IEEE Transactions on Nuclear Science, Vol. NS-27, No. 1, February 1980

AN AUTOMATED TLD SYSTEM FOR GAMMA RADIATION MONITORING*

P.C. Nyberg (1), J.D. Ott (2), and C.M. EdmondsEnvironmental Monitoring and Support Laboratory-Las Vegas

U.S. Environmental Protection Agencyand J.L. Hopper

Reynolds Elecrical and Engineering Company, Inc.

Summary

A gamma radiation monitoring system utilizing acommercially available TLD reader and unique micro-computer control has been built to assess the externalradiation exposure to the resident population near anuclear weapons testing facility. Maximum use of themicrocomputer was made to increase the efficiency ofdata acquisition, transmission, and presentation, andto reduce operational costs. The system was testedfor conformance with an applicable national standardfor TLD's used in environmental measurements.

Introducti on

While the measurement of radiation exposure topersonnel working in nuclear facilities is not new, itis only in relatively recent times that sufficientlysensitive systems have been developed to measure theexternal radiation exposure in the environment out-side. The increasing interest of scientists and lay-men alike in the possible health effects of low levelionizing radiation has created strong pressure formore precise measurement of ever lower radiationlevels which, in turn, has required more sophisti-cated data handling and analysis techniques.Fortunately, developments in electronic instrumenta-tion have advanced to the point where equipment is nowavailable which can perform such tasks with speed andease. The use of microcomputers has revolutionizedthe instrumentation field, and radiation measurementinstruments have just begun to be affected.

For many years thermoluminescent dosimeters(TLD's) deployed in a large network, have been used tomeasure the background radiation exposure in the areasurrounding the U.S. Department of Energy's NevadaTest Site. Located in south-central Nevada, the TestSite and its exclusion area form a rough rectangle 150km north-to-south and 125 km east-to-west that isprimarily used for the underground testing of nuclearweapons. Under a memorandum of understanding, theU.S. Environmental Protection Agency and previouslythe U.S. Public Health Service have conducted the off-site radiological monitoring program. This consistsof continuous and systematic efforts to assess theradiological impact of the nuclear testing program onthe nearby offsite environment and its residents, manyof whom live in the small communities and far-flungranches which dot this vast area. Portions of westernUtah, Nevada, and eastern California are routinelymonitored by the TLD network for external gammaradiation.

Although TLD's had been used for this monitoringsince 1965, the original system was labor intensiveand lacked the desired sensitivity for backgroundlevel measurements.1 In order to minimize the man-power required for the routine handling of the TLD's

* This work was performed under Memorandum of Under-standing No. EY-76-A-08-0539 for the U.S. Department of Energy.

(1) Current affiliation: Office of Radiation Programs,Las Vegas Facility, EPA

(2) Current affiliation: General Atomics, Inc., SanDiego, CA

and the processing of the data, some type of automaticsystem seemed appropriate. The Harshaw* Model 2271Automated TLD System was purchased in 1973. Sincethe background radiation levels in much of the areaare generally lower than the nationwide average, thehighly sensitive phosphor, dysprosium-activated cal-cium fluoride (CaF2:Dy), designated TLD-200 byHarshaw, was selected as the thermoluminescent (TL)material . Handling and processing techniques werethen developed and tested, and the first field datawas obtained in late 1974. The system was acceptedfor routine use in July 1975.

As confidence and experience in the use of thisTLD system was gained, it was felt that more attentionshould be paid to the data handling activity, whichstill required considerable manual effort. Therefore,in 1978 an improved reader control system and datahandling scheme based on a microcomputer was designed,and the necessary additional equipment was purchased.The new system has been in routine use since mid-1979and performs the TLD readout, quality control checks,and preliminary report preparation locally. Data arethen formatted for transmission to a larger hostcomputer for more sophisticated analysis.

TLD System

The reader is a Harshaw Model 2271 Automated TLDSystem, shown in Figure 1, which has been described indetail elsewhere.2 It is unique in that the solidthermoluminescent material, in the form of a 3.2- by3.2- by 0.89-mm "chip", is permanently mounted to a43- by 31-mm aluminum card onto which is glued a Mylarmask covering a punched hole array which makes up theindividual card identification number. This code isoptically scanned for each card during readout.Furthermore, the card contains two chips, each ofwhich is sandwiched between thin sheets of Teflonplastic and suspended in two 9.5-mm holes punched inthe card as shown in Figure 2. Up to 250 of thesecards may be loaded into a tray from which they aresequentially fed into the reader. At the operator'sdiscretion it can be programmed to read one or bothchips at a selectable temperature for a selectabletime. A typical readout time is 10 seconds per chipto a maximum temperature of 3000C. Each chip issequentially heated by a solenoid-driven hot fingerequipped with thermocouple feedback control for tem-perature reproducibility. The actual current integra-tion and timing is performed by a Harshaw Model 2000BIntegrating Picoammeter.

The main advantages of the system are the machinereadable identification on each dosimeter, the auto-matic reading of many dosimeters in a carefullyprogrammed heating cycle, the generation of computer-compatible output, and the capability for externalprogram control. These are the essential features ofany automated system, as they allow the use of com-puter control for flexibility and data processingtechniques to produce several levels of sophisticationin dosimetry reports. The lack of flexibility in bothdosimeter configuration and operating regimen is a

* Use of brand name does not imply endorsement by theU.S. Environmental Protection Agency

U.S. Government work not protected by U.S. copyright 713

Figure 1 Harshaw 2271 Automated TLD System

disadvantage of the system. Only the specially de-signed dosimeter cards can be read by the system, thusrestricting the actual dosimeter to the 3.2- by 3.2-mmchip. Furthermore, since the chip is permanentlymounted to the card by the thin Teflon sheet, it maynot be possible to optimize the heating cycle for thephosphor without damaging the plastic.

As the TLD system was designed primarily for usein personnel dosimetry, several adaptations were re-quired to make it suitable for environmentalmeasurements. Ie TLD material most commonly used,lithium fluoride, was not considered sufficientlysensitive for this application, and TLD-200(CaF2:Dy), a more sensitive phosphor, was selected.This, in turn, strongly influenced the design of thedosimeter card holder because an energy com-pensation shield was required to minimize the over-response of the TL material to photons in the 100- to150-keV energy range. Sheets of 1.2-mm thick cadmiummetal (99.99% pure) were chosen for this shield forboth their physical characteristics and economy.Cadmium was considered a reasonable compromise sincethe dosimeters were not intended for use in any areawith a significant neutron flux. A sheet of 0.1-mmpolyethylene was placed between the dosimeter card andthe metal shield to protect the card from abrasion.The result was an energy response which is uniformwithin ±25% from 80 to 1000 keV, but which dropsrapidly below that range. This was considered ac-ceptable as the data were intended to assess the wholebody dose to the population, and low energy pho-tons in typical background spectra contribute anegligible amount to the whole body dose.3 The en-ergy response of the dosimeter is shown in Figure 3.To provide a weatherproof, light-tight, and protectivecovering, a two-piece high impact polystyrene plasticholder was also designed. The holder is relativelytamperproof, but easy to assemble and disassemble.For identification, the dosimeter card number is writ-ten on a colored adhesive label and applied to theoutside of the holder at the time of assembly.

All operations on the dosimeter cards, i.e., de-dosing before use, post-irradiation annealling, andreadout, must be done by the reader in some mod-ification of the readout sequence. Oven annealling ofthe cards is not recommended. Therefore, when it was

714

Figure 2 TLD Card

found necessary to employ a post-irradiation annealprior to the readout in order to minimize the well-known fading characteristics of the TLD-200 phosphor,considerable experimentation was required. Heatingthe individual chips to 160°C for a 10-second periodprior to the final readout minimized the signal con-tribution from the fast fading, low temperature traps.

1.0

.8

.6

I.

cI

.2.

.1-

.08-

.06-

.04.

.02-

l'Measumed /Response II,

rI Calculated

; "'t Response7

I

0.010 0.iO 1.0

Photon Energy (Mev.)

Figure 3 Dosimeter Energy Response(3.2- by 3.2- by 0.89-mm CaF2:Dy chipswith 1030 mg/cmZ cadmium shield)

To date only the TLD-200 dosimeters have been usedroutinely at this Laboratory but, as mentioned before,the Harshaw 2271 Automated TLD system has been usedelsewhere for personnel monitoring. For that purposelithium fluoride has been more commonly used, either

.N

in the form of lithium enriched in the 6Li isotope,which is sensitive to both neutrons and gamma rays, orlithium enriched in the 7Li isotope, which isinsensitive to neutrons. With appropriate phosphors,packaging, and shielding, TLD cards can be used forbeta, gamma, and neutron dosimetry.

While the 2271 system has generally performedwell, it is certainly not without problems, some ofwhich are as unique as its design. For example, theglue which holds the Mylar mask to the aluminum cardwas a continuing source of annoyance until the man-ufacturer found a type which would neither dry out norcreep. Tiny drops of glue oozing from the early modelcards were enough to foul the narrow clearances withinthe handling mechanism, causing a fault conditionduring processing. The newest cards use a superiorglue which does not ooze, eliminating many of thoseproblems. The close tolerances of the card handlingmechanism presented other problems when a new batch ofcards was found to be slightly oversize, requiringsome modification of the card feeding chute. Becausethe TLD-200 material is not commonly used with thissystem, the manufacturer used his own readout re-gimen during the acceptance testing of a new batch ofcards made for this laboratory; however, the resultswere different from those obtained by the purchaser,and considerable effort was required to resolve thedi screpancy. While many improvements have beenincorporated into the system design since its incep-tion, there is no doubt that the weakest aspects areits mechanical features. Very few serious electronicproblems have thus far been encountered.

Microcomputer Control

The 2271 system is designed to be used alone witha simple programmed reading sequence and visual faultindicators, or under external control with a moresophisticated programming device setting the oper-ational sequence. We chose to implement the externalcontrol using a Northstar Horizon Microcomputer, con-sisting of a Z-80 microprocessor, 32 kilobytes ofread/write memory, and dual Minifloppy magnetic diskdrives. Parallel interface ports are available forconnection to the 2271 reader and the 2000B integrat-ing picoammeter, while a video terminal, 120-cpsprinter, and 1200-baud modem are connected via twoRS-232C EIA-compatible serial interface ports. Ablock diagram of the system is shown in Figure 4.

Figure 4 TLD System Block Diagram

An extended version of the BASIC programminglanguage is available for this microcomputer and waschosen for the TLD Reader Operating System because ofits simplicity and free data entry format. While aFORTRAN language compiler is also available, for ourpurposes its use would be unnecessarily complicatingand the data entry format overly rigid. It is likely,however, that FORTRAN implementation would result infaster program execution. The programs that con-stitute the TLD Reader Operating System may be roughlydivided into two categories: first, the data acquisi-tion programs, in which the microcomputer and readerinteract to produce the basic TLD readout and qualitycontrol data; second, the data manipulation programs,in which the microcomputer acts alone to perform aseries of calculations and error checks on the datapreviously acquired and produces several types of re-ports and data files. Both of these are detailed asfollows.

Data AcquisitionThe most important program in the TLD Reader Oper-

ating System controls the readout of the TLD cards,and is designated TLDR. This BASIC program opens adata file on the disk, initializes the 2271 and 2000Bcomponents, prompts the operator to enter the date andother identification information, provides for theacquisition of quality assurance information, andstarts the actual readout sequence by a command to the2271. TLDR controls the readout sequence whileutilizing many of the capabilities of the 2271. Theseinclude measurement of photomultiplier tube dark cur-rent (or background) and a built-in light source (orreference light) for quality assurance. TLDR al sointerprets the fault interrupt logic of the 2271 toprovide both audible and visual indications of anyreader error, e.g., a jammed dosimeter card, areversed card, a heater malfunction.

Under normal operation with microcomputer control,the TLDR program initiates the readout cycle for thefirst card, which drops by gravity until the firstchip is in the reading position. This portion of theprogram is written in Z-80 assembly language becausethe 2271 operation is asynchronous and some eventspass too quickly to be handled under BASIC programcontrol. The chip is then heated in a carefully con-trolled time and temperature sequence, and theintegrated photomultiplier tube current is measured bythe 2000B. At the end of the heating cycle, signaledby a logic pulse from the 2000B, control is returnedto the BASIC language program, which stores the ac-cumulated reading on the disk file and also prints itout (if desired). The card then drops to the secondchip position and the sequence is repeated. The carddrops once more, and the coded identification numberi s read and stored on the disk file. The card thenpasses into the receiver tray and the next card isprepared for readout. Thus for each card a minimum ofthree pieces of information are stored: the TL re-adout of the first chip, the TL readout of the secondchip, and the card identification number.

The 2271 has the internal program capability forboth reading andrereading each dosimeter chip duringthe readout sequence. This capability, coupled withthe need for post-irradiation annealling of the TLDchips, led to the only significant modification thatwas made to the 2271 reader. Originally the post-irradiation anneal required every card in a readoutgroup to be processed through the reader and each chipto be heated to a maximum temperature of 1600C. Thecards were then manually reloaded, and the operationwas repeated with a maximum temperature of 3000C.With the addition of a duplicate maximum temperaturecontrol and some switching logic to allow the "read"and "reread" functions to be accomplished at different

715

maximum temperatures, the post-irradiation anneal andthe readout now happen in a single processing of thecard through the reader. A schematic diagram of themodification is shown in Figure 5.

Because the dual temperature modification com-plicated the readout sequence, separate data fileswere established on the disk to record the "read"function (post-irradiation anneal) and the "reread"function (the actual readout). While the former isnot routinely used in further analysis, we have found"read" function data to be of value for diagnosticpurposes when problems develop. At present the datafiles have a maximum length of 250 TLD card readings,but the length is somewhat arbitrary. The readoutnormally continues until all the cards in the trayhave been read, at which time the data files areclosed. If at any time during the readout a faultshould develop, program TLDR will interpret the faultlogic of the 2271 to determine the failure mode whilesimultaneously closing the data files to protect thedata already written. Manual intervention is usuallyrequired to correct the fault, and the program then al-lows the data file to be reopened and the readoutsequence to continue.

Figure 5 Dual Temperature Modification toHarshaw Model 2271

Microcomputer control of the 2271 has made it pos-sible to perform operations that were previously bothtedious and difficult. The dual temperature heatingcycle, for example, became practical only when twodata files could be established to store the annealand readout data separately. Where it was previouslynecessary to insert a special blank dosimeter cardinto the tray to obtain background and reference lightreadings, program control now does this with ease.The microcomputer allows greater efficiency during thededosing operation as well. While several passesthrough the reader formerly were necessary to dedosethe dosimeter chips adequately before issue, programcontrol now repeatedly applies the standard heatingcycle to each chip until it gives an acceptably lowsignal. If this does not occur after a predeterminednumber of cycles, an alarm sounds to indicate an unac-ceptable dosimeter. This has saved considerable oper-ating time while minimizing wear and tear on both thereader and dosimeter cards.

Data ManipulationThere are three general types of files used by the

microcomputer for data storage and report generation.As previously outlined, a file is created to store thedata from the dosimeter readouts and is formattedspecifically to accept the output from the 2271 sys-tem. A second type of fi l e i s used to store thesensitivity correction factors for each serially

numbered TLD card in the inventory. Before it is usedfor field measurements, each dosimeter is givenseveral controlled exposures from which a sensitivitycorrection factor is calculated for each chip. Thisprocess is repeated periodically over the life of thecard. The use of correction factors tends to minimizethe variance of the corrected readouts and isessential for exposure calculations of backgroundlevels. The third type of file was developed to storethe location, date and time information from the fieldcycle. These elements are necessary for the calcula-tion of the average exposure rate. Field data cardsprovide such information as the monitoring location,TLD identification numbers, issue and collect dates,and any other pertinent observations, and are com-pleted by the individuals who issue and collect theTLD's. A custom-designed key-to-disk program uses thevideo terminal and a prompting scheme to simplify thefield data entry.

When a group of dosimeters is prepared for issue,a sample is set aside in a low background area in thelaboratory for calibration. Midway through the fieldcycle, half of the sample is exposed to a knownamount of radiation. The calibration dosimeters areread with the dosimeters returned from the field.Once the TLD data and the field data have been storedon the disk file, the operator should make severalchecks to see if the system is performing properly.The PM tube background and reference light readingsindicate the stability and sensitivity of the reader.Readings of the calibration dosimeters provide afactor relating the amount of light output from theTLD to its radiation exposure. A program calledCALFAC calculates the mean and standard deviation forboth the PM tube background and the reference lightreadings. The program also computes the radiationcalibration factor from the intentionally exposeddosimeters and the controls. This factor should befairly constant with time if the other readout para-meters are stabl e, and thus serves as a check on theentire system. CALFAC produces a simple report con-taining all this information used by the health phys-icist to assess the system's performance.

Each time a group of dosimeters is sent to thefield, certain ones are designated transit controls.These are carried throughout the exchange and are readwith the collected dosimeters. Following the de-termination of the calibration factor, the programGROSS1 retrieves information from the readout file,sensitivity correction factor file, and field datafile. The information is matched by the dosimeternumber, checked for logical integrity, and arranged bylocation identification code. Another code identifiesthe transit control dosimeters from which the averagetransit exposure rate is calculated. An error reportis then prepared that lists any unmatched TLD's, un-matched field data, or logical inconsistencies in thefield data.

Following the completion of GROSS1, program GROSS2is automatically loaded and executed. This programuses the data previously assembled to calculate foreach location the exposure adjusted for transit timeand the average exposure rate. Si nce there are twochip readings per dosimeter and usually threedosimeters per location, the standard deviation of theexposure at a location is also calculated.

The preliminary report of adjusted gross readoutsarranged by monitoring location is produced by a pro-gram called P-REPORT. Using data assembled by GROSS2,P-REPORT prints out a title page, an error report,listings of the readout and field data files, and aformatted listing of the exposure, exposure rate, andmonitoring dates for each location. An exampl e of

716

FRELIMINARY REPORT -- 0731791240

DOS I MET ERSTATION NUME8ER

004STAO1O 492349744943

I SSUEDIATE

0412679150004267915000426791500

COLLECTDATE

072079151507207915150720791515

REAEIOUTDATE

073179124007317912<400731791240

003STA106 4952 0428791200 0720791630 07317912404920 0428791200 0720791630 07317912404936 0428791200 0720791630 0731791240

003STA240 4998 0426791900 0720791230 07317912404910 0426791900 0720791230 07317912404902 0426791900 0720791230 0731791240

003STA245 4900 04287914004607 04287914004926 0428791400

0721790930 0731791240072-1790930 0731.79124007121790930 0731791240

CHIF' 1RE:SULT27.302?7 . 8424.92

CHIP 2RESULT28.5029.0526.01

10.99 12.1811.02 12.6912.56 13.61

21.2420.2820.05

22.4921.1220.97

19.80 20.5320.18 21.7823.54 22.46

097404140 4971 0426791900 0720791300 0731791240 16.96 18.48

545647304 4982 0426791900 0720791230 0731791240 19.69 20.18

STATIONMEAN MR/DAY

27.27

12.17

21,02

.320

.146

.248

21.38

17.7219.93

.209

.235

PERCENTSTD. [EV.

5.708

8.385

4.097

6.825

6.065

1 .7:;A

Figure 6 Preliminary TLD Reportthis listing is shown in Figure 6. In addition, aspeci al ly formatted output file called TEMP2 isproduced and is transmitted to a Control DataCorporation CDC-6400 computer for the compilation of afinal report. The TEMP2 fi l e can be sent vi a a1200-baud telephone data link using a specially de-signed communications program from the microcomputer.Once this data file is in residence on the CDC-6400system, it can be merged with other similar files andused to produce the final periodic report. Usually 7to 10 such files must be merged in this fashion toproduce a typical quarterly report.

The historical data files for the offsite dosime-try network are stored on the CDC-6400 and are used inthe production of the final periodic report. This isnecessary because the environmental radiation back-ground at a given location cannot be determined from asingle measurement4,5. The background must be de-termined independently of the present measurement inorder to find out if an exposure above background hasoccurred. There are several schemes6 that can beused to calculate the background, including the as-sumption of invariance with location, the assumptionof invariance with time at a given location, or apredictive model based on past performance and currentconditions. While the third is by far the most ac-curate, it is also the most costly and complex. Sucha model has been used on data from this network7,but at present the assumption of invariance with timeat a given location is employed. This allows the de-termination of net exposures above background in therange from 0.6 to 8 mR at the 95% confidence level fora typical quarterly measurement cycle. A FORTRANprogram has been written to produce the finalquarterly report for the network and to update thehistorical data files automatically. Other programsare also available to compile and list selected datafrom the historical files for additional analysis whenrequired.

mance as uniformity, reproducibility, photon energydependence, and sensitivity to environmental insults.Table I summarizes the results of tests on thedosimetry system described here. The system exceededthe minimum acceptable levels set by ANSI N545-1975except for energy dependence. Although the energyresponse of these dosimeters decreases rapidly below100 keV, this has been considered adequate in the pastfor the assessment of whol e body dose. It may bepossible to improve the response to low energy gammarays by adjusting the energy compensation shielding,but this is not currently being contemplated.

Conclusion

A gamma radiation monitoring system utilizing acommercially available automatic TLD reader, uniquemicrocomputer control, and automatic data process-ing techniques has been designed and implemented bythe U.S. Environmental Protection Agency to assess theexternal radiation exposure to the resident populationin the vicinity of the Nevada Test Site. The system'smaximum use of automatic equipment has reduced opera-tional costs and increased the efficiency of datapresentation, storage, and retrieval. The system isdesigned to accommodate over 100 fixed locations withmonitoring periods of 1 month or longer. It is mostuseful where continuing data storage is a requirementand where rapid access to historical data is desir-able. Although this particular system was designedspecifically for gamma radiation monitoring, the useof other thermoluminescent phosphors and packagingtechniques could allow the measurement of the absorbeddose from beta radiation or from neutrons. The micro-computer carries out the data acquisition, qualitycontrol checks, and preliminary exposure calculationsthrough programs written in BASIC, while the prepa-ration of the final report and historical data storageare done on a larger computer with FORTRAN programs.

Dosimeter System Performance

In order to judge the adequacy of a particulardosimetry system for making environmental radiationmeasurements, a performance standard8, ANSI N545-1975, was drafted by the American National StandardsInstitute. This voluntary standard provides a guidefor anyone making environmental radiation measurementsto check his performance against a satisfactory leveldetermined by a group of knowledgeable experts. ANSIN545-1975 addresses such aspects of dosimeter perfor-

717

ADJUSTED GROSS REAIIOUT VALUES HOPPER RTE 2ND OTR 79

Performance EPA 2271/TLD-200Aspect ANSI N545-1975 Requirement Dosimeter System

Uniformity Response of TLDs within a batch shall not differ from 24%each other by more than 30% at an exposure of 20 mRfor quarterly monitoring.

Reproducibility Responses for a single badge given repeated exposures 8%and readouts shall not differ by more than 10% for anexposure of 20 mR.

Length of Field The ratio of the response obtained for the field cycle 0.96Cycle period to twice that obtained for half the field cycle

period shall not be less than 0.90.

Energy Response shall not differ from that obtained with the ± 25% fromDependence calibration source by more than 20% for photons with 80 keV - 3 MeV

energies from 80 keV to 3 MeV.

Directional Response averaged over all directions shall not differ 6%Dependence from the response in the usualy calibration position

by more than 10%

Light The results of the light shielded TLDs shall not differ no differenceDependence from those obtained for the non-shielded TLDs by more

than 10%.

Moisture The response of the TLDs exposed to moisture shall not no differenceDependence differ from those not exposed to moisture by more than

10%.

Self-Irradiation The self-irradiation of the dosimeters shall not exceed no measurablea rate equivalent to 10 UR per hour for the duration of self-the field cycle. irradiation

Table 1. Dosimeter Performance Compared to ANSI N545-1975

References

1. Fitzsimmons, C.K., and Horn, W.H., "EnvironmentalMonitoring with Thermoluminescent Dosimeters"USPHS Report SWRHL-58r, October 1969

2. Cox, F.M., and Lucas, A.C., "An AutomatedThermoluminescent Dosimetry System for PersonnelMonitoring (Part I)," Health Physics, Vol. 27(October), pp. 339-345, 1974

3. Becker, K., Rosa, H.L., and Wing, P-S., "Environ-mental and Personnel Dosimetry in TropicalCountries," Proc. of Third Intern. Conf. onLuminescence Dosimetry: Riso, 1971, Riso ReportNO. 249

4. Burke, G. de P., "Thermoluminescence Dosimetry:Environmental Monitoring Near Nuclear ReactorSites," IEEE Trans. on Nucl. Sci., NS-23, No.3,1224-1236, 1976

5. Brinck, W.L., Gross, K.C., Blanchard, R.L., Kahn,B., "Natural Radiation Measurements for Environ-mental Surveillance at Nuclear Power Stations,"Proceedings of the Tenth Midyear TopicalSymposium of the Health Physics Society, 1976

6. Burke, G. de P., "Variations in Natural Environ-mental Gamma Radiation and Its Effect on theInterpretability of TLD Measurements Made NearNuclear Facilities," USAEC Report, HASL 289, 1975

7. Lantz, M.W., "A Predictive Model of BackgroundRadiation Rates," Master's Thesis, San Diego StateUniversity, 1977

8. American National Standards Institute, "AmericanNational Standard Performance, Testing, andProcedural Specifications for ThermoluminescenceDosimetry (Environmental Applications)," ANSIN545-1975, 1975

718