Pietro Di Giuseppe, Vincenzo Zaccari
ENEL - Nuclear Plant Department - Rome, Italy

Jene N. Vance
Vance & Associates - Ruidoso - NM, USA

In 1986-87, the LWR Italian nuclear plants, Trino, Garigliano and Caorso were permanently shutdown by public referendum. Prior to their closure, the low level waste that was generated in the operation of the plants was placed into storage at the individual plant sites because of the absence of a low level waste disposal facility in Italy. Currently, Italy still does not have a low level waste disposal facility in operation. However, the regulations for the characterization of low level waste, which in many respects are similar to the U.S. 10CFR61 regulations, have since been promulgated. These regulations, like 10CFR61, require the classification of waste according to the concentrations of radionuclides, several of which are difficult-to-measure (DTM) using standard gamma spectroscopy techniques. The problem this posed for ENEL was that the wastes were already generated and packaged, making it very difficult to institute a sampling program to develop scaling factors for DTM radionuclides. In 1996, ENEL contracted with Vance & Associates for the use of two computer codes, 3R-STAT and SF-STAT, to be used in the characterization effort of the Italian low level wastes. 3R-STAT was used to determine the release rates of I-129 and Tc-99 from reactor fuel into low level waste. SF-STAT was used to develop scaling factors for Sr-90 and the transuranic radionuclides. The operation of the two computer codes requires the input of the concentrations of the gamma radionuclides of the five short-lived iodines (I-131, I-132, I-133, I-134 and I-135) and the two cesium isotopes of Cs-134 and Cs-137 that had been routinely measured in reactor coolant. The data were input and the codes were used to develop release rates and scaling factors for the individual plants. This paper presents additional details on the characterization effort and the results of the computer analyses for the Italian plants.

In Italy four Nuclear Power Plants were in operation: Trino (PWR, 270 MW), Garigliano (BWR, 160 MW), Latina (Magnox, 160 MW) and Caorso (BWR, 860 MW). The commercial activity started in the period '64-65 for the first three and in '81 for Caorso and were operated by ENEL (the Italian Utility for Electric Power ).

In 1987 there was a public referendum on Nuclear Energy acceptance in the future National Energy Plan and had negative results. As consequence all the NPPs in refueling or in maintenance outage during 1986-87, were permanently shutdown and subsequently put in preliminary decommissioning stage (present situation).

Up to 1985 no specific regulations existed in Italy for Radioactive Wastes Management, therefore, prior to this date, the relevant regulations varied from one plants to another according to the operation requirements and to the rules enforced by Licensing Authority on a local basis. Therefore the wastes generated in the NPPs were processed in different ways for packaging and for ultimate Low Level Waste disposal.

This issue was dealt with in the Technical Guide n. 26 (TG 26) of ENEA/DISP (the former Italian Regulatory Authority, now ANPA) setting forth the criteria and the requirements needed for proper waste management.

The essential principles on which TG n. 26 are based can be summarized as follows:

• Radiological and environmental protection
• Minimization of waste generation
• Classification criteria of the radioactive wastes.

About the classification criteria, TG 26 is similar to 10CFR 61 because radioactive wastes are classified into three categories, according to their radioisotopic inventory and corresponding concentrations.

The first category concerns the wastes from non-nuclear activities (such as from biomedicine) and containing radionuclides whose half-life is lower than one year.

The second category concerns those wastes that are primarily generated from Nuclear Power Plants. They are characterized by initial radioisotopic concentrations requiring a period of a few decades to a few centuries to decay to a final concentration of a few hundreds of Bq/g.

The third category includes high-level wastes from fuel reprocessing and from other research or industrial activities and/or containing also alpha and neutron emitters. For our purpose we are chiefly interested in the aspects pertaining to the second category wastes. They are characterized by radioisotopic concentrations not higher than those shown in Table I and they must be conditioned in a solid matrix with appropriate waste form characteristics. Dry Active Waste (DAW) consisting of slightly radioactive or contaminated materials, whose radioisotopic concentrations are not higher than those indicated in Table II, do not need conditioning. In case of radionuclide mixture the above values are met if the sum of the fractions, obtained dividing each nuclide concentration by related limit, is not greater than 1.

Conditioned wastes shall possess such physical, chemical, and radiological characteristics as to make them suitable for land disposal. The minimum requirements for conditioned wastes and the relevant tests are listed in table III.

Reference to national or international standards is in some cases made for specific requirements or testing methodologies (UNI, ASTM, ANSI, etc.).

The required tests shall be performed within a Quality Assurance Program.

ENEL stored radioactive waste at plant site only owing to the absence of LLW disposal facility. In this context several problems, due to the waste volume generated, were

encountered and various campaigns of volume reduction have been carried out [1] especially for Caorso plant (Incineration, Supercompaction, Wet-Oxidation).

A general plan has been performed for the activities related to the first stage of Plants decommissioning and in particular for those aimed at waste radiological characterization. The present situation of the radioactive wastes generated by ENEL plants results both from wastes processed or stored as generated in the first period (prior 1985) and from wastes processed in respect to TG 26 requirements. An outlook is shown in Table IV for LWR, as in this work the Magnox reactor is out of the scope.

The major problem facing ENEL in the waste characterization effort is that the radionuclides that are important in waste disposal are not readily measured by typical gamma spectroscopy techniques. They must be measured by complex and costly radiochemistry techniques which require that the waste be sampled. Virtually all of the low level waste had been processed and packaged for storage at the plant site making sampling and measurements a very expensive proposition.

As an alternative, ENEL contracted with Vance & Associates for the use of two computer codes, 3R-STAT and SF-STAT, to be used in the waste characterization effort. These codes and the results of the characterization are described in the next sections.

As highlight above different practices were used in the Italian NPPs for LLW processing and storage, nevertheless the radiological characterization of the waste packages was performed in similar mode.

In the common procedure the first step is the sampling of a representative part of the waste to be characterized and subsequently to perform the gamma spectrometry in the considered sample.

The sampling process is the result of various tests and experience gained during the operation to make it significant in respect to the batch in progress.

As well known such procedure presents problems in the estimation of the homogeneity and in the evaluation of the self-shielding effects in the packages; therefore several overestimation of the activity inventory occurred.

In the case of solidified or supercompacted wastes obviously the sampling is difficult to take therefore other measuring methods were developed [2],[3].

One of said methods has been developed by FIAT/SE.P.IN (see Fig. 1) for characterization of packages produced in Trino and Caorso plants and includes the following components:

• Drum elevator with rotating plate at constant speed for various size drums (maximum weight: 600 kg).
• Hyperpure Germanium detector in horizontal configuration with the following characteristics:

• Detector shielding with 100 mm lead to reduce the radiation background
• Lead collimator to allow a step by step detection every 100 mm on the drum surface.

The detector is interfaced with a multichannel system provided by a computer and managed by adequate software. The spectrometry is executed by sweeping the gammas on the drum surface at ten levels during the rotation. Algorithms has been implemented for the handling of the devices and to evaluate, in particular, the self-absorption of the medium and the statistical elaboration of the casual errors.

The minimum detectable activities for the more common radionuclides are shown in the Table V and the relevant values represents the average calculated across the ten levels of detection.

The system calibration has been carried out on the basis of the following parameters:

• Drum dimension (60, 220, 320 liters)
• Certified liquid sources (Ba133, Cs137, Co60, Y88)
• Four source geometry (static point-source, rotating point-source, rotating expanded source, homogeneously distributed source)
• Five medium matrix (air, water, wood, metal scraps with two apparent density).

The V&A computer code, 3R-STAT, was used to determine the quantities of I-129 and Tc-99 contained in the low level waste generated at the three Italian plants since the startup of the reactors. The 3R-STAT code uses as input to the code, the concentrations (mCi/g) of the following radionuclides routinely measured in reactor coolant:

Reactor coolant may be sampled as frequently as daily or every other day or as infrequently as once per week. Based on these inputs, the code calculates the reactor core conditions to determine the fractions of I-131 released from:

1. Tramp fuel

and

2. Defective fuel rods

and the fractions of I-131 released by the following mechanisms:

1. Recoil
2. Diffusion
3. Knockout

In addition the codes also calculates the:

1. Burnup of the defective rods

and

2. Escape rate coefficient for the defect size in the cladding

Once the reactor core conditions are determined, the code calculates the release rates of I-129 and Tc-99 in units of curies per second or curies per megawatt-day.

For the three Italian plants the reactor coolant gamma spectroscopy data were retrieved from plant files and archives for the following fuel cycles:

At the time of the project, the Trino plant was still retrieving data from storage and so only the fuel cycles shown above were included within the scope of this project. Once the data had been retrieved it was input to the 3R-STAT computer code. The data were analyzed on a fuel cycle basis where the data set used in each analysis spanned the entire fuel cycle.

A summary report was generated by the code for each fuel cycle showing the average release rates of I-129 and Tc-99 for the selected fuel cycle. The Table VI summarizes the results of the 3R-STAT analyses for the three Italian plants.

The V&A computer code, SF-STAT, was used to derive scaling factors for Sr-90 and Pu-239 keyed to the gamma-emitting radionuclides of Co-60 and CS-137. The scaling factors for these two radionuclides are derived for specific waste streams as designated by the User in the setup files for the code. The code uses the same input of the short-lived iodines isotopes, the two cesium isotopes and Co-60 as described above for the 3R-STAT computer code. Based on the input, SF-STAT determines the release rate of I-131 from tramp fuel by a knockout mechanism. Strontium-90 and Pu-239 are both produced only from tramp fuel by a knockout mechanism. Therefore, once the release rate of I-131 from this source is determined then the release rates of Sr-90 and Pu-239 are determined based on their inventories in the tramp fuel as compared to the inventory of I-131.

The release rates of Cs-137 and Co-60 are determined from their measured concentrations in reactor coolant. Factors are applied to account for the deposition of plutonium on in-core surfaces. Scaling factors for purification resins and filters are thus determined by their respective release rates and the applied factors. Scaling factors for other waste streams are determined based on the relative behavior of the radionuclides as they migrate from the primary system to other plant systems.

The relative behavior is based on empirical factors derived from several hundred samples from virtually all of the U.S. reactors. Scaling factors for the three Italian reactors were derived for each fuel cycle as described above. A set of scaling factors for the waste streams identified for each plant was determined for each fuel cycle for a total of 12 sets of scaling factors.

Table VII shows a typical set of scaling factors determined by SF-STAT for the Caorso plant

The remaining transuranic radionuclides of Pu-241, Cm-242, Am-241, Pu-238 and Cm-244 were scaled to Pu-239 based on nearly 2300 samples from the U.S. industry database. These scaling factors are relatively constant for all reactors for all waste streams.

The scaling factors for the activation products of C-14 and Ni-63 were derived on a waste stream basis for each plant type, BWR and PWR.

These scaling factors were based on approximately 700 BWR samples and 900 PWR samples from the U.S. industry database.

These scaling factors are relatively constant in the plant waste streams because of the similarity in their production and in their chemical behavior.

Therefore, the scaling factors derived for a plant type are assumed to be valid for the Italian reactors.

The concentration of the gamma radionuclides in packaged low level waste will be determined by direct measurement. The contents of the of radionuclides I-129, Tc-99, Sr-90 and the transuranics in low level waste in storage at the Italian reactors will be determined based on the analyses performed by the 3R-STAT and SF-STAT computer codes. Empirical relationships for the activation radionuclides of C-14 and Ni-63 were derived from the U.S. database. Thus the content of the difficult-to-measure radionuclides was determined without having to open and sample any of the low level waste already in storage at the Italian reactors.

1. A. AGOSTINELLI, C. D'ANNA, P. DI GIUSEPPE, "Experience achieved on volume reduction of radioactive waste and on final product qualification", Joint International Waste Management Conference, Oct. 23-28, 1989, Kyoto (Japan).
2. REHER ET AL., "A gamma ray scanning device for the metrology of heavy radioactive waste containers", Nuclear Instruments & Methods in Physics Research, A312 (1992), 273 ¸ 277.
3. FIAT/SEPIN, "Sistema di caratterizzazione radiologica di fusti contenenti rifiuti radioattivi mediante spettrometria gamma", Internal Report, Feb. 23,1993.

BACK