Optimizing demineralized water production
20% reduction in annual pump electricity costs and 25% reduction in annual maintenance spending.
Challenge
The customer’s existing operating procedure involved maintaining the demineralized water tank at maximum capacity to prevent shortages during increased consumption or membrane washings. This practice, however, led to the membranes/filters being cleaned less frequently than manufacturer recommendations, causing them to be replaced more often due to the perceived risk of water shortage.
Solution
Arundo developed a suite of machine learning models to enhance the demin water plant's operation and maintenance. This included a consumption forecast model to maintain consistent production rates, a predictive model for membrane cleaning and replacement scheduling, and a forecast model to ensure demin water needs were met during the cleaning process. A user interface was also created, providing real-time information and notifications, along with connecting the model outputs to the Distributed Control System (DCS) for automated pump regulation.
Impact
Implementing Arundo’s solution resulted in a 20% reduction in annual pump electricity costs and a 25% reduction in annual maintenance spending, significantly improving operational efficiency and reducing expenditures.