Moreover, because their fissile content or isotopic composition determines their strategic value as well as their value for nuclear reactor applications, it is important to measure isotopic compositions as well as absolute amounts of these elements. For example, natural uranium contains 0.7% of the fissile isotope 235U, with the remainder being the fertile isotope 238U. Both fissile and fertile isotopes will fission when bombarded with neutrons, but fissile isotopes are the important ingredient in nuclear fuels because they fission with high probability when bombarded by lowenergy neutrons, whereas fertile isotopes will not fission unless they are bombarded by high-energy neutrons. Plutonium239 is the principal fissile isotope in all plutonium fuels. This isotope is produced in nuclear reactors through the reaction

Uranium-238 is called fertile because it produces the fissile isotope 239Pu by neutron capture. Other fertile isotopes are 240 Pu, which produces fissile 241 Pu by neutron capture, and 232Th, which produces 233 U. Thus safeguards measurements are concerned with determining the presence of a large number of fissile and fertile isotopes of uranium, plutonium, and thorium.

In this article we discuss nuclear NDA methods based on detecting either the natural radioactivity of these materials (passive assay) or the radiations produced when these materials are bombarded by external sources of neutrons and gamma rays (active assay). The methods are rapid, usually requiring only a few minutes to complete a measurement. They are nondestructive in the sense that the materials can be measured without removing them from their containers and that the measurement is based on observation of the

decay or transmutation of a negligible number of nuclei relative to one mole. Moreover, they interrogate the entire bulk of the material rather than just a sample and are therefore capable of accurately measuring heterogeneous materials such as scrap and wastes from processing.

One of the earliest, most critical needs for NDA methods was for the measurement of nuclear process scrap and waste. Chemical analysis was not reliable because of the problem and expense of obtaining representative samples from these characteristically heterogeneous materials. Some large inventory discrepancies (for example, the problem encountered at the NUMEC facility in Apollo, Pennsylvania, in 1965) have been difficult to resolve because of deficiencies in such measurements. Good measurements of scrap also were needed to assess the material's value for recycle, and NDA methods for the screening and assay of low-level waste were needed to dispose of the materials safely and economically.

NDA techniques were also required. to extend the ability of safeguards inspectors to check facility operator compliance with safeguards requirements-in particular, to check that measured material inventories matched declared "book" inventories within the limits of measurement uncertainties. To accomplish this objective, inspectors carry out detailed auditing of records and procedures; perform on-the-spot inventories of containers of nuclear materials, which involve weighing and verifying seals; collect random samples for chemical analysis; and examine analytical laboratory procedures. With portable and in-plant NDA instrumentation, together with certified standards, an inspector can independently verify the types and amounts of materials in a facility, including materials held up in process lines.

International Atomic Energy Agency

(IAEA) inspectors present special challenges for measurement instrumentation because they work under difficult political and physical constraints. The terms of the safeguards agreements between the country being inspected and the IAEA define the political constraints. For example, their terms limit materials accessibility and inspection time. Further, the inspector must work in a foreign environment without the normal supporting services, such as calibration standards, that are available for domestic safeguards. Equipment therefore must be lightweight, reliable, rugged, and easy to calibrate.

The LASL Safeguards Program has studied and continues to study these and other challenging measurement problems. In most cases, the problems have been identified by domestic and IAEA safeguards inspectors, Department of Energy (DOE) field offices, DOE facility operators, DOE safeguards systems analysts, and designers of integrated safeguards systems for new fuel cycle facilities, as well as the operators of LASL facilities, principally CMB Division staff. Because we want to gain widespread acceptance of our new technology, we consider test and evaluation of a fully engineered prototype instrument in the operating environment of a host nuclear facility to be the most important phase of the development of a measurement method. We also prepare comprehensive design and performance documentation to facilitate the work of instrumentation vendors and potential users.

In the remainder of this article we describe (a) techniques for nondestructive assay of fissionable material including some applications of major methods, (b) the status of NDA technology development and its implementation in the fuel cycle and DOE facilities, and (c) some future challenges in this field of measurement technology. Many laboratories over the world have

contributed significantly to the development of this technology, but rather than present a comprehensive review, we will focus on LASL-developed instrumentation and methods.

Characteristics of a Nondestructive Assay System

No single NDA instrument or method will suffice for the assay of the diverse forms and containments of nuclear materials in the fuel cycle. Some typical samples are shown in Fig. 1. Typical "feed" materials for input to the fuel fabrication process are UF, in metal cylinders with capacities for up to ~104 kg, depending on enrichment; plutonium nitrate solution in 10-L plastic bottles (120 cm long); and uniform fuel blends in few-liter cans. Reactor fuel materials include pellets, plates, rods, and bundles. Recoverable scrap, such as defective product material and calcined process residues, is stored in few-liter containers. Slightly contaminated wastes, such as paper, rags, and rubber gloves, are often stored in 120- and 220-L drums.

Clearly then, the capability to "view" bulk materials quantitatively is essential for most applications; hence, for both active and passive assay the interrogating beams and radiation. signatures of the nuclear materials being measured must be highly penetrating. Neutrons and gamma rays are obvious choices from the domain of low-energy nuclear physics, as opposed to less penetrating alpha and beta rays. Additionally, calorimetry, which is based on the measurement of heat generated by radioactivity, can be used to measure plutonium with high precision, provided the isotopic composition of the plutonium is known from independent mass spectrometry or gamma-ray spectrometry measurements. (Uranium does not generate enough heat to be measured with calorimetry.)

The characteristics that are common to all practical NDA systems are:

1. an assay principle based on the detection of suitably copious fundamental signatures of the isotopes or elements to be measured, that is, neutrons or

or from interrogation with an external

source;

2. a detection system optimized for the signature selected and the types of samples to be measured;

3. electronic and mechanical systems for data acquisition and assay controls; and

4. a consistent means for obtaining the desired mass of the isotope or element from observed counting data.

Detectors for NDA instruments are essentially solids or gases that ionize as they interact with incoming neutrons or gamma rays. The amount of ionization is proportional to the energy deposited in the detector by each interaction. The generated charge forms an electronic pulse that is then amplified and processed and either counted or analyzed by electronic and microprocessor units. Gamma-ray detectors used in NDA instruments are solid crystals of NaI and germanium, standard tools of the gamma-ray spectroscopist. In Nal detectors, the ionization in the crystal produces light (scin tillation), which is converted to an electronic pulse by a photomultiplier tube. In germanium and other solid-state detectors, the charge produced is converted directly into an electronic pulse. The neutron detectors most commonly used for NDA instruments are gas proportional counters, filled with 'He, BF3, or "He; however, plastic or liquid scintillators also can be used.

The conversion of counting-rate data. to mass determinations or isotopic abundances of SNM involves the most subtle and difficult problems. Among these are calibration of the instruments and corrections for absorption and scattering of the signature radiation by the sample being measured. The sample includes not only the particular isotope or element of interest, but also "matrix" materials in which the material of in

terest is embedded. In an active method, one also has to account for the absorption and scattering of the interrogating beam of photons or neutrons as it penetrates the sample. If the chemical and isotopic compositions of samples are well known, as with identical cans of PuO, powder, then the correction problem may be circumvented simply by calibrating the measurements with a set of standards that closely cover the range of unknowns. However, we prefer a method whose response is as independent of sample matrix materials as possible, to reduce the number of calibration standards. It is even more important to reduce matrix dependence in measurements of poorly characterized scrap and waste materials. We use the sophisticated radiation transport codes developed over the years for defense projects and reactor design to guide the design of an assay system.

General Methods for Passive Gamma-Ray Assay

Each isotope of SNM is radioactive and decays at a known rate through alpha or beta decay (for example, 235U 231Th + a + y's and 239 Pu 235U+ ay's). These spontaneous decays produce a characteristic spectrum of gamma rays, specific in both quantity and energy. Thus a direct measurement of the amount of SNM in a sample can be obtained by counting the gamma rays it emits at specific energies.

The pulse-height spectra of gamma rays from high-enriched (93% 235U) uranium and from reactor-grade (15% 240 Pu) plutonium are shown in Figs. 2 and 3, respectively. The isotopic abundances of the reactor-grade plutonium. are: 238 Pu 0.15%, 239Pu 81.6%, 240 Pu 15.2%, 241 Pu 2.4%, 242 Pu = 0.66%, and 241 Am 0.8%. These data were taken with a high-resolution germanium gamma-ray detector and a multichannel pulse-height analyzer. The

=

=

=

plutonium spectrum is very complex, with contributions from several hundred gamma rays from 238 Pu; 239Pu; 240Pu 241 Pu, and the daughters or products of 241 Pu decay; 237U; and 241 Am. (In Fig. 3, a few of the peaks are labelled with their energies and isotopic origins.) Furthermore, the gamma-ray spectrum of plutonium changes with time because of the relatively short half-lives of 241 Pu and 237U.

a

An assay is based on determining the areas under one or more peaks in the spectrum, usually calculated with a minicomputer that is an integral part of dataacquisition system. We use either Nal or germanium gamma-ray detectors, but we prefer germanium because its higher (30X) resolution permits cleaner separation and more accurate analysis of the gamma-ray spectrum peaks. On the other hand, for field measurements, such as the assessment of material holdup in process equipment, portable assay units. comprising a Nal detector and singlechannel electronic analyzers set to bracket specific peak and background regions of the gamma-ray spectrum are satisfactory.

multichannel-analyzer,

For measurement of uranium, we use the prominent 186-keV gamma ray from 235U; 4 X 10 gamma rays are emitted per gram 235U per second. Similarly we use the 414-keV gamma ray from 239Pu, which has a comparable intensity, to measure plutonium. To convert counting rates of these isotopes to total uranium or plutonium, their isotopic abundances. must be known or measured independently. In all applications except spent-fuel recovery, uranium isotopes other than 235U and 238U are present in such small amounts that they can be neglected. Uranium-238 can measured using as the signature the 1.001-MeV gamma ray from the decay of its daughter 234mPa. However, the measurement must be made at least 3 months after chemical purification of