Performance, monitoring and modeling of hydrogenotrophic sulfate reductiona study in a gas-lift bioreactor

Abstract

SO2 is a contaminant harmful to the environment that is mainly emitted from fossil fuels combustion, and amongst its associated problems are the acid rain and the formation of particulate pollutants. SO2 treatment can be performed through SONOVA/ENSURE process that aims at treating the flue gas emissions targeting sulfur recovery that could allow its valorization in other economic sectors. It consists of three global steps: gases absorption, biological conversions, and elemental sulfur recovery. This thesis focused on the biological step and, particularly, on the sulfate reduction to sulfide, using H2 as electron donor and CO2 as carbon source. In that sense, this study was developed in three main facets. The first consisted in the experimental study of the biological sulfate reduction. Herein, preliminary experiments were performed to enhance the hydrogenotrophic sulfate reduction using an inoculum from a full-scale UASB reactor that treated sulfate-rich wastewater from a pulp and paper factory. The microbial community was enriched in hydrogenotrophic sulfate reducing bacteria (H2-SRB) through sequential operation of two stirred-tank reactors (STR) and a gas-lift reactor (GLR) fed with H2 and CO2, in which methane production was avoided but low sulfate reduction rate (SRR) was obtained. Therefore, the GLR was set up in a sequential batch operation aiming at increasing the SRR, in which no methane production was observed, and acetate was only produced when the sulfate loading rate (SLR) was low; thus, an optimal and stable operation was achieved at a SLR of 902 mg S-Sulfate L-1 d-1 with a 93±7 % (w/w) sulfate removal efficiency. The second part of the research consisted in manufacturing a sulfide online-monitoring system (S-OMS), in which it was possible to measure total dissolved sulfide (TDS) in a range of 1.5 to 30400 mg TDS L-1, and the repeatability and reproducibility tests showed a suitable RSD in the range 1.8-5.3 %. The results of the GLR monitoring for nine cycles showed that the S-OMS measurements correlated satisfactorily with measures from a commercial sulfide ion selective electrode. The last step of this thesis consisted in the development, calibration, and validation of a mathematical model to describe the GLR operation. The model was calibrated and statistical analysis for the model validation showed good results for sulfate and acetate with less accuracy for TDS that was explained from the large experimental deviations of TDS values. Different simulations were performed to evaluate the model predictions under different operational scenarios. These simulations showed that sequential batch operation with shorter cycle duration or higher liquid volume exchange could improve the specific sulfate reduction rate (s-SRR) and thereby, less solid would be accumulated without affecting the SRR. It also predicted that the system could operate in continuous mode with SRR comparable to the sequential batch operation and higher s-SRR. Overall, the selection and growth of H2-SRB and the maximization of the SRR through a sequential batch operation was demonstrated together with a good description of the experimental values through a mathematical model. Also, a good approach was achieved for the implementation of a S-OMS for the real-time monitoring of sulfide concentration.