Enhance of Polymer Efficiency using Box-Behnken Design 

 

 

 

Efficiency was improved reducing variability and maintaining end product quality. 

 

 

The Case

When lean six-sigma deployment started as CEO initiative in the organization, the binder consumption was one of the vital items on the agenda. In production operations, the total cost of binder was more than 4 million euro and the overall efficiency was 64%. I was project member in this project and in charge of all binder efficiency improvement projects. 

 

Objective

The objective of the project was to improve binder efficiency maintaining essential properties of the end products in the selected range of products. 

 

Work Done

In define phase the CTQ and y selected as output. The y was binder efficiency. Thus, the scope of the project was very broad and time schedule was long like eight months. Inputs and outputs are defined for the process.

In measurement phase, it was noticed that there are missing approved protocols and procedures for process control. These procedures were created in accordance with CE. In this process the validation of the standard operating procedures made via Gage R&R and bias studies were made.

In the project, an intensive measure, analysis and improve circle applied. The essential improvement was made with the enhancement ideas of process owners and associates on the affinity diagram in the two brainstorming sessions.

The number of variables reduced in measure phase from 154 to 42, after analysis phase reduced to 8 and in improvement phase only 2 variables focused on in design of experiment to lead an essential improvement.

One of the associates shared his observations in one cleaning day of the production. This idea illuminated our way and we decided to focus on this parameter and designed the experiment as Box-Behnken with three factors, two levels and one output. The output was the most important characteristic of the end product that must be maintained.

 

 

Conclusion and Results 

At the end of the experiment, that we made trial in whole production operations, the most important factor was found out. As in the following graph the amount of binder in x-axis does not influence the output. It was the one that stems from the associate’s observations. The potential impact of the improvement was to increase efficiency by 12 percent. The solution was troubleshot, fine-tuned and delivered as the solution.

 

Furthermore, the standard operating procedures created in the course of the project decreased variation in overall operation. The process owners were able and comfortable to predict the outcome more precisely at selected parameters.

 

The efficiency improvement was proposed as from 64% to 76%. Moreover, the variation is reduced with operating procedures and its trainings.

The financial impact of the project was declared by finance sub-committee 240k Euro asset value per year per factory, which forecasted as 1.2 million Euros per year in all operations.