Affordable Cleanup?: Opportunities for Cost Reduction in the Decontamination and Decommissioning of the Nation's Uranium Enrichment Facilities (1996)

This appendix supplements chapters 2 and 3 on the radionuclides at the gaseous diffusion plants (GDPs), the characterization process, instrumentation for radionuclide detection, and regulatory requirements on contamination levels.

Table 2-4 in Chapter 2 lists the radionuclides present at the GDPs and the ionizing radiations by which these radionuclides can be detected. Detailed decay schemes, including other radiations that are emitted at very low intensity but may be useful on occasion, can be found in nuclear data tables (Kocher, 1977). The radionuclides present at the GDPs consist mostly of naturally occurring uranium—uranium-238, -234, and -235 (²³⁸U, ²³⁴U, and ²³⁵U)—with their short-lived progeny. Some material fed to the GDPs was uranium recycled after use in nuclear reactors. This material contains a number of radionuclide contaminants, as illustrated in Table E-1.

Technetium-99 (⁹⁹Tc) is a high-yield (6 percent) fission product. Some ⁹⁹Tc accompanies uranium during reprocessing of spent reactor fuel and forms a gas during fluoridation. Hence, recycled uranium is contaminated with ⁹⁹Tc. In the cascade, the relatively light ⁹⁹ Tc moves toward the enrichment end. Traces of plutonium-239 (²³⁹Pu) and neptunium-237 (²³⁷Np) accompany recycled uranium and are present near the feed point of the cascades.

The radioactive impurity shipments to Paducah given in Table E-1 overstate the amounts in the cascade because only about 85 percent of ⁹⁹Tc and 25 percent of ²³⁷Np and ²³⁹Pu accompanied the uranium feed (Smith, 1984). At Oak Ridge also, only about 25 percent of the ²³⁷Np and 1.5 percent of ²³⁹Pu present in the recycled uranium entered the cascade. Moreover, many of the contaminants deposited by 1976 were removed during the cascade improvement and upgrade effort during the 1980s (Ritter et al., 1990).

Uranium salts have been deposited within the cascades on surfaces as thin films and in bulk at cool locations and when moisture enters. These salts exist outside the cascades due to leaks or to seals being breached for repairs or changes. Moisture changes the chemical form of the uranium gas by hydrolysis from uranium hexafluoride (UF₆) to uranyl fluoride (UO₂F₂). The ²³⁸U:²³⁵U:²³⁴U ratio in natural uranium feed is 1:0.0072:0.000054 by weight and 1:0.047:1 by activity (decay rate); the ratios of ²³⁵U and ²³⁴U to ²³⁸U increase with enrichment in the cascade.

Uranium purification and conversion to gas in preparation for enrichment by diffusion remove the radioactive progeny of the ²³⁸U/²³⁴U series and the ²³⁵U series. In time, the direct progeny with relatively short half-lives (see Table 2-4) approach radioactive equilibrium again and then decay at the same rate as the parents. Uranium-236 (²³⁶U) is produced by neutron activation of ²³⁵U (competing with the fission process). Its immediate radioactive progeny is long-lived and would not have accumulated noticeably.

Traces of other radionuclides can be estimated from the amounts of observed radionuclides in Table E-1. The first long-lived progeny in the uranium chains are thorium-230 (²³⁰Th, with an 80,000 year half-life) following ²³⁸U/²³⁴U and proactinium-231 (²³¹Pa, with a 32,800 year half-life) following ²³⁵U. For these, the average in growth can be calculated to be 0.693 times the ratio of the average in growth period to the half-life; if the period is taken to be 23 years, the fraction of ²³⁰Th to ²³⁸U is 0.00020, and of ²³¹Pa to ²³⁵U 0.00050. Relative to the values in Table E-1, ²³⁰Th and each of its progeny would amount to 140 curies (Ci), and ²³¹Pa and each of its progeny to 16 Ci.

Small amounts of other long-lived fission and activation products, including strontium-90, cesium-137 (¹³⁷Cs), various uranium and plutonium isotopes, americium-241 and curium-244, may also have accompanied recycled uranium. ¹³⁷Cs has been detected at Paducah and ²³²U at Portsmouth.

An initial or scoping characterization report must be prepared to plan the decontamination and decommissioning (D&D) program by mapping both the uncontaminated areas and the extent of contaminated surfaces and materials. Much of this information can be compiled from available reports, although some additional characterization will undoubtedly be needed to fill information

gaps or improve detection sensitivity. In some instances, performing new surveys may be simpler and more reliable than interpreting obscure documents.

The initial scoping measurements will be concerned with uranium within the cascades that can be located by portable detectors that measure gamma rays emitted by ²³⁵U and ^234mPa or neutrons emitted by the uranium isotopes. The amount of uranium within the cascade must be estimated from the outside for complex source-detector geometries. Uranium contamination on building surfaces is measured by moving alpha-particle monitors in close proximity to the surfaces. Beta-particle monitors detect ^234mPa under thin layers and at greater distances, and gamma-ray monitors can detect ²³⁵U and ^234mPa within solids. Hence, in situ characterization depends on analyzing data from all three types of detectors. Laboratory analysis provides calibration for in situ monitoring, better delineates horizontal and vertical radionuclide distributions, and detects individual radionuclides that are obscured by others in field monitoring.

The minor radionuclides ⁹⁹Tc and ²³⁹Pu generally cannot be detected in situ in the presence of much larger amounts of uranium; their presence at specific locations is known from earlier characterization studies using laboratory analyses. Gamma rays emitted by ²³³Pa can be detected at locations where ²³⁷Np has become concentrated.

A second cycle of characterization guides decontamination and maintains control of radioactive materials while ensuring radiation protection for workers and others in the environment. Progress in removing uranium must be monitored for cascade components in situ and, after disassembly, for recyclable materials, building surfaces, and scrap.

Sufficient information is available from plant upgrading and from the Capenhurst decommissioning to prepare plans for characterization during cascade disassembly, component decontamination, and site cleanup, with associated waste processing and radiation protection of personnel.

Decontamination of building surfaces, particularly floors, by washing, scraping, or scabbling requires a complex characterization effort because the radiation to be detected comprises alpha particles, low-energy beta particles, and gamma rays. Alpha particles are hidden by paint and by wash solutions but can be detected on bare surfaces and in airborne dust. All three types of detectors are necessary to trace the movement of uranium, ensure its removal from surfaces, assay the concentration of resulting wastes, ensure worker protection, and measure effluent release rates.

The third and last cycle is compliance characterization to ensure that materials and equipment are suitable for free-release, continued use within restricted areas, or disposal as waste and to ensure that locations can be opened for uncontrolled access by the public. Measurements must be sufficiently sensitive to demonstrate that radionuclide levels meet regulatory limits and sufficiently comprehensive to represent all materials and locations.

The buildings that contain the cascades are most effectively monitored after they have been emptied. The effects of surface-covering materials, such as paint or sealant, in preventing detection of covered alpha particles must be compensated for by use of gamma-ray detectors. Covered areas, as well as locations at which penetration by radionuclides beneath surfaces is suspected, require depth sampling for analysis in the laboratory. Rubble from surface-decontaminated and dismantled buildings must also be sampled because surface analysis is not sufficiently informative. The vast expanse of the cascade buildings presents a challenge for applying innovative monitoring techniques. For surface monitoring, robotic monitors that can move independently across floors at a rate controlled by the collected radiation count rates and that process measurements for data analysis and mapping could be highly cost-effective. For collecting numerous samples, sampling patterns should be designed to give as complete coverage as possible using a reasonable sampling and analysis load.

Detecting ionizing radiations—alpha and beta particles and gamma and x rays—is the conventional method for measuring the radionuclides present at the GDPs. Contamination amounts are given in radiation units of curies (Ci) or subunits (e.g., picocuries, pCi), although mass units are commonly used at the GDPs.

The more familiar chemical analysis techniques can be applied only to radionuclides that have half-lives so long that they have measurable masses at pCi levels. For ²³⁸U and ²³⁵U, 0.33 pCi and 2.2 pCi, respectively are equivalent to 1 microgram. The masses of shorter-lived radionuclides with correspondingly higher ratios of activity per mass—6.4 pCi/nanogram for ²³⁴U, for example—may be measured by techniques utilizing mass spectrometry.

The heavy elements also emit neutrons at very low rates due to the fission process and the F-19(,n)²²Na reaction with the emitted alpha particles that interact with fluorine. The frequency with which neutrons are emitted by spontaneous fission per million alpha-particle disintegrations is 1.1 for ²³⁸U, 0.000022 for ²³⁴U, 0.0024 for ²³⁶U, and much less for ²³⁷Np and ²³⁹Pu. Neutrons of average energies just below 1 MeV are generated at higher rates by the (,n) reaction in UO₂F₂; estimated rates per million alpha particles are 1.4 in ²³⁸U, 0.79 in ²³⁴U, 0.46 in ²³⁵U, and 3.1 in ²³⁶U. Neutron emission rates in UO₂F₂ are about 1.6 per minute per gram of natural uranium but above 100 per minute per gram of highly enriched uranium, mostly due to ²³⁴U (Reilly et al., 1991).

Uranium levels have been monitored with the instruments listed in Table E-2 for many years, with some improvements and new developments. The approximate lower limits of detection listed in Table E-2 were estimated to indicate the applicability of specific detectors for various characterization efforts, particularly for checking uranium levels at release limits. In situ measurements are performed to locate uranium accumulated within the cascade and contaminating its components and surroundings and to estimate radionuclide levels of contaminated areas in real time. Laboratory analyses of samples collected from monitored objects and sites are more sensitive and accurate and can distinguish more effectively among several radionuclides that may be present; however, they are much more time consuming and costly.

Uranium within large, thick-walled cascade components (i.e., converters) is measured with fast-neutron detectors that are large and shielded against thermal neutrons. ²³⁵U amounts are inferred from neutrons generated mostly by ²³⁵U alpha particles in fluorine. In smaller cascade components, such as pipes and converters at the high enrichment end of the cascade, germanium (Ge) detectors with portable spectrometers are effective for measuring ²³⁵U and ^234mPa gamma rays. Uncertainty about the extent of gamma-ray attenuation circumscribes application of this technique for ²³⁵U to relatively thin uranium deposits and walls. Where difficulty in access prevents using a Ge detector with associated cryostat and Dewar flask, a dosimetry survey instrument with a sodium iodide (NaI[T1]) detector can be substituted.

Gamma-ray detectors are also useful to evaluate the progress of decontamination processes, survey areas for contamination, and check equipment and materials for free-release or transfer to waste repositories. Uranium is detected on the surface and within solids and liquids, both nearby and at a distance. By discriminating electronically against the full background energy spectrum and measuring only the characteristic gamma rays, the spectrometer detects smaller amounts and also provides isotopic analysis.

Alpha-particle detectors are most sensitive for measuring surface contamination. Because the alpha particles are stopped by 0.1-mm-thick solids or liquids or by 4 cm of air, application of these detectors is limited to unobstructed surfaces viewed in close proximity. Beta-particle monitors are used to measure the energetic beta particles emitted by ^234mPa. These beta particles are monitored more conveniently than alpha particles because they are not as readily attenuated. The low-energy beta particles emitted by ²³⁴Th and ²³¹Th, on the other hand, are strongly attenuated. The detection limit is poorer for these beta particles than for alpha particles because the background is higher.

Surface contamination is characterized as removable or fixed by rubbing approximately 100 cm² of the surface with a paper or cloth "smear" or "wipe." The "smear" is counted with an alpha-particle, beta-particle, or gamma-ray detector in the field or, for greater sensitivity, in the laboratory.

Sample sizes for laboratory measurements of samples by gross alphaor beta-particle activity are limited, because samples must be thin for counting with gas ionization and solid scintillation detectors or must be dissolved in small volumes for liquid scintillation counting. The sensitivity of laboratory analysis can be improved by processing a sample of several grams to separate uranium from the bulk of the sample medium and other radionuclides and then measuring alpha particles with a solid-state detector or liquid scintillation system and spectrometer for as many as 1,000 minutes. The effort of chemical separation can be avoided by counting gamma rays from a kilogram sample in a calibrated container with a Ge detector and a spectrometer.

Neutron activation analysis provides sensitive laboratory detection capability if an intense neutron flux is available. ²³⁵U can be determined by measuring any conveniently detected fission product; ²³⁸U by measuring the neutron activation decay product, ²³⁹Np. The sensitivity values in Table E-2 were estimated for a thermal neutron flux of 10¹³ neutrons/cm ²-s and measurement by gamma-ray spectrometer.

Uranium is analyzed chemically by measuring the fluorescence of a sodium fluoride melt in a platinum dish. The isotopic constitution of the sample must be known to convert mass units to the pCi units in which limits are specified.

Measurement in situ of ²³⁹Pu and ⁹⁹Tc is not feasible because any emitted radiations would be attributed to uranium or its progeny. ²³⁷Np may be detected by measuring the ²³³Pa gamma ray. Only prior knowledge and inferences from combined alpha-particle, beta-particle and gamma-ray monitoring can suggest whether ²³⁹Pu or ⁹⁹Tc are present. Smear samples and washing tests may differentiate ⁹⁹Tc by its chemical behavior. ²³⁹Pu, ²³⁷Np, and ⁹⁹Tc are identified and quantified by laboratory analysis. ²³⁹Pu and ²³⁷Np may be characterized with the same detection limits as uranium by radiochemically separating the elements and counting them, by distinguishing alpha particles by their energies with a spectrometer, or by distinguishing x and gamma rays with a spectrometer. ⁹⁹Tc is measured by counting it with gas-filled ionization or liquid scintillation detectors. The detection limit is approximately 1 pCi.

Some recent developments in radiation monitoring are listed in Table E-3. An inductively coupled plasma mass spectrometer can detect the equivalent of 0.2 pCi ⁹⁹Tc, 0.007 pCi ²³⁷Np and 0.6 pCi ²³⁹Pu per gram sample and is much more sensitive for uranium analysis. Although expensive, this analytical instrument may well be cost-effective for the numerous samples expected in this program.

A robotic radiation detection instrument carrier, programmed to survey large areas of floors, walls, or ceilings automatically and uniformly, can reduce work force needs and improve data uniformity. Conceptually, the data can be accumulated and promptly processed for conversion to activity per unit surface area and for preparing a radionuclide contamination map. The robot can be programmed to move around obstructions, but it is particularly effective for the large open areas and radiation fields of simple geometry that are expected at these plants. Robots can also survey radiation fields within vessels, pipes, and ducts and in narrow or remote areas, if designed to operate under dimensional restrictions.

The laser fluorescence monitor is more sensitive for uranium detection than a gamma-ray spectrometer, but its response depends on the chemical form of the uranium salts. This type of monitor could be useful as a field instrument even if additional information on chemical forms were required for quantitative analysis.

In long-range alpha-particle detection, the ions generated by alpha particles in air are measured in the air swept from the source to the detector. The technique is useful for detecting radionuclides that emit alpha particles in spaces that are not accessible to an alpha-particle detector; however, it is subject to error due to detecting ionization generated by other processes.

Values promulgated as decommissioning limits or guides will not only control the extent of decontamination but can significantly affect disposal decisions and characterization procedures. Characterization instruments must identify and measure radionuclides at the concentration limits specified by regulatory agencies.

Guides for decommissioning have been published by the Nuclear Regulatory Commission and the U.S. Department of Energy (DOE). Cooperative efforts to revise them on the basis of current radiation protection concepts are underway by these two agencies and the U.S. Environmental Protection Agency (EPA). The International Atomic Energy Agency (IAEA) is also developing guidance for releasing radioactive materials.

The Nuclear Regulatory Commission (Nuclear Regulatory Commission, 1974) published standards for acceptable levels of surface area contamination, as shown in Table E-4, for decommissioning nuclear reactors. This agency does not now have regulatory oversight for the GDPs, but will have this responsibility in 1996. The currently responsible agency is DOE; it has published the same limits in DOE Order 5400.5 (DOE, 1993, Figure IV-1), except for omitting the second line in Table E-4, covering transuranics. Exposure limits for members of the general public were 500 mrem/yr at that time.

The Nuclear Regulatory Commission recently issued the draft regulatory guide (Daily et al., 1994)—the first document to be issued in the current cooperative effort—associated with amending regulations in Title 10 of the Code of Federal Regulations, Part 20 (10 CFR 20), with the values given in Table E-5. The concentrations in soils and on surfaces that would achieve an annual dose of 15 mrem to exposed persons were derived on the basis of the listed scenarios. Compared with Table E-4, the surface concentration values, except for ²³⁹Pu, are lower than would be expected at the lower dose limit.

The EPA is preparing draft radiation site cleanup regulations for soil and plans to develop such regulations for residual structures, groundwater, waste, and recycled materials (EPA,

1994). The soil concentrations given in Table E-6 are estimated to lead to an annual dose equivalent of 15 mrem by three exposure scenarios. The residential pathway values are about threefold lower than those in Table E-5 for uranium and ⁹⁹Tc, similar for ²³⁷Np, and higher for ²³⁹Pu.

Both drafts establish the goal of decontaminating to radiation background but will accept contamination leading to an annual radiation dose of 15 mrem for unrestricted public access or higher radiation levels for restricted access. Both agencies use calculational models that estimate the radionuclide levels that would result in the specified dose rate to humans in selected pathway and exposure scenarios. Neither draft considers maximum and removable contamination versus average contamination as shown in Table E-4, or the extent to which individual measurements should be averaged for comparison with the limit. The agencies recommended that exposure scenarios be evaluated for the specific site.

Although regulatory agencies have not yet established the annual radiation dose limit, for planning purposes the ratios of radionuclide concentration per dose for the GDPs should be developed promptly with site-specific scenarios based on regulatory agency models. The process will be simple because radioactive contamination is due mostly to uranium. Note that uranium differs from most other radionuclides considered in decontamination guides because its concentrations in nature are not far below the limits; its retention by humans is controlled in part by its mass; and its chemical toxicity may be controlling in establishing intake limits.

Other important local factors in developing decommissioning guides for these plants are the contaminated areas contiguous to the GDP structures, the valuable metals on site, and the large structures. For example, the maintenance of access restrictions at a contiguous site may negate the rationale for decontaminating parts of the plant for free access; the radionuclide limit for freely releasing metals may be driven upward by their value or downward by concerns about subsequent radiation-sensitive applications; decisions on retaining structures versus turning them into rubble may depend on the potential value of the former and the disposal cost of the latter.

The concentrations proposed in an IAEA draft report (IAEA, 1993) for unconditional clearance—that is, for release for reuse or as waste—are related to a radiation dose equivalent to exposed persons of 1 mrem/yr. (See Table E-7). They are described as representative values that are generally within a factor of 100 of the reviewed published estimates. The uranium levels are similar to those in Table E-6, while the ⁹⁹Tc and ²³⁷Np values are higher; only the ²³⁹Pu values are lower in accord with expectations for the 15-fold lower dose criterion.

Material sent to land disposal facilities for radioactive waste must meet criteria in 10 CFR 61. These include, for Class A waste, concentration limits of 0.3 Ci/m³ for ⁹⁹Tc and 10 Ci/m³ for long-lived alpha-particle-emitting transuranic nuclides, and, for Class C waste, 10 times these limits. Class C waste has more rigorous requirements for waste-form stability and protection against inadvertent intrusion at the burial site.

Waste materials that include radioactively contaminated hazardous substances are categorized as mixed waste. They must be evaluated for disposal in terms of both EPA and Nuclear Regulatory Commission regulations. The complexity of this process suggests the desirability of radioactive decontamination before disposal of hazardous substances as waste.

The decontamination process also must meet regulations for protecting radiation workers and controlling release of radioactive effluent from nuclear facilities. The limits are based on dose equivalent limits of 5,000 mrem/yr to radiation workers and 50 mrem/yr each from airborne and liquid effluent to members of the public according to 10 CFR 834 (in draft) and 10 CFR 835, respectively. Limits are similar for the uranium isotopes, higher for ⁹⁹Tc, and lower for ²³⁷Np and ²³⁹Pu. Facility operators also must protect persons in compliance with the requirement that doses be "as low as reasonably achievable."

Characterization at regulatory limits is feasible only if concentrations of radionuclides attributed to the facility can be distinguished from the background. The detection limit is the net amount of radionuclide of interest that can be distinguished reliably from the selected background value, and depends on the variability in these background values. Natural uranium concentrations in rock, soil, and concrete typically range from near zero to several pCi/g. ²³⁹Pu from fallout in surface soil is approximately 0.01 pCi/g. If the limits are near these values, direct characterization may not be feasible. Only analysis by isotopic content, physical characteristics, or chemical behavior may distinguish between contaminant and background at this level.

Daily, M. C., A. Huffert, F. Cardile and J. C. Malaro. 1994. Regulatory Guide on Release Criteria for Decommissioning: Nuclear Regulatory Commission Staff's Draft for Comment. Washington, D.C.: U.S. Nuclear Regulatory Commission.

DOE (U.S. Department of Energy. 1993. U.S. DOE Order 5400.5, Radiation Protection of the Public and the Environment (chg 2). Washington, D.C.: DOE.

DOE. 1994. Decommissioning Handbook. DOEW/EM-0142P (Section 7.0). Washington, D.C.: DOE.

EPA (U.S. Environmental Protection Agency). 1994. Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil (Review Draft). Washington, D.C.: EPA.

IAEA (International Atomic Energy Agency). 1993. Recommended Clearance Levels for Radionuclides in Solid Materials (draft). Vienna, Austria: International Atomic Energy Agency.

Kocher, D.C. 1977. Nuclear Decay Data for Radionuclides Occurring in Routine Releases from Nuclear Fuel Cycle Facilities. ORNL/NUREG/TM-102. Oak Ridge, Tennessee: Oak Ridge National Laboratory.

Nuclear Regulatory Commission. 1974. Termination of Operating Licenses for Nuclear Reactors. Regulatory Guide 1.86. Washington, D.C.: U.S. Nuclear Regulatory Commission.

Reilly, D., N. Ensslin and H. Smith, Jr., eds. 1991. Passive Nondestructive Assay of Nuclear Materials. NUREG/CR-5550, LA-UR-90-732, pp. 339, 345, 413. Washington. D.C.: U.S. Nuclear Regulatory Commission.

Ritter, R. L., L. D. Trowbridge and S. E. Mainers. 1990. Neptunium Experience at PGDP [Paducah Gaseous Diffusion Plant] . K/ETO-30. Oak Ridge, Tennessee: Martin Marietta Energy Systems for DOE.

Smith, R.F. 1984. Historical Impact of Reactor Tails on the Paducah Cascade. Paducah Gaseous Diffusion Plant Report KY/L-1239.