MEMS-based OBN: Lessons Learnt from the Largest OBN Survey Worldwide

Despite a recovery in the number of towed-streamer surveys being conducted, OBN (Ocean Bottom Node) seismic projects continue to take an increasing market share over towed-streamer surveys. In OBN acquisition, each node is equipped with a pressure sensor (hydrophone) and three motion sensors (typically, geophones). The nearly-a-century-old geophone technology has, however, certain inherent shortcomings that degrade the recorded signal. Geophone performance deviates from reference specifications due to manufacturing tolerances, ageing and changes in temperature. As an example, for 15-Hz omnitilt geophones, as commonly used in OBN acquisitions, the variation in response reaches 3 dB in amplitude and 10 degrees in phase within their range of manufacturing tolerances. These uncertainties in sensor response prove particularly difficult to model and correct for in practice and result in final data sensor artefacts. The insensitivity of geophones to the gravity field also requires the use of additional tilt meters for the verticalisation of the 3C with resulting issues related to the relative orientations of these two pieces of equipment. Today, MEMS (Micro-Electromechanical Systems)-based digital seismic accelerometers have proved to be the high-fidelity alternative to geophones. Their specifications are not affected by temperature, ageing or manufacturing tolerances, making the recorded signal accurate in phase and amplitude with the seismic signal over the entire seismic bandwidth. As MEMS can detect the gravity vector, the integration of this sensing technology into OBN has demonstrated that 3C MEMS provide, without pre-processing, seismic signal with true verticality, and a vector fidelity error (error in orthogonality between the three sensors) that is an order of magnitude lower than for 3C geophones. The excellent low-frequency performance of the latest, third-generation MEMS is also ideal for reaping the full benefit of novel low-frequency sources (Ronen 2017), and in this way pushing back further the limits of FWI. This, along with other MEMS properties, makes this sensor a strong driver for the growth of OBN acquisition – especially for sparse or blended acquisition, where sensor fidelity matters more than ever. At the time of writing, the world’s largest OBN survey is continuing in the Middle East and is starting to deliver a promising dataset from the 23,000 MEMS-based OBNs deployed. Observations from this mega-survey, as well as from a previous experimental survey that includes direct comparisons with geophone-based OBN, are presented and discussed in this article.