RFour Energy · Field Notes

Pore Pressure — Reading the Rock Before the Bit Arrives

Q: What causes overpressure?

Two families of mechanism. The first is disequilibrium compaction, also called undercompaction: sediment is buried faster than its pore fluid can escape, so the fluid carries part of the load that the grain framework would otherwise carry and porosity stays abnormally high for the depth. The second is fluid expansion, where the pore fluid volume increases or is added to after burial, through aquathermal expansion, hydrocarbon generation, clay diagenesis such as smectite to illite, or charging from another zone. The distinction matters because the two leave different signatures in velocity and therefore require different prediction equations.

Q: What is the difference between the Eaton and Bowers methods?

Eaton's method compares a measured log response, usually sonic or resistivity, against a normal compaction trend and raises the ratio to an empirical exponent to obtain pore pressure. It assumes a single relationship between the log response and effective stress, which is valid when undercompaction is the cause. Bowers introduced a pair of velocity-versus-effective-stress relations: a loading curve for normal compaction and undercompaction, and a separate unloading curve applied inside velocity reversal zones created by fluid expansion. Where unloading has occurred, using the loading curve alone reads an effective stress that is too high and therefore a pore pressure that is too low.

Q: What is the centroid effect?

In a dipping permeable body encased in shale, the fluid inside the body is connected, so its pressure follows the light gradient of that fluid, while pressure in the surrounding shale follows the steeper trend set by compaction. The two lines cross at one depth, the centroid. Above the centroid the pressure inside the body exceeds the pressure in the adjacent shale, so a well drilling into the crest meets a higher pressure than a shale-based prediction suggests. Below the centroid the reverse is true and the risk becomes losses rather than an influx.

Q: How is a pore pressure prediction calibrated?

Against anything that measured pressure directly or responded to it. Direct measurements come from formation testers and drillstem tests in permeable beds. Indirect evidence comes from the mud weight history of offset wells, kicks and their shut-in pressures, connection and background gas, the drilling exponent corrected for mud weight, the shape and volume of cavings at the shakers, and leakoff or formation integrity tests which constrain the fracture gradient rather than the pore pressure. A prediction that has not been reconciled with these is a model, not an estimate.

Every well is drilled against a number nobody has measured yet. Predicting that number is one of the few subsurface tasks where being wrong is not an academic disappointment — it is a kick, a stuck string, or a well that never reaches its objective.

RFour Energy · Geoscience & Rock Mechanics Series

Pore pressure prediction is usually taught as a set of equations — Eaton, Bowers, equivalent depth — and that framing quietly hides the only decision that matters. The equations are not interchangeable tools for one job. Each one assumes a mechanism, and applying the wrong one to the wrong rock does not produce a slightly worse answer; it produces an answer that is wrong in a predictable and dangerous direction. So the real workflow is not "pick a method and calibrate it". It is: work out why this rock is overpressured, then choose the method that matches, then calibrate against everything the well and its neighbours have already told you.

The Principle Underneath Everything

Rock at depth carries the weight of everything above it. That total vertical stress S_v is shared between the grain framework and the fluid in the pores. Terzaghi's effective stress principle states the split: σ′_v = S_v − P_p. Every velocity-based or log-based prediction method in use exploits one consequence of this — rock properties respond to effective stress, not to depth. Compaction, velocity, resistivity and density all track how hard the grain framework is being squeezed.

Normal compaction follows from it directly: as burial proceeds, fluid escapes, effective stress rises, porosity falls along a trend that Athy described as an exponential decline with depth.^[1] A prediction method works by detecting a departure from that expected behaviour and converting the departure into pressure.

Every method measures the same thing indirectly: how compacted the rock is compared with how compacted it should be. The disagreements are all about what to do when the rock got there by an unusual route.

Two Mechanisms, Two Signatures

Overpressure has many named causes, but for prediction purposes they collapse into two families with genuinely different consequences.^[2]

Disequilibrium compaction (undercompaction). Sediment is buried faster than its pore water can escape — typical of rapidly loaded, low-permeability sequences. Fluid takes up load the grains would have carried, so effective stress stops rising. The rock is frozen at a shallower compaction state: porosity abnormally high, velocity abnormally low for its depth. Crucially, the rock has never been more compacted than it is now.
Fluid expansion (unloading). Pressure is added after the rock was already compacted — by aquathermal expansion, hydrocarbon generation, clay diagenesis such as smectite to illite, or charging from another zone.^[3] Effective stress falls. But the rock does not un-compact: the grain framework keeps most of the stiffness it earned at its maximum past stress. Porosity stays low while pressure climbs.

That asymmetry is the whole problem. Under disequilibrium compaction, velocity and effective stress lie on one curve. Under fluid expansion, the rock returns along a different, much flatter path — a hysteresis loop. Reading a fluid-expansion rock on the compaction curve is the classic and most consequential error in the discipline.

FIG. 01Why one equation is not enough. Bowers's method uses a pair of velocity–effective-stress relations: a loading curve for normal compaction and undercompaction, and a separate unloading curve applied inside velocity reversal zones caused by fluid expansion.^[4] The worked point shows the cost of ignoring the second: a velocity that truly corresponds to σ′ = 1500 psi is read off the loading curve as 3247 psi, so the predicted pore pressure comes out about 1750 psi too low — an error in the direction that gets wells killed rather than merely delayed.

The Methods, and What Each One Assumes

Normal compaction trend and Eaton's exponent

The workhorse. Establish a normal compaction trend (NCT) through the clean shales where pressure is known or assumed hydrostatic, then measure how far the observed log departs from it. Eaton's formulation converts that ratio directly into pore pressure through an empirical exponent, applied to sonic or resistivity.^[5] It is simple, transparent, fast, and it is what most software runs by default.

Its assumption is the part to keep in mind: one single relationship between log response and effective stress. That is exactly right for undercompaction and exactly wrong for unloading. The exponent is also not a constant of nature — it is a fitted parameter, and fitting it to a well in one basin and exporting it to another is where a great many bad predictions begin.

FIG. 02The departure method. Below the top of overpressure the log stops following the compaction trend, and the size of the gap is converted to pressure. The diagnostic worth learning is the shape of the departure: a gradual divergence is consistent with undercompaction, while a sharp reversal — especially one where sonic slows but density does not fall correspondingly — points to unloading, and tells you to reach for the second curve rather than a larger exponent.^[4][6] Circles mark calibration points, without which the whole curve is a hypothesis.

Seismic velocity, before there is a well

Pre-drill, the only volume-filling measurement available is seismic velocity, and prediction from it is a mature but demanding discipline.^[7] Two cautions carry most of the risk. First, processing velocities are not rock velocities: stacking velocities smoothed for imaging are far too coarse for pressure work, and the analysis needs velocities derived for the purpose — tomography or inversion — with the rock-physics link stated explicitly. Second, the vertical resolution of seismic is tens of metres at best, so a thin, highly overpressured interval can be real and still invisible. A pre-drill profile is a basin-scale expectation, not a plan for a casing shoe.

The Centroid — Where Shale-Based Prediction Misleads

Almost every method above predicts pressure in shale, because shale is what compaction trends describe. Reservoirs are not shale, and in a dipping permeable body the difference is not academic.

Inside a connected sand, the fluid column behaves like a fluid column: pressure changes with depth at the light gradient of the fluid it contains. In the bounding shale, pressure follows the much steeper trend that compaction dictates. Two lines with different slopes cross at exactly one depth — the centroid.^[8] Above it, the sand is more pressured than the shale beside it; below it, less.

FIG. 03The centroid. Sand pressure follows the light gradient of its own connected fluid; shale pressure follows the steeper compaction trend. For the geometry drawn, the crest of the sand sits about 750 psi above the adjacent shale and the base about 750 psi below. A prediction built on shale trends alone therefore underestimates pressure exactly where wells usually target — the crest — and overestimates it downdip, where the real hazard becomes losses rather than an influx.^[8][9]

Calibration, and What Counts as Evidence

A pressure curve is only as good as what it has been reconciled against. The evidence falls into a clear hierarchy, and it is worth being strict about which is which.

Evidence	What it actually constrains	Strength
Formation tester (RFT/MDT), DST	Pore pressure — directly, in permeable beds only	Definitive, but only where there is permeability
Kick with recorded shut-in pressure	Pore pressure at that depth, a hard lower bound	Very strong
Mud weight history of offset wells	A bracket: pressure was below the mud that held it	Strong as a bound, weak as a value
Connection and background gas	Balance is marginal — direction, not magnitude	Indicative
Corrected drilling exponent (d_c)	Trend break as the bit enters overpressure	Indicative, real-time
Cavings volume and shape	Wellbore instability — often stress, not pressure	Diagnostic, easily misread
Leakoff / formation integrity test	Fracture gradient — the upper bound, not P_p	Definitive for the other side of the window

Two entries deserve underlining. A leakoff test says nothing about pore pressure — it constrains the top of the mud weight window, and conflating the two is a persistent source of muddled models. And cavings are ambiguous: angular, splintery cavings usually indicate shear failure driven by stress and mud weight, not necessarily an underbalanced pore pressure. Reading them as a pressure signal has caused more than one crew to raise mud weight and make the hole worse.^[10]

The Estimate Narrows — It Never Becomes a Number

The honest way to carry a pore pressure prediction is as a distribution that tightens as evidence arrives, not a line that is either right or wrong.

FIG. 04What a prediction really is. The pre-drill estimate is wide because it rests on seismic velocity and analogue trends; offset wells narrow it; measurements while drilling narrow it further and shift the centre; the post-well analysis closes it — for that well. The band never collapses to zero, and a casing design built on the centre line rather than on the band is a design with no margin in it.

Validation

The effective stress framing, the classification of overpressure mechanisms and the treatment of pressure at depth follow Zoback's Reservoir Geomechanics.^[6] The exponential porosity–depth compaction relationship that underpins every normal trend is Athy's.^[1] The re-evaluation of overpressure-generating mechanisms, and the case that disequilibrium compaction and fluid expansion must be distinguished rather than lumped, follows Osborne and Swarbrick.^[2][3] The departure-from-trend equation with its empirical exponent is Eaton's.^[5] The paired loading and unloading velocity–effective-stress relations used in Fig. 01, and the observation that a second relation is required inside velocity reversal zones caused by fluid expansion, are Bowers's.^[4] Prediction from seismic velocity, and the requirement that velocities be derived for pressure work rather than for imaging, follows Dutta.^[7] The centroid and lateral transfer treatment in Fig. 03 follows Traugott^[8] and Yardley and Swarbrick.^[9] Drilling-derived indicators — the corrected d-exponent, connection gas, cavings and their ambiguity — follow Mouchet and Mitchell.^[10]

Choose the method that matches the mechanism. Everything else in pore pressure prediction is calibration.

References

Athy, L.F. (1930). Density, Porosity, and Compaction of Sedimentary Rocks. AAPG Bulletin 14(1), 1–24. The exponential porosity–depth relation underlying normal compaction trends.
Osborne, M.J. & Swarbrick, R.E. (1997). Mechanisms for Generating Overpressure in Sedimentary Basins: A Reevaluation. AAPG Bulletin 81(6), 1023–1041.
Swarbrick, R.E. & Osborne, M.J. (1998). Mechanisms that Generate Abnormal Pressures: An Overview. In Law, B.E., Ulmishek, G.F. & Slavin, V.I. (eds.), Abnormal Pressures in Hydrocarbon Environments, AAPG Memoir 70; doi:10.1306/M70615C2. For the relative magnitude each mechanism can generate, see also Swarbrick, R.E., Osborne, M.J. & Yardley, G.S. (2002), Comparison of Overpressure Magnitude Resulting from the Main Generating Mechanisms, in Huffman, A.R. & Bowers, G.L. (eds.), Pressure Regimes in Sedimentary Basins and Their Prediction, AAPG Memoir 76, 1–12.
Bowers, G.L. (1995). Pore Pressure Estimation From Velocity Data: Accounting for Overpressure Mechanisms Besides Undercompaction. SPE Drilling & Completion 10(02), 89–95; SPE-27488-PA, doi:10.2118/27488-PA. Introduces the paired loading / unloading velocity–effective-stress relations; the unloading relation applies inside velocity reversal zones caused by aquathermal pressuring, hydrocarbon maturation, clay diagenesis and charging from other zones.
Eaton, B.A. (1975). The Equation for Geopressure Prediction from Well Logs. Fall Meeting of the Society of Petroleum Engineers of AIME, Dallas; SPE-5544-MS, doi:10.2118/5544-MS, 11 pp. See also Eaton, B.A. (1972), The Effect of Overburden Stress on Geopressure Prediction from Well Logs, SPE-3719.
Zoback, M.D. (2007). Reservoir Geomechanics. Cambridge University Press, Cambridge, 449 pp. ISBN 978-0-521-77069-9. Pore pressure at depth, effective stress, and the interaction of pressure with the stress field.
Dutta, N.C. (2002). Geopressure Prediction Using Seismic Data: Current Status and the Road Ahead. Geophysics 67(6), 2012–2041; doi:10.1190/1.1527101.
Traugott, M. (1997). Pore/Fracture Pressure Determinations in Deep Water. World Oil 218(8), 68–70. Centroid and deepwater pressure/fracture-gradient practice.
Yardley, G.S. & Swarbrick, R.E. (2000). Lateral Transfer: A Source of Additional Overpressure? Marine and Petroleum Geology. Overpressure redistributed along a dipping carrier bed — the mechanism behind the centroid geometry in Fig. 03.
Mouchet, J.-P. & Mitchell, A. (1989). Abnormal Pressures While Drilling: Origins, Prediction, Detection, Evaluation. Elf Aquitaine, Boussens. Drilling-derived detection: d-exponent, gas shows, cuttings and cavings.

Frequently Asked Questions

What causes overpressure?

Two families. Disequilibrium compaction — burial outpaces fluid escape, so fluid carries load the grains would have carried and porosity stays abnormally high. Fluid expansion — pressure is added after compaction, by aquathermal expansion, hydrocarbon generation, clay diagenesis or charging from another zone, so effective stress falls while porosity stays low. They leave different velocity signatures and need different equations.

What is the difference between the Eaton and Bowers methods?

Eaton compares a log against a normal compaction trend and raises the ratio to an empirical exponent — one relationship between log response and effective stress, valid when undercompaction is the cause. Bowers uses a pair of relations: a loading curve for normal compaction and undercompaction, and an unloading curve for velocity reversal zones caused by fluid expansion. Using the loading curve inside an unloading zone reads effective stress too high, and therefore pore pressure too low.

What is the centroid effect?

In a dipping permeable body encased in shale, the connected fluid inside follows its own light pressure gradient while the shale follows the steeper compaction trend. The two cross at one depth — the centroid. Above it the body is more pressured than the adjacent shale, so a well into the crest meets higher pressure than a shale-based prediction implies; below it the reverse, where the hazard becomes losses.

How is a pore pressure prediction calibrated?

Against formation testers and DSTs where there is permeability, kicks with recorded shut-in pressures, offset mud weight histories as bounds, connection gas, the corrected d-exponent, and cavings — read carefully, since cavings often indicate stress-driven failure rather than underbalance. Leakoff tests constrain the fracture gradient, not pore pressure. A prediction not reconciled with these is a model, not an estimate.