Pipeline monitoring using radar sensors for the detection of oil leaks

Uche, Elijah Oko (2025). Pipeline monitoring using radar sensors for the detection of oil leaks. University of Birmingham. Ph.D.

Uche2025PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (10MB)

Abstract

This research explores the use of radar sensors for real-time remote monitoring of oil and gas pipeline infrastructure, enabling the detection of oil leaks, spills and unusual activities while ensuring prompt escalation to authorities for early response. This helped to evade a flooding or spillage occurrence and thereby curtailed the wastage attributed to issues of pipeline spillage/leaks. The research also focused on the theoretical development and simulation of rough surface models with a view to understanding and analysing the reflectivity of various rough surfaces and how this reflectivity varies, when the dielectric properties of these rough surfaces are modified. The modification amongst other parameters can be due to oil spills and leaks. In particular, the radar cross section (RCS) of various soil samples (with oil and without oil) was analysed. The varied reflectivity results from these soil samples enabled us to understand and appreciate the difference between a soil that is soaked with oil and another that is not. The oil in the soil sample soaked with oil could be due to spills and leaks from oil pipelines, so an understanding of these differences in reflectivities was utilised in improving the detection and identification of oil spills/leaks in pipeline infrastructure which is the major question this research addresses. The radar parameters modified in this project for the RCS analysis of varied soil samples include frequency, aspect angle, polarisation as well as the dielectric properties of the material with the obtained results extensively analysed and discussed. After showing and analysing the simulation results of the rough surface model, an experimental investigation of this RCS analysis was also conducted as means of validation of the theoretical / analytical results. For the theoretical and analytical phase, rough surface models which can be likened to various soil surfaces on which these pipeline networks are laid upon were developed in MATLAB and imported into CST for the simulation of the radar cross section (RCS) / reflectivity analyses under various dielectric properties and parameters. The validity of the rough surface models was tested through simulations with varied soil samples and dielectric properties. In the simulations, the impact of oil as well as other dielectrics on the reflectivity or RCS analysis of various soil samples was presented for the first time. Furthermore, an in-depth analysis was conducted on the measurement parameters, their relationships with dielectric constants, and their impact on the reflectivity of various soil samples. This has given us an idea of what the reflectivity of oil, emanating from the pipelines because of oil spills / leaks, could be when, for example, the pipeline infrastructure is laid on a dry or wet soil, such that my radar sensor is able to discriminate between these oil spills/leaks from other obscurants like water etc which might have seeped into or floats on the soil under investigation. In addition, the results of the rough surface models have been compared with literature. The validity of the developed theoretical rough surface models was tested and applied to experimental, proof-of-concept data in a controlled environment using radar instruments and the experimental results, which were thoroughly analysed and extensively presented for the first time, agreed with the analytical or theoretical results. The results of the theoretical or analytical RCS of the rough surface models were compared with the experimental results and the results have been presented. The results show that there is 10 dB difference in RCS from a 14 cm rough surface without oil and a similar surface that has been soaked with at least 7.5 L of oil. This implies that a litre of oil results in at least 1 dB change in RCS from a rough surface of 14 cm rms height. In addition, there is about a 4 dB difference for that of 7 cm surface roughness for a 9L oil that leaks onto the rough surface. This also implies that 2 L of oil results in a dB change in reflectivity from a rough surface of 7 cm rms height. This 4 dB difference is significant because it is about half the difference obtained for the 14cm rough surface profile and the 14 cm rough surface profile is double the 7 cm rough surface profile. The results obtained via experiment were also verified via simulations and both results tally and align with each other as have been extensively discussed and analysed. Distributed and back scattering as well as back reflections from the sandy soil was also investigated during the experiments and the results have been thoroughly analysed. Finally, the rough surface model was integrated with the radar system and the parameters measured from this system were thoroughly examined and extensively analysed. In addition, the performance of the developed theoretical and experimental models and systems was compared to specific requirements by the oil and gas industry, successfully meeting expectations in line with global best practices. The methods used, theoretical and experimental results obtained as well as the key findings and research output have been analysed, validated and presented as appropriate.

In summary, this research successfully developed and validated a radar-based remote monitoring system for oil and gas pipeline infrastructure, enabling real-time detection of leaks, spills, and anomalous activities. A key achievement was the theoretical development and experimental validation of rough surface models to analyse radar cross-section (RCS) variations in soil samples with and without oil contamination. The study demonstrated, for the first time, the impact of dielectric properties on soil reflectivity, showing a measurable correlation between oil volume and RCS variation – such as a 10 dB difference for a 14 cm rough surface soaked with 7.5 L of oil. These findings enhance the ability to discriminate oil spills from other environmental factors, improving detection accuracy. The results, validated through simulations and experiments, align with industry standards, positioning this work as a novel contribution towards enhancing oil spill detection and response in pipeline monitoring systems

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):