Unloading and the Centroid

Pressure Without New Load

Undercompaction traps pressure by adding load faster than it drains. But pressure can also rise with no new load at all, through unloading mechanisms that generate fluid inside a sealed rock. Heat a trapped pore fluid and it expands against its container (aquathermal pressuring). Mature the kerogen in a source rock and it cracks to oil and then gas, and the volume expansion of that reaction (gas generation) can pressure a seal violently. Dehydrate smectite clay to illite as it heats past about 100 degrees and it releases bound water into the pores. None of these adds overburden; all of them add pressure. The distinction matters because unloading, unlike undercompaction, can lower effective stress that was previously higher, and a rock whose stress path retraces backward behaves differently on a velocity log, the subtlety the Bowers method of two sections on exists to handle.

The figure isolates the most operationally dangerous unloading geometry, the centroid effect. Take a permeable sand tilted within a sealing shale. Fluid in the sand is connected along its dip, so it equilibrates to a single, nearly flat pressure gradient, the fluid's own light gradient. The surrounding shale follows its steeper, near-lithostatic gradient. The two gradients cross at one depth, the centroid; above it the sand is over-pressured relative to its neighboring shale, below it under-pressured. The crest overpressure is the gradient contrast times the height above the centroid: $\Delta P = (G_{shale} - G_{fluid}) \times h$fluid)timesh. For a sand with 300 m of relief above its centroid and a gradient contrast of about 2.4 MPa per kilometer, the crest carries roughly 0.7 MPa of extra pressure that the local shale never warned of.

Why the Crest Is a Trap

That modest-sounding number is a real hazard, because pore-pressure prediction usually reads the shale, which is where the velocity trends live, and then assumes the adjacent sand shares it. At the crest of a tilted reservoir that assumption fails: the sand is systematically more pressured than the bounding shale, and a well that drills into the crest expecting shale pressure can take a kick it did not plan for. The effect scales with relief, so it bites hardest in exactly the high-relief structures that make the best traps. A large tilted sand can carry several megapascals of crest overpressure, enough to swing the required mud weight by a full pound per gallon or more between the flank and the crest of the same reservoir. The lesson the course carries into calibration: predict pressure in the shale, but correct it toward the sand crest by the centroid offset before trusting a mud plan, because the fluid, being connected, does not respect the local trend the way the shale does.

References

• Traugott, M. (1997). Pore/fracture pressure determinations in deep water. World Oil, 218(8), 68-70.
• Swarbrick, R. E., & Osborne, M. J. (1998). Mechanisms that generate abnormal pressures: An overview. AAPG Memoir 70, 13-34.
• Zoback, M. D. (2007). Reservoir Geomechanics. Cambridge University Press.