If the refiner plates contact each other during the refining process, significant damage to the plates and the motor could occur. Katahdin needed a plan so they could avoid potential loss of life or serious injury that could occur if the plates were to catastrophically fail, not to mention their desire to avoid the costly damages involved, potentially ranging up to $100,000 for replacement of damaged equipment. Their final concept utilized vibration sensors mounted to the machine to measure any such events. They determined that using overall vibration signals, and converting them to a 4-20 mA output signal, was the best course of action to prevent such damages.
Why Use 4-20 mA Signals
Today’s modern systems offer flexibility in sensor selection, and use standard 4-20 mA current loops for most applications. Process control provides a wide variety of monitoring options, time based trending, and control applications to keep machines performing efficiently and running at their required capabilities. 4-20 mA current loops are inherently low in noise and signals can be transmitted over long distances, making an ideal combination for industrial applications. Sensor outputs converted through vibration signal conditioners are proportional to current with 4 mA representing a zero level, and 20 mA representing a maximum level over a given range. The 4-20 mA signals are integrated into a controller system, which can be set up to alert operators if refiner plate contact occurs, and can also be integrated into a shutdown system, which will turn off the equipment prior to catastrophic damage.
Vibration Considerations & Mounting Locations
There were two vibration sources of interest on Katahdin’s refiners: the inboard bearing, and the outboard housing. Inboard Bearing – The inboard bearing was selected for monitoring because unbalance was the primary fault that could be detected. It is a low speed (600 RPM), low frequency bearing point, with velocity being the main measurement type, and 1x running speed measurements of primary interest. A low frequency, 500 mV/g accelerometer was selected for mounting to the bearing housing (see Figure 2).
Outboard Housing – For the outboard measurement, axial measurements were the focus, with particular interest in detecting when the refiner plates were starting to make contact. A general, multi-purpose 100 mV/g sensor was selected for this measurement point. What made this measurement unique is the location – the accelerometer was mounted on the housing around the plates (Figure 3), and acceleration was measured in order to detect the force of the clashing plates.
Monitoring & Alarming Considerations
The ability to access the raw vibration data was very important. The use of the vibration signal conditioners permitted easy access to that vibration data, while providing a 4-20 mA analog output to the controller system in the plant, which allowed for continuous monitoring. A signal conditioner enclosure was utilized to consolidate multiple signal conditioners into one NEMA 4X enclosure that allowed for easy access to the vibration data, as well as a protected environment for the signal conditioners. Further consideration was given to adding a junction box to the system to allow for the easy collection of time waveform data from the sensors without needing to access the signal conditioner enclosure.
Based on historical vibration data from previous trending and analysis, alarm levels were set for the equipment being monitored. The alarms were set up for an increase of up to 50% of overall vibration from a baseline vibration signal based on the historical vibration data. The use of historical data was very important in identifying these alarm points. Too low of an alarm would cause either the shutdown or the unnecessary analysis of the equipment; whereas too high of an alarm might not alert the operators or maintenance staff that there was a potential issue. Because the signal is displayed on the main floor display, the alarm levels are also accessible to the operators, alerting them to any vibration issues.
Sensors – Sensor selection was determined by type of equipment being monitored - accelerometers were used for the selections. For the inboard bearing location, where the running speed was very low, a low frequency, side exit 500 mV/g accelerometer was selected for mounting to the bearing housing. For the outboard housing, a standard side exit 100 mV/g accelerometer was selected.
Cabling – Due to the environment in which the refiners are located, the cable connecting the sensors to the signal conditioner enclosure needed to be robust, chemical resistant, moisture resistant and reliable. A low cost composite connector with a silicone o-ring and threaded locking ring provides the seal required to protect against the environment. Due to its low cost and high performance, a flexible, Teflon jacketed twisted shielded cable was chosen to carry the signal from the accelerometer to the enclosure.
Enclosure – A junction box provides convenient, direct access to the vibration data for the analyst, as well as a protected housing for the signal conditioners (see Figure 5 and 6). A NEMA 4X enclosure with water-tight cable entry into the enclosure was recommended to ensure that the water would not collect inside the enclosure. Additionally, a secondary junction box was utilized for ease of access to the vibration data (see Figures 7, 8 and 9). This also allowed the signal conditioner junction box cover to remain closed at all times, reducing the effects of environmental issues inside of the enclosure.
Signal conditioners can also be used in conjunction with standard dynamic accelerometers, piezo velocity sensors, or displacement probes. The signal conditioner accepts the dynamic input and converts it to a proportional 4-20 mA output for the PLC, DCS or SCADA system. This type of application has many benefits. The signal conditioner can be adjusted in the field so that the scaling and filters match the application. The dynamic vibration signal is available via the standard BNC connection on the front of the signal conditioner, or as an optional output from the terminal block.
Modern control schemes like the PLC, DCS, and SCADA systems integrating multiple sensors, inputs, and outputs in operations centers, offer the flexibility in sensor selection. For the Katahdin Mill, a controller system was utilized to take the 4-20 mA analog output from the signal converter in order to fulfill the monitoring and alarming requirement.
Katahdin Paper Company utilizes 4-20mA solutions to protect equipment against catastrophic failures by incorporating the outputs into their controls systems to shut down high speed, critical equipment if there is a significant vibration issue or event.
If potential problems are identified through process monitoring, the fault can be investigated and addressed in more detail. Vibration analysts can access the dynamic signal for detailed analysis using industry standard vibration analyzers and portable accelerometers, or in conjunction with the dual output loop power sensors, or signal conditioners, the analyst can use the dynamic vibration signal available to them.
Ed Nisbett is a vibration analyst that has 24 years of experience in the predictive maintenance field. Ed has attended numerous vibration courses presented by SKF, and has been on the forefront of many product improvements, developments and releases.
Tim Gilliss is a process engineer at Katahdin Paper Company, and is responsible for the implementation of the 4-20mA output signals into Katahdin’s control system.
Tom LaRocque is the Engineering Manager, for Connection Technology Center, Inc. He has been involved in the design, manufacturing and quality of vibration analysis hardware for 11 years, and is currently the research and design engineering manager at CTC. He is a Certified Vibration Analyst Category III from the Vibration Institute, and graduated with a Bachelors of Science degree in Engineering from Clarkson University. Tom can be reached at 585-924-5900 Ext. 817 or firstname.lastname@example.org
“R.A.I.” the Reliability.aiTMChatbot
You can ask "R.A.I." anything about maintenance, reliability, and asset management.