A growing number of companies are moving toward condition-based maintenance (CBM) on assets such as motors and pumps. This concept is based on the idea of matching maintenance to the state of the equipment, rather than time based preventive maintenance tasks. When the maintenance manager has insight into the real-time condition of a machine, he can adjust his maintenance strategy optimally.
Sensors collect data, which is used to determine the condition of the assets. For example, velocity sensors or accelerometers measure vibration. For each application there are several commercially available sensor options - some more robust or accurate than others. One common drawback with sensors, however: they must be mounted on a machine to make measurements.
Alternatively, a new method to determine machine condition is by measuring a high frequency on the cables of motors or pumps. A deviation in the power supply can indicate a mechanical or electronic problem to the asset itself. For example, vibrations disrupt the electromagnetic field in an electric motor, which can be read in the data that generates the current. Every failure mode has a specific signature or fingerprint that can be measured in the current or voltage.
This alternate method of measurement still drives toward the goal of condition monitoring. A large difference with other “traditional” methods is where the measurement takes place. The current measurement does not have to be determined at the asset; it can take place in the control cabinet. Installing sensor modules in a control cabinet (rather than in the field) can provide many advantages.
Field Sensors Are Subject to External Factors
Sensors in the field are subject to local conditions. In the food sector, for example, strict hygiene and quality requirements apply. Rooms, surfaces and materials are cleaned often. The equipment and the sensors must be resistant to water or high humidity.
In other sectors, external factors such as extremely high or low temperatures, pressures, or contaminants may be present. This can lead to problems when sensors are not sufficiently robust. A defective sensor does not provide reliable data. Sensor modules mounted in a control cabinet are in a stable, conditioned, dry room - ideal conditions for sensors to do their job.
Field Sensors Need Energy Sources
To collect and send information, sensors must be provided with an energy source. In the past, this was done via cables. However, laying cables is a time-consuming and expensive affair. In recent years, wireless sensors have gained a lot of ground. These sensors usually run on batteries. Depending on the type of sensor and the application, a battery will run out quickly or less quickly. Sending data once a day or every fifty milliseconds is a big difference. When batteries are drained, a mechanic will have to replace them – no battery, no data.
One solution for this problem is energy harvesting, a process in which energy (for example heat) is extracted from the immediate environment to supply the sensor with energy. However, this is currently not yet possible with all sensors or in all situations. With sensor modules placed in a control cabinet, the energy problem disappears. In a control cabinet, a sensor module can easily be connected to the mains.
Field Sensors Have Higher Installation Costs
A third advantage of sensor modules in a control cabinet is in installation. Motors or assets whose condition needs to be measured are often spread over a facility. In a baggage handling system, enormous distances exist between motors of a conveyor belt. Installation of sensors on each individual motor would take a lot of installation time.
Moreover, installing sensors at the right place in the asset is not always easy. Certain sensors must be located very closely to their source, while other sensors must be installed in places that are difficult to reach. Getting to motors integrated in larger machines, or submerged pumps can often prove problematic.
At some locations, there may be a flammable ATEX environment. Sensors here must meet certain ATEX or Classified flammability certifications. The installation of the sensors themselves in this environment also requires extra measures. Installing sensors in the field is therefore often difficult, cumbersome or expensive.
In the case of sensor modules in switch cabinets, the above issues do not play a role. The power supply of several assets comes together in one central location: the switch cabinet. If possible, this is always placed outside a potentially explosive area. It must always be accessible or easily accessible. As a result, installation costs can be reduced.
Furthermore, preparations such as pulling cables can be done ahead of schedule, which means that installations only need be out of operation for a short period of time.
Some have already installed sensor modules in their control cabinets. Two examples are Vopak Vlaardingen, a storage company; and Kaak Group, a manufacturer of machines for the bakery industry.
Marcel Kool, Maintenance Engineer at Vopak:
“Vopak Vlaardingen temporarily stores products from sea-going vessels in storage tanks. The product is then further distributed in trucks or lighter barges. About two hundred pumps take care of the loading at this location. For us it is important to monitor our equipment better so that we can increase service to our customers. We want to increase the predictability of maintenance.”
Vopak Vlaardingen chose not to install sensors on the pumps themselves, but centrally in a control cabinet. Kool: “The pumps are not positioned far from each other, but they are insulated so you can't reach them directly if you want to install a sensor. ATEX was not an issue at our location, but it would be an extra factor to take into account at other locations.”
Installation & Baseline Determination
Vopak started a pilot program on ten pumps. “The installation of the sensor modules in the control cabinet went quite smoothly. No special procedures were required, which resulted in a great deal of installation advantage. The installation of the modules was followed by a period in which the machine learning programmes were worked into and a baseline measurement was performed, a sort of starting position of the pump.
“Almost immediately after this phase, we received two indications based on the data. The dashboard indicated that the pump was almost failing. Mechanics examined these two pumps in the field and what the system indicated appeared to correspond with the findings in the field. This has increased the confidence in the system, which certainly offers perspective for the future.”
Justification for system extension to all 200 pumps may not make sense. “The pumps we selected for the pilot run regularly. We only use a small number of pumps on site sporadically. These pumps would take a little longer to learn, simply because they are used so little. If you let go of a business model on these pumps, the outcome may be that it is not sufficiently profitable to monitor them, whatever sensor technology you choose. For the majority of the pumps, condition monitoring is an option to consider.”
Kaak Group has also chosen to install sensor modules in their control cabinet. Marcel Trapman is team leader at iBakeware, which builds software for monitoring and analysing the bakery line and the baking process:
“Our bakery lines that are with customers consist of a combination of a number of machines. There are critical, large motors in these lines that are built to customer specifications. If such an engine breaks down at a certain moment, this means that the line must be stopped, and a delivery may not be possible. To avoid a long standstill, a spare motor must be stocked. This can be prevented by monitoring motors using sensors. If a discrepancy is found, an inspection can be carried out in time, spare parts can be ordered and maintenance can be scheduled.”
The bakery lines currently contain many different types of sensors and can be equipped with optional sensors at the customer's request. An example of such an optional sensor is that of Semiotic Labs.
“At a later stage we want to be able to give the customer the choice whether he wants to monitor the motors using data and sensors. For us, therefore, the possibility of retrofitting sensors in a fairly simple manner is of great importance. With the sensor module in the control cabinet, we do not need to reach the motor at all. We therefore see it as an added value for the customer to be able to retrofit it fairly easily. That's why we chose this type of system.”