Inspecting the pipelines that carry oil and gas over thousands of km has been a monumental task frought with error - until now.
By Zheng Kai, tech journalist
While sensing technologies deployed on oil and gas pipelines aren’t new, they tend to be plagued by issues such as false positives, false negatives, and misidentification. Huawei has taken the optical technologies it uses in fiber communications and applied fiber sensing, resulting in the Huawei OptiXsense EF3000 — Huawei's first intelligent optical sensing product.
Petroleum is one of the most important economic resources in the world today. However, uneven geographical distribution and the physical properties of petroleum mean that transporting the resource to where it’s needed presents a tough logistical challenge.
Oil and gas are typically transported by rail, road, water, air, or pipelines. Of these, pipes are the least restrictive method. However, inspections are notoriously difficult on pipe networks that can traverse tens of thousands of kilometers. A 200-km to 300-km stretch of pipeline typically requires a dedicated team of three inspectors in one vehicle, assuming navigable terrain free of obstacles like reservoirs, lakes, forests, and steep slopes.
At HUAWEI CONNECT 2021, David Wang, Executive Director of the Board and President of ICT Products & Solutions, said that digital infrastructure of the future would need to be hyper secure, reliable, and deterministic, as well as equipped with efficient data circulation and computing power capabilities. Huawei has taken the optical technologies that it usually uses in fiber communications, applied them to fiber sensing, and built the Huawei OptiXsense EF3000, Huawei's first intelligent optical sensing product.
The challenge of inspecting 90,000 km of pipeline
China's oil and gas resources are mainly concentrated in the northeast and northwest of the country, but the major fuel consumers are located in the cities of the southeast coast and south-central area. Oil and gas must be transported over vast distances, with pipeline transportation providing the best solution, but one that presents the most challenges when it comes to inspections.
China Oil & Gas Pipeline Network Corporation (PipeChina) builds and operates various infrastructure, including oil and gas trunk pipeline networks as well as gas storage and peak shaving. It also works on interconnecting trunk pipeline networks and docking with local pipelines, operating and scheduling nationwide networks, and ensuring fair and open user access to the infrastructure. As of 2021, the total length of PipeChina's oil and gas pipelines exceeded 90,000 km, and by 2025, the total length is expected to exceed 120,000 km.
PipeChina employs a pipeline inspection team of about 15,000 people. However, as the pipelines become longer, inspection management and operations are becoming increasingly difficult.
To address this problem, PipeChina plans to combine both manual and automatic pipeline inspections and use digital technologies to automate the entire inspection process, from detection and analysis to decision-making and warnings. Zero false negatives and minimal false alarms are expected, and reliance on manual inspections will be gradually reduced. Other problems will also be solved such as losses resulting from issues like inadequate repairs to damage caused by other construction projects, inefficient pipeline management methods, and poor accuracy of early warnings.
Yu Hao, a senior expert at PipeChina's Digitalization Department, said, "Huawei has been actively involved in the pipeline industry and has innovated technologies such as optical sensing and big data. These technologies have shown us the possibility of automated unmanned inspection of pipelines. We hope that Huawei continues to invest in this area to contribute to ensuring safe, reliable, and efficient pipeline transportation in China."
Three core technological capabilities
Some digital technologies have already been applied to pipeline inspections such as sensor technologies. However, problems still remain such as false negatives where damage goes undetected, damage is detected where there is none, or misidentification occurs where construction is mistakenly categorized.
Some technologies, such as aerial drones used for inspecting power grids, are ill-suited to oil and gas pipeline scenarios, because most pipelines are built underground.
Huawei’s OptiXsense is an optical-fiber vibration sensor with best-in-class identification accuracy. Three major capabilities set the OptiXsense EF3000 apart from rival products:
Signal collection: Huawei's enhanced oDSP module has a strong, built in correction algorithm for blind spots, which can correct and shape the phase of weak signals and raise the effective signal acquisition rate to 99.9%.
Accurate identification of incidents: Huawei's exclusive vibration ripple identification engine can analyze construction incidents across multiple dimensions. For each construction vibration point, a minimum of 32 pieces of phase information are obtained. Multiple features can be extracted, including voiceprint, frequency, space, time sequence, and duration. Samples are identified and compared through multi-dimensional deep convolution, raising incident identification accuracy to 97%.
Fast learning: Huawei OptiXsense EF3000 can perform iteration based on new construction behavior data and different geological environment scenarios in the database. Huawei works with universities that provide large amounts of geological data – up to 1,000 new incident samples each day. Huawei uses this data to improve the sensing and warning accuracy of its OptiXsense products.
With these three technical characteristics, we believe that OptiXsense products can help digitalize and automate the E2E inspection process with zero false negatives and minimal false alarms.
All-optical innovation: Technology and use cases
Huawei has drawn on its 30 years of cumulative expertise in optical technology to apply all-optical sensing technology to the pain points that have existed in oil and gas pipeline transport for years.
OptiXsense EF3000 marks Huawei's first foray into fiber sensing beyond communications networks. We plan to develop additional optical technologies, such as the distributed sensing of optical fiber vibration, temperature, stress, and water quality, for many more industries, including electric power, transportation, government, and sanitation. In conjunction with big data and GIS mapping, these technologies will support differentiated, multi-dimensional, and intelligent detection and warning solutions.
Shandong Jihua Gas was one of the first companies to adopt the OptiXsense EF3000. The company operates gas pipeline networks at various levels, comprising more than 3,000 km of piping that supplies gas to 920,000 households and more than 3,000 industrial, boiler, and public welfare users in Jinan city. To ensure that the fiber sensing and adaptive AI identification solutions can be adopted in pipeline construction, Huawei has worked with Shandong Jihua Gas to deploy a 20-km underground pipeline that traverses complex environments such as suburbs, national highways, and rural areas. Construction samples from excavators, rammers, ditchers, and manual excavation were collected onsite for training and model creation. Based on these, 56 tests were conducted on different road segments. The initial phase of technical verification has been completed, proving the efficacy of Huawei's first intelligent optical sensing product.
As major energy arteries of the world economy, oil and gas pipelines will inevitably become digitalized. And in the future, fiber sensing technologies will continue to evolve, benefiting more industries by addressing particular requirements.
