During migration of petroleum, either as phases separate from water or in water solution, an important role is assigned to pore fluid movements which result from compaction. The common view is that progressive burial of the sediments results in their compaction with consequent expulsion of pore fluids. These fluids are pictured as moving upward toward the depositional surface, even though the pathways (in actual detail) may include some lateral and downward movement. This commonly accepted view is often incorrect.

In the early stages of basin subsidence and sedimentation, the flux of water with reference to the depositional surface is downward, even though the flux of fluids expelled by compaction is upward across stratigraphic units. In later stages, a deep, subsiding basin contains a more or less constant volume of water. As subsidence, sedimentation, burial, and compaction continue, the sediments can be visualized as slowly moving downward through a fixed volume of water. Relative to a stratigraphic marker, the fluids move upward, but for the most part, they do not move to shallower positions relative to the surface of deposition.

When source sediments move downward into the "thermal window" for hydrocarbon generation, some of the hydrocarbons formed go into water solution. Subsequent migration and release of hydrocarbons from solution depend on the fluid flux and the positions of the isotherms. Exsolution occurs where the temperature of the solution falls below the saturation temperature (geologists have proposed that large volumes of deep, hot water are physically transported to shallower, cooler zones); however, exsolution occurs if the waters retain their position relative to the depositional surface while the isotherms are depressed.

Results of a Gulf Coast migration study indicate a formation temperature drop of about 50°F since early Pliocene time. In this case, a few thousand feet of section below the 200°F or 250°F isotherms could have exsolved hydrocarbons equivalent to the total known oil and gas in the area.

Quantitative modeling shows that some upward movement probably occurred when fluids from high pressure shales leaked through thin sands or along faults; however, migration by this mechanism is small in areas where the section retains abnormally high pressure and above-normal porosities.