Glass curtain walls have gained significant popularity as building enclosure structures, owing to their exceptional performance. Consequently, it becomes imperative to reduce energy consumption during the glass tempering process to ensure sustainable development within the construction industry. In this study, an experimental design was conducted using the response surface methodology (RSM). The particle count and cooling energy consumption were selected as the response value, while the influencing factors included four process parameters, namely, the spray distance (H), mist load fraction (f), spray pressure (p), and final cooling temperature (Tz). The second order model fitting of the RSM was used to obtain the polynomial regression equation between each response and each factor. Under the optimized conditions of H = 200 mm, f = 1.03, p = 0.5 MPa, and Tz = 150 °C, the number of fractured glass particles was found to be 234, and the cooling energy consumption was 0.00048 kW·h. Further analysis indicated that under these spray conditions, the net present value was 523.90 k$, and the dynamic payback period was 1.1 years. Compared to traditional air-cooled tempering, spray tempering demonstrated superior economic performance.