弁財天

ゴフマン「専門家を信じるのではなく、自分自身で考えて判断せよ」

原発からの放射性希ガス放出

RADIOACTIVE NOBLE GASES IN EFFLUENTS FROM NUCLEAR POWER STATIONS

H. E. Kolde, W. L. Brinck, G. L. Gels, and B. Kahn
Radiochemistry & Nuclear Engineering Laboratory
U. S. Environmental Protection Agency
National Environmental Research Center
Cincinnati, Ohio 45268

Abstract

The discharge rates, isotopic composition, and major in-plant pathways for radioactive krypton and xenon in effluent gases at commercially operated nuclear power stations in the United States are reported. Specific information was obtained in the course of radiological surveillance studies at two BWR and two PWR stations, and periodic average discharge rates are reported by station operators. Techniques for radiochemical^ analyzing these radioactive gases are described: in-plant samples at relatively high concentrations are analyzed directly by gamma-ray spectrometry; low-level samples, especially those collected in the environment, are first concentrated, and may also be treated to separate krypton and xenon from each other and from other gases. Also discussed is the measurement of radionuclide concentrations in air by radiation dosimetry.

INTRODUCTION

Radioactive noble gases are major waste products of nuclear reactor operations. Krypton and xenon are generated along with many other radionuclides by nuclear fission. Although practically all fission products remain in place within the fuel, a small fraction leaks through minute cladding imperfections into the reactor coolant. Radioactive gases from the coolant, at present, are held for partial radioactive decay, and then are discharged to the atmosphere. Being relatively non-reactive, they remain in the air, and may be measured with very sensitive instruments for some distance downwind of the release point, until atmospheric dispersion processes dilute them to a level difficult to detect. In many cases, detection limits are set by natural background radiation. Local populations are exposed to the radiations emitted by the plume, although such exposures have been small fractions of the annual allowable limits permitted by governmental regulations (Rogers, et al, 1971). Based on this experience, and the concept of lowest practicable population exposure, guidelines are being considered to restrict radioactive reactor effluents to levels that would keep exposure to persons living near stations to five percent or less of average natural background radiation (AEC NR4-30, 1973).

The use of large reactors and multiple reactor sites for commercial electrical power generation has increased significantly during the past decade. To meet needs for more detailed information on reactor-discharged radioactivity and possible environmental effects, studies were begun in 1967 by the U. S. Public Health Service, and are now being continued by the U. S. Environmental Protection Agency. Many of the discharges and environmental measurements at that time were being reported by plants only on a gross activity basis. The studies identify internal plant pathways leading to the discharge of specified radionuclides, determine the degree of dispersion afforded by local meteorology, test the applicability of mathematical models and sampling and analytical techniques, and measure the radiation dose in the environment.

Studies were conducted initially at the Dresden 1 boiling water reactor (BWR) (Kahn, et al., 1970) and Yankee pressurized water reactor (PWR) (Kahn, et al., 1971), and later at the larger Haddam Neck (Connecticut Yankee) PWR (Kahn, et al., UP-a) and Oyster Creek BWR plants (Kahn, et al., UP-b). Dresden 1 and Yankee operate at approximately 200 megawatts-electrical (MW[e]) output; the latter two, at approximately 600 MW[e]. All studies were performed in cooperation with the state health or environmental protection departments, the U. S. Atomic Energy Commission (USAEC), and the station operator. At Oyster Creek, the USAEC also undertook many in-plant measurements.

GASEOUS WASTE HANDLING SYSTEMS

Of the more than 30 krypton and xenon radionuclides produced by fission, most change rapidly within reactor fuel to stable or radioactive nuclides of other elements. The noble gases of potential significance as gaseous wastes are those with longer half-lives, as listed in Table 1 with their fission yield, generation rate, and radioactive progeny. The generation rate, a convenient indication of relative importance, was estimated from fission yield and half-life values for a unit MW[e] (~ 3 megawatt-thermal) power operation. All except 85Kr possess half-lives of less than 5.3 days. Although generated in minor amounts, 85Kr is of interest because its discharge adds to the world-wide inventory of gaseous radionuclides. Fission-produced noble gases are accompanied by such other radioactive noble gases as 37Ar, 39Ar, and 41Ar produced by neutron activation of air in reactors. Other radioactive gases include 3H, l3N, 14C, and radioiodine. The relative abundance of all gaseous radionuclides in reactor effluents depends on reactor type and power level, fuel cladding, waste treatment system, and plant operating practices.

Gaseous waste pathways and treatment are different at BWR and PWR plants due to their distinct coolant systems. In the direct-cycle BWR, air that continuously leaks into reactor coolant water must be removed to maintain operation. The purged air contains the radioactive gases swept from the reactor with the steam. A PWR reactor primary system is sealed, resulting is gas volumes small enough to be accumulated, stored, and released in batches. The resultant shorter holdup times at BWR's produce higher effluent radioactivity levels than PWR's, as shown in Table 2, due to a preponderance of short-lived radionuclides. Average BWR release rates during 1972 were on the order of 104 /uCi/sec, based on discharge reports from six stations operating longer than one year. Average releases from seven PWR's operating before 1972 were lower, and more varied, ranging from 6 x 10-1 to 6 x 102 /aCi/sec (AEC, UP).

The Oyster Creek gaseous waste disposal system shown in Figure 1 illustrates typical pathways in BWR's built by the General Electric Co. The gases entrained in the reactor-produced steam are removed after passage through the turbines by air ejectors at the steam condensers. The off-gas is passed through a holdup system that normally requires approximately 72 minutes for passage at Oyster Creek (23 minutes at Dresden 1). This achieves reduction of radioactivity through decay of short-lived radionuclides. A particular benefit is the decay of 90Kr before the effluent gas is filtered, so that no significant amounts of its 90Sr daughter can be formed in the environment. The off-gas is passed through high efficiency particulate air (HEPA) filters before entering the stack. The gas is diluted by a factor of about 2000 at the base of the stack by ventilation air exhausted from the turbine, reactor, and occasionally other buildings. A tall 112-meter stack at Oyster Creek (91-meter at Dresden 1) provides additional dilution by atmospheric dispersion before the plume reaches the ground.

Other BWR pathways that contribute in minor ways to effluent radioactivity include: (1) steam leading through turbine gland seals from which gas is evacuated and pumped to the stack through a 2-minute holdup line; (2) leaks from various reactor components, pipes, etc., into building ventilation air; (3) turbine building air exhausted partially through roof vents during warm weather; and (4) radioactive gas emanating from liquids contained in waste tanks.

In the PWR, radioactivity from activation or leaking fission products accumulate in the primary coolant, until adjustments of coolant volume or composition are made, resulting in partial removal of gases. Figure 2 depicts the pathways observed at Haddam Neck (Yankee has similar pathways). Removed gases at Haddam Neck are collected in a sphere where they are held for decay. The gas is discharged 2 to 4 times per year to a 53meter high vent stack at Haddam Neck (46-meter at Yankee), diluted with building ventilation and outside air by large fans, and discharged to the atmosphere.

Other sources include the contamination of secondary coolant by leakage of the radioactive primary coolant through faulty heat exchanger tubes. Gases in the secondary coolant are removed continuously by air ejectors on the condensers and vented directly to the discharge stack. When the reactor or its vapor container are opened for refueling or major maintenance, gases accumulated in these vessels are purged with large volumes of air and exhausted to the vent stack. Other releases to the stack occur when aliquots of primary coolant are depressurized and collected for analyses, or as gas separates from liquids stored in waste tanks.

These descriptions of gas treatment systems do not include changes being planned to meet the proposed USAEC regulations for "as low as practicable" levels of effluent radioactivity. Most new systems feature increased holdup times by adsorbing gases on charcoal or molecular sieve beds at ambient or lower temperatures. An alternate method is cryogenic distillation of noble gases and storage in gas cylinders. In many cases, the new systems will be incorporated in major radioactivity pathways, and plant discharge of gases by secondary routes may continue at current levels.

MEASUREMENTS IN GASEOUS WASTE HANDLING SYSTEM

Gases within the reactor plant are collected for measurement in sample containers of sizes determined by radioactive concentration and analytical requirements. Gases highest in radioactivity are obtained in sealed glass serum bottles 4 to 15 ml in volume, while those of lower concentration are collected in evacuated metal 2liter bottles. Gases that emit photons are identified and measured with Ge(Li) or NaI(Tl) detectors coupled to multi-channel analyzers. Aliquots of large samples are transferred for counting to sealed evacuated 200-ml volumetric flasks. Samples are counted for periods ranging from 1 to 1,000 minutes, and counts are repeated at intervals.

Krypton-85, at relatively low concentrations, is processed through a gas separation apparatus developed by the Las Vegas National Environmental Research Laboratory, EPA. Since aliquots are usually of small volume, the apparatus at this laboratory has been modified to incorporate a 83mKr tracer with each sample to determine separation yield. The Kr fraction is transferred to 25-ml vials, containing approximately 15 ml of 1mm plastic scintillator spheres, and measured by conventional liquid scintillation counters (Stevenson, et al" 1971).

Results of radionuclide measurements in the internal pathways at the Oyster Creek BWR and Haddam Neck PWR are summarized in Table 3. These data represent annual releases based on average observed concentrations normalized to duration or frequency of various operations per year ― such as, number of days of reactor operation, stored gas release volume, how often the reactor is opened for refueling or maintenance, etc. The data confirms that PWR releases tend to be long-lived, mostly 133Xe and 85Kr. BWR effluents contain similar amounts of long-lived radionuclides, but also short-lived radionuclides in much greater amounts. Gaseous effluents other than noble gases are mostly 10-minute 13N at BWR's, and tritium as water vapor at PWR's.

Relative emissions through principal BWR pathways are indicated in Table 4 in terms of 133XE and l35Xe measurements at Oyster Creek. Practically all atmospheric discharge resulted from gas removed by the air ejectors from reactor steam after passing through the turbines. Gas escaping from turbine seals and contaminated ventilation exhaust account for less than one percent.

From observations at Haddam Neck and Yankee, no single PWR pathway appears to predominate. Typical Haddam Neck pathways, and their estimated annual contributions, are given in Table 5; the two most abundant gaseous radionuclides are used as indices. Most of the 133Xe discharge resulted from gas leaking from the reactor into the secondary coolant system and from contaminated ventilation air. Purging the gases accumulated in the reactor vapor container resulted in discharge of much of the longer-lived 85Kr. At the Yankee PWR, most plant discharge, on an annual basis, was long-lived, and resulted from vapor container discharge. The major pathway for short-lived noble gases, although a small fraction of the total, was losses occurring during primary coolant sampling.

MEASUREMENTS IN THE ENVIRONMENT

PWR discharges ― mostly the longer-lived radionuclides 85Kr and133Xe ― lead to maximum radiation doses of approximately 1 mrem/year. BWR release rates, usually 100 to 1000 times higher than PWR's, but discharged from tall ( ^ 100-meter) stacks, result in maximum doses of about 10 mrem/year. The maximum exposure applies usually to small groups of people living near the reactor exclusion boundary in prevalent downwind directions; whereas most people, residing at greater distances or in less prevalent wind directions, receive much less dose. By comparison, these dose levels are small fractions of the average per capita dose of 130 mrem/year from natural radioactivity in this country, or the 500 mrem/year allowed for operations licensed by the USAEC. The stringent proposed criteria of five percent or less of average natural radiation, however, will require the aforementioned gas treatment in many instances.

Highly accurate and sensitive techniques are needed to measure ambient reactor radioactivity, and to distinguish it from natural background radioactivity, which fluctuates at a relatively low intensity. Often, short-term sampling is conducted only when meteorological, or other conditions, are favorable. Typical environmental concentrations and doses from BWR and PWR releases are given in Table 6. The PWR example applies during release of stored gas ― usually the only occasion when its plume is detectable.

To determine the effects of a planned reactor, radiation to populations within 80 km of the site is estimated from expected atmospheric dispersion of effluents, taking into account local wind and atmospheric stability patterns. Environmental studies by this laboratory are made to confirm such dose estimation techniques, and to seek sensitive methods for monitoring exposure.

Plumes in the environment at BWR's can usually be detected readily with 5 x 5-cm NaI(Tl) detectors coupled to count-rate meters. The plume at a PWR during a stored gas release was located with a large thin (2-mm thick by 13-cm diameter) NaI(Tl) probe developed for the detection of low-energy photons ― in this case, the 81 keV gamma ray from 133Xe. Such instruments can provide semi-quantitative exposure rates, if they are calibrated with reference, for example, to an ionization chamber.

Short-term measurements, to test dose estimates from plume gamma radiation, are obtained with a Shonka tissues-equivalent ionization chamber coupled to a sensitive vibrating capacitor electrometer. The instrument can measure an increase of approximately 0.1 rad/hr, which corresponds to a steady exposure of 1 mrad/year. Each reading requires 30 seconds to 10 minutes, depending inversely on the radiation intensity. The addition of a chart recorder allows continuous instantaneous readings (Gustafson, et ai, 1964).

Long-term exposure monitoring is accomplished with a commercially available, high-pressure ionization chamber and a continuous recorder. Its sensitivity is similar to the Shonka ionization chamber. Simultaneous monitoring at many locations is performed with thermoluminescent dosimeters (TLD's). TLD's must be carefully selected for low intrinsic radioactivity, and, at best, are less sensitive than ionization chambers. Optimum sensitivity of the EG&G Model TL-15 CaF2(Mn) type, for example, is approximately 10 mrad/year. Taking the TLD reader to the reactor site eliminates the need to account for the sizable dose accumulated in shipping TLD's (Beck, et al, 1972).

Samples for determining ambient concentrations of radioactive gases in the plume are obtained by compressing air into 34-liter metal bottles, with rated capacities of 0.9 m3 each. Xenon-133 is a useful radionuclide for analysis because its discharge rate is relatively high at both BWR's and PWR's and its halflife is conveniently long. Xenon-133 is concentrated for analysis by passing 100-liter aliquots through a 450-ml bed of charcoal immersed in a dry ice-acetone refrigerant bath. The charcoal is transferred to a sealed 450-ml container, and analyzed by a gamma-ray spectrometer. Minimum detectable concentration, using a 10- x 10cm NaI(Tl) detector, is 400 pCi/m3 for analysis 5 days after sampling.

Atmospheric dispersion values can sometimes be determined by passing large volumes (> 1 m3/min) of air through particulate filters for sampling the 17.8-min 88Rb and 32-min 138Cs progeny of noble gases. Their short half-lives, however, require that multichannel analyzers be nearby for immediate counting. Since the concentration varies as the plume shifts, sampling durations must be restricted to less than an hour to minimize errors in decay calculations.

Sampling ground-level air in the environment, as well as stack effluents, is a direct method for determining the effects of local meteorology and topography on plume dispersion. Samples of airborne particles may also give this information, but require rapid analysis and some interpretation of results. The Shonka tissueequivalent ionization chamber yields direct readings of exposure rates with high precision, but the instrument must be handled carefully in the field, and is sensitive to adverse weather conditions. The high-pressure ionization chamber is useful for measuring plume radiation in the environment for long periods of times. Simultaneous dose measurements at many locations is obtained economically with low background TLD's, when high sensitivity is not necessary. For long duration measurements, monitoring, to integrate fluctuating natural radiation contributions during the period, must be performed to obtain the net dose from the plume.

REFERENCES

Beck, H., et al., 1972), New Perspectives on Low Level Environmental Radiation Monitoring Around Nuclear Facilities, Nuclear Technology 14, 232-239.

Gustafson, P. F., J. Kastner, and J. Luetzelschwab, (1964), Environmental Radiation: Measurements of Dose Rates, Science 145,44.

Kahn, et al., (1970), Radiological Surveillance Studies at a Boiling Water Nuclear Power Reactor, U. S. Public Health Services Rept. BRH/DER 70-1.

Kahn, B., et ai, (1971), Radiological Surveillance Studies at a Pressurized Water Power Reactor, U. S. Environmental Protection Agency Rept. RD 71-1.

Kahn, B., et ai, (UP-a), Summary Report ― Field Trips to Haddam Neck Nuclear Power Station, U. S. Environmental Protection Agency Rept., to be published.

Kahn, B., et al., (UP-b), Radiological Surveillance Studies at the Oyster Creek Nuclear Generating Station ― Progress Report No. 2, U. S. Environmental Protection Agency, not published.

Martin, M. J., Nuclear Data Project, Oak Ridge National Laboratory. Untitled data, to be published.

Meek, M. E., and B. F. Rider, (1972), Compilation of Fission Product Yields, General Electric Co. Rept. NEDO-12154.

Rogers, L., and C. C. Gamertsfelder, (1971), USA Regulations for the Control of Releases of Radioactivity into the Environment in Effluents from Nuclear Facilities, Environmental Aspects of Nuclear Power Stations, International Atomic Energy Agency, Vienna, 127-144.

Stevenson, D. L., and F. B. Johns, (1971), Separation Techniques for the Determination of8SKr in the Environment, Rapid Methods for Measuring Radioactivity in the Environment, International Atomic Energy Agency, Vienna, 157-162.

LJ. S. Atomic Energy Commission, (AEC, UP), Report on Releases of Radioactivity in Effluents and Solid Waste from Nuclear Power Plants for 1972, to be published.

U.S. Atomic Energy Commission (1973), News Releases 4, No. 30 (AEC, NR4-30).

投稿されたコメント:

コメント
コメントは無効になっています。