Low Level Wastes generated and received in the Nuclear Technology Development Centre - Centro de Desenvolvimento da Tecnologia Nuclear / CDTN, Belo Horizonte / MG - Brazil, have been solidified through the cementation process. A 200 litre (L) cementation batch unit was designed and built in 1986 by CDTN research workers. In 1987 inactive tests were made and in 1989 began the operation with radioactive wastes. The performance of the system was not good, so it was decided to make some modifications in order to improve the operation and consequently the waste form obtained. The changes were planned in 1993, consisting mainly on the installation of new equipment in the 200 L system and design and construction of a 20 L prototype cementation system. Prototype allowed savings of money and carrying out a large number of tests for studying important operational parameters, such as mixing paddle and time necessary for mixing. This paper describes in details the steps of the optimization.

The Nuclear Technology Development Centre - Centro de Desenvolvimento da Tecnologia Nuclear / CDTN is one of nuclear research centres of the National Nuclear Energy Commission - CNEN. CDTN was created in 1952, is located in Belo Horizonte / MG - Brazil, and is engaged in research and development of activities related to nuclear fuel cycle, radioisotope production, materials science, environmental control, radioprotection and waste management. At CDTN Waste Management Program, the cementation process was established to solidify radioactive wastes generated at the Centre and received from other Institutions.

The cementation process for nuclear waste immobilization, with and without additives, has been commonly used on an industrial scale for several years in different countries. The main reasons for using cement are [1] :

• the relative simplicity of handling;
• extensive experience in civil engineering;
• the availability of raw material;
• the relatively low cost;
• the high density (shielding effect) and the mechanical strength of cement products;
• the compatibility of water with the matrix material.

Basically, the process consists of mixing the cement with the waste, be it as a solution, slurry or solids. After homogenization the mixture is allowed to set. The mixing of the cement with the various radioactive wastes, i.e. sludge, resin beads, etc., affects the properties of the waste form. In all cases the waste form under consideration consists of two or more phases, and uniformity of dispersion of either the waste or individual radionuclides in a matrix may affect the mechanical or physicochemical characteristics of the waste form.

At CDTN only low-level radioactive wastes are generated [2]. Portland cement and bentonite, a special clay as additive, have been used for immobilization of these wastes. In a first step, the optimum materials, for example type of cement, additives and working conditions are determined in laboratory tests with real waste samples. The properties of the cemented waste forms such as the mechanical strength, bleeding-separation of free standing water, viscosity, the setting time and leach behavior are examined. The best conditions obtained are used for the treatment in the 200 litre system [3].

The 200 L cementation batch unit was completely designed and built at CDTN, using Brazilian equipment and materials. Operation began in 1987 with inactive tests and in 1989 began the operation with radioactive wastes. Problems in the operation, such as impossibility to do an adequate cleaning of the plant and an unsuitable mixing paddle, required changes in the project.

In 1993 a technical visit was made at Nuclear Research Centre of Karsruhe (Kernforschungszentrum Karsruhe - KfK) in Germany [4], sponsored by International Atomic Energy Agency. The changes were discussed with the German specialists and it was decided to buy new equipment to improve the plant performance.

It was also decided to construct a 20 L cementation prototype system, similar to the 200 L, in order to save money and to carry out a large number of tests. In this prototype the flow rate of cement/additive and waste can be varied according to the test to be performed. The power requirement and dimensions of the components were decreased proportionally to the 200 L cementation plant. The first tests were carried out to study the homogenization of waste form. Two paddles were constructed - one of them was similar to the existing paddle in the 200 L unit, and the other paddle was a model recommended by the literature [5], an helical one.

A 2^k factorial design [6,7,8,9] was conducted to analyze the performance of the 20 L cementation prototype system. Initially four variables were studied: time of mixing, type of paddle, water/cement ratio and bentonite/cement ratio. The experiments showed that the helical paddle had best performance and it is being scaled up to be installed in the 200 L unit. The next step will be to determine how many minutes are necessary to achieve a good homogenization.

The various cementation processes can be classified according to how the mixing of waste and cement is achieved. At CDTN in-line mixing process has been used. In the process, cement/additive and waste are separately metered into one end of the mixing tank. The cement/additive is fed with a screw feeder while the waste is fed by gravity. An agitator, specified for solid/liquid mixing [10], keeps waste homogenization while it is being introduced into the mixing tank. After homogenization, the mixing is released into a metallic drum through the bottom of the tank.

The CDTN 200 L cementation unit occupies an area of 10 m² and is compound mainly by a waste tank, a cement/additive silo, a mixing tank, a control operation panel and accessories. As mentioned in introduction, the operation presented problems that conducted to implementation of modifications in the design of the unit. The main ones will be described below.

The 200 L should be able to reproduce laboratory tests, in order to obtain waste forms suitable for disposal. To assure accuracy on dosage of cement/additive and waste, an automatic dosage system was installed in the mixing tank.. This system doses the components at the same time they are being introduced into the tank and when the amount reaches pre-set values, then feed is stopped.

Changes were made in the tank to allow the installation of dosage system, to improve the level of decontamination of internal surface through covering by a plastic liner and to decrease secondary waste volume through cleaning with water under high pressure.

This valve made it possible automatic control for feeding of waste in the mixing tank. When the amount of waste reaches the pre-set value, the dosage system commands automatically shutting of the valve. This valve acts against accidental operation errors since the release of waste to mixing tank is made only if the valve is switched on.

To survey indirectly the viscosity of cement/additive while waste mixing is being done, an ammeter was installed in the mixer. It made possible to follow the mixing process in order to interfere immediately in case of improper functioning of the mixer.

This device improves the quality of waste form produced through programmable inversion of mixer rotation. It is possible to choose how many minutes the mixer will rotate for each direction.

The flowsheet presented below is the CDTN cementation process after modifications.

Prototype system is a copy of 200 L cementation unit. The capacity is 20 L and it was designed with the following objectives:

• to make possible to improve operation parameters of 200 L cementation unit, saving money and carrying out a large number of tests;
• to present suitable versatility in order to vary, in so far as necessary, operation parameters;
• to treat small volumes of wastes with higher level of radiation or wastes that be incompatible with others.

Prototype components are the same of 200 L cementation unit with exception of the dosage system and presents the following characteristics:

• the flow of the cement/additive and the waste can be varied according to the test to be performed;
• power of motors and reducers was decreased proportionally;
• changes of paddle were easily made. This allows carrying out randomized experiments.

In Figure 2 the prototype is shown.

Experience showed that waste forms obtained by the operation of the 200 L cementation system are not completely homogeneous. So it was decided to design and compare two paddles - one of them similar to the existing in the 200 L unit and another a model as indicate in literature (an helical one). In the Figure 3 the paddles are shown.

It was decided to apply statistical design for the tests. Four variables were selected to be studied in a 2^k factorial design:

• type of paddle;
• mixing time;
• water / cement ratio;
• bentonite / cement ratio.

Initially it was necessary to carry out 18 tests to adjust some parameters of the prototype, such as speed of the screw conveyor system and suitable level of variables.

After these adjustments the variable levels considered in the study are shown in Table I.

The sixteen tests were performed as following:

• test of air tight sealing;
• adjust the time for automatic reversion of the mixer rotation;
• feeding of water into the mixing tank;
• feeding of cement and turn on the mixer;
• feeding of bentonite. Addition of water, cement and bentonite was made separately to maximize the effect of the paddle.;
• after the time defined for mixing, released it into a 20 L drum. The 20 L drum was constructed maintained the same proportion height / diameter of the 200 litre standard drum to ease posterior comparison;
• in the following day, the 20 L drum was dismounted and the specimen was let to dry cure;
• after seven and twenty eight days of curing time, fifty measurements of the Rebound Number R were taken in various parts of the specimen. This number was used to evaluate superficial hardness of the sample [11]. Through this variable, the compressive strength of hardened pieces or structures of concrete and cemented pastes can be evaluated.

The technical principle for our test is that the more homogeneous is the cemented specimen, the less is variance of the measurements. The variance was calculated for the measurements of each level and each sector considered in the Figure 4. The results of the experiment are shown in Table II.

The shown results were obtained based on the following statistical model:

y_ijkl = m + t _i + b _j + g _k + q _l + (t b )_ij + (t g )_ik + (t q )_il + (b g )_jk + (b q )_jl + (g q )_kl

+ (t b g )_ijk + (t b q )_ijl + (b g q _)jkl + (t b g q )_ijkl + x _jkl , where :

• , b , g , q are effects of type of paddle, water/cement ratio, bentonite/cement ratio and mixing time;
• (t b ), (t g ), (t q ), (b g ), (b q ), (g q ) are effects arising of interaction of two variables;
• (t b g ), (t b q ) and (b g q ) are effects arising of interaction of three variables;
• (t b g q ) are effects of interaction of four variables;
• term x represents the experimental error.

The conditions of the model are that experimental error presents normal distributions, with mean zero, variance constant and they are independents (x _jkl ~ N/ 0, s ²) .

The conclusions are summarized as follows:

1. Equipment installed in 200 L cementation unit improved a lot operation efficiency and consequently the waste forms obtained.
2. The tests carried out have shown that the utilization of the helical paddle results in the production of a more homogeneous waste form, in all experimental conditions, compared with the similar 200 L paddle.
3. The 34 tests carried out in prototype presented a estimated cost of US$3,500. If they were carried out in 200 L cementation unit the cost would be about US$11,000.

In next step of our work, we will study the effect of other variables on process performance. Since the bentonite/cement is usually kept on the 10%, it will be now important to develop a model to characterize the influence of mixing time and water/cement ratio on the homogeneity of the waste form.

The authors thank Mrs. Eduardo Saraiva, Antônio Temóteo, Fernando Pugliese and all persons who cooperated with this work.

1. International Atomic Energy Agency. "Conditioning of low and intermediate-level radioactive wastes". IAEA Technical Reports Series N, 222. Vienna, Austria, 1983.
2. CENTRO DE DESENVOLVIMENTO DA TECNOLOGIA NUCLEAR. "Radioactive Waste Management Program" - PGRR / CDTN Rev.1 - 5 Section. Belo Horizonte, MG, Brasil, 1993.
3. C.C.O. TELLO "Preparation of Cemented Mortars and Specimens Containing Bentonite and Simulated Waste". Procedure TQ-028. Nuclear Technology Development Center, Belo Horizonte, MG, Brazil, 1989.
4. L.C.A. REIS, "Technical Visit at Karlsruhe Nuclear Research Center (KfK) in Germany". Mission Report CT3 - RM - 01/93. Belo Horizonte, MG, Brasil, 1993.
5. H. PERRY, C.H. CHILTON, "Chemical Engineers` Handbook - Multiphase Contacting and Separations". ISBN 0-07-049478-9. Japan, 1973.
6. D.C. MONTGOMERY, "Design and Analysis of Experiments". New York, United States of America, John Wiley 4^th edition,1996.
7. S. AGUIAR; M. C. C., WERKEMA, "Variance Analyzes: Comparison of Various Situations"- Tools of Quality, Vol. VI, Fundação Christiano Ottoni, Belo Horizonte, MG, Brasil, 1996.
8. S. AGUIAR; M. C. C., WERKEMA, "Design and Analysis of Experiments: How to Identify the Main Influents Variables in a Process"- Tools of Quality, Vol. VII, Fundação Christiano Ottoni, Belo Horizonte, MG, Brasil, 1996.
9. S. AGUIAR; M. C. C., WERKEMA, "Process Optimization and Statistic: How to determine the Operation Condition of a Process in Order to Reach a Improvement Goal"- Tools of Quality, Vol. IX, Fundação Christiano Ottoni, Belo Horizonte, MG, Brasil, 1996.
10. L.R. TERRON, "Fundamentals Concepts on Liquids Agitation". Brazilian Magazine of Chemical Engineer. v.8, n. 4, p. 5-29, 1986.
11. Technical Standard Brazilian Association. "Hardened Concrete - Superficial Hardness Evaluation through the Rebound Number R". ABNT- NBR 7584. Brazil, 1982.

BACK