Performance of mobile survey solutions for natural gas pipeline leaks under different soil type and moisture conditions

Abstract

Advanced leak detection (ALD) solutions for natural gas (NG) play a crucial role in swiftly identifying leaks to mitigate risks, costs, and methane (CH₄) emissions. Despite recent advancements in CH₄ leak detection, current ALD solutions often overlook the influence of subsurface and atmospheric conditions on gas behavior and, therefore, leak detection success. To address this gap, we conducted a series of controlled leak detection experiments under varying soil moisture levels and soil types, with two leak rates: low (0.5 slpm) and high (10 slpm), using walking, driving, and simulated UAV (UAV_sim) surveys. The performance of these methods was evaluated using the probability of detection (POD). Results indicated that the POD of walking, driving, and UAV_sim surveys is highly influenced by soil moisture and type. Low moisture conditions have a 10–45 % higher POD than high moisture conditions at higher leak rates, depending on the soil type and survey method. Higher permeability soils result in a 10–18 % higher POD than the lower-permeability soil tested within 5 m of the pipeline centerline. As the distance from the centerline increases, the magnitude of the impact grows, resulting in PODs 40–60 % lower than those along the centerline itself. Similarly, at low leak rates, the impact of high moisture and low permeability becomes more pronounced. Based on averages across all soil types and survey deployment platforms, the POD under dry soil conditions at a high leak rate is 20–40 % higher than at a low leak rate. Results further demonstrate the influence of the detection threshold, especially for low leak rates, where a 0.1 ppm increase in the detection threshold can reduce the POD by 50–60 % under moist soil conditions. These findings underscore the importance of considering the interplay between soil and operational conditions to enhance CH₄ leak detection success. Emission measurement solutions can therefore use this knowledge to establish performance metrics that account for variations in soil conditions.