Predictive Maintenance Ready To Go

Wind Turbines using Datumize Intelligent Auto Predict Tool.

Wind turbines are industrial systems controlled by PLC devices (1) and usually collecting and integrating its metrics in SCADA (2), which is being used as a control centre where some business intelligence systems input data from conventional databases. However, are the usual 10 ´ SCADA metrics enough to enhance the use of predictive maintenance?

Currently the most common maintenance models include run to failure / reactive or preventive techniques, despite they may result in e.g wind turbines getting broken or inefficient post-mortem analysis (due to lack of data).

Reactive maintenance is mainly known as the “run it till it breaks” maintenance mode. According to O&M Best Practises Guide (The Federal Energy Management Program's Operations and Maintenance, United States (3)) more than 55% of maintenance programmes are Reactive. This would be the “easy” option due it requires a smaller investment or less staff involved than within other models. However, within the medium and long term cost will increase due to unplanned equipment downtimes, overtime labour costs of repairing/replacing the equipment or inefficient use of the resources.

Preventive maintenance accounts around 31% (O&M Best Practices) and is commonly understood as those actions performed on a time- or machine-run-based schedule that detect, or mitigate degradation of a component or system with the aim of controlling degradation to an acceptable level. This may be cost effective in many capital-intensive processes and flexible to adjust maintenance flexibility along the entire life cycle. On the other hand, it mainly prevents a failure or a replacement, but it does not predict beforehand what is likely to happen, which can still result in catastrophic failures. It also includes unneeded over maintenance or workers performing recurring tasks without any supporting data.

Why to use predictive maintenance:

  1. Increased component operational life/availability.
  2. Allows for preemptive corrective actions.
  3. Decrease in equipment or process downtime.
  4. Decrease in costs for parts and labor.
  5. Better product quality.
  6. Improved worker and environmental safety.
  7. Improved worker morale.
  8. Energy savings.
  9. Estimated 8% to 12% cost savings over preventive maintenance program.

Predictive maintenance strategy is not widely spread within the industry for different reasons:

  1. The number of available metrics is very limited and mainly restricted to those data collected in SCADA.
  2. Big Data and Analytics teams are not yet ready to store more data than SCADA.; Big Data platforms are only available for the “big ones”.
  3. Advanced maintenance predictive models require from algorithmic modelling huge amounts of data.

Datumize helps to move from obsolete preventive and reactive maintenance methods to predictive strategies. How?

  • Datumize capturing all available metrics from the network.
  • Data metrics are processed and filtered in real time.
  • Datumize is not intrusive, so it does not affect SCADA or PLCs.
  • Data is stored as requested in the customer Big Data platform.
  • Data is available to be exploited by the customer predictive maintenance teams.

In conclusion, Our predictive Auto diagnostic tool reduces over maintenance costs by a 3rd and offers clients real time Data at the right time. The Dark Data that is running through the PLC and SCADA is identified, captured and unlocked, then pushed to your analytics dashboard for a full scale visual operation.



written by Adrian Hinrichsen