We offer to implement root cause analyses on tenacious cases of process impairment in pulp mills based on mill operating data, machine learning and deep learning techniques. Previous experience with root cause analysis of kraft pulp mill processes include projects on poor black liquor combustibility, impaired dregs settling in a green liquor clarifier, high fluctuations of chloride and potassium in ESP ash, organic deposit formation in foul condensate stripping systems, sludge dewatering and anaerobic wastewater treatment issues, and others. Some of the main challenges of root cause data analysis can be poor data quality and the fact that highly influential parameters are not always measured frequently. In those cases additional chemical analyses may be necessary.
We offer to identify various strategies for process optimization. Focus of optimization may be a decrease of the standard deviation of key process variables to ensure a high-quality product at all times, minimize the downtime for equipment cleaning, or the time during which the product is outside of quality specifications. The ultimate goal is to efficiently produce a high-quality product at a minimum cost, and by maintaining a small environmental footprint.
We offer classes for mill personnel on basic and advanced data analysis methods to identify the causes of process issues. The tools for these analyses include multivariate data analysis and python-based machine learning techniques.
Each of the previous projects helped the mill gaining more insight into the causes of the process impairment, and ultimately save costs in the long run. For root cause analysis and optimization projects the compensation for the work is variable, whereas a smaller part is based on the number of hours spent on the project, and a larger part is dependent on the success of the analysis.
DMIP
Copyright © 2022 DMIP - All Rights Reserved.
Powered by GoDaddy
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.