S.R. Bose and R. U. Mulder
Department of Mechanical,
Aerospace and Nuclear Engineering
University of Virginia
Charlottesville, VA 22903
C.A. Sondhaus and D. Silvain
Radiation Control Office
University of Arizona
Tucson, AZ 85724
M. E. Wacks
Department of Aerospace and Mechanical
Engineering
University of Arizona
Tucson, AZ 85721.
ABSTRACT
The assessment of pre-operational benchmark radioactivity in and around a nuclear facility is necessary to determine whether an observed change in the radioactivity levels is due to the operation of the facility. A detailed investigation of radioactivity levels in soil samples obtained from sites at the University of Arizona (UA) campus and at the off-campus Page Ranch Radioactive Waste Repository (now closed) was carried out during 1994-95. The on-campus soils were analyzed to evaluate a potential influence on the radioactivity in the local environment due to research application of isotopes at various laboratories at the University and operation of the UA TRIGA reactor. Unfortunately, no background measurements were made prior to operation of either facility. In the absence of baseline data, off-campus soils at and near the repository were analyzed to compare radioactivity in these samples as well as with the University samples in reference to soil samples collected on the UA campus.
The variation in natural radioactivity found was attributed to normal ranges due to soil and rock type affected by natural erosion, weathering and rock ageing processes. The results indicate that reactor fission or activation products are not present in the soils obtained from the UA either on- or off-campus. However, measurable levels of 137Cs fallout were observed in both on- and off-campus samples. These variations are not related to the research reactor activities but to natural processes such as rainfall variations, wind patterns and surface water flow.
INTRODUCTION
Assessment of benchmark local environmental radioactivity is a reasonable prerequisite for any nuclear facility. The levels of operationally-induced radioactivity in soils of the area in and around the facility must be monitored to determine an increase of radioactivity due to facility operation in a timely manner. Therefore, radioactivity data must be obtained under pre and post operational conditions, as well as during operation.
The University of Arizona (UA) has a pool-type research reactor (TRIGA), for which the licensed steady state power is 100 kW. This reactor has been in operation since December 1958 and is used for instructional reactor operations (steady state and pulsed experiments) as well as research and development (R&D) programs. During normal reactor operation, corrosion products of the materials in the system and trace impurities present in the charging water are activated and removed in a dual bed ion exchange system. A possible source of mobile radioactivity are the fission products in the fuel. They may be released during fuel elements failure in abnormal situations such as dropping an element when installing it or changing its location.
During normal operation, the soil adjacent to the reactor pit (if the pool or tank is small) may capture fast and thermal neutrons escaping from the reactor system. The UA reactor core is 6.4 m distant from the soil below the reactor vessel. This soil is irradiated by a very weak neutron flux. Therefore, this soil may contain trace activation products at levels depending on soil composition, reactor power history and isotopic half-life. Generally, the isotopes in activated soil are short half-life beta emitters. They do not present an environmental hazard. Additionally, the local climate is arid and the structure limits groundwater access. Also, they are present in negligible concentrations.
The UA owns a closed shallow-land burial site located at the Oracle Agricultural Center (Page Ranch) in Pinal County, about 45 km from the University. Different types of radioactive waste were disposed of there by the Radiation Control Office (RCO) of the UA from 1962 until 1986. In the past, the RCO collects waste on a monthly basis, and transported it to the burial site via pick-up and stack-bed truck. This waste included dry waste, liquid waste, animal carcases, animal wastes and scintillation vials.
The major radionuclides in the waste include 14C, 36Cl, 60Co, 137Cs, 244Cm, 3H, 63Ni, 90Sr, 133Ba, 22Na, 226Ra, natural U, depleted U, natural Th, 210Pb, and 55Fe. In addition, the following radionuclides are present in trace amounts: 241Am, 39Ar, 45Ca, 57Co, 134Cs, 54Mn, 65Zn, 109Cd, 204Tl, and 124Pm. Some low-level activation products (less than 1% of the total activity) generated by the UA TRIGA facility were also disposed at this site, but due to their short half-lives, those radionuclides have decayed. The waste was packed in metal, plastic or cardboard containers and deposited in pits later backfilled. The soil was compacted by a backhoe passing back and forth over the pits.
Over the years 1962-1986, UA generated about 1426 m3 of low-level radioactively contaminated wastes. The total activity of these wastes in January 1996 was estimated (1) at about 1.1x10 4 mCi. Of this total activity, the contributions of 3H and 14C were 79% and 20%, respectively. The remaining radionuclides constituted only 1% of the total activity.
The disposal site is in a mesquite-grasses-cholla ecosystem. The climate is semi-arid with low rainfall, high evapotranspiration, and low humidity. Rain occurs as thundershowers in July and August that are characteristically of high intensity and short duration, and as rains in December and January which are gentler, of longer duration, and more prone to infiltration. The disposal site rests on 240-270 m of alluvial deposits consisting of unconsolidated and semi-consolidated clay, silt, sand, and gravel.
Whether radioisotopes originate from artificial or natural sources, they may eventually reach society through different pathways depend upon isotropic half-lives, release conditions and transfer mechanism. Human exposure is classified as either internal or external. To cause internal radiation exposure, radioisotopes deposited in soils, water, vegetation, aquatic species etc., must enter the human body via ingestion of food, or through inhalation of dust. External exposure results from gamma-ray radiation emitted by radionuclides present in soils and building materials. Both internal and external exposures are caused by a combination of natural and artificial sources.
Among the artificial sources of isotopes, 131I, 90Sr and 137Cs are recognized as the most important. Strontium-90 and 137Cs, which are generally present in the environment as a result of fallout (weapon testing), are biologically important because of their i) high yield in the fission process (5.76% and 6.13% respectively), ii) long physical (about 30 years) and biological half lives, and iii) easy entry into the human body. Iodine-131 has a short half-life (T1/2= 8 day) and is a problem only for a short period (perhaps 3 months) after release. Iodine is accumulated in the thyroid and is of concern only in the event of serious nuclear accidents concurrent with fuel damage and fission product release.
The natural sources of radioactivity in the environment are generally classified into two groups; primordial and cosmogenic. The present investigation involves the primordial group, which includes 40 K and a majority of heavy elements isotopes belonging to the three naturally occurring decay series of 238U, 235U and 232Th. Environmental radioactivity is due principally to the presence of these series of nuclides in the earth, either because their half-lives are comparable to the age of the earth or because they are being produced continually through the decay of unstable parent nuclides. Isotopes in the 238U and 232Th series are comparably more important because the 238U and 232Th abundances are significantly higher than the 235U abundance. Uranium is a mixture of three isotopes consisting of 238U (99.3%), 235U (0.7%) and a trace quantity of 234U. The specific activity of 238U in natural U is 12.2 x 10-3 Bq kg-1. Thorium-232 is present as 100% of natural Th, and has a specific activity of 4.0 x 10-3 Bq kg-1. In the course of time, both 238U and 232Th reach near radioactive equilibrium with their respective daughters (2).
Uranium may enter the food chain via plants, at levels depending mainly on the soil matrix, plant metabolism and 238U availability. Uranium-238 daughters play a significant role in the exposure of man and his environment through naturally occurring radioisotopes. For example Radium-226 is chemically similar to Ba or Ca and is easily absorbed from soil by plants. Thus, the main sources of environmental "contamination " from the U and Th series are due to diffusion of radon gas into the atmosphere in living and work places and other series decay products formed at the time of daughters decaying. Certain areas of the world, such as the monazite sand beaches of Brazil and India, have background radiation levels on the order of 10 times the average worldwide background (3-5) due to the high U and Th content of the sand.
The naturally occurring 40K isotope is radioactive, having a
half life of 1.3 x 109 years; its abundance in natural K is 0.0118%
(6), imparting a specific (
-)
activity of about 30 Bq kg-1K. Because of sufficient abundance and
energetic beta emission (1.3 Mev), 40K is the predominant
radioactive component in everyday foods and in human tissues. It is, therefore,
responsible for most of the internal radiation dose. Since
40K is ordinarily present in all soils and emits a high energy
gamma-ray (1.46 Mev), it also contributes about one third of the external
radiation dose to man from natural background sources.
A detailed study was carried out during 1994-1995 to investigate the levels of natural and fallout radioactivity in soils obtained from sampling sites at and near the University of Arizona (UA) campus and at the off-campus Page Ranch (the closed radioactive waste repository). The UA soils were examined for increased radioactivity due to UA reactor operation. The Page Ranch soils, obtained from the waste area as well as from a nearby reference site, were compared similarly to see if the operation of the sites had affected the local radioactivity of the environment. Samples from all sites were collected four times a year, in the months of January, April, July and October for two years, in order to assess possible seasonal variation of activity levels. Levels of activity due to Th daughters, U daughters, 40K and 137Cs in the soil samples of the above locations were determined and compared.
MATERIALS AND METHODS
Sample Collection and Preparation
Soil samples were collected from the following on-campus locations: Engineering Building, Harvill Building, Football field, Franklin Building and Graduate Library of the University of Arizona (Fig. 1a). In addition, samples were collected also from the Page Ranch waste area, which is located about 48 km north of the University of Arizona campus (Fig. 1b). The Page Ranch repository is located within the University of Arizona's Oracle Agricultural Center and is situated at an elevation of about 1.5 km above sea level while the University is at an elevation of 1 km above sea level with Public access to the facility being restricted. The repository area is about 152 m by 30 m in size, and is surrounded by a fence to limit access to authorized personnel.
Surface soil samples were collected randomly from different points within the repository on a routine basis. Reference soil samples were collected from the east side of the repository. All soil samples were collected at the same time of the year from a depth of about 0 to 60 mm and processed for measurement under identical conditions. The samples were air dried at atmospheric temperature, ground, stored, weighed and sealed in aluminum cans, and then counted using an HPGe detector. The aluminum cans used had a holding capacity of approximately 425 g of soil or ~425 ml of water, and were supplied by the Central State Can Co, U.S.A. Soil sample dry mass in the cans was adjusted to 425 g for all measurements.
Fig. 1a. Soil sampling sites of the
University of Arizona.
Fig. 1b. Soil sampling sites of the
Page Range Radioactive Waste Repository.
Determination of Gamma-ray Activity
Each 425 g sample of dried soil contained in an aluminum can was "counted" using a shielded HPGe detector for about 15 h. For measurement of sample background, an aluminum can containing 425 ml of distilled water was counted for 15 h, and the spectral analysis was made under identical conditions used for the soil sample analysis. The background count of each water blank sample was subtracted from the collected soil sample count to obtain the true count rate due to radionuclides present in the sample. All count data were obtained using an energy calibrated detector. In order to determine counting efficiency of the detector, a standard reference sample (7) was counted in the HPGe detector under identical geometrical conditions. The counting efficiency of the detector as a function of gamma energy is shown in Fig. 2.
Fig. 2. Detector (HPGe) efficiency as
a function of gamma-ray energy.
Detailed information about equipment and operating instructions may be obtained from the Instruction Manual (8). The spectral analyses of the count data were made using the program "GDRMAIN" version 4.6 provided by Quantum Technology (9). The spectrometric analysis was verified using a different program, "PCA-II", obtained from Oxford Instruments (10). The prominent gamma-ray energies observed in the spectra for different radionuclides are presented in Table I. All the net counting rates were converted into specific activity (Bq kg-1) using the efficiency data of Fig. 2.
Table I Relevant Prominent Gamma-ray Energies
Counting Errors
The counting errors for individual radionuclide were calculated at a (±
) standard deviation, and the
values are expressed as percentage of the net counting rates. The average
activity for each radionuclide was calculated using the total number of samples
employed for the analysis. The error for the average value was obtained by
dividing the total error (total ) for each radionuclide by the total number of
samples. The Below Detection Limits (BDL) values for all the isotopes were
calculated in accordance with the standard literature formula (11).
The counting errors associated with the measurement of 228Ac, 212Bi, 212Pb and 208Tl ranged from 1.6% to 2.2% (average 1.9% ), 2.8% to 4.5% (average 2.9%), 0.5% to 1.0% (average 0.7%), and 1.0% to 1.7% (average 1.2%), respectively. The error for 232Th activity computed from the total error () for each radionuclide ranged form 1.5% to 2.0% with an average of 1.8%. The counting error on U daughters were observed to be 3.0% to 4.0% (average 3.8%) for 226Ra, 0.5% to 0.9% (average 0.7%) for 214Pb, and 0.9% to 1.2% (average 1.0%) for 214 Bi. The error for 40K had a range of 19% to 21% with an average of 20%. For 137Cs the counting error was found to vary from 0.1% to 0.6% with an average of 0.4%.
RESULTS AND DISCUSSION
The levels of activity for each radionuclide including its counting error (%), and BDL (Bq kg-1) values for different radionuclides in the soil samples collected at the UA Engineering Building, Harvill Building, Football Field, Franklin Building, Graduate Library and Page Ranch area during 1994-95 are presented in Table II. The BDL values for a counting time of 54000 sec for 228Ac, 212Bi, 212Pb, 208Tl, 226Ra, 214Pb, 214Bi, 40K and 137Cs were estimated to be 9, 18, 4, 7, 17, 4, 5, 91 and 0.70 Bq kg-1, respectively. The observed levels of radioactivity are discussed below:
In accordance with the Th decay series (12), the activity of 232Th may be determined from daughters 228Ac, 212Bi, 212Pb and 208Tl. The Th daughters are assumed to be in secular equilibrium in the decay chain (13-15). It has also been reported (16) that the thorium daughters in any disturbed soil reach near-secular equilibrium if the soil is stored in a closed container for at least one month after collection. All samples in this measurement were stored in aluminum cans. The cans containing dry soils were sealed and counted after a minimum of one month's storage. Therefore, the daughter activity should provide a good basis for an approximation for the activity of the parent 232Th. The contribution of 208Tl is about 36% of its parent 212Bi in the decay series. On the basis of this assumption, the measured activity of 208Tl needs to be multiplied by 2.78 in order to get the activity of its immediate parent 212Bi (15,16).
Table II Moisture Content (%) in the University of Arizona on and
off Campus Soils Collected During 1994-95
The direct measurement of 226Ra in soil is difficult. The gamma-ray emission of 235U and 226Ra are at 185.75 keV and 186.10 keV, respectively. The energy resolution of the HPGE detector (full width at half maximum) used for the measurements was in the range of 2 keV. Therefore, it was not possible to discriminate these isotopes spectrometrically. The measurement of activity in the 186.10 keV spectrum included the activity of both 226Ra and 235U. The use of the 186.10 keV line for the measurement of 226Ra may lead to an error of about 25% in the 226Ra activity due to the presence of the 185.75 Kev line of 235U. Therefore, all the 226Ra data were corrected by assuming a 25% contribution of 235U (16), which introduces an additional uncertainty of about ± 1% in the 226Ra determination.
The activities of 214Pb and 214Bi, are relatively easy to measure without any interferences. They are daughters of 226Ra, with abundant gamma emissions. In undisturbed soils, the short lived daughter products of 226Ra are expected to be in secular equilibrium with 226Ra. Since the half life of 222Rn is 3.82 days, it may be assumed that in disturbed soils, a considerable amount of 222Rn will escape, once it is produced in the decay chain, into the environment by diffusion through the soil. This 222Rn diffusion disturbs the equilibrium ratios between 214Pb and 214Bi in the decay chain. As a result, secular equilibrium among the U daughters in the decay series, i.e, the daughters after 226Ra, may not exist. In a medium of undisturbed surface soils, however, the activity of 238U is reported to be in secular equilibrium with its daughter 226Ra (16,17). It has been reported that 222Rn and decay products normally remain trapped in the crystal lattice structure of the soil matrix, and equilibrium is only disturbed through chemical processes such as acid digestion or heating (18). Van Cleef (1) in 1994 reported that the activity of 226Ra can be measured by the determination of its daughters 214 Pb and 214 Bi. Therefore, measurement of the activity of 226Ra in an undisturbed soil can provide an estimate of the activity of 238U in the same sample. The activity levels of different radionuclides in the samples collected from different UA on- and off-campus locations are discussed in the next section and presented in Table III.
Table III Radioactivity (Bq kg-1 dry soil ± percent error) in
the University on and off campus soils collected during 1994-95
Activity of Thorium Daughters
The observed activity levels of 228Ac, 212Bi, 212Pb and 208Tl in the UA on- and off- campus soils range from 37 ± 1.6 to 176 ± 2.4 Bq (average 73 ± 1.8 Bq), 35 ± 2.8 to 188 ± 4.5 Bq (average 74 ± 2.8 Bq), 38 ± 0.50 to 174 ± 1.0 Bq (average 69 ± 0.7 Bq), and 36 ± 1.0 to 161 ± 1.7 Bq (average 68 ±1.2 Bq) kg-1 dry soils. The level of 232Th (i.e. the average activity of the Th daughters), ranges from 47 ± 1.6 to 117 ± 2.0 Bq with an average of 71 ± 1.7 Bq kg-1 dry soil. When the on- and off- campus soils were analyzed separately, the following results were obtained. For on-campus soils, the activity of 232Th ranges from 47 ± 1.6 to 66 ± 1.7 Bq with an average of 55 ± 1.6 Bq kg-1 soil. In the case of off-campus (i.e. Page Ranch) soils, the activity ranges from 105 ± 2.0 to 117 ± 2.0 Bq with an average of 111 ± 2.0 Bq kg-1 dry soil. It is observed that the activity of 232Th for off-campus soils is higher by a factor of two than that of on-campus soils, probably due to different mineral forms in the soils. The literature survey shows that the levels of 232Th in the UA soils compare well with those reported for Virginia (16) (range 5 to 84 Bq kg-1 dry soil) and Louisiana (20) (range 11 to 62 Bq kg-1) soils. The activity of 232Th in Spanish soils as measured from the 911 keV peak of 228Ac was reported (21) to be widely variable, ranging from 5 to 254 Bq with an arithmetic mean value of about 41 Bq kg-1 soil. This average value is comparable to those observed in the UA on-campus soils, but lower than the off-campus (Page Ranch) soils. The world average concentration of 232Th was reported to be 25 Bq kg-1 dry soil (22). In our study, the observed results indicate that the levels of 232Th are higher than the world average by a factor of about two for on-campus soils, and a factor of four for off-campus soils, respectively.
Activity of Uranium Daughters
The soil activity data indicates that 226Ra was not in secular equilibrium with its daughters. Therefore, the average activity of Ra daughters may not represent well the activity of 238U. The activity of 214Pb and 214Bi averages about 60% of the activity of 226Ra. Since 226Ra is close to its parent 238U in the decay chain, the activity of 226Ra may provide an estimate of the activity of 238U (16,17). The activity of 226Ra in the UA soils ranges from 50 ±3.0 to 135 ±4.0 Bq with an average of 82 ± 3.8 Bq kg-1 dry soil. On the basis of on- and off-campus soils, the concentrations of 226Ra range from 54 ± 3.0 to 103 ± 3.6 Bq (with an average of 74 ± 3.3 Bq kg-1) for on-campus, and 75 ± 3.4 to 135 ± 4.0 Bq (with an average 102 ±3.6 Bq kg-1) for off-campus soils. The activities of 214Pb and 214Bi in the UA soils range from 31 ± 0.6 to 96 ± 0.9 Bq (average 48 ± 0.7 Bq), and 31 ± 0.9 to 98 ± 1.2 Bq (average 50 ± 1.0 Bq) kg-1 soil, respectively. The observed range of 226Ra, 214Pb, and 214Bi activities (range 8 to 310 Bq, average 51 Bq) kg-1 soil appear to be consistent with those reported for surface soils (2, 13,14,23). In comparison to the world average (22) (30 Bq kg-1) the levels of 226Ra in on- and off- campus soils are higher by factors of two and four, respectively.
Natural 40K Activity
The levels of 40K in different sampling sites of the UA range from 872 ± 19 to 1568 ± 20 Bq kg-1, with an average of 1113 ± 18 Bq kg-1 dry soil. Since the activity of 40K in both on- and off-campus soils shows a similar trend, no clear interpretation regarding differences in activities between these sites can be made. However, some variation of activity levels among different sampling sites was observed. From the average result, it may be seen that the 40K activity in the soils obtained from the Franklin Building site appears to be higher than those obtained at other sampling sites. The reason for such variation may be attributed to the particle size of K bearing minerals, soil weathering conditions, soil compositions, and leaching/migration of K containing particles due to rainfall. The observed data compare well with previously reported values in the literature (16,21). The world average activity of 40K is reported (22) to be 370 Bq kg-1, which is about three times lower than the average result of our observation.
Fallout 137Cs Activity
The activity of 137Cs in the UA soils varied widely, having a range of BDL ± 0.1 to 44.2 ± 0.6 Bq with an average of 7.0 ± 0.4 Bq kg-1 dry soil. In general, the activity in the samples of Page Ranch and Graduate Library are higher than those obtained in the rest of the sampling sites. It may be seen that the highest level of 137Cs was obtained in the reference sample (Page Ranch reference), and the lowest in the sample obtained from the Franklin Building site. The average results indicate that 137Cs activity within the Page Ranch radioactive waste repository is lower than that observed in the reference site (Page Ranch reference (outside)). The level of 137Cs in the sample obtained from the site of Engineering Building is very similar to those observed in other on-campus sites, except for the location of the Graduate Library where the sample registered a higher activity. The deposition of fallout 137Cs on soil surface depends on atmospheric wind direction and rainfall. Cesium is a monovalent element and is quite soluble. Therefore, the variation of 137Cs levels among different sampling sites is mainly due to variation of concentration levels of this element in the atmospheric air, and its ultimate fallout on the neighboring ground soils by precipitation and rainfall. The observed results compare well with the 137Cs levels (range BDL to 71 Bq kg-1 surface soils) reported earlier (2,14,17,24).
CONCLUSION
The soil samples collected from the vicinity of the UA Engineering Building, during the period of January 1994 - January 1995 did not show any fission or activation products. The results for 137Cs clearly indicate that there is no 137Cs impact on the on-campus soils due to operation of the UA TRIGA reactor, although a detectable level of fallout 137Cs was observed in both on- and off-campus soils. These levels of 137Cs activity are most probably the result of atmospheric fallout due to nuclear weapon tests (25-30) . The results indicate that secular equilibrium among the thorium daughters exists in all soil samples considered for this study and that the assumption of secular equilibrium between U and its decay products cannot be made. The reason for this inconsistency cannot readily be explained without further detailed studies. In order to satisfy this point a separate study, which was not the main purpose of the present investigation, may eventually be carried out. The activity of Th and U daughters, on the average, is observed to be higher in the off-campus, Page Ranch, than the on-campus soils.
The results of the overall study may be used as reference for assessment of environmental radioactivity in the areas surrounding the on-campus TRIGA reactor, as well as the off-campus radioactive waste repository. The on-campus information will be needed before eventual decommissioning of the reactor, and will also be useful in the event of actual or alleged dispersal of radionuclides from the reactor.
ACKNOWLEDGMENTS
Dr. John Williams, Professor, of the UA NEE Department, kindly supported the principal author during this work in summer of 1994. Dr. Williams made many valuable suggestions useful for the completion of this study. His time, advice and financial support is deeply appreciated. The assistance of the RCO staff in the completion of this work is also gratefully acknowledged.
REFERENCES