Comparing RSM-BBD and GEE models to optimize urban stormwater runoff treatment in dolomite-modified porous concrete

With rapid urbanization exacerbating stormwater pollution and stressing limited water resources, developing efficient and cost-effective treatment technologies to transform contaminated urban runoff into a usable water source has become a pressing environmental challenge. This research investigated the effectiveness of a porous concrete matrix containing dolomite adsorbent for treating urban stormwater runoff. Incorporation of dolomite adsorbent into the porous concrete matrix resulted in a 20% decrease in hydraulic conductivity and a 54.78% reduction in porosity were observed. Conversely, compressive strength increased by 34.85%, and tensile strength improved by 21.56%. To optimize removal of chemical oxygen demand (COD), total dissolved solids (TDS), total suspended solids (TSS), and lead ions (Total Pb), the study employed a combined approach of response surface methodology utilizing a Box-Behnken design (RSM-BBD) and a generalized estimating equation (GEE) model as a function of six operational parameters: 63 < COD < 146 mg L^-1, 320 < TDS < 1600 mg L^-1, 66 < TSS < 937 mg L^-1, 2 < LTotal Pb < 5 mg L^-1, 15 < contact time < 360 min, and 5 < dolomite levels < 15%. The RSM-BBD approach outperformed the GEE model, demonstrating its ability to effectively capture non-linear relationships between input variables and removal efficiencies. The sensitivity analysis, based on Wald Chi-square values, showed that the removal of TDS, COD, TSS, and total Pb was primarily influenced by the initial concentrations of TDS, TDS, TSS, and the dolomite adsorbent dosage, respectively. The system achieved complete removal (100%) of Total Pb, along with substantial reductions of 83% for TSS, 78.91% for COD, and 64.68% for TDS. These remarkable removal efficiencies indicate the potential of the porous concrete-dolomite adsorbent system for treating urban stormwater runoff.