The Normal Compaction Trend
Learning objectives
- Run the backbone workflow in order: constrain overburden and hydrostatic, pick mudstones, fit the normal compaction trend, predict, then calibrate and loop
- Pick mudstones with a gamma-ray cutoff, and let the cutoff vary with depth when the background radioactivity drifts
- Fit the trend only over the interval known to be hydrostatic and extrapolate it; fitting through the overpressured section hides the anomaly it should expose
- Calibrate the prediction against direct pressure measurements (MDT, RFT) and state where the vertical-stress family stops: reverse regimes need mean effective stress
The Step Every Method Stands On
Eaton and Bowers, the next two sections, are conversion formulas: they turn a velocity anomaly into a pressure. But an anomaly is a departure from a reference, and the reference is not given, it is built. That construction is the normal compaction trend (NCT), the backbone step of every conventional prediction, and real projects run it as a fixed sequence. First constrain the overburden and the hydrostat: integrate the density log for as in Part 1, and lay down the brine hydrostat of section 4.1. Second, pick your mudstones. The compaction methods apply only to mudstone: sands do not follow a simple compaction trend, and, as the centroid taught last section, a connected sand carries pressure from wherever its far end reaches, so its velocity is not a local gauge. The pick is made with a gamma-ray cutoff, mudstone is the radioactive lithology, and seasoned interpreters rarely trust one global number: the background radioactivity of a rock column drifts with depth, so a cutoff that picks clean mudstone at one kilometer can silently discard it, or admit silt, at three. Third, fit the trend, but only where you are certain the pressure is normal: the shallow section, above every seal, where the hydrostat of 4.1 is not in doubt. Extrapolate it below. Fourth, predict. Fifth, calibrate against every direct measurement the well affords and loop, and this last step is the subject of the closing panel below.
The figure is the working crossplot: mudstone velocity against the effective stress the depth would imply if pressure were hydrostatic, shaded by depth, with the fitted trend solid across the fit window and dashed where it is extrapolation. Where the deep points peel below the dashed line, the rock is announcing effective stress below its depth's due, and Eaton, next section, converts exactly that shortfall to pressure: the right panel grades the result against two MDT measurements, and with the sound workflow the canon reservoir comes back at 35.3 MPa at 3 km. Then commit the two classic sins. Drag the fit window down into the overpressured section and watch the trend rotate toward the anomaly it was supposed to detect, under-predicting both MDTs at once: the trend must be fit where pressure is known, never where it is the question. And freeze the cutoff to one global value: the drifting baseline carries the deep mudstones below it, the picks vanish, and with them every deep control point, which is precisely why the depth-varying cutoff is standard practice among pore-pressure specialists.
The Trend Has a Form, and a Family
The straight line in the figure is not an arbitrary ruler. It is the Bowers velocity-effective-stress form, , drawn with . In practice the mudline velocity is not a free parameter: it is pinned near the speed of unconsolidated water-saturated sediment, about 5000 ft/s, and the fit adjusts and alone, the pair a real study calibrates to offset wells and direct pressures. Invert the form and the prediction falls out: , then . The figure lets the intercept float precisely so you can watch a bad window corrupt it. Hold the fit to the hydrostatic section and the readout's fitted pair barely moves, with the 3 km prediction steady within half a megapascal; drag the window to 4 km and the intercept climbs from about 1610 to nearly 1770 m/s while falls by a fifth, and the prediction slides almost seven megapascals low. The parameters are the prediction: that is why they are treated as calibration targets, never as constants borrowed from a paper.
Nor is velocity the trend's only clothing. The Bowers form is one member of a family that all make the same three moves with different compaction proxies: Hubbert fits an exponential normal trend to porosity, the geotechnical tradition fits void ratio against the logarithm of effective stress, Butterfield works in specific volume, and Eaton, next section, sidesteps the inversion by working in ratios against the trend. At the Eugene Island 330 field, the classic calibration ground for this family, all five variants fit the same mudstone data about equally well, with correlation coefficients between 0.92 and 0.96. The choice among them is convention and data availability; the discipline is what transfers. Choose a compaction proxy, fit its normal trend where the pressure is known, and read the anomaly as effective stress.
Calibration, and Where the Whole Family Stops
Two closing disciplines separate a textbook exercise from a field study. First, the trend's parameters are local facts, not constants. Before a well is drilled they come from offset wells; once it logs, they are refit; and wherever the well yields a direct pressure, a wireline formation test (MDT, RFT) in a permeable stringer, a drillstem test, even a kick, those measured pressures outrank every inference, and the trend is adjusted until the prediction honors them. Predict, measure, update the components, predict again: the loop never closes, it just tightens. Second, know the family's boundary. Everything in this part reads the vertical effective stress, which is the right lever where the vertical stress drives compaction, the normal-faulting basins of Part 5's coming taxonomy, and it is the standard workflow there. But in a reverse-faulting regime the horizontal stresses are the largest loads on the rock, their work on compaction is not negligible, and a velocity-to-vertical-stress recipe mis-reads the physics: there the profession moves to a mean effective stress formulation, tying velocity to the full stress state rather than to alone. Carry the workflow, carry its boundary, and Eaton's formula, next, becomes what it always was: the easy step, standing on the hard one you just built.
References
- Zhang, J. (2011). Pore pressure prediction from well logs: Methods, modifications, and new approaches. Earth-Science Reviews, 108(1-2), 50-63.
- Flemings, P. B. (2021). A Concise Guide to Geopressure: Origin, Prediction, and Applications. Cambridge University Press.
- Bowers, G. L. (1995). Pore pressure estimation from velocity data. SPE Drilling & Completion, 10(2), 89-95.
- Mouchet, J.-P., & Mitchell, A. (1989). Abnormal Pressures While Drilling. Elf Aquitaine.
- Zoback, M. D. (2007). Reservoir Geomechanics. Cambridge University Press.