Robust optimization for designing air quality monitoring network in coal ports under uncertainty

Abstract

Air quality monitoring network (AQMN) is an important management tool to mitigate the impact of air pollution from coal port operations on coastal atmospheric environment and public health. The location of monitoring stations in the AQMN is influenced by wind conditions and emission intensity of air pollutants, which is highly related to the efficiency of coal port operations. Therefore, this paper aims at solving the problem of how to design the AQMN with consideration of uncertain operational efficiency and wind conditions. This study proposes a robust optimization method to address the impact of uncertainty on the design of AQMN in coal ports. Firstly, the deterministic AQMN design problem is summarized as a maximum weighted covering location problem (MWCLP) and formulated as a gradual covering model with the cooperative covering strategy. Then, a linear approximation method is applied to the non-linear model to improve computational efficiency. Secondly, a method for constructing uncertain scenarios is proposed by combining sampling method, port emission inventory, and atmospheric diffusion simulation. After that, a robust optimization model is developed based on the concept of p-robustness to ensure that the relative regret value of each uncertain scenario is less than a specified threshold. Finally, we take a coal port in northern China as a case study. Results show that the average monitoring capacity of robust layout is 5.25% and 1.57% higher than that of conventional layout and deterministic layout. The robust layout ensures a relative regret value of no more than 3.41% for each scenario. The proposed method can provide port and environmental authorities with robust decision support of AQMN design under limited budget and uncertainty of operational efficiency and wind conditions.