Big Data against gas leaks

Gas leaks in the natural gas industry usually have severe consequences. Gas is an important energy source and will become even more important as a substitute for coal in the medium to long term. However, the transport and especially the fine distribution of natural gas to households is a technical challenge and has a significant impact on public safety and the environmental compatibility of this fossil energy source.

Leaks in gas pipelines occur frequently and mean damage to the infrastructure which, due to their location in often densely populated conurbations, can only be found with a correspondingly large amount of effort. The odorization of natural gas began in the 1950s. Since then, gas suppliers have been able to count on their own customers to detect leaks by "smelling" them and alerting operators to a gas leak.

Technological developments over the last thirty years have led to measuring instruments that can detect methane as the main component of natural gas. In many places, gas pipelines are periodically walked - on the one hand to meet regulatory requirements, and on the other because of the difficulty of detecting underground leaks. This is very cost-intensive and the gas detectors used only allow very limited spatial monitoring of the gas network.

Conventional gas leak detection requires a measurement directly above the affected gas pipeline, which is usually located under a road or the pavement. The efficiency of this method, which depends on the position of the gas pipeline, is highly dependent on the experience of the inspector and the local conditions of the natural gas network. In the Anglo-Saxon world, this method of leak detection is called the "asset-based method" - the "asset" being the natural gas pipelines of the gas distributor.

Gas leak detection: from asset based to area based

Better and more efficient methods are needed to reduce the large number of odor calls. Firstly, such a technology should have the advantage of a mobile platform in order to be able to check the hundreds of kilometres of gas pipelines efficiently and at regular intervals. Secondly, this mobile platform must contain a much more sensitive detector than conventional measuring instruments in order to be able to locate leaks that are not in the immediate vicinity of the measuring probe. The last point rules out a method based on the position of the gas pipeline (asset-based), since gas networks are never fully accessible.

It is the aforementioned disadvantages of asset based leak detection that the Picarro Surveyor significantly improves, revolutionizing the search for gas leaks. The Surveyor offers a method that no longer depends on the position of the detector in relation to the gas network, but focuses on the entire gas infrastructure within an area. This area-based approach makes it possible to find gas leaks that are not directly accessible with a conventional measuring device.

Focusing gas leak detection on an area rather than the immediate position of the gas pipeline allows leaks to be located even if the geographical position of the pipeline is not precisely known or the pipeline is not accessible due to structural conditions.

In area-based leak detection, a zone of the gas infrastructure is systematically traversed, with the part monitored by the surveyor being recorded in real time on a map (Google Map). Unlike conventional measuring instruments, the wind direction and speed are also measured to show the part of the gas network actually covered. Depending on the geographical characteristics and wind strength, access lines and difficult-to-access domestic installations are also recorded.

The method only requires a systematic driving style within the selected zone. It is therefore much less demanding than a leak detection that has to take place directly above the affected gas pipeline. This reduces the differences that arise due to different working methods and experience of the inspectors.

Adopting an area-based method to detect leaks from a distance requires an accurate understanding of local transport conditions in the atmosphere and computer algorithms that can distinguish the unique signature of natural gas leaks from those of other methane sources. This is a critical factor in avoiding false positive indications while still ensuring efficient and complete leak detection.

Gas network operators who have integrated the area based leak detection method into their organisation can carry out the monitoring of the infrastructure much more efficiently and thus increase the frequency of leak detection, which in the longer term contributes to an improvement in the safety and sustainability of the gas network.

"Big Data" in gas infrastructure management

The area-based leak detection method contains a wealth of data that goes far beyond the actual gas measurements. In addition to regulatory control of the infrastructure, it can also be used for other applications.

The Picarro Surveyor Method offers a response to a changing and increasingly networked market environment (Internet of Things), which among other things also requires the processing and secure handling of very large amounts of data (Big Data). The Pcubed platform collects the gas concentration measurements, wind and weather information and integrates the geographical information (GIS) about the gas infrastructure on a cloud-based platform.

In addition to regulatory monitoring of the gas infrastructure, this comprehensive data on Pcubed can also be used for further applications in the maintenance of a gas network operator. Based on the analysis of many thousands of gas leaks detected in over one million inspection runs since 2014, a model for the risk weighting of gas leaks has been developed in cooperation with a large US gas network operator. This analytical software evaluates the dangerousness of a gas leak and assigns it a ranking based on the calculated risk potential, which serves as a basis for planning further work processes (localization).

This cloud-based smart analytics software assigns a priority level to detected gas leaks based on their risk potential, so that maintenance resources can be deployed, for example, in pipeline sections where leaks occur frequently. In addition to recording the number of gas leaks within a network section, the analytical software can also be programmed to prioritize leaks with the greatest emissions. This setting of risk-based leak detection helps to reduce the risk of an accident and can also be used as an additional application for mitigating greenhouse gas emissions.

The efficiency of area-based leak detection in combination with smart data processing allows two to three times more potentially hazardous gas leaks to be detected compared to conventional methods. Due to the increased efficiency, the share of costs for leak detection (compliance survey) decreases and allows (with the same budget) more funds to be used for repairs and maintenance. These investments in infrastructure increase the quality and safety of the gas network, which in the longer term leads to a reduction in dangerous gas leaks and the associated odor calls. (Source: AQUA & GAS/ "Wasserspiegel")