Predictive Maintenance

Predictive maintenance is a new and important trend. Predictive, plannable maintenance is possible in conjunction with continuous plant monitoring (condition monitoring) to forecast the wear and condition of plant and machinery. Malfunctions can then be detected early, system availability increased and maintenance costs reduced.

Condition monitoring is already a routine requirement for plant and machinery manufacturers and operators. For this purpose, sensor data is continuously collected, stored in an intelligent database system and evaluated in real time. Older generations of plant and machinery with a diversity of interfaces can be retrofitted and connected to powerful data loggers such as the Expert Logger.

• Vibration Monitoring: Gears, bearings and shafts, e.g. in generators and turbines, are systematically monitored and alarms automatically triggered. Shaft vibration analysis can also be used to monitor rotating shafts equipped with journal bearings. Expert Vibro can perform frequency analysis on even very large volumes of data and is therefore capable of determining the causes of unusual vibration occurrences.

• Online processing of high-resolution long-term data: Long-term monitoring of vibration and process data requires the acquisition of very large volumes of data which then requires storage and instant online evaluation. Intelligent data acquisition devices with software channels (virtual channels) enable the creation and evaluation of high-resolution frequency spectra (FFT) that provide vital information on the condition of plant and machinery.

• Linking vibration and process data: Delphin systems provide data acquisition at 24-bit resolution and sampling rates of up to 50kHz per channel and are therefore capable of simultaneously recording and intelligently evaluating vibration and process measurement data, e.g. temperatures, in a single system. Faster and more reliable diagnosis is thereby provided.

• Connection to control technology: Interfaces such as Profibus, Modbus TCP or OPC UA are available to transmit data. The monitoring of plant and machinery can be integrated into existing process control technology when required.

Gradual implementation of predictive maintenance

• Analysing complex relationships: For predictive maintenance, highly complex relationships in large volumes of data need to be analysed. Algorithms can reliably identify relationships and patterns. This requires in-depth knowledge of the application and metrology expertise.

• Extending condition monitoring: An existing system of plant and machinery monitoring can be gradually extended to incorporate predictive maintenance. As a first step, pre-engineering by Delphin analyses evaluation options and serves as a basis for the development of an application, which Delphin can also supply as a single provider.