Quantifying the Minimum Fault Throw Resolution using Ant Tracking and High Resolution Mega Mapping (2019)

Deze MSc scriptie is geschreven door oud-stagiair Yoel Beaucaire.

With the Masterplan Aardwarmte, the Dutch government has set goals to increase the role of geothermal energy in the Dutch energy mix. Currently, geothermal energy (2-4 km deep) is still relatively unexplored in the Netherlands, with only 22 geothermal doublets in the Dutch onshore subsurface. Smaller companies and operators are trying to participate in this new and upcoming market, but do not have the same facilities as larger operators. Therefore, most of their analyses are based on the freely available digital ground model (DGM-Deep) published by the Dutch geological survey, part of TNO. The grid resolution of these key geological maps is 250x250m. Because the resolution of the available seismic data on which these horizons are typically interpreted is 25m, many detailed structural features might not be seen on these horizons which can have crucial consequences when planning the well site or trajectory. As faults could pose a drilling hazard, as they have might constitute pressure barriers, missing faults when planning a new well location can potentially lead to blowouts.

During this study, as part of a 5 month internship at EBN B.V. (Energie Beheer Nederland), detailed seismic interpretations were conducted for three key stratigraphic intervals for the West Netherlands Basin: the base Upper North Sea Group, the base North Sea Group, and the base Chalk Group. This interpretation makes use of the full grid resolution of 25x25m, which is defined by the 3D seismic data available. In addition, Schlumberger’s Ant Tracking software (advanced Petrel functionality to automatically detect faults) was used for fault analysis, building on the results of Kortekaas and Jaarsma (2017). After careful manual interpretation, 50 representative faults were selected and measured in detail to quantify what the minimum fault throw detection limit is for the different mapping methods. This analysis shows that when viewing horizons with a 25x25m resolution, using optimized display settings, faults with a vertical fault throw of just 1 ms (~ 1 m in depth) can already be detected at a depth ~1 km. On the coarse DGM grids, the fault detection threshold is 10 ms at best. Ant tracking delivers fault maps quickly and results are similar to the high resolution interpretation, with faults showing up starting from 1.5 ms fault throw. The consequences are that many faults, most notably when there is a second set of (conjugate) faults, are not visible on the DGM horizons. It is therefore strongly recommended that well planning makes use of high resolution surfaces, rather than the DGM surfaces.