Pressure or Fluid?

The Question a Repeat Survey Asks

Shoot a seismic survey over a field, produce it for a few years, then shoot the same survey again. The difference between the two, the time-lapse or 4D signal, is where the reservoir changed. But a change in the seismic is not self-explaining. Two very different things happen during production, and both can dim or brighten the same reflection. The pressure can move, as fluids are withdrawn or water is injected to maintain it, and the fluid itself can move, as gas comes out of solution or oil is swept by water. The central problem of 4D is that these two look alike in the most commonly measured quantity, the P-wave impedance. Reading a 4D map without untangling them is reading it wrong.

Three Scenarios on One Sand

Make it concrete on a single soft sand, quartz grains at porosity 0.25, and start from a brine-filled base at effective pressure 20 MPa: it has $V_P = 2.884$ km/s, $V_S = 1.508$ km/s, and a P-impedance $I_P = 6.47$. Now run three production stories on that same rock. Depletion: produce fluid, the pore pressure falls, effective pressure rises to 30 MPa, the frame stiffens, and both velocities climb, $V_P$ to 2.976 (up 3.2 percent) and $V_S$ to 1.605 (up 6.4 percent), with impedance up 3.2 percent. Gas breakout: hold the pressure but let gas replace the brine, and $V_P$ drops hard to 2.353 (down 18.4 percent) while $V_S$ actually rises to 1.592 (up 5.6 percent), because gas is light and lowering the density speeds the shear wave; impedance falls 26.9 percent. Injection: inject water, the pore pressure rises, effective pressure falls to 12 MPa, the frame softens, and both velocities drop, $V_P$ to 2.778 (down 3.7 percent) and $V_S$ to 1.392 (down 7.6 percent), impedance down 3.7 percent.

The Separator

Line the three up and the trap is plain. Gas breakout lowers the P-impedance by 26.9 percent; injection lowers it by 3.7 percent. A weak 4D brightening from a little gas and a modest brightening from a pressure rise both read as impedance down, and if P-impedance is all you carry, injection and gas can be confused. But look at the two velocities together and they split cleanly. A pressure change moves $V_P$ and $V_S$ the same way: depletion raised both, injection lowered both. A fluid change moves them in opposite directions: gas dropped $V_P$ while lifting $V_S$, because replacing brine with light gas softens the rock to compression but lightens it for shear. That opposition is the fingerprint of a fluid, and its absence is the fingerprint of pressure. So the way out of the ambiguity is to bring in the shear information, from a converted-wave survey or, more often, from the amplitude-versus-offset behavior that carries $V_S$ implicitly. Read $V_P$ alone and pressure and fluid blur; read $V_P$ and $V_S$ together and they separate. The final section asks whether either signal is even large enough to see through the noise.

References

• Landro, M. (2001). Discrimination between pressure and fluid saturation changes from time-lapse seismic data. Geophysics, 66(3), 836-844.
• Mavko, G., Mukerji, T., & Dvorkin, J. (2009). The Rock Physics Handbook (2nd ed.). Cambridge University Press.