Abstract - The hydraulic properties of faults in the subsurface are mostly determined by characteristics which are below the resolution limit of seismic data. Recourse to accessible analogues is therefore necessary for prediction of the structure and properties of subsurface faults. The structure and content of the Moab Fault zone are described for 37 transects across the fault zone where throws range from less than 100m to c. 960m. The 45km long fault trace intersects a sedimentary sequence containing a high proportion of sandstones with good reservoir properties, interspersed with numerous mudstone layers. The Moab Fault therefore provides a good opportunity for testing existing techniques for fault seal prediction in the subsurface, which are based on the presence or absence of shaley gouge and shale smears. Both the complexity of fault zone structure and the spatial variation of shaley gouge content along the Moab Fault, support the use of empirical methods employed by industry, as opposed to the more deterministic methods which are sometimes advocated. The Moab Fault also reveals features that can lead to errors in fault seal prediction.
Typically, the fault zone is bounded by two external slip zones with the fault zone components separated by up to nine internal slip zones. Fault zone components are tabular lenses of variably deformed sandstones and sandstone cataclasites and breccia, with a wide size range, usually enclosed in a matrix of shaley fault gouge containing mm to m scale entrained sandstone fragments. Neither the structure nor the content of the fault zone can be predicted by extrapolation over distances as little as 10m. Where fault throws are less than c. 200m, fault relays are preserved but they are breached at higher throws with the formation of large lenses (up to 500mx50m) of rotated sequences and variably deformed fault rock. The extreme fault zone heterogeneity is attributed mainly to two processes which affect fault zones generally: tip-line bifurcation, by which a fault surface is split into two fault segments separated by a relay ramp, and asperity bifurcation by which asperities on the external boundaries of a fault zone are progressively removed.
Except where mudstone is < c. 20% of the faulted sequence, shaley gouge always occurs in the fault zone but its thickness is highly variable and unpredictable. The presence or absence of shaley gouge is therefore reliably predicted by the existing algorithms, which are now routinely used by reservoir evaluation and production teams. The empirical nature of these algorithms, in which seal prediction for a target fault is based on quantitative data from similar faults ideally located in the same area, incorporates implicitly the expected but unpredictable complexity of fault zones and so provides a rational basis for risking of fault seal predictions. The Moab Fault data show that one of the principal sources of error in fault seal prediction is failure to take account of sub-seismic resolution fault relay structures, which are likely to exist even though a subsurface fault may be imaged as a continuous structure. Given the limited resolution of seismic data it is tempting to rely on the much finer scale data provided by wells, but given the complexity of structure and content of fault zones as demonstrated by the Moab Fault data, it is evident that any method for fault seal prediction should not rely on extrapolation from well data even over distances of a few metres.
Abstract of talk given to:
Modern and Ancient Analogues: Uses and Abuses in Hydrocarbon Studies, Aberdeen, September 1998