How to optimise carbon storage monitoring with marine robotics

For a long time, while carbon capture and storage (CCS) in offshore underground reservoirs had been widely regarded as a major way to reduce carbon emissions, it failed to attract the up-front investment needed to make it work. That’s now changing.

The challenge

In today’s far more climate conscious world, sentiment and interest in CCS has very much changed. Significant projects are now being planned. Projects are moving forward in Norway, Netherlands and the UK. Carbon storage licenses are being awarded and wells are being drilled specifically for carbon capture and storage.

Since 1996, CCS projects have been relatively small, yet their potential is vast. On the UK Continental Shelf alone there’s at least 78 gigatonnes of CO2 potential storage capacity – some 200 times the UK’s 2016 emissions*.

From capture to transport by pipeline and injection into a suitable geological formation offshore, there’s a lot to process. But the challenges do not end there. What happens to the CO2 once injected? How will we know if it finds a leak path to the surface?

To answer these challenges and achieve the visions that operators from Equinor in Norway to BP in the UK are promoting, increased capability for marine robotics is required.

The solution

Thanks to the Energy Technologies Institute (ETI) funded three-year research programme back in 2014, this challenge for increased capability of marine robotics has been attained. The project was delivered by a consortium of experienced companies including Fugro, National Oceanography Centre (NOC), British Geological Survey (BGS), Plymouth Marine Laboratory (PML) and ourselves, Sonardyne.

Along with increasing the capability of marine robotics for successful CO2 storage, four key technology elements for large carbon storage and monitoring projects were identified.

The first is a low-power and hence long-endurance autonomous underwater vehicle (AUV). This is required for cost-effective wide-area coverage surveys during baseline and repeat environmental surveys. We found using a combination of our Solstice side scan sonar and chemical sensing worked extremely well.

Second and third elements are seabed landers capable of detecting and monitoring any leakage at high-risk locations. These consist of two different landers, one using an active sonar and the second combining passive sonar and chemical sensing.

The active sonar lander, based on our Sentry integrity monitoring system (IMS), gives sensitive and reliable automated leak detection capability across a wide area. For instance, around an injection well, Sentry can monitor an area of over 2.3 million square metres, to help visualise that’s equivalent to around 325 football pitches. The passive sonar and chemical lander, uses the smarts from our underwater acoustics capabilities. It’s capable of both detection of leaks, but offers improved verification and has the potential to estimate leak rates at shorter ranges.

The fourth and final element is a surface gateway to enable communication between a shore-based monitoring office and the underwater systems. Such a gateway can be deployed from a fixed platform, from a moored buoy or from an uncrewed surface vessel (USV), many variants of which are now readily available in the market for over-the-horizon data harvesting missions.

We have a range of payloads suited specifically for use on operator’s USVs for their requirements. We also offer our own end-to-end data-harvesting service, when you just want the data without the worry about the interfaces involved in getting it.

Our system of systems approach to CCS was tested on the ETI project. Wideband acoustic communications between the underwater landers and a buoy on the surface was used to forward all data via satellite communications to a server. This type of set-up is well-proven and used globally on tsunami monitoring systems. Display and interpretation of the monitoring data can be simply integrated into a third-party system to allow non-expert users access via a web portal. From here they can see data visualizations and run reports.

The leak target was deployed in the North Sea, east of Bridlington. The NOC’s Autosub Long Range (ALR) was deployed from the small port at Bridlington and towed a short distance off the coast. After the ALR performed a series of tests to demonstrate safe navigation, the leak – a small CO2 leak – was turned ‘on’ with a flow rate of between 16 and 20 litres per minute of gas at depth, depending on the state of the tide.

With the leak “on”, ALR performed a series of different wide-area and fine-area search patterns over five days to seek out the leak. The sensor hub on the vehicle processed in real-time a complex set of Solstice sonar, physical and chemical sensor data, into useful information.

Automatic target recognition algorithms were used to identify any leaks or regions of interest. The system then scored these regions of interest and saved a small “snippet” of the sonar image data. At regular intervals throughout the survey, ALR would surface and send back data via satellite, including navigation data, chemical and physical sensor data and details of snippets of sonar data from detected leaks – an example of which can be seen below.

All of the uploaded data was simultaneously transferred to an internet server which allowed for presentation and interpretation using Fugro’s Metis software. This is an intuitive data delivery platform that allows metocean, vehicle navigation, chemical and sonar snippet data to be combined and displayed. This allowed data sharing across a wide team and supported operational decision making.

During the five days of testing, the ALR travelled a total of 270 km and could have surveyed 54 sq km of seabed in normal operation. However, for the purposes of the demonstration, a total of 16.1 sq km was actually surveyed. Throughout its mission, the ALR was remotely controlled from the shore, mostly from the NOC’s control room in Southampton.

The results

The ETI project consortium demonstrated a functional “system of systems” which can provide operators of offshore CO2 storage sites with a high level of confidence in their safe operation and assist in the provision of regulatory compliance.

We’ve proven it is possible to conduct shore-to-field-to-shore environmental survey operations using a long-endurance AUV. We’ve also shown it’s more than possible to operate well in excess of normal AUV deployments.

This method of working makes it possible to rely on a small local deployment team for CCS projects. The small team can then be supported by remote shore-based operations and a data interpretation team. This cuts both the time and cost of CCS operations considerably.

It is also entirely possible, and has been demonstrated elsewhere, that a further reduction of human decision making can be achieved to reduce operator intervention.

The ETI project demonstrated that it is possible to build highly cost-effective and autonomous sensing systems with on-board intelligence. These systems are both simple to deploy and operate and are very cost competitive with vessel-based or vessel supported AUV survey operations.

The project members have also developed two flexible seabed lander packages capable of extended duration deployments of six months to a year. These can provide localised and still also wide-area monitoring, automated processing of data subsea and communication of that information to surface.

Looking beyond carbon capture, the potential applications of such integrated marine robotic and intelligent remote sensing technologies are many and varied across ocean science, renewables, security and naval domains.