Optical and even quantum sensing is being assessed to help monitor the subsea oilfields of the future as part of remote sensor networks. Elaine Maslin reports on work in the field by the University of Aberdeen.
A deep water environment/biological landers, used by the University of Aberdeen. Photo from University of Aberdeen.
The amount and complexity of subsea production equipment is growing. In 2013, about 20% of UK offshore production, or 280,000boe/d was from subsea developments, a figure set to rise to about 400,000boe/d by 2016. Subsea separation is growing in use and subsea gas compression is nearing its first commercial use.
As these developments come online, in ever deeper, or harsher waters, where local, manned structures may not be possible, or are too costly, the need for remote monitoring, especially leak detection, is also increasing.
While operators and contractors work out how to make variable speed drives work in 3000m water depth, others have been focusing on future monitoring systems, including the University of Aberdeen.
The university is researching the different types of sensor systems required for remote subsea developments, under its work as part of the Scottish Sensor Systems Centre (S3C), a collaboration between eight Scottish universities, running 2009-March 2014.
Richard Neilson, Director of Research and Commercialization for the College of Physical Sciences, at the university, says future subsea sensor system networks will need to operate remotely, providing high-quality data, including operational data (flows, pressures, and temperatures), condition data (corrosion and actuator condition) and environmental monitoring data. Sensor systems will need to provide early leak detection, with automated leak warnings, as well as hydrocarbon flow monitoring and pipeline integrity monitoring. “Environmental leakage, without having to introduce tracers into the flow, is going to be one of the biggest areas,” Neilson says. “The other holy grail is pipeline inspection,” Neilson adds. “If you could fly an AUV along the length of the pipe to detect wall thickness, corrosion pitting, hydrate formation or waxing, instead of putting a pipeline inspection gauge through the pipeline, with the potential to have it jam and interrupt production, which would be ideal.”
The S3C program assessed various sensor types, including spectroscopy, for leak detection, and has developed a new technique, quantum sensing, for pipeline inspection. It also looked at what is needed to support subsea sensor networks, such as wireless sensor and communication networks and subsea power.
Molecular sensing, using optics, is one way to examine well fluids and to detect well fluids released to the environment without using tracers by detecting molecular species, i.e. hydrocarbons and other compounds, such as nitrogen, carbon, and sulfur dioxide.
“Using laser diodes you can see what is reflected back and what is being adsorbed,” Neilson says. “At a molecular level, you can look at not-visible light to either measure absorption or reflection and use the information to quantify the presence of hydrocarbons.”
Options using optics were researched by Johannes Kiefer, an honorary professor and former senior lecturer at University of Aberdeen’s School of Engineering. They include infrared and near-infrared spectroscopy, Raman and fluorescence spectroscopy, and UV/LED absorption and fluorescence (see panel at bottom of page).
While each of the above offer benefits, the most beneficial sensor network would use a combination of technologies, Neilson says.
“Each technique has its specific pros and cons, for example, regarding the types of molecules that can be detected, the sensitivity/limit of detection that can be achieved, the possibility for multispecies measurements,” Kiefer says. “Hence, it is beneficial to combine several methods in a sensor network in order to obtain a complete picture, e.g. of the chemical composition and the thermodynamic state. Which techniques should be selected and how they can be integrated depends on the required specifications of the sensor system.”
The University of Aberdeen’s Quantum sensor. Photo from University of Aberdeen.
Another technique being developed by Aberdeen under the S3C program, is quantum-based interferometry led by Charles Wang from the university’s physics department. A recent breakthrough in quantum physics has enabled the use of ultra-cold atom interferometers—sensitive instruments able to measure acceleration, rotation, gravity, and magnetism. They have, to date, been used in space to look at gravity waves.
This technology has the potential to be used to measure pipeline integrity from outside the pipeline, Neilson says. A laboratory-based demonstrator quantum sensor is being built at the university to test some of the theories for its use.
“Their (quantum-based interferometry sensors) capabilities make them an ideal tool in the next generation of gravimeters, enabling gravity anomaly detection to be carried out for subsea sensing and inspection using autonomous underwater vehicles (AUVs), remotely operated vehicles (ROVs), including pipeline condition monitoring using magnetic fields,” Wang says. “Quantum technology is expected to improve measurement sensitivity by at least three orders of magnitude over existing methods, with less power and size.”
To create a smart subsea sensor system, you would need to create a network of sensors around key installations, to “sniff” for small hydrocarbon releases, and detect them before any major release, as well as monitoring flow assurance and asset integrity, using wireless communications and underwater wireless sensor networks, Neilson says.
“Wired systems offer access to power supply, shielded cabling, and high data rates, but typically have a single point failure, and high installation and maintenance cost,” Neilson says. “Wireless networks would have lower installation cost, and would not require the use of deepwater wired connectors. However, underwater wireless sensor networks could suffer from interference, data rates are slower, and power is limited,” Neilson says.
The ideal system would be low cost, with low power consumption, offering secure data transfer, re-configurability, and robustness in the underwater environment, Neilson says. Such a system would also need to have adaptive routing, to avoid single point failures, reduce power consumption (by signals finding the shortest route), and secure connections. Systems of this type would enable AUV communication, and back up, or remove the need for, existing copper of fiber optic communication channels.
Available wireless networks for subsea are optical, electromagnetic, and acoustic. Each have their own capabilities and limitations, for example, acoustics have low attenuation and long range but a low bit rate. Electromagetic (commonly called radio frequency) communication subsea has high bit rate but short range. Optics has high attenuation, but very high data rates, at up to 10-150Mbps in up to 100m water.
Just as using a combination of sensors, depending on the application, is most effective, using a combination of wireless networks is also the most likely solution.
For such a system to operate, it would be best to be part of an all-electric system, which would provide more diagnostic data for the electric actuators and well-understood faults, and direct control of power, rather than electro-hydraulic, Neilson says.
But, increasing subsea power requirements, for subsea boosting or compression, could mean AC power is not desirable in terms of losses, and that high voltage DC (HVDC) power will be needed, due to better transmission efficiency. However, to date HVDC is mostly used point to point. EU and Engineering and Physical Sciences Research Council funding is supporting research, led by Prof Jovcic, on some of the issues around HVDC, including HVDC circuit breaker technology to protect HVDC networks and HVDC DC-DC conversion to allow direct voltage step-up and down, Neilson says.
To be autonomous, subsea systems would need to be able to make decisions and report. But, autonomous systems, which are able to make decisions and then act on them, could raise problems around employee and public perception, as well as trust in the system and choice of decision making scheme—using ruled bases, Bayesian reasoning, Fuzzy logic or a neural network, Neilson says. Making sure the knowledge base/domain for decision making is complete, is also critical.
Neilson, however, points out that such systems already exist in others sectors from simple, e.g. anti-lock braking systems (ABS), through engine management systems, and automatic parking in the automotive sector, to commercial autopilot systems and the likes of the Typhoon Eurofighter flight control system in aviation.
Finally, how an autonomous subsea field sensor network then reports is a further challenge, but one which is being addressed. “There are a number of data mining companies that take data and look for trends. But there are also further developments. Data2Text (a University of Aberdeen spin-out recently acquired by Arria) is able to take the data, analyze it and put the result into a readable document,” Neilson says. Such a system would be able to, in plain English reporting, identify alerts which need an expert’s attention and potentially allow an engineer to immediately formulate an action plan to correct the cause of alerts.
*Oil & Gas UK interpretation of Wood Mackenzie data.
Infrared (IR) spectroscopy - an absorption technique using molecule specific absorption in the mid-infrared spectrum, where the vibrational frequencies of molecules can interact with radiation, e.g. each hydrocarbon species exhibits a fingerprint IR spectrum, so the IR spectrum of a mixture can be used to quantitatively analyze the chemical composition. Moreover, IR spectroscopy is sensitive to the chemical environment of a species, allowing the investigation of effects like emulsion stabilization, for example in oil/water systems.
Near-infrared (NIR) spectroscopy - an absorption technique using molecule specific absorption in the near-infrared spectrum, where the overtone and combination vibrations in molecules can interact with radiation. NIR is typically less species-specific than IR, but may be cheaper and easier to implement, Kiefer says.
Raman spectroscopy – a technique complementary to IR. Some molecules (e.g. hydrogen) are not IR active but they are Raman active. As in IR, the interaction of radiation with the fundamental vibrational frequencies of molecules are used. However, in Raman it is not an absorption but a scattering process (typically of incident laser light).
Fluorescence spectroscopy - selected target molecules (typically aromatic compounds or tracer dyes) are excited by visible or ultraviolet radiation (e.g. from a laser) and subsequently emit fluorescence. Fluorescence offers detection limits in the order of ppm and below, Kiefer says.
UV/LED absorption and fluorescence - similar to IR, the absorption of radiation in the visible and UV spectrum, where selected molecules can be electronically excited, can be used. Like fluorescence, the technique can be highly sensitive and allow traces of specific molecules to be detected and quantified.
Neilson presented the S3C’s work at the Institute of Mechanical Engineer’s Subsea Engineering conference, in Aberdeen, in May.
A new university collaboration project is partially taking over from the S3C this year, the Centre for Sensors and Imaging Systems (CENSIS), based in Glasgow. The University of Aberdeen is involved in this, as well as the newly formed Oil & Gas Innovation Centre (OGiC), established to accelerate the development of innovative technology, systems and processes.