A novel pareto based multi-objective simulation-optimization model for in-situ bioremediation of contaminated groundwater

•Two simulation–optimization models are proposed for multi-objective bioremediation.•The simulation is performed using meshless MLPG and MWS methods.•A hybrid HDEWOA optimizer with non-dominated sorting methodology is implemented.•The models are demonstrated using field-scale aquifer contaminated wi...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Journal of hydrology (Amsterdam) Ročník 662; s. 134052
Hlavní autoři: Das, Sanjukta, Eldho, T.I.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.12.2025
Témata:
Groundwater modelling
Hybrid optimization
Meshless methods
Multi-objective bioremediation
Pareto solutions
Simulation optimization
Groundwater modelling
Pareto solutions
Hybrid optimization
Simulation optimization
Meshless methods
Multi-objective bioremediation
ISSN:0022-1694
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•Two simulation–optimization models are proposed for multi-objective bioremediation.•The simulation is performed using meshless MLPG and MWS methods.•A hybrid HDEWOA optimizer with non-dominated sorting methodology is implemented.•The models are demonstrated using field-scale aquifer contaminated with BTEX.•Total and partial aquifer restoration scenarios are designed using SO models. Bioremediation is an effective technique for restoring aquifers contaminated with organic compounds, particularly petroleum products. The design of bioremediation systems typically involves multiple, conflicting objectives. While significant progress has been made in the single-objective bioremediation, studies addressing the multi-objective designs remain limited. In this paper, a multi-objective Simulation-Optimization (SO) model is presented for bioremediation design of Benzene-Toluene-Ethylbenzene-Xylene (BTEX) in field-scale aquifers. The simulation involves solving coupled groundwater flow, contaminant transport, oxygen transport, and instantaneous reaction using Meshless Local Petrov Galerkin (MLPG) and Meshless Weak Strong (MWS) methods. For a field type problem, in comparison to generally used MODFLOW-RT3D model, the MLPG and MWS solutions of BTEX concentrations are in excellent agreement with 2.105% and 2.697% deviation. Further, a novel multi-objective hybrid optimization algorithm, named Non-dominated Sorting Multi-Objective Hybrid Differential Evolution and Whale Optimization Algorithm (NSMO-HDEWOA), is specifically developed for groundwater remediation problems, and integrated with MLPG and MWS simulators. The performance of NSMO-HDEWOA is compared with Non-dominated Sorting Genetic Algorithm-II (NSGA II). The SO models determine the optimal locations and corresponding rates of injection and extraction wells across various management periods. The models are applied for bioremediation of field-type aquifer under scenarios of total and partial aquifer restoration. For the total aquifer remediation scenario with 100% BTEX destruction, the MLPG-NSMO-HDEWOA and MWS-NSMO-HDEWOA outperform MLPG-NSGA II and MWS-NSGA II by identifying designs with 15.4% and 12.4% lower remediation costs, respectively. Thus, this study highlights the potential of presented SO models in addressing real-world groundwater contamination challenges.
ISSN:0022-1694
DOI:10.1016/j.jhydrol.2025.134052