In Canada, the sawmill industry is a key manufacturing sector in lumber production for Canadian use and export.
It must constantly improve its competitiveness through operational control of its activities, especially by maximizing the availability of sawmill equipment and reducing costs. Maintenance based on prognostic models has become a major tool for improving and maintaining equipment performance and thereby improving a company’s overall equipment effectiveness.
The proactive (prognostic) maintenance approach involves a two-part process. First, it involves condition-based monitoring of the assets through a sensor network installed on critical components, such as vibration sensors, thermography, cutting quality, etc. Second, it involves developing algorithms for predicting component degradation based on the data collected. Prognostic maintenance, with its proactive approach, prevents unnecessary additional costs, unplanned and chronic breakages, as well as any unwanted events of corrective and reactive maintenance.
An Effective Methodology
During recent work at a member company, FPInnovations was asked to apply the proactive (prognostic) maintenance methodology to anticipate irregularities on a sawmill line. The goal was to measure deterioration in critical equipment’s component.
To reach operational performance objectives, the method is structured and sequenced in a time based; through:
- Sawing process surveillance to monitor and improve performance on the corresponding production line;
- An indication of the performance level achieved during each activity in the sawing process across the entire line;
- The development of diagnostic and prognostic failure models for each part (component) of critical equipment; and
- The development of appropriate maintenance plan to maintain the expected level of performance on the production line throughout its expected life cycle.
Some strategies have been developed to meet the definitions and achieve the main objective of proactive (prognostic) maintenance. These are:
- Sawing process modeling
- Overall equipment effectiveness (OEE) rate calculation
- Vibration data collection and analysis
- Preventive and Predictive maintenance plan development
- Asset observation and documentation
- A FMEA-type format design based on stakeholder and vendor’s expertise
- Analysis and valuation of data generated by the Computerized Maintenance Management System (CMMS) software
In the proactive (prognostic) maintenance domain, understanding trends in anomalies affecting equipment used during operations is important. The methodology therefore provides for the analysis of failures affecting certain equipment components by monitoring the outcome of their behaviour based on the various sources, including BPFI (Ball Pass Frequency of the Inner race), BPFO (Ball Pass Frequency of the Outer race), FTF (Fundamental Train Frequency) and BSF (Ball Spin Frequency).
Lastly, results are generated through failure modelling using machine learning techniques based on advanced statistical calculations.
Solid results for all types of mills
In the example of the member mill whose capacity is of the order of 150,000 MFB annually, gains arising from the strategies recommended by modelling are also significant and beneficial for all operations. These gains can be summarized as follows:
- Operational performance: potential average gains that may reach 2,50% per day and more
- Lost opportunity costs and benefits: potential gains exceeding $70,000 annually with an annual return on investment of 1.5
- Reduction of spare parts: potential gains of $75,000 annually and more.
Proactive (prognostic) maintenance can be an effective and lucrative approach for many mills in the wood processing or pulp and paper sectors.