Whether it’s new builds such as kindergartens, hospitals or loop line construction projects: whenever specific sections of a pipeline have to be monitored either temporarily or permanently to detect potential third-party damage, PipeMon+ comes into play.
The system uses cathodic protection (CP) test stations already installed along the pipelines. At these CP test stations, PipeMon+ uses sensors which, if they are employed on a permanent basis, require an off-grid power supply which is provided by marker posts equipped with small solar panels. The sensors send their data to software in the cloud, which analyses it using artificial intelligence (AI). This makes PipeMon+ a solution that is also very well suited for temporary use. The system can be installed and removed without any additional civil engineering work being required.
Once PipeMon+ is installed, the AI is trained for its purpose – and then put into operation.
In 2005, our technical department thought about how best to detect third-party damage to our pipelines as quickly as possible.
I had the idea of using the electrical current applied to the pipeline for cathodic protection (CP) for this purpose. In the event of a fault between two CP test stations, the system would generate a differential value as a fault signal. Initial tests at an institute in Freiberg confirmed the feasibility of this approach. However, the instrumentation available at the time was not suitable for field use. It was not until about seven years later that the sensor technology was ready.
With cathodic protection (CP), OGE provides electrochemical protection around the pipe. This direct current-based protection prevents metal dissolution on the surface. Because the current detects coating holidays, the sensors can show exactly where the damage has been caused. PipeMon+ processes the measured data from the sensors and triggers an alarm when a defect is reported.
Not all construction activities are announced beforehand. Quite often, the excavator operator does not even know that there is a pipeline in the ground, or the exact location of the pipes or cabling is not known. If, for example, a buried pipe was installed several decades ago, it is quite possible for something to be damaged accidentally during excavation work. Such damage is not always reported by the parties responsible.
In the sections of the pipeline network monitored by PipeMon+, no real damage has been reported so far.
The signals that are measured are erratic, for example if there is interference from other cathodic protection systems in the vicinity. Motorways or railway lines in the vicinity of a pipeline can also change the signals. If the measured signal exceeds a specific limit, an alarm is triggered. AI is used to filter out false alarms. At OGE, the AI system is trained to distinguish false alarms from real alarms more and more effectively with deliberately generated false measurements.
With PipeMon+, there is no one single AI. Each project, each pipe, each monitored section gets its own AI. The material characteristics and age of the pipelines as well as the soil conditions vary from site to site. The AI has to be adjusted and set up exactly as needed for the conditions on site. So, before PipeMon+ can start measuring, the respective AI is trained and fed with data. During PipeMon+ operation, OGE regularly provides feedback to the AI and corrects it. This largely automated process also takes into account, for example, changing temperature conditions that influence the AI.
PipeMon+ can currently map up to 600 sensors and collect and process their data. Up to 2.2 terabytes of data are collected from 600 sensors in 14 days alone. Performance, scalability and data analysis are made possible thanks to a cloud-native architecture and detection software set up in cooperation with Microsoft.
PipeMon+ measures pipe flows in much the same way that a cardiologist measures cardiac current flows during an ECG. But despite the laboratory conditions under which medical ECGs are made, the measuring device used in PipeMon+ is more accurate in harsh construction site environments. The currents on buried pipelines are recorded about 1,000 times more precisely than the cardiac currents on an ECG. It’s not obvious at first glance what is responsible for this outstanding measuring performance, but it is the MiniTransPlus receiver, which is well hidden in the typical yellow measuring signposts near the pipelines.
PipeMon+ is an example of how the benefits of digitalisation are used in practice at OGE. The existing infrastructure is modernised in a future-proof way using digital tools such as artificial intelligence – with noticeable added value for safety and the general public.
Patrick Krzyzak
