Sealing Faults and Reservoir Compartments By John K. Davidson, Principal Geologist/Geophysicist, Stochastic Simulation, 19 Apr 2016

It appears industry geologists generally consider fault seal to be dependent on acceptable shale/sand ratios across faults regardless of Earth stresses. On the contrary, the primary control is whether the fault surface is essentially open or closed, that is, emphasis on post migration stress history.

The western margin of Australia is a typical ‘passive margin’. It is experiencing Pliocene to Recent compression along its length, the exception being the Timor Sea area where post Eocene loading by over-riding of Timor has caused extension and normal fault seal breach of Jurassic traps.

Fig.1. Click here to view enlargement

Figures 1 and 2 are fault seal graphs generated from patented Stress from Seismic software. Figure 3, from the Vulcan Sub-basin, shows the Oliver Fault bounding a 170 m Jurassic, Plover gas discovery of Oliver-1ST1 and the Cockell Fault bounding Cockell-1 with only paleo shows.

Fig.2. Click here to view enlargement

The red line on Figures 1 and 2 indicates ‘Paleo Stresses’, the stress state at the time of deposition of each mapped isopach. To the right on the horizontal Fault Seal Confidence axis, reverse fault stress state is the most conducive to fault seal, contributing to fault plane gouge formation and also to holding the fault zone closed. The addition of further fault gouge formed by post reservoir compressional events can be indicated by the dashed red lines.

Fig.3. Click here to view enlargement

Fault seal is 48 to 52% or P50 chance at the Callovian-Ryzanian regional seal for the Oliver Fault and nearer 43 to 53% at Cockell-1 (the 79% point at the Ry is suspect being derived from data near the edge of the small 150 km2 survey).

Fig.4. Click here to view enlargement

In most basins this would be the end of the exercise, but are there effects caused by the post Eocene extension? In both cases the red line tends to 10% (P10) near the surface which indicates almost no chance of shallow sealing on either fault. Normal faults are extensional and can leak if steeply dipping (see top of Figure 4) but are concurrently compressional and potentially capable of sealing sand against sand deeper in the section as the fault plane dips decrease and maybe become listric if shaley (also see the deformation ellipse in Figure 1).

The black line traces the post Eocene ‘Last Extension’. The Oliver Fault approaches dips of 77⁰ (see bottom axis) near the surface with seal risk of 21% (as with the red line) and would probably leak, even shale juxtaposed to shale. Across the Upper Plover (Ry-Cl) regional seal the Normal Fault dip is nearer 52⁰ with a Fault Seal Confidence of 55%, an increase over the Paleo Stress at that depth from 50%. Over the same interval the Cockell Fault actually steepens with depth from 60⁰ to 71⁰ and has a Fault Seal Confidence decreasing from 35% to 29%. Oliver retained its gas and Cockell’s was lost, strongly corroborating the varying effects of fault seal due to the normal fault loading of Timor.

A method now exists which can predict, at the prospect stage, the occurrence of reservoir fault seal before a well is drilled and at any stage during the drilling/production evaluation process. The current emphasis on large development costs for LNG plants indicates the potential of sealed/non-sealed compartment detection.