Equinor has played an important role in the last decade in the testing and development of ultradeep azimuthal resistivity (UDAR) measurements both for look-ahead and look-around applications. Today, UDAR technology is applied in more than 70% of Equinor’s high-angle or horizontal wells. In this paper, the authors will review the use of UDAR in Equinor over the last decade and highlight both successful use and real-time challenges related to the interpretation of the inversion results. UDAR technology and inversion algorithms have been very powerful for reservoir mapping to geosteer or geostop according to plan. However, we forget far too often the fact that we need a good understanding of the reservoir to interpret and evaluate the uncertainty in the inversion result. The number one mistake in a real-time setting is to interpret a resistivity contrast as a specific layer in the reservoir (for instance, top reservoir) and hold on to that same interpretation, even if we drill away from that contrast and may cross multiple layers as distance to the observed contrast increases. Other challenging real-time UDAR exercises relate to uncertainties in the prediction of resistivity inside the reservoir and reservoir thickness from inversion results when still drilling above the reservoir. A third mistake often seen in real time is the detailed interpretation of one-dimensional (1D) inversion results, even when other indicators are pointing towards two-dimensional (2D)/three-dimensional (3D) complexities in the reservoir. Equinor and other operators have pushed for more and more advanced inversion solutions, leading to 3D mapping capabilities for more complex reservoirs. The UDAR advances over the last few years are important for Equinor’s planned roadmap ahead. However, 1D through 3D inversion results can result in bad decisions if the uncertainty in the inversion result is not managed correctly. We see a need to investigate how to best exploit UDAR technology and inversion results without extending assumptions beyond an acceptable uncertainty level. Better handling of uncertainties in geosteering operations will become increasingly important for the well economy with smaller targets, complex geological settings, and varying sweep efficiencies. How can we best handle the uncertainty in inversion results in real-time operations to avoid inaccurate decisions that can potentially destroy well economy? This is an important question that will be addressed and should be handled in the future if UDAR technology is to continue its important role in well placement in the next decades.

Equinor has played an important role the last decade in testing and development of ultra-deep azimuthal resistivity (UDAR) measurements both for Look-Ahead and Look-Around applications. Today, more than 70% of Equinor high angle or horizontal wells are drilled with UDAR-technology. In this paper, the authors will review the use of UDAR in Equinor the last decade and highlight both successful use and real-time challenges related to interpretation of the inversion results. UDAR-technology and inversion algorithms have been very powerful for reservoir mapping to geosteer or geostop according to plan. However, we forget far too often the fact that we need a good understanding of the reservoir to interpret and evaluate the uncertainty in the inversion result. The number one mistake in a real-time setting is to interpret a resistivity contrast as a specific layer in the reservoir (for instance top reservoir) and hold on to that same interpretation even if we drill away from that contrast and may cross multiple layers as distance to the observed contrast increase. Other challenging real-time UDAR-exercises relate to uncertainties in the prediction of resistivity inside the reservoir and reservoir thickness from inversion results when still drilling above the reservoir. A third mistake often seen real-time is detailed interpretation of 1D-inversion results, even when other indicators are pointing towards 2D/3D complexities in the reservoir. Equinor and other operators have pushed for more and more advanced inversion solutions leading to 3D mapping capabilities for more complex reservoirs. The UDAR advances the last few years are important for Equinor’s planned roadmap ahead. However, 1D-3D inversion results can result in wrong decisions if the uncertainty in the inversion result is not managed correctly. We see a need to investigate how to best exploit UDAR-technology and inversion results within its limits, but also ensure assumptions are not extended beyond an acceptable uncertainty level. Better handling of uncertainties in geosteering operations will become increasingly important for the well economy with smaller targets, complex geological settings, and varying sweep efficiencies. How can we best handle the uncertainty in inversion results in real-time operations to avoid wrong decisions that can potentially destroy well economy? This is an important question which will be addressed and should be handled in the future if UDAR-technology is to continue having an important role in many of the wells to be drilled the next decades.

The Troll oil field has been one of the largest oil producers on the Norwegian continental shelf for the last 25 years and is now moving into late life. The remaining oil column is thin, and the fluid contacts vary due to production effects. To extend field lifetime and secure the last reserves, enhanced well placement and increased drilling efficiency is needed to reduce cost and increase recovery in the long multilateral horizontal wells. Due to thin oil column and low reserves number, every meter correctly placed in the reservoir counts. To investigate these challenges a technology development project was initiated between Equinor and Baker Hughes to develop automatic interpretation of the oil-water contact (OWC) based on inversions, and automatic steering advice for taking faster geosteering (RNS) and downlink decisions placing the wellbores at the optimal distance to the OWC. Automating log interpretation is a complex task, but solvable given a known environment. As an engineering problem it must be split up into multiple smaller tasks that can be independently solved and when combined, solve the greater task. Logging while drilling (LWD) deep azimuthal resistivity data is run through inversion processing which provides a resistivity profile from which the OWC position is identified. The inversion input model constraints are set based on field/area specific data and is run automatically as drilling progresses. To assess the quality and validity of these results, several flag curves are computed and used. This automatic quality control of the OWC points enables the creation of a forward projection of this boundary. A steering advice is calculated, giving a recommendation on how to achieve the desired stand-off and inclination above the OWC as efficiently as possible. All the output from the automatic interpretation is published to a central datastore and is immediately available for the geoscientists to optimize operational decisions. Close co-operation between the operator and the vendor during the development and testing of the service has proven beneficial for identifying areas for further improvements. The service has been tested for well placement in actual producers. Several loops have been made in the development between the different tests and the learning curve has been steep in both companies. Based on the experiences and results from the actual wells, the project has moved into a new phase for further optimization of the steering advice and linking the automated steering advice to an Automatic Drilling Control (ADC) system to deliver a more automated closed loop service.