Available online at www.sciencedirect.com

Procedia 7 ( 2012 ) 178 – 183

ATALANTE 2012 International Conference on for Sustainable Fuel Cycles Implementation of separation from a PUREX raffinate

Marie-Jordane Bollesteros, Jean-Noël Calor, Sylvain Costenoblea*, Marc Montuir, Vincent Pacary, Christian Sorel, Fabien Burdet, Denis Espinoux, Xavier Hérès, Catherine Eysseric

CEA, Nuclear Energy Division, Radiochemistry and Process Department, Marcoule Research Centre, BP17171, F-30207 Bagnols sur Cèze

Abstract

Recovering of minor from spent is investigated for heterogeneous recycling in Generation-IV reactors. After , americium is the main contributor to residual heat of long term which determines waste ©density 2012 Elsevier within B.V...Selectiongeological repository. and/or peer-reviewSelective americium under responsibility separation of (EXAm the Chairman process) of theby ATALANTEliquid-liquid 2012extraction from a ProgramPUREX Committee raffinate was studied. Two experiments were performed in ATALANTE facilities. The first test, on surrogate solution, validated the americium extraction performance. The second trial was carried out in the high-level shielded process line from a genuine PUREX raffinate. Tools used to manage the process are introduced to show performance process achievement.

© 2012 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of the Chairman of the ATALANTE 2012 Program

Keywords : ATALANTE ; americium ; liquid-liquid extraction ; EXAm ; PUREX raffinante

1. Introduction

In the back-end cycle, the PUREX (Plutonium EXtraction) process is industrially implemented in the AREVA NC plant at La Hague for spent fuel reprocessing and recycling. This process allows mainly plutonium recovery thanks to different liquid-liquid extraction steps and MOX fuel elaboration. A

* Corresponding author. Tel.: +33 4 66 79 13 81; fax: +33 4 66 79 16 48. E-mail address: [email protected]

1876-6196 © 2012 Elsevier B.V...Selection and/or peer-review under responsibility of the Chairman of the ATALANTE 2012 Program Committee doi: 10.1016/j.proche.2012.10.030 Marie-Jordane Bollesteros et al. / Procedia Chemistry 7 ( 2012 ) 178 – 183 179

PUREX raffinate is then obtained which contains americium, and others fission products. Its treatment is currently based on vitrification process and repository. As part of the French Act of June 2006 on sustainable radioactive waste and waste management, the recovering of minor actinides from spent nuclear fuel is investigated for heterogeneous recycling in Generation- IV reactors. Within minor actinides, americium is, after plutonium, the main contributor to residual heat of long term radioactive waste. The americium recycling in future nuclear reactors could decrease toxicity and residual heat which determines the waste density within geological repository. For nuclear fuel manufacturing, it would be better, for the same reason, to separate americium from curium in spite of some implementation difficulties. A liquid-liquid extraction flowsheet called EXAm was then developed to recover quantitatively and selectively americium from a PUREX raffinate. In the purpose of demonstration of this americium separation step, two experiments were carried out. The aim was to demonstrate the selective americium separation (americium recovery rate over 99% and decontamination factor of Am versus Cm higher than 500) from a PUREX raffinate following two stages: x Americium extraction, at high acidity, with some , x Selective desextraction of this minor by complexation thanks to polyaminocarboxylic acid at low acidity in the aqueous phase. This paper describes the implementation of these two experiments in the ATALANTE facility, one in glove boxes with a surrogate feed solution and the other in the high-level shielded process line CBP using a genuine PUREX raffinate.

2. Experimental methods

2.1. Process devices

The flowsheet consisted of PMMA mixer-settler extractor batteries where the aqueous and organic phases flow countercurrently as represented on Fig. 1.

Aqueous Aqueous phase phase

Organic Organic phase phase STEP Mixer Organic phase flow Settler Aqueous phase flow (a) (b) Aqueous phase tank

Fig. 1. (a) Pictures and (b) principles of PMMA mixer-settler extractor batteries

Stainless steel perforated paddle impellers were used for mixing with rotation speeds set between 2000 to 2200 revolutions/min. Interphases in settler compartments were adjusted thanks to Teflon slides located on aqueous phase weirs. Other weirs were settled at each battery outflow to regulate flow rate variations and to allow on-line spectrophotometric measurements thanks to optic probes. All reagents were introduced by rotary piston pumps located outside the confinement barrier. Flow rate measurements were performed thanks to Coriolis mass flowmeter (Quantim®) and controlled by a specific supervision application. Aqueous and organic flows between mixer-settler batteries were assured thanks to other rotary piston pumps. For the hot test, the high-level shielded process line CBP was used. This facility is shielded by 1 meter of high density concrete with 1 meter leaded glass windows. Experiments were carried out thanks to remote manipulators.

180 Marie-Jordane Bollesteros et al. / Procedia Chemistry 7 ( 2012 ) 178 – 183

2.2. Analytical measurements

On-line and in situ UV-visible spectrophotometric measurements were implemented to follow americium and neodymium concentrations. The light source was constituted of several light emitting diodes which reconstitute a homogeneous beam on a 350-850 nm wavelength range. A multichannel absorption UV-visible spectrophotometer (Acton 500i from Roper Scientific) with a 500 nm focal length was used. It was equipped with a diffraction grating and a CCD sensor with 1340 x 1300 pixels (Fig. 2a). Software has been developed from a Labview® application and allows spectrometer control, real time spectra treatment and data processing to gain access to the actinides and lanthanides concentration evolution during the trial. The spectra mathematical treatment was based on a factor analysis method called PLS (Partial Least Squares) which determines concentrations from a single spectrum thanks to calibration curves for each studied elements. Spectrophotometric probes were designed in order to be used in mixer-settler extractor. These probes (Fig. 2b) are 5 mm diameter with a 1.6 mm screw drilled for two optical fibres and a mirror. The light is carried via a first incident fibre from the light source to the probe. Then the light goes through the studied media for the first time until the mirror probe which sends back the attenuated light through the studied media for the second time.

VisualisationSpectra visualisation des spectres for BundleOptic fibre de bundle fibres connected optiques de chaqueeach sonde probe connectéesto sensor aux capteurs Ø5 mm 1 1 1 probe 2 ArrangementOptic fibres deson thefibres 2 Worm 3 sur la fenteslot of d’entrée the du 2 3 4 spectromètrespectrophotometer 5 4 Lock-nut 3 5 4 Worm 5

R Lock-nut Body Mirror CCD Lock-screw (a) (b)

Fig. 2. Diagram of the UV-visible spectrophotometer (a) and pictures of probes used for on-line spectrophotometric measurements (b)

Moreover, samplings of solutions in mixer-settlers were performed during the test in order to support this on- line technique and to gain access to complementary information by different analytical techniques carried out in the shielded analysis line.

2.3. Flowsheet of the EXAm process

The flowsheet of the complete EXAm process is illustrated in Fig. 3. In the case of the first experiment in glove boxes, only the extraction, scrubbing and Ln stripping functions were used so as to confirm the Am/Cm selectivity in the extraction/scrubbing operation. This flowsheet was elaborated thanks to laboratory experiments and data which were implemented in the PAREX code [1,2]. This code is a tool to model systems of actinide extractions such as distribution of species, chemical reactions in homogeneous phase, kinetics of interphase mass transfer, hydrodynamics in the extraction apparatus and heat transfer.

Marie-Jordane Bollesteros et al. / Procedia Chemistry 7 ( 2012 ) 178 – 183 181

The main operations are: x Extraction at high nitric acidity. Americium and lighter lanthanides extraction is based on Am/Cm selectivity of DMDOHEMA reinforced by a complexing agent (TEDGA, Pharmasynthese, purity over 99%) in aqueous phase, x Selective Mo and Ru back-extraction thanks to citric acid in buffer media at pH 3, x Americium back-extraction with HEDTA and citric acid at the same pH value, x Lanthanides and back-extraction by the use of a mixture of TEDGA and oxalic acid in nitric acid media. x AX : Solvent AF : PUREX raffinate AS (DMDOHEMA, HDEHP, TPH) (Am, Cm, PF) (TEDGA, HNO 3 )

EXTRACTION SCRUBBING

AW (Cm, Eu, Sm, Gd) LS BX : Solvent BX LX : Solvent (Citric, pH 3) (HEDTA, Citric acid, pH 3)

LW Mo stripping Ln scrubbing Am stripping (Pd, Mo, Ru) 30°C 45°C

LS' (NaOH) BP CW : Solvent (Am) Ln stripping

CLn CX (TEDGA, HNO , H C O ) (Nd, Ce, Pr, La, Fe) 3 2 2 4

Fig. 3. Flowsheet of the EXAm process implemented in the shielded process line

3. Experimental results

3.1. Results of the EXAm process in surrogate solutions (glove boxes test)

The experiment was based on a surrogate feed solution which was constituted with every extractable inactive fission products (Ln, Mo, Zr) spiked with small amounts of americium and curium. This test lasted 64 hours during which the transitory and equilibrium phases were followed by neodymium on-line spectrophotometric measurements and ICP/MS titrations carried out in the shielded analysis line (results are compared in Fig. 4 showing a good agreement between these two techniques). Alpha and gamma spectrophotometric assays were performed to follow americium and curium concentrations at each aqueous outflow of the process. These measurements allowed to demonstrate an americium recovery rate at 97% (lower than the expected values) and decontamination factor of Am versus Cm higher than 1000.

182 Marie-Jordane Bollesteros et al. / Procedia Chemistry 7 ( 2012 ) 178 – 183

3500 Back scrubbing aqueous phase (on-line spectrophotometry) Back scrubbing aqueous phase (ICP/MS measurements) Ln stripping aqueous outflow (on-line spectrophotometry) 3000 Ln stripping aqueous outflow (ICP/MS measurements)

2500

2000

[Nd] (mg/L) 1500

1000

500

0 01/12 01/12 01/12 01/12 01/12 01/12 02/12 02/12 02/12 02/12 02/12 02/12 03/12 03/12 03/12 03/12 03/12 00:00 04:00 08:00 12:00 16:00 20:00 00:00 04:00 08:00 12:00 16:00 20:00 00:00 04:00 08:00 12:00 16:00 Date/Hour

Fig. 4. On-line neodymium spectrophotometric monitoring (-) and ICP/MS measurements (Ƈ) during the EXAm process experiment in glove boxes

3.2. Results of the EXAm process in genuine solutions (shielded process line test)

In order to optimize and ease the mixer-settler implementation in the shielded process line, a mockup in a horizontal rail was realised at the same scale as the experimental one. All inactive reagents were introduced from glove boxes located on the roof of the shielded line. The PUREX raffinate (the genuine feed solution) was stored in a special tank and previously characterized before the test. After a hydrodynamic test phase with inactive solutions in the shielded line (which allow to check the proper functioning of mixer-settler batteries and the accuracy of the flow rates), the test started in the same conditions as the glove boxes test conditions. The aim was to achieve the same performance and then to modify some flowsheet parameters (thanks to the pump monitoring system and PAREX calculations) in order to reach the expected results (americium recovery rate over 99% and decontamination factor of Am versus Cm higher than 500). The monitoring of the process was based on americium on-line spectrophotometric measurements as illustrated on Fig. 5 for different stages of the scrubbing mixer-settler extractor. These curves show: x A good reliability of the technique by comparison with ICP/AES and gamma spectrometry measurements from samplings realised during the trial, x A disturbance occurrence noticed during the test by the decrease of the americium concentration in all the stages of the extractor. Therefore, adjustment was performed before the end of the test, x An equilibrium state attainment for americium at the end of the trial by the stability of the actinide concentration. After 54 hours, the test was stopped and several samplings in the aqueous and organic phases were carried out to characterize the equilibrium state and the process performances with alpha and gamma spectrometry and ICP/AES analysis. The results showed that more than 99% of americium was extracted and the decontamination factor versus curium was about 500.

Marie-Jordane Bollesteros et al. / Procedia Chemistry 7 ( 2012 ) 178 – 183 183

700 Back-scrubbing aqueous phase Stage 6 scrubbing Stage 12 scrubbing 600 Back-scrubbing aq. phase (ICP/AES) Back-scrubbing aq. phase (gamma spectrometry)

500

400

300 [Am] (mg/L) (mg/L) [Am]

200

100

0 30/03 30/03 30/03 30/03 31/03 31/03 31/03 31/03 31/03 01/04 01/04 01/04 01/04 07:00 12:00 17:00 22:00 03:00 08:00 13:00 18:00 23:00 04:00 09:00 14:00 19:00 Date / Hour

Fig. 5. On-line americium spectrophotometric monitoring in the scrubbing mixer-settler extractor and comparison with titrations performed by ICP/AES and gamma spectrometry

4. Conclusion

The EXAm process which allows to recover quantitatively and selectively americium from a PUREX raffinate was demonstrated (americium recovery rate over 99% and decontamination factor of Am versus Cm higher than 500) thanks to a hot test performed in the shielded process line of the ATALANTE facility. Prior to this experiment and on the basis of modeling and simulation, a trial on a feed surrogate solution in glove boxes confirmed the Am/Cm selectivity in the extraction/scrubbing operation. Several monitoring tools such on-line and in situ spectrophotometric measurements were implemented in this corrosive and irradiative media. These techniques, complemented with analytical measurements in the shielded analysis line, are noticeable advantages to process managing and quick decision-making.

Acknowledgements

We wish to acknowledge AREVA NC (France) which financially supported the present work.

References

[1] Montuir M, Sorel C, Pacary V, Hérès X, Roussel H, Baron P, Dinh B. Parex simulation code: an efficient tool to model and simulate solvent extraction operations. Proceedings ISEC 2011, Santiago, Chile, October 2011. [2] Montuir M, Sorel C, Pacary V, Hérès X, Roussel H, Baron P. Parex simulation code to model and simulate solvent extraction operations such as the EXAm process. Récents progrès en génie des procédés, n°101, Ed SFGP, Paris, France, 2011.