Thermal denitration of sodium nitrate in a fluidized bed reactor: Optimization of process parameters and application of S-statistics for fluidization state monitoring

• Published in: Chemical engineering journal advances 2022-11, Vol.12, p.100390, Article 100390
• Main Authors: Badiwal, Ankit, Bhowmick, Sandip, Mukherjee, Debmalya, Singh, K.K., Saha, Shilpi, Shenoy, K.T.
• Format: Article
• Language: English
• Subjects: Liquid-sprayed fluidized bed; Pressure fluctuations; S-statistics; Thermal denitration
• ISSN: 2666-8211; 2666-8211
• DOI: 10.1016/j.ceja.2022.100390

•NaNO3 bearing liquid waste from Nuclear Reprocessing Plants is treated via fluidized bed thermo-chemical denitration route.•Sucrose is used as carbon source to reduce denitration temperature, agglomeration tendency and increase conversion.•Reconstructed attractor comparison technique based on pressure fluctuations measurement inside fluidized bed is applied for computing a parameter ‘S’.•Parameter ‘S’ is successfully tested for early detection for agglomeration in fluidized bed during thermal denitration of NaNO3. A voluminous liquid waste stream is generated as sodium nitrate (NaNO3) solution in Spent Nuclear Fuel Reprocessing Plants. Fluidized bed thermal denitration is a strong contender among various disposal options for this stream. In the present work experimental investigations on thermochemical denitration of NaNO3 have been carried out in an 80 NB laboratory scale fluidized bed reactor. Carbon based reductant was utilized to convert NaNO3 to sodium carbonate (Na2CO3) as well as to reduce denitration temperature. The operating parameters were optimized for maximum conversion of NaNO3 and 99.13% denitrification has been achieved. Pressure fluctuation across bed was also recorded and analysed in the state space domain employing attractor comparison technique. This technique computes a statistical metric ‘S’, as a measure of the fluidization state in the fluidized bed reactor and can be used for early detection of agglomeration in the bed. The sensitivity and early agglomeration detection capability of S-statistics are illustrated with experimental results for a liquid-sprayed gas fluidized bed.