th
5 MEGAPIE TECHNICAL REVIEW MEETING, Nantes, France, 2004 : Initial start-up operations chemistry analysis for MEGAPIE, J-L. Courouau
1/10
Initial start-up operations chemistry analysis for MEGAPIE. J-L. COUROUAUa, S. SELLIERa, F. BALBAUDb, K. WOLOSHUNc, A. GESSId, P. SCHUURMANSe, M.OLLIVIERa, C. CHABERTa a
: CEA Cadarache DEN/DTN/STPA/LCP F-13108 Saint-Paul-Lez-Durance cedex Tel.: +33-4 42 25 32 66 Fax: +33-4 42 25 72 87 e-mail:
[email protected]
b
: CEA Saclay DEN/DPC/SCCME/LECNA : Los Alamos National Laboratory, US-DOE d : Brasimone research center, ENEA e : Mol research center, SCK-CEN c
Abstract During the initial start-up of any facilities to be operated in lead-bismuth eutectic alloy (LBE) such as the MEGAPIE spallation target for instance, the main source of pollution is identified as being the oxygen, so that the super saturation of the LBE is assumed as almost certain. Oxygen comes from the LBE ingots, from adsorbed gases in the structure, as well as from the residual traces of oxygen and moisture in the circuits before filling. Start-ups are known to present the highest chance for air ingress due to failure or misconception of systems. The initial melting and start-up operation should then control the contamination risk (clogging) to ensure the proper conditions for the Integral Tests (MITS), in order to ensure that the LBE is always kept below the oxygen saturation threshold. Considering the narrow section of some pipes, as well as the low operating temperature that set very low oxygen saturation, this risk is assessed as being not negligible. The analysis of some of the available experience from various loops allowed drawing a number of observations and recommendations on the level of procedures and/or process control that could benefit to the MEGAPIE experiment to increase its proper achievement with the highest confidence during the initial fill/drain tests.
1
Introduction
The development of the heavy liquid metal chemistry control and monitoring is one of the issues that is critical for nuclear systems using lead-bismuth eutectic (LBE) alloy either as a spallation target such as MEGAPIE or as a coolant. Indeed, the chemistry interacts with at least 3 operating specifications for any nuclear system: -
Contamination control, to ensure stable hydrodynamics and heat transfer during service-life time (avoid lead oxide clogging or even corrosion products plugging due to mass transfer in a non isothermal system, avoid deposits that reduce the heat transfer capacity, etc.);
-
Corrosion control, to ensure sufficient resistance of the structural materials during the expected service life-time (self-healing oxide protection layers that requires oxygen control, coatings, corrosion inhibitors, etc.);
-
Radio-activation control, to ensure a safe management of the operating and maintenance phases;
Depending on the potential source of pollution, on the operation (start-up, restart after maintenance or repair, normal operation) and the operating conditions (temperature, temperature gradient, velocity…), one specification could become critical when compared to the others. Indeed, the low operating temperatures that were specifically chosen as well as the small service life of the target could justify neglecting the corrosion control. In addition, as the target will not be maintained nor repaired, but only disposed off after the operations, the radio-activation control could be simplified. However, the contamination issue could prove to be the most critical issue for the good achievement of the Integral Tests if not well taken into account. Indeed, deposits of slags on pipe walls of heat exchanger, or even the electromagnetic pump duct, may affect the temperature and power control calibration. Difficulties for draining or filling the lead-bismuth eutectic may even happen due to an excessive presence of oxides, requiring a higher temperature for filling or draining for instance. This was observed on various facilities during the initial stage of operations. The excess of oxygen proved difficult to recover from, and mostly time consuming. For the MEGAPIE Integral Tests [1] [2], because of the tight schedule and because of the attention of the international community that will necessarily be focused at this moment on these tests, it appears that this control is important to get it right at first with a high
th
5 MEGAPIE TECHNICAL REVIEW MEETING, Nantes, France, 2004 : Initial start-up operations chemistry analysis for MEGAPIE, J-L. Courouau
2/10
confidence. Difficulties at this stage could possibly be interpreted as a technology failure, whereas it would only result from design choice to comply with the tight schedule and the restricted budget. The analysis of some feedbacks that is now available from the first filling operation of various loops may help to sort out the critical points when compared to theoretical and ideal behavior, and how is could be controlled with the easiest way. Indeed, these operations, that was considered at first as quite straight forward, based on the liquid metal experience like the sodium or lead-lithium eutectic, or based on the available information from the Russians, proved to be a quite complex operation, especially in case of loss of control of the chemistry specification, as it proved to be very difficult to recover from a large pollution. The objective of this paper is then to analyze what was achieved for the start-up of 5 loops, DELTA (LANL Los Alamos), LECOR and CHEOPE III (ENEA-Brasimone), as well as CICLAD (CEA-Saclay) and STELLA (CEACadarache). These start-up operations can be considered as representative of the MEGAPIE target by the sparse chemistry control that was available then, except for the CICLAD and BOR-60 loops. It allows confronting the theory for filling and transfer operation, in order to define what is the minimum requirement for such an operation, and what recommendations could be given. The review of the potential pollutions sources that could be expected in a LBE system is given first, then applied to the initial start-up operation of the MEGAPIE target. Then, the analysis of the 3 loops will be given to sort out some basic recommendations and hints to achieve properly this operation, safely and according to the planning.
2
Identification of the pollution sources for start-up operation
One of the main impurities [3] [6] is obviously the oxygen as it forms solid lead or bismuth able to build up within the system, but it is not the only one to look after, as there could be many effects from traces present in the coolant that could have macroscopic effect on the operation: the local accumulation of corrosion products able to cause clogging on the long term, or the effect of activation of the system able to cause operational or design difficulties (activation of traces of Ag, shielding of the cover gas system, deactivation process before maintenance, etc.). The other main pollution sources are identified as the corrosion products (Fe mainly, Ni, Cr, etc.) expected to be generated continuously at a rate depending on the operating temperature, liquid metal flow rate, materials, etc., as well as the spallation and activation products (Po, Hg, Tl, etc. ), without speaking of the proton beam itself and the light particles production as hydrogen (including tritium) within the reactor core. The sources of impurities can be addressed by analyzing the system at various operating steps: initial start-up before the first irradiation, normal operating mode, incidental/accidental conditions. The present analysis is limited to the start-up operation, so that corrosion is negligible and no radio-activation is expected. The initial pollution sources refer to pollution present in the coolant or in the system before the first irradiation of the system: -
Intrinsic pollution of the coolant itself;
-
Pollution during the melting operation;
-
Gases adsorbed on the inner surface of new structure;
-
Traces of impurities (O2, H2O) remaining in the circuits before filling;
-
Casual impurities (cutting, oil, grease…);
-
Incidental pollution (air, oil, mercury…);
These are further detailed hereafter. Intrinsic source: The intrinsic source of pollution refers to the traces of impurities present in the coolant itself. They might be due to the use of materials for the LBE mixing and casting that could contaminate slightly the liquid metal, especially if the material is not devoted to LBE only. This was illustrated by the LBE quality supplied by the SOGEMET Company, which specialty is tin, which appeared to be contaminated by 600 to 800 ppm in tin [3]. On the opposite, the Metaleurop Company used a specific new tank as well as new pipes, and this proved to be efficient for not contaminating the alloys during the mixing and casting procedure.
th
5 MEGAPIE TECHNICAL REVIEW MEETING, Nantes, France, 2004 : Initial start-up operations chemistry analysis for MEGAPIE, J-L. Courouau Metaleurop Recherche (µg/g)
3/10
LiSoR sample (µg/g)
Sn
