A Multi-Criteria Decision Analysis of NOx Abatement Options for Refineries

Abstract

The research is conducted with the case of Shell Netherlands Refinery (SNR), which has already implemented certain NOx reduction measures such as steam injection in gas turbines and Low NOx Burners in furnaces. The decisions to implement these technologies were primarily based on the level of NOx emissions and financial considerations. This thesis aims to support decision-makers by providing a comprehensive comparison of multiple NOx abatement systems based on multiple criteria.
SNR has various emission sources for NOx, such as burners in furnaces with the emission point being the stack. Each emission point and NOx source within SNR has its own technical limitations and possibilities, which have been carefully assessed to determine the appropriate NOx abatement systems for each emission point.
The Multi-Criteria Decision Analysis (MCDA) approach was employed to answer the above research question. MCDA is a structured approach that evaluates and compares different options or alternatives based on a set of criteria. It enables decision-makers to make informed and objective decisions by considering multiple criteria and weighting their relative importance.
The study concludes that combustion-based systems, specifically ULNB for furnace burners and (U)DLN burners for gas turbines, are more advantageous than end-of-pipe systems for NOx reduction in SNR. However, in cases where combustion-based technologies are not technically feasible, the integration of end-of-pipe systems is necessary. In such scenarios, the ClO2 wet scrubber is recommended due to its high final utility score. SCR, SNCR, and LOTOX technologies are generally not advised due to their lower final utility scores but SCR or LOTOX could be viable when high NOx emission reductions are required (>90%).
According to outcomes of MAUT, the ULNB was identified as a particularly effective NOx abatement technology, showcasing a consistently high final utility score at numerous emission points, specifically O, P, Q, N, L, I, J, E, M, F, D, B, and H. in contrast, at emission points R, A, E, G, K, S, and W, the most efficient solution was found to be the ClO2-based wet scrubber system, which achieved the highest final utility score. For the final emission points, namely T, U, and V, the (U)DLN burners scored the highest final utility score, thereby earning the recommendation as the preferred NOx abatement technology for these emission points.
However, this study had its limitations. The cost analysis conducted was somewhat superficial. A more in-depth exploration of the CAPEX and OPEX, coupled with the use of complexity factors per emission point, could refine the MCDA outcomes. It is recommended that future studies integrate Cost-Benefit Analysis (CBA) into the MCDA for a more rigorous examination.
Another constraint was the limited scope and quality of data used. It is recommended that subsequent research take a more rigorous approach to data collection, including engaging factories that have already implemented NOx abatement systems. Such an approach could yield a more comprehensive dataset, enhancing the reliability of MCDA outcomes.