that a developer is going to build
houses on?  That's the problem that we're stuck in.

    MR. RATHBUN:  I have a  question along that same line.  I have sort of lost
track of where we are in the  scheme of modeling these days.  I know that UDAD
has gone through a certain  amount of transition, and I hear people talking
about AIRDOSE and MILLDOSE  and I  wonder how all these things are related and
what the status is now.

    MR. SCOTT:  Offhand, it would seem that things like your MILLDOSE and
AIRDOSE are not really particularly relevant to this problem.  The difficulty
is not radon coming in from outside, but radon from inside trying to get out.
As to the question of modeling houses, I don't know how far that is along.
Again, there are so many parameters that it's difficult to know where to
start.  This indeed is the  unsolved mystery.  So what impels the stuff to come

    MR. BAILEY:  This is sort of  what I was trying to lead to with the
question to Gustav and Carole.  My second question was:  Do you have a limit
for a new structure?  Do you  allow the new house to be built and then go in
and measure it and then say,  "Well, you're past the limit, you're going to
have to do something about  it."?  Or, are you developing a method for looking
at a piece of property and  saying, "Most likely remedial action will be
necessary in this type of a structure."?  If so, what parameters are you
looking at as flag parameters to  decide whether there should be modifications?

    MR. AKERBLOM:  In fact, we are looking for areas with high radon
concentrations in the soil.  It's not only high radon concentrations; it's how
much air is available to get  into a house.  We need to have some way to tell
if it's a big problem.  In  my opinion, all houses should be built a little bit
off the ground surface.  Even on  normal ground radon comes in, as shown today
from many of the reports on ground problems of radon in soil.  You have to
really, from the beginning, ask,  "Are the building techniques what they should
be?  Are they proper to meet  the  requirements of health inside the houses?"

These concerns must be impressed upon the  people who are building them.  You
have to implement some building regulations.   Otherwise, we'll have problems
elsewhere.  We'll have it there and there  and  there.

    MR. SCOTT:  The problem to some extent is  probabilistic.  If you remember,
I showed a view graph of the distribution  of means  in Elliot Lake.  Now, this
is an area where undoubtedly there is a radon  problem by our definition of a
radon problem.  On the order of 20 percent of  the houses were above .02 WL.
Now, Elliot Lake is a relatively small and compact  town.  But, suppose we turn
it around.  We're saying in fact 80 percent of the  houses don't have a
problem.  The soil, to all intents and purposes, is the same over this area.
It's about as uniform as you're likely to  get  anywhere in that kind of area.
It isn't like some areas such as down in Florida where the soil has been
carefully resegregated, and it is probably even more uniform, though high.
Eighty percent of the houses don't have a  problem.  Interestingly enough, we
have some areas where there's an attempt to do radon proof structures, and 50
percent of the houses have the problem.

    So the real difficulty is that there are some subtle effects taking place
in the soil.  We've even had different working levels, considerably different
working levels, from one half of a duplex  to another.  If you cannot get
uniformity over a distance of 20 feet in the parameters that control the entry
of radon into houses, I think it is almost a vain attempt to try and talk of
large areas except in the vaguest of terms.

    If you could just paint the curbstone  with zinc sulfide and find your way
home at night, it might not be a good area to  build on.  But, even places that
are essentially at the low end of the scale, less than one picocurie per gram
radium, still have some houses in them that exceed  .02 WL on the long-term
average.  So, one can only make probabilistic  statements.  It is well-known,
now, that neither the public nor the regulators understand matters of

    We seem to have drifted off a little bit from the original question, which
was:  Is the underlying theme, or has the  underlying theme of the meeting
been, environmental-type measurements or measurements in houses, and should
we, if we are thinking about the underlying theme,  think about it more,
perhaps?  Suppose, in fact, we take for granted that really the main interest
of the meeting has been matters of environmental homes.  Then there are a
couple of questions that spring naturally  out  of this, and one of them is:
Are the current devices you've got good enough? I  mean, should we, in fact,
ostracize anyone who comes around for a grant  to improve an instrument on the
grounds that he's purely wasting the taxpayers' money?  Or, should we, in
fact, say that the present instrumentation is  so inadequate that anyone who
comes around proposing action on past measurements  should be ostracized
because he clearly can't determine anything useful? Or, indeed, is the
situation in between the two?

    Would anyone possibly on the user side like to  say something about
instrumentation and its present adequacy?

    MR. HESS:  I notice that McDonald Wrenn  is  not here and Henry Spitz is not
here, so I'll say a few nasty words about  Wrenn meters.  They have a
difficulty with humidity, which has been written about and published by Andy
George and others.  They also have the property that when you set them down
vigorously some of the parts inside come undone.  So you need to be somewhat
experienced in repair of equipment in order  to  make field measurements with
them.  I would say that having a rugged electronic device, having one that's
rugged enough to be able to be taken from  one lab to another, ought to be one
of the criteria applied before a device is suggested for common use in field
studies; it ought to be able to travel around in an automobile and be bumped
and so on.  I guess that's not the usual degree of strength that you'll find
in laboratory equipment at universities.  So, "ruggedization," I think, is one
of the things that I'd like to see in some of the equipment.  Maybe that makes
it more expensive, but I think it's well worth  it.

    The other thing that I'd like to encourage  people to do is take meters
from one lab to another and try to get multiple calibrations.  I've found that
very useful and also I learned a great deal  by  doing it with the Wrenn meter.

    As far as the Track Etch cups go, I think I'm reasonably satisfied with
the way they behave.  I would really like  to see, though, some type of a track
etch device or something analogous to it that would be perhaps a factor of
three or four cheaper.  That would really  open  up surveys of a thousand
houses, which would be quite nice.  If something could be done to have a
reliable detector that works like a track  etch  detector does but that would
sell for perhaps $5.00 apiece or something like that, then that would be a
very valuable contribution, I would say.

    MR. PHILLIPS:  I'd be interested in a  good  passive working level
monitor—one that sells for under a thousand dollars and reads down to .005
working levels.

    MR. SCOTT:  As a user, I'd be interested in just getting a good reliable
air sampling pump that was small enough and  light enough to be carried around
by a five foot two, 18-year-old girl, preferably with blond hair and blue eyes.


    MR. BAILEY:  In following what Arthur  was just talking about, a year or
two ago, Mr. Pai, up in the Canadian area, was  working on a little,
aquarium-like pump with a track etch intimately placed to a filter system.  I
got a letter or two about it, but has it gone anywhere, or has it gone into
the fish tank?

    MR. SCOTT:  No, no.  It's been saved from the fish tank.  This is Lynn Pai
of the Ontario Ministry of Labor, which is responsible for radiation
monitoring.  What he has is an adaptation  of the French CAE nuclear track
dosimeter for working levels.  It isn't passive; it's active.  That's why the
pump is required.  It uses cellulose nitrate film; and the energy sensitivity
of cellulose nitrate is used in this in conjunction with the absorber so that
you can, in fact, get the total number of  radium A atoms, radium C prime, and

also thorium C prime of that deposited on the  filter.   It's hoped that it will
shortly be put into use in Elliot Lake, as a matter of  fact, to monitor new
construction for compliance with the new construction limit of .02 working

    MR. PRICHARD:  While we're in the business of wishing for new and improved
equipment, a little gadget I'd like to see on  the market-i-please excuse me if
it's there and I haven't seen it—is a simple  and rugged and inexpensive pulse
logger.  I don't think we have so much a shortage of devices that are capable
of translating an alpha particle into an electronic pulse, but, at least in
the realm of field equipment what I have seen, there is a real shortage of
handy little gadgets to take a pulse and store it until you're ready to ask it
how many pulses it saw.

    MR. SCOTT:  Well, I was unable to provoke  many flights of fantasy there.
In the operational field, and, of course, most people here are not involved in
quite that sense, we do have a continual problem that I touched on this
morning, and that is how well the short-term measurements represent the
long-term situation.  Now, to get an annual average you must make measurements
for a year.  But one of the topics—if anyone  has any views on it I'd like to
hear it discussed—is the question of variability, in the short run.  You know,
maybe a few measurements in the best of all possible worlds would be
sufficient to give us an estimate of the long-term average.  Can I stimulate
any interest here?

    No, clearly not.  Well, we've heard one heart-felt  cry regarding
calibration and inter-laboratory comparisons.  I think  everyone who has
participated in one has found it instructive,  particularly if they came out in
the middle.


    Yes.  Well, that's useful and instructive.  It's instructive if you were
one of the outliers.  The general theme seems  to be we  can't have enough of
them.  Would anyone like to comment on those kinds of things?  Are there
sufficient numbers of these things?  And what  about the crude and untutored
users of the majority of a lot of this equipment?  I think, particularly, of
people in mines.  Is there sufficiently wide publicity  given to
intercomparison to persuade them to send in an air pump and sealer?  Maybe one
of the laboratory people might say something.

    Yes, please.

    MR. HOLUB:  You mentioned pumps; it is really a bad problem.  The best
seems to be a self-regulating device.  It costs  about $300 and can control its
flow to pressure up to 50 inches of water. We had the  best experience with
these.  All of the others failed, and this one eventually failed, too, in the
mine.  But it was definitely the best.  I can  provide details on the pump to
anyone who wishes them.

    MR. SCOTT:  Though, this doesn't quite answer my question.

    MR. HOLUB:   Oh,  I'm sorry,  I  thought you asked about the pump.

    MR. SCOTT:   No,  I'm saying, what about the user?  It's all very well to
have a good pump,  but do the  users ever get encouraged to actually send in a
pump and sealer to compare  the  working levels of the climax mine with the
working levels  on  your test mine?

    MR. HOLUB:   Definitely  not, no.  They don't really care.

    MR. SCOTT:   Indeed, a question of accuracy in a vacuum, so to speak.

    MR. RATHBUN:  I  think some  of this is done, although probably not enough
of it, especially  with the  air  pumps.  I formerly worked for the Mine
Enforcement and Safety Administration in Denver, which worked quite closely
with the Bureau of Mines there  and the people did send in those pumps and
sealers—all sorts of equipment—to be calibrated by the government.

    MR. HOLUB:   I  have to add that these recommendations are usually ignored.
I am not complaining, just  stating.

    MR. SCOTT:   The  reality is, everyone checks.  A given organization often
checks its sealer  with a source that's been around for five or seven years and
has got someone's  initials  scratched in on the active side.  Unfortunately, it
so often seems  there is a tremendous gap between what we know we can do and
what we actually do.  But,  still, there seems to be no place to get
self-confessions from the assembled multitude.

    I see Andy  George here  sitting pensively, so I would like to direct a
question to Andy.  With all the instruments that come walking through his
laboratory, when do  you think we  will have enough?

    MR. GEORGE:  I think we have  enough already.  It's just a matter of
calibrating and using them  properly.  We have every conceivable instrument
that you need to determine  exposure, concentrations, and so on.  So I don't
understand why  you need to  develop any more.  We just have to take what we
have and try to do some quality control calibrations and start using them.  I
think our immediate  goal is to  go out and make a lot of measurements to find
out what is the exposure throughout the United States.  There are questions to
be resolved. For  example,  Jim  believes geological formations cause the high
levels in Pennsylvania.  I  don't  think so.  We found radium concentrations in
soils about 1.2 to 1.5 picocuries per gram.  He can hardly say that's
abnormal.  So,  there must be  many, many other locations like that throughout
the United States.

    So as far as instrumentation, I think we should go out and take numerous
samples and identify these  areas  in the country.  We do have instrumentation,
if we learn how to use it properly.  All the instruments I work with are
usually built in our laboratory and I can vouch for them.  But I've seen a lot
of instruments  come  in that are not working; and they are supposed to be
working.  They  are off by factors of 2 to maybe 22, as much as that.  So it
depends on who  builds the instruments and how much he puts into calibrating

his instruments.  As far as I know,  no commercial devices have been calibrated
at all.  Sometimes they put a calibration factor from  theory, but that doesn't
work.  For example, when you calibrate an instrument that uses a TLD
dosimeter, you have to calibrate the whole system,  including the TLD and
everything.  So, when you are using  calcium fluoride or lithium fluoride, it
is only as good as you can calibrate it.   I mean the whole system—not just
the PERM or the passive device or the active device—has to have the TLD
included.  Everything has to be included.

    MR. PHILLIPS:  I hate to disagree with Andy because he's certainly the
expert.  But I don't think we have a working level  instrument with which we
can make long-term measurements in a large number of homes without expending
an exorbitant amount of manpower and money.  Now, I think we can do it for
radon, but, unfortunately, guides are not set in terms of radon.  Even though
we might all say that the equilibrium factor is .5, you still, at some point,
have to measure the working level; and I  don't think we have those instruments
yet.  At least, I'm not familiar with them.

    MR. SCOTT:  Would anyone in the  audience like to comment on measuring the
environment?  Instead of talking, getting out there measuring?  Oh, well
here's a surprise, Mr. McGregor.

    MR. MCGREGOR:  It looks like Arthur is pointing in my direction.  It's our
agency that has perhaps made the most extensive use of grab sampling
techniques—and I was quite interested to hear Arthur's comments regarding his
synthetic grab sampling with his replica  sampling procedures.  To date we have
measured in 19 communities now, and  about thirteen  and a half thousand homes.
But they're all single measurements.  We  feel that  this has given us some idea
of what the general background levels might be in various communities and
perhaps in various geological environments.

    We have also done perhaps another five communities across the country to a
much more limited basis of sampling.  But it's rather  interesting to hear
Carole Wilson's comments regarding the work in Sweden  using the geological
maps that they have prepared for their aerial survey work.  We have also noted
in Canada some areas, Castle Guard,  British Columbia particularly comes to
mind, where there is a town down the Columbia River called Trail and there is
a definite, significant difference in the average levels of radon and radon
daughters in these two communities.   Castle Guard is being affected by a large
uranium potential ore body.  It's about .6 percent  0303 up the hillside
which has undergone valley glaciation, at least twice, with the detritus being
deposited in the valley and contributing  to the higher levels in Trail—oh,
sorry, in Castle Guard as compared with Trail. This has also been noted in
geological differences in the Bancroft area for radon  concentrations in homes
and in northern Saskatchewan.

    Speaking of Saskatchewan, there  is a  study currently underway with the
provincial authorities using the extensive aerial survey maps to try to
delineate high risk areas for radon  in the community of northern
Saskatchewan.  They are using these  maps  and then going into the very small
villages and towns and doing house surveys.

    MR. SCOTT:   Thank you very much.   I think this shows that you can in fact
determine quite a lot if  you  actually  go out and measure, rather than sit
around and think about how best  to measure.

    MR. AKERBLOM:  While  I have  the opportunity, I want to ask if anyone has
been doing radon measurements in underground homes, which seems to be a
popular way of  building homes in the States and is starting to be in Sweden.
But, I think,  if you have a problem on the earth's surface, digging yourself
down into it will create  a much  greater problem.  I would be very happy to get
in contact with people who have  done actual measurements in underground homes.

    MR. TOOHEY:  We have  done some preliminary measurements; we've done six
underground homes.   I don't have the data with me, but I can easily send it to
you.  I do have the average with me, which is an arithmetic mean of 4.5
picocuries per  liter for  the  six homes.  We also have done some preliminary
work on solar powered homes;  specifically, a few which have essentially put a
big rock somewhere in the home as a heat sink and reservoir.  In those, we
didn't see much out of the ordinary.   But, on the average, the underground
homes have been higher.

    MR. SCOTT:   Well, as  we seem to have run out of inspiration, I think we
will just conclude the meeting so that everyone can get a cup of coffee before
they depart for the laboratory.  I'd like to thank all our panelists and all
you participants and turn over the meeting to Chick Phillips for the
concluding remarks.


    MR. PHILLIPS:  Thank  you, Arthur,  and the panelists.  I'd just like to
thank all of you for your participation in the meeting, and, particularly, I'd
like to thank Cathy, Marca, and  Bob for their assistance.