How UPM Wisaforest uses condition monitoring for improved asset management.
This article explores the use of condition monitoring at the UPM-Kymmene's Wisaforest pulp and paper mill in Finland. Production capacity is 800,000 air dried tons per annum (ADt/a) of pulp and 180,000 ADt/a of kraft and sack papers.
The WISA 800 REC (RECovery) project entailed a single line to replace the mill's two previous recovery lines. A new sawdust cooking line was added, providing more flexibility in the raw material base by allowing the use of sawdust as feedstock for selected pulp qualities. This delivered both environmental advantages (fewer emissions) and process advantages by allowing the mill to draw from a larger pool of locally available sources for pulp feedstock.
The project scope was extensive. The main equipment suppliers for the recovery island were Andritz Corporation (process equipment), Siemens AG (main turbo-generator) and Metso Automation (process control system). GE Energy was chosen to supply a condition monitoring solution.
The Condition Monitoring System
Employees working with the WISA 800 REC project defined the scope of the CM system based on machinery criticality (see Table 1).
Large, high-speed turbomachinery generally warrants a conventional rack-based continuous monitoring system, and the plant chose proximity probes coupled with a machinery protection system. With exception of the lime kiln, all other machinery used rolling element bearings and was more appropriately addressed by a monitoring system using a scanning architecture. Both hardware solutions were linked to monitoring software to allow a common, connected platform.
Project Execution
One of the first items that needed to be defined was the total number of measurement points, which would be used to determine the number of dynamic scanning modules (DSMs) required. Ten DSMs were needed, reflecting the appropriate balance of wiring costs, hardware costs and scanning times. The locations for these DSMs were identified based on wiring topology, availability of power and network connections, and other factors.
The machines in the facility vary drastically from one another, including operating speed (1 to 3,000 rpm), drive mechanism (direct, belt and gear), and operating mode (constant speed, variable speed, constant load and variable load). Suppliers and plant personnel needed to cooperate to determine and document the correct values for all settings, and then enter these values into the software's configuration screens.
Taking the New Recovery Unit into Operation
One of the most crucial times for the CM system is when machines are tested and brought online for the first time. Problems that may not have been apparent at the factory may surface, or the installation of the machine may have introduced problems. Consequently, an important aspect of the project was to ensure that the CM system was configured and ready as each machine was started.
Through careful advance planning and schedule coordination, the CM system was ready to begin monitoring. As other machines were subsequently brought online, the system's configuration and commissioning were coordinated to coincide with their start-up dates.
When the new recovery unit officially began full-time operation, all measurement points had been collecting data for several weeks. The team turned its attention to fine tuning alarm levels and other system configuration settings. During these adjustments, no machine failures could be allowed and faulty operating conditions needed to remain visible. The company and supplier worked collaboratively to successfully accomplish these objectives in a timely fashion.
During the start-up phase, the CM system identified a number of machinery problems, which allowed proactive intervention and remedy before the entire plant went live. This early payback of the system and its usefulness during start-up activities had been a high priority for the Wisaforest project team and was part of the justification for installing the system. After full-time operation commenced, the system continued to deliver value by logging many other machinery saves.
Case Histories
Case History #1
Problem: Bearing Lubrication
Machine: Secondary Air Fan
Unit: Recovery Boiler
The secondary air fan is a 600 kW direct-driven overhung fan that is critical for the recovery boiler operation. Shortly after start-up, abnormal changes in trends of the high frequency data from the inboard bearing accelerometer were noted. Figure 1 is taken directly from the CM software, showing a two month trend of high frequency data from the accelerometers on the inboard and outboard bearings.
The elevated levels on the inboard bearing (blue) compared to the outboard bearing (orange) are readily apparent.
Spectral analysis suggested that the bearing's outer ring was wearing prematurely, and the root cause was ultimately traced to problems with the bearing lubrication system. The prominent dips in the trend plot correspond to intermittent operation of the lubrication system, showing a marked decrease in vibration for the inboard bearing when lubrication was flowing properly.
Even though the root cause was identified, implementing the changes to the lubrication system was a lengthy process, and the machine was required to operate in the interim. Although the bearing had to be replaced twice during the first six months, the CM system proved useful in scheduling these replacements, allowing the plant to monitor bearing degradation closely and intervene at the right times before catastrophic bearing failure and collateral machine damage occurred. Outages could be planned, allowing the bearing change-outs to be performed when impact to production was minimized.
Case History #2
Problem: Resonance
Machine: Lime Kiln Driver
Unit: Lime Kiln
The lime kiln is a large machine with slow rotational speeds (as low as 5 rpm). Two drivers provide rotational power. Depending on production conditions, the kiln must run at different operating speeds. When the kiln ran at higher speeds, higher vibration levels were noted, occurring predominately at 2X. This led plant personnel to initially conclude it was an alignment problem, but realignment of the drivers did not correct the situation. The data was re-examined, this time by looking at phase and rpm levels in addition to amplitude and frequency (Figure 2 and Table 1).
Table 1
Table 2
Figure 2 is an Amplitude/Phase/Time (APHT) plot where the horizontal axis is time. If the machine speed is changing markedly with time, an APHT plot can show a resonance response. The classic features of resonance are two-fold: First, the filtered (1X, 2X, etc.) amplitude will increase to a maximum at a rotational speed that excites the resonance, and then will decrease as the machine speed goes above this frequency. Second, the phase lag will undergo a 180 deg shift, generally passing through approximately 90 deg at the point of resonance.
While Figure 2 does not label each individual data point with its corresponding rpm, the CM software can provide this information as a tabular output. The points clustered between 10 a.m. and 2 p.m. on the plot illustrated the most dramatic shifts in amplitude and phase, and coincided with a change on the kiln from low-speed operation to high-speed operation and back again. A tabular output of the data points in Figure 2 was generated, and a subset of this data is summarized in Table 3, clearly showing the correlation between amplitude/phase changes and running speed, and helping to confirm a structural resonance at approximately 36 Hz.
During subsequent maintenance on the unit, the supports for the drivers were stiffened and strengthened, raising the resonant frequency of the structure and eliminating the vibration problems.
Case History #3
Problem: Faulty Coupling
Machine: White Liquor Pump
Unit: Recausticizing
At Wisaforest, 15 pumps were connected to the CM system. Increased vibration levels were noted on the motor driving the white liquor pump. Further analysis revealed that the rubber element in the coupling had deteriorated, allowing metal-to-metal impacting. The coupling was repaired and vibration levels returned to normal.
Case History #4
Problem: Bearing Deterioration
Machine: Mixer Adjacent to Rotary Filter
Unit: Recausticizing
Rotary filters are one of the most difficult machines to monitor since their rotational speed can be extremely low-as little as 0.5 rpm. The bearings are fitted with accelerometers, and acceleration enveloping is one of the signal processing techniques used to help identify degradation and other anomalies.
The CM group noticed increased vibration levels on the filter's inboard bearing, observable in both the enveloped amplitude and the high-frequency amplitude trends (Figure 3).
The enveloped acceleration time base (Figure 4) clearly showed evidence of periodic impacting, but examination of the spectral components did not yield any frequencies corresponding to the bearing geometries or rotative speeds of the filter or its gearbox.
A visual examination of the filter gave the reason: it was not the filter at all, but rather a separate mixer, located below the filter, with a damaged bearing. Although the mixer was totally unmonitored, the impact vibrations occurring from its faulty bearing were being mechanically coupled into the accelerometer on the filter bearing, located nearby. The root cause was found to be broken lubrication piping feeding the mixer bearing, which was subsequently repaired. However, the bearing had been irreparably damaged and needed to be replaced.
This case history is particularly noteworthy in that it demonstrated the sensitivity of the monitoring system to detect changes in unmonitored machinery, which was an unexpected, and valuable, benefit.
Payback
For Wisaforest, the CM system translated to economic benefits, which exceeded expectations. Several of the machines highlighted in these case histories have critical roles and represent substantial lost production costs if they go offline. Plant personnel have avoided several total outages, resulting in an estimated payback time for their investment of just eight weeks.
The maintenance and production departments view the system as more than just a tool for strengthening preventive maintenance capabilities; it is viewed as a tool for increasing the mill's total productivity. Strong credibility has been established with management regarding the value of condition monitoring and its role in ensuring plant uptime, leading the plant to consider expanding the system to additional equipment.
Summary
Wisaforest is achieving ongoing success with their CM system for several reasons:
-
Proven, quality technology from a knowledgeable supplier was chosen as the basis for the plant's CM program.
-
The plant enlisted the assistance of the supplier to help install and implement the technology correctly.
-
Adequate transducers were installed where feasible, including speed/phase, rather than just vibration, allowing confirmation of faults that would have been difficult or impossible to isolate when limited to only amplitude and frequency data (e.g., Case History #2).
-
Start-up activities were coordinated to include the CM system as "must have" capabilities before a machine was brought online.
-
Plant anagement made certain that everyone understood the CM program's goals and objectives, and that there was buy-in from all parties. This helped ensure that the system would be used proactively and consistently.
-
Results were documented, allowing the users to quantify the system's value to management and other stakeholders in the plant.
Consequently, the WISA 800 REC project has led to not only a world-class facility, but also world-class asset management practices and results.
Pumps & Systems, October 2009