Exploring correlation between specific energy consumption (SEC) and variables of the plastics injection molding process

Abstract

To combat the emission of GHGs and the depletion of fossil fuels, research must explore methods to reduce energy consumption in different sectors of the economy. Within the industrial sector, rubber and plastics production is an energy-intensive activity. Therefore, this investigation focused on increasing energy efficiency in one of the most energy intensive processes in the plastics industry; injection molding. Consultations with plant managers at plastics injection molding companies revealed that energy consumption could highly vary during injection molding activities under relatively stable conditions. The reason for these variations in energy consumption were unknown to plant managers and had not been investigated before in scientific literature. Subsequently, a list of variables which could possibly cause these fluctuations in energy consumption was compiled from previous research and interviews with plant managers. Three main categories of predictor variables were identified; shifts, machines, and products. In the data collection phase, energy consumption measurement of 8-hour shifts over a seven-month period was done for 15 injection molding machines, which in this period produced a total of 142 different products. Within the three main variables, other sub variables were determined. For each of the three main categories, multiple regression analyses were carried out to determine the correlation directions (which were compared to initial hypotheses) and sizes between the found sub predictor variables and energy consumption. For one of the three main variables (machine variables), not enough sample data was available to carry out a multiple regression analysis. Therefore, not all identified predictor sub variables could be included. However, standardized correlation values between the remaining predictor variables and energy consumption were compared. The obtained results indicate that among the included predictor variables, the amount of mass produced per cycle has the largest correlation with energy consumption. Optimizing this process parameter – for example by increasing mold cavities, and thus identical products produced per cycle – could theoretically yield a ±20% increase in energy efficiency. Other predictor variables also showed significant correlation with energy consumption, but as some of the predictor variables showed intercorrelation, a trade-off effect between predictors exists, meaning that optimizing one predictor variable would affect the ability to optimize another predictor variable. Consequently, a primary recommendation for plant managers is to optimize the predictor variable with that highest correlation with SEC; Mass produced per cycle, for example by increasing the number of mold cavities. For further research, it is interesting to investigate the influence this optimization has on specific energy consumption, but also on other important factors such as product quality. Another recommendation for researchers with intentions of conducting further investigation, is to keep in mind the sample size required for statistically significant research.