Response Surface Methodology (RSM) Approach for Optimizing Waste-Based Acoustic Panels in Generator Set Noise Reduction at PT X

Generator set operation at PT X produced noise levels above the workplace threshold limit, reaching 89.9 dBA in 2025. This study aimed to optimize generator set noise reduction using waste-based acoustic panels made from bamboo, cassava peel, and sawdust through Response Surface Methodology (RSM). The analyzed factors were material type, panel thickness, and testing duration, while the response was noise reduction in decibels. The experimental data were processed using Design Expert software with a D-Optimal design. The results showed that the linear model was the most suitable model, with a significant p-value, non-significant lack of fit, and good predictive ability. ANOVA results indicated that material type was the most dominant factor, followed by panel thickness and testing duration. Among the three waste materials, bamboo composite produced the best damping performance. The optimum condition was achieved using bamboo composite with a thickness of 2 cm, resulting in a predicted noise reduction of approximately 12 dB. The optimized result was then applied to a generator set enclosure design using SketchUp to support practical visualization and technical planning.