New prototype software from Siemens known as Marine Image Processing could soon make oil and gas exploration cheaper and more efficient. The software is based on image analysis algorithms used by Healthineers, and is designed to automate the process of evaluating ultrasound data gathered by an exploration vessel. To do this, the data is used to generate a 3D model of a water column between a vessel and the ocean floor and identify gas bubbles in that column – a sign of gas. Experts from Siemens and Shell recently presented the technology at a conference in Paris.
The Future of Oil and Gas
Digital Oil Prospecting with Ultrasound
Siemens has brought together image analysis technologies, familiar in the world of medicine, so they can be used to detect sources of oil and gas below the sea floor. The result: "the software is better than experts".
Bubbles with a Message
A significant amount of the world’s gas reserves is located below the sea floor. The search for underground hydrocarbons in offshore basins starts with geological considerations and evaluation of geophysical and seismic data often acquired using expensive exploration vessels. Even when the data suggest a favorable environment and geological structures to trap hydrocarbons, in lightly explored or frontier areas the presence of a working petroleum system can be a major risk. High resolution ultrasound surveys can then be conducted to identify gas plumes in the water column that can lead to taking better hydrocarbon samples. This step is designed to ensure against the risk of drilling a dry well, which can cost more than €100 million.
Until now, analyzing subsea ultrasound data, or multi-beam echo sounding (MBES) data was expensive and time consuming. Typically, a ship heads to a promising site, directs ultrasound waves to the sea floor and records the reflected energy. The vessel then returns to land, where experts evaluate the data. They create a 3D model of the water column between the ship and the sea floor, and geolocate the data using the ship’s position and the topography of the sea floor. They also correct for various effects, such as the fact that the ship travels a certain distance between transmitting and receiving a signal, and that currents cause gas bubbles to drift away from their initial escape point. Using the adjusted images, specialists manually identify the elusive gas bubbles. Based on the resulting analysis, a decision is taken as to whether a second expedition equipped with an underwater sampling system should be ordered.
Deep Sea Diagnostics
The Siemens Oil & Gas market development board brought together experts from Siemens Corporate Technology and Siemens subsidiary Dresser-Rand who worked with exploration geoscientists from Royal Dutch Shell to develop algorithms to automate this evaluation process and deliver results promptly once the data is recorded. The software generates a 3D model and links it with GPS positions, ship speed, and the topography of the sea floor. In addition – and this is where the expertise built up by Siemens in analyzing medical ultrasound images comes into play – the software corrects the images to minimize errors caused by factors such as shadows and, most importantly, automatically identifies gas bubbles.
In tests with Royal Dutch Shell on high resolution MBES data sets, Marine Image Processing located 95 percent of the locations where gas was escaping from the sea floor – a much better result than the rate of less than 80 percent achieved with manual analysis. The prototype software is also up to four times faster. This means that decisions can be taken on follow-up steps with a high degree of certainty while still on board ship. Considering that survey vessels can cost tens of thousands of euros per day to operate, using the new tool to search for oil and gas will not only be more efficient but will result in substantial savings.