Project

COMPRESS

Condition Monitoring for Predictive Maintenance Adapted to Geothermal Electric Submersible Pumps

DigitalisationEnergy

Reliable feed pump technology is a basic requirement for the development of deep geothermal energy—especially in the context of the planned project to convert the existing district heating projects in the Ruhr Metropolis. The objective of COMPRESS is to minimise the immense costs generated by the frequent substitution of pumps, which in turn lead to long downtimes in the production facilities.

Starting Point

Reliable feed pump technology is a basic requirement for the development of deep geothermal energy—especially in the context of the planned project to convert the existing district heating projects in the Ruhr Metropolis. Due to the environmental conditions in which these pumps operate, their efficiency and durability are severely reduced through increased wear and formation of deposits, for example. Additionally, the numerous breakdowns in the sensor technology diminish their efficiency. Consequently, it is imperative to carry out scientific studies of computer-aided optimisation of maintenance intervals and improvements in the sensor technology of deep geothermal energy.

Our Solution

The objective of COMPRESS is to minimise the immense costs generated by the frequent substitution of pumps, which in turn lead to long downtimes in the production facilities. The specified sources of errors should be reduced through monitoring of the operating pumps in combination with computer-based prediction models for planning optimised maintenance intervals. To that end, it is essential to characterise the relevant operating conditions and wear parts of the submersible pumps and to monitor the operation of the individual pump components using sensors or through the analysis of the operating data. The prevalent environmental conditions present great challenges to the hardware, and for this reason fibre-optic sensors are needed. The innovative core of this project is the technical implementation of intelligent pump monitoring connected to a condition monitoring system, which delivers statistical predictions about the condition of a well pump using machine-learning technology.