M. A. Medrano, And N. Zarate
Gerencia De Servicios A Laguna Verde
Instituto Nacional De Investigaciones Nucleares
Mexico

R. Ramirez
Proteccion Radiologica
Central Laguna Verde
Comision Federal De Electricidad
Mexico

Mexican standards (NOM-004-NUCL-1994 and NOM-018-NUCL-1995) establish criteria and methods for the classification of wastes generated in Mexico. As no specialized laboratories are available in Mexico to perform radiochemical analysis which are required to obtain a complete inventory of the Laguna Verde Nuclear Power Plant wastes, 13 samples of selected wastes were send to be charaterized in a US laboratory. In addition there are almost 180 isotopical analysis for most wet wastes from unit 1 generated since 1989 to 1996. Due to constant and homogeneous inventories, in almost all the batches, Laguna Verde NPP has developed an analytical methodology based on the fractional abundance of Co-60, Identifying the next four ranges # 65 %, ( 65 % ; 75 % ), ( 75 % ; 80 % ) , and > 80 %, to which scaling factors were obtained.

Mexican regulations have recently incorporated standards establishing criteria and methods for the safe handling of wastes generated by the increase in industrial and medical activities. Those include standards that are based mostly on recommendations given by international organizations, as well as from USNRC regulations. One of the most important standards under the acronym of NOM (Mexican Official Standard) is NOM-004-NUCL-1994 Awaste classification@ (reference 1) based on the US 10CFR 61 part 55 (reference 2). Therefore the nuclide inventories and limits contained in tables 1 and 2 of that regulation, are the same as the Mexican standard for waste classification. Moreover methods and criteria contained in the USNRC Branch Technical Position (BTP) on radioactive waste classification (May 1983), (reference 3), are included in the NOM-018-NUCL-1995 "Methods for activity concentration and total activity assessment in radioactive waste packages@, (reference 4).

Laguna Verde Nuclear Power Station (LVNPP) is a two 675 Mwe BWR-5 units from General Electric with turbine-generators supplied by Mitsubishi. Construction started in 1976 in the Laguna Verde area, about 450 km. East of Mexico City on the coast of the gulf of Mexico. Unit 1 was connected to the grid since 1989 and started commercial service in July 1990, while unit 2 started commercial operation in April 1995. As of yet there are no waste disposal facilities in Mexico for final storage, LVNPP has decided to temporarily store waste generated as a result of plant operations. Also the activities related to waste classification, based on national standards as those mentioned above, have been performed in order to have all documentation required for final disposal available in a timely manner. In this context the waste classification program described in this paper plays an important role.

An acceptable method for waste classification (see reference 4), is based on individual determination of waste inventories in selected samples, by means of comprehensive radiochemical analysis, to establish proper scaling factors as a tool to classify routine waste batches. Those scaling factors are obtained as the ratio of nuclides difficult to measure (DTM), as H-3, C-14 or Tc-99, to key nuclides as Co-60, Cs-137 or Ce-144 easily detected using gamma spectrometry. Typically, DTM activation product radionuclides are related to the activity of Co-60 whereas DTM fission product radionuclides are related to the activity of Cs-137, and DTM transuranic elements are related to the activity of Ce-144.

Since routine data does not provide information on nuclide concentration and quantities of the wastes linked to the power plant Ns operational history, the scaling factors must be calculated repeatedly whenever the plant condition changes substantially. For this reason it is required to establish confirmatory analysis programs based on criteria related to the applicability of reference sample results to routine wastes.

Correct scaling factor determination methods requires costly and sophisticated radiochemical analysis in specialized laboratories; it is not feasible to install such kind of laboratories in countries like Mexico, whose needs of this type of analysis are very limited in size. To cope with this problem LVNPP is developing a classification program based on a minimum set of samples to be send to foreign reference laboratories with no loses in results confiability.

As no specialized laboratories are available in Mexico, 13 samples of selected wastes were sent to be characterized in a US laboratory (reference laboratory), being such samples representatives of the LVNPP main waste streams such as spent resins, cleanup sludges, waste sludges and concentrates. In addition there are almost 180 isotopic analysis for most wet wastes from unit 1 generated from 1989 to 1996. Even though wastes produced in unit 2 have shown similar composition as in unit 1, to date there is not enough information for the application of the same methodology and criteria. Results from the 13 waste samples analyzed in the reference laboratory included the suggestion of classifying them in 3 main routine waste streams (see reference 6): RWCU sludges, spent resins and waste sludges. In addition, scaling factors recommended for those waste streams were provided by the reference laboratory.

Based on the above, a detailed assessment was performed of those characterized wastes and the whole set of results of the isotopic analysis (only beta-gamma emitters), resulted in the following interesting results:

Due to differences in dates between waste generation and isotopic analysis in routine batches, and the existence of short lived radionuclides in waste inventories, and based on the fact that final disposal requires identification only of radionuclides whose half life are important in the long term waste behavior, it is appropriated to decay correct isotopic analysis results in order to minimize discrepancies in results analysis due to those non significant short lived radionuclides. For this reason decay corrections were evaluated in routine wastes, from 1 month through 4 years, concluding that a two years decay correction is appropriated to avoid results misunderstanding due to short lived radionuclides. Same decay correction in characterized samples implied no significant changes in inventories due to higher time periods between sampling and analysis in the reference laboratory.

It should be mentioned that if decay correction is applied in reference and routine analysis results there will be not analytical impact in the final results of the classification process. In the case of LVNPP wastes, this decay correction corresponds with the minimum on-site temporary storage before being shipped for final disposal.

Comparing the whole set of routine isotopic analysis, after 2 years decay, a constant and homogeneous nuclide inventory was identified, as shown in table I, which contains the two years decay corrected average values of the fractional abundance of the radionuclides identified in the four LVNNP routine waste streams. In that table Co-60, Mn-54, Ag-110m, Zn-65, Co-57 and Co-58 can be observed, having average fractional abundances ranged respectively from 74%, 24%, 0.7%, 0.6% , 0.4% down to 0.2%.

An important statement contained in reference 3 reads as follows: any radionuclide beta-gamma emitter with half life lower than 5 years, is considered significant for classification purposes, if it is contained in the waste in concentrations higher than 7 m Ci/cm³. Based on the maximum batch concentration obtained in the period 1989 -1996, in the most conservative case (93 m Ci/cm³ ), in terms of fractional abundance it is required that any radionuclide have at least 7 % of fractional. Based on this considerations, and the results shown in table I, only Co-60 and Mn-54 are significant radionuclides in the LVNPP unit 1 wastes.

In order to apply scaling factors to routine batches it is necessary to compare the fractional abundance of main radionuclides between routine and reference batches, to assure both are taken during the same plant condition. Analyzing the whole set of waste batches, based on the reference laboratory suggestions, it was observed that fractional abundance of Co-60 and Mn-54 in the four waste streams ranged widely. In spent resins it varied between 60-90 % and 5-35 % ; 70-90 % and 15-30 % in cleanup sludges; 70-90 % and 10-30 % in waste sludges; as well as 70-95 % and 10-30 % in concentrates, while average fractional abundances in the four streams ranged between 60-95 % and 5-35 % for Co-60 and Mn-54 respectively.

Statistical analysis of the four waste streams histograms shows that fractional abundance of Co-60 can be adjusted to normal distributions, whose mean values and variances are almost the same, based on the chi-square and F tests. For this reason it could be assumed that the four waste streams belong to the same population, and they can be taken as a single stream.

Examination of fractional abundance of Co-60 and Mn-54 across the time, shows a temporary variation, having that yearly Co-60 averages has an increase rate of 1 %, while Mn-54 decreases in the same rate, although both radionuclides are corrosion products, originated in valves and pipes internals, their concentrations in the reactor coolant primarily depend on the composition of structural materials, wear resistance, their corrosion rates and transport behavior, as well as filtering systems characteristics.

Total concentrations in the whole set of isotopic analysis ranged from 1E-05 through 1E+01 m Ci / g, with most of the batches falling in the range from 1E-04 through 1 E+00, ( with first and thirth quartiles varying between 2 or 3 orders of magnitude). Average concentrations in the four waste streams are as follows: 5.46 m Ci / g for precoat sludges, 1.57E-01 m Ci / g for spent resins, 8.56E-01 m Ci / g for waste sludges and 8.53E-02 m Ci / g for concentrates. This means that waste concentrations distribution are narrow and well shaped.

As an important result of the 13 characterized samples (see reference 6), it was identified that waste samples contain corrosion products (Mn-54, Co-58, Co-60, Fe-55, Fe-59, Ni-63, Zn-65), some activation products (H-3, C-14), and some amounts of Sr-89, Sr-90 and Tc-99. However no evidence exists of Cs-137 and transuranic elements. This is concordant with indications that no fuel leaks had occurred in unit 1.

In order to compare the homogeneity of the waste inventories among the 13 characterized samples, Euclidean distances of the scaling factors between each two samples were assessed, In this analysis we found distances ranged between 1 and 10, due to the obtained Fe-55 scaling factors, as this factor varies from 0.4 through 8.4, while the contribution of the other nuclides scaling factors to the Euclidean distance produce small variations.

Taking into account the above analysis, four waste types have been established, based on Co-60 and Mn-54 fractional abundance and on Fe-55 scaling factors obtained in the 13 characterized samples. Therefore the scaling factors obtained for wastes were those whose fractional abundance of Co-60 are lower than 65 %, are in the ranges of 65 - 75 % , 75 - 80 %, and those higher than 80 %.

As a secondary result of samples characterization it is possible to compare LVNPP isotopic results (beta-gamma emitters) and those obtained by the reference laboratory, in order to have an informal intercomparison with such kind of specialized laboratory. In this sense inventories and results were very similar: Specifically in both sets of analysis Cs-137 was not observed, meaning that no fuel leaks have occurred in the time period when samples were taken. Moreover, due to the absence of Cs-137, as shown in table I, it can be confirmed that LVNPP unit 1 has not had fuel leaks.

Scaling factors determination process is based on the election of the most accurate, defendable and reasonable values. The method used by LVNPP is a result of the combination of several methods and takes into account the 9 techniques or rules described in references 5 and 7.

Due to the lack of evidence of the existance of Cs-137 and Ce-144, scaling factors in the 13 characterized samples are based only on Co-60 concentrations, rather than Cs-137 and Ce-144 for fission products and TRU elements. This situation gives more confidence in the use of the method described in this paper, to obtain scaling factors based on the fractional abundance of Co-60 in routine batches.

Based on the analysis performed above, it was possible to implement the methodology described in this section to characterize routine wastes produced during the period 1989 - 1996 in unit 1:

• It is required for the waste batch to contain only corrosion products, as significant nuclides, as those shown in table I.
• Routine Isotopic analysis report is decay corrected for a period of 2 years.
• Fractional abundance of Co-60 and Mn-54 in the waste sample are calculated.
• Based on the fractional abundance of Co-60, routine waste are classified in one of the four intervals, # 65 %, ( 65 % ; 75 %] , ( 75 % ; 80 % ] , and > 80 %.
• Proper scaling factors are used to project the nuclides difficult to measure in the routine wastes.
• Due to the fact that in some cases there are batch mixtures in the same container, for instance, high integrity containers (HIC), it is required to obtain a corrected waste inventory of the mixture before classification.
• Waste batch is classified in accordance with nuclides and limits contained in the Mexican Official Standard (NOM-004-NUCL-1994.)

During the period 1989 -1996 LVNPP unit 1 has generated about 180 waste batches whose radionuclide inventories are constant and homogeneous, based mostly in Co-60 and Mn-54, due this fact it has been possible to develop an analytical methodology for waste classification, based on the fractional abundance of Co-60.

Since waste streams, spent resins, precoat sludges, waste sludges and concentrates have almost the same inventories and histograms in respect to the fractional abundance of Co-60, scaling factors have been obtained to classify routine waste batches in the next four intervals of fractional abundance of Co-60: lower than 65 %, between 65 - 75 % , 75 - 80 %, and higher than 80 %..

A simple method was established to classify almost 180 waste batches generated in LVNPP unit 1 in the period 1989-1996, based on the results of 13 samples characterized in a specialized laboratory. It could be possible to use this methodology for batches generated in unit 1 after 1996 and possibly to those generated in unit 2 if the inventories and criteria described in this paper are applied.

1. Clasificación de los Desechos radiactivos, Norma Oficial Mexicana, NOM-004-NUCL-1994, Nov. de 1995.
2. US NRC 10 CFR part 61. Licensing Requirements for Land Disposal of Radioactive Waste. Final rule, 47 FR 57473, Dec 1982.
3. USNRC low-level Waste Licensing Branch Technical Position on radioactive Waste Classification, May 1983.
4. Métodos para Determinar la Concentración de Actividad y Actividad Total en los Bultos de Desechos Radiactivos, Norma Oficial Mexicana, NOM-018-NUCL-1995, Agosto de 1996.
5. Assay of Long-Lived Radionuclides in Low-Level Wastes from Power Reactors, J.E. Cline, et al., NUREG/CR-4101, April 1985.
6. Central Laguna Verde 10CFR Part 61 Program Scaling Factor Determination, REPORT # 1, 011R-93-007, Scientech, Ago. 1993
7. Scaling Factor Selection Method, A Program to Select Scaling factors for Waste Characterization, rev. 5, J.E. Cline, SCIENTECH Inc., February 1990.