Tilbake til søkeresultatene

DEMO2000-Prosj.ret tekn.utv. petro.virk

Pipeline Inspection using Underwater Hyperspectral Imager - Qualification Study

Alternativ tittel: Rørlednings inspeksjon ved bruk av hyperspektral avbildning under vann - Kvalifiserings studie

Tildelt: kr 2,9 mill.

Det finnes flere tusen kilometer med olje og gassrørledninger i havet utenfor Norge og andre steder i verden. Disse rørledningene må inspiseres jevnlig for å sikre at de er i god stand og for å bestemme vedlikeholdsbehov. Dagens metoder baserer seg i hovedsak på video og stillbildeinspeksjon utført med ROV, sammen med akustiske undersøkelser. I dette prosjektet er det utviklet et forbedret undervanns hyperspektralt avbildningssystem med programvare som automatisk detekterer forskjellige materialer av interesse langs rørledningene og rapporterer dette. Det er også utviklet software for å oversette materialdeteksjonene til relevante eventer (for eksempel "skade på rørledning, eksponert bart metall") som følger Equinor sin standard for rørinspeksjonsrapportering. Disse automatisk genererte rapportene kan sammenstilles med de manuelt genererte rapportene for å øke den totale kvaliteten på inspeksjonen. Prosjektet har avdekket at det er behov for mer forskning på hvordan oversettelsen mellom materialdeteksjon og inspeksjonsevent kan gjøres for å gjøre resultatene av det automatiske genererte rapportene mer relevante.

The project has provided a new automated tool for pipeline inspection that complements the manual inspection and can contribute to higher quality and lower uncertainty in the inspection reports. It is also a steppingstone to further research and development to achieve even more relevant automated event detection that could bring reduced cost and increased quality to the pipeline inspection workflow. The hardware development done in the project is bringing the UHI technology one step further regarding resolution and light sources. An upgraded UHI based on the project results is already made available in the market and benefits new users of the general UHI technology within other applications such as environmental mapping and monitoring. The custom low-power collimated light source prototypes being tested and improved during the project has now been scheduled for the final design iteration. This light source makes it feasible to use UHI on AUV because of the reduced power consumption.

There are thousands of kilometers of subsea oil and gas pipelines installed on the Norwegian continental shelf and in the rest of the world. These pipelines require regular inspection to ensure their integrity and to optimize maintenance cost. Current methods include visual (video or still photos) inspection from ROV and acoustic survey using ROV, AUV or ship, at regular intervals. The overall objective of this project is to implement and finalize development of an UHI system tailor-made for automatic detection of pipeline features, providing a functional hardware and software solution with required image quality at operational ROV survey speeds. The main motivation with this project is improved quality and reduced cost of pipeline inspection and integrity surveys. Another important motivation is the potential for the camera to detect features or anomalies (events) that are difficult to observe manually (i.e. small features). The overall goal is to develop an autonomous method for pipeline inspection, reducing the need for manual visual inspection onboard the vessel during survey operations. The main R&D Challenges are as follows: - Upgrade, test and implement a complete hardware system which is fully Integrated with existing Pipeline Inspection payload on survey ROV and vessel. - Develop a functional software for data handling and analysis, including identification and classification of events and ensure these are incorporated in the Pipeline Inspection protocols and reporting system. - Demonstrate complete hardware and software system and protocols ready for approval at a TRL Level of 8 or 9.

Budsjettformål:

DEMO2000-Prosj.ret tekn.utv. petro.virk