Data driven approach boosts productivity

Edited by the Engineeringtalk editorial team May 3, 2006

Process Monitoring software produces tangible results for Abitibi-Consolidated, a global leader in newsprint and commercial printing papers.

Abitibi-Consolidated is a global leader in newsprint and commercial printing papers as well as a major producer of wood products, serving clients in some 70 countries from its 45 operating facilities.

Abitibi-Consolidated is the largest recycler of newspapers and magazines in North America, diverting annually approximately 1.9 million tonnes of waste paper from landfills.

It also ranks first in Canada in terms of total certified woodlands.

Sheet-breaks are a significant contributor to lost paper machine efficiency, productivity as well as greater operating costs.

Consequently, sheet break reduction is a major area of focus for most paper operations.

Like many mills in the industry, the Fort Frances paper machine 7 suffered from deteriorating production.

PM7 is a 210in, 1920s-vintage, Valmet Bel-Baie III paper machine, rebuilt in 1984, and makes value added paper for communication.

It was running at an average of 905m/min, suffering an average of 2.3 wet end breaks per day, with an overall equipment efficiency (OEE) of 69.4%.

The average downtime for 2004 was 75min per day.

In spite of the mill's various improvement initiatives, it could not sustain its performance gains, and the benefits gradually reduced over time.

"Abitibi was looking for a new sustainable method to improve the paper machine production", said Jorg Ruppenstein, Process Control Engineer at Abitibi.

"We were looking for a quantifiable data driven approach".

"This would enable us to compare our performance before and after the project, so that we could measure a true economic benefit".

Abitibi partnered with Matrikon to perform a joint sheet-break reduction initiative.

Abitibi provided all the necessary plant data and plant operations experience.

Matrikon provided the tools and expertise to perform the advanced data mining and analysis.

The core technology for this effort was the Matrikon .

The project was broken into four distinct phases: interview and knowledge discovery; data collection; data mining and analysis; and defining improvement projects.

The data analysis used 8 months of data that were already stored in the mill's process data historian.

This data included information from the paper machine's automation systems (distributed control systems, quality control system, and programmable logic controllers) drives and other sources.

The analysis enabled engineers from Abitibi and Matrikon to make the following conclusions.

The root causes were classified into three major process areas: stock prep and stock proportioning, the steam system, and the draw system.

30% of the breaks were related to changes in the broke system.

15% of the breaks were related to dryer section differential pressure variability along with dryer motor current variability.

15% of the time a second sheet break followed the first sheet break within 30 minutes and 15% of the breaks were preceded by large changes in draws.

The team identified 50 improvement opportunities.

The top five improvements were: a change in the stock proportioning logic so that broke only displaces virgin groundwood and not virgin kraft; changes to the broke addition logic and piping from three parallel streams into a single stream; repairs to the dryer controls and instrumentation; recommissioned the white water silo temperature control system; and mechanical repairs to eliminate debris hung up and release after a break.

As the goal of the project was to deliver sustainable improvements, the team calculated all measurements over a nine-month period after the project.

The results were impressive.

The average wet end sheet breaks per day reduced by over 40% (from 2.30 to 1.36 breaks per day).

The average production rate increased by 12% (from 303 to 338t per day).

Average downtime reduced by 37% (from 75 to 47min per day).

OEE (overall equipment efficiency) increased by 8.3% (from 69.4 to 77.7%) "The Matrikon Process Monitoring software helped our Fort Frances paper mill realise over $1 million in savings during the first 9 months alone", said Ruppenstein.

"We have a sustainable, quantifiable data driven approach to improve paper machine production".

"It was truly a team effort between operations, engineering, maintenance and Matrikon".

Abitibi has future plans to deploy this solution throughout its mills.