| Abstract: | In this study, regression and cost analyses based on both mass balance and cost are applied to optimize the recycling of processing water for a 6-inch semiconductor wafer manufacturing plant. The analyses can also be applied to obtain the optimum reuse rate if the plant is expanded to produce 8-inch or 12-inch wafers. A mass balance diagram is first developed for three systems: processing system, wastewater system and recovery system. The information contained in the diagram is then modified to include the total costs of various water treatment unit processes for analyzing the costs for the above three systems. These costs are then used as the bases for optimizing the water recycling operation. Regression analyses to show the influence of various parameters e.g. the recovery rate of processing water, water usage, operation and maintenance costs and quantities of pollutants, on the operation costs of the above systems, have also been carried out using the Statistical Package for Social Science (SPSS) software. The resulting optimum recovery rates are 77% for the 6-inch wafer plant, 80% for the 8-inch wafer plant and 86% for the 12-inch wafer plant. Thus, when the wafer size is upgraded, the water-recycling rate should also be raised for achieving the most cost-effective water reuse. The analyses also evaluate the water recycling practice for various assumed unit water costs. Using the 6-inch wafer plant as an example, the optimum water-recycling rate should be raised to 74%, 77%, 78%, 81% and 84% for water costing $0.20, $0.36, $0.51, $0.71 and $1.00 per tons, respectively. Applying the cost analysis methods proposed in this study, the optimum water-recycling rate could be determined in response to variations of wafer manufacturing operations. Additionally, the r regression analysis results have demonstrated that the wafer size, costs of the three aforementioned systems, and water cost significantly influence the optimized recycling rate of semiconductor processing water. |
