Geospud, I did a lot of this kind of work while working on...

Geospud, I did a lot of this kind of work while working on projects in the Gulf of Mexico for a major US oil and gas company. They were similar shale-sand environments to those in VN. As you know, seismic traces indicate density difference in the rocks at their interface. If the rocks are all shale, there will be no density difference and thus no seismic indication. If there are sand formations, there can be 3 outcomes: shale-water sand-shale, shale-gas sand-shale, shale-gas sand-water sand-shale. (I'm not including oil-sand because the density difference between it and water-sand is small.) In the shale-water sand-shale case, the density difference is small and thus the seismic return will be small (no amplitude). In the shale-gas sand-shale case, the density difference is large and thus larger seismic amplitudes are generated. If the formation is thick and completely gas filled, a large amplitude will be generated at the top interface and a equal but opposite amplitude will be generated at the base of the sand. Using the time between the two amplitudes and the predicted density of the gas sand, a thickness can be calculated. As the gas-sand thins to shale or there is a salt water sand below, the seismic wave from the top and bottom interact to produce a lesser seismic return. A good seismic modeler can use the seismic returns to calculate the thickness and lateral extent of the gas accumulation.

When a well is drilled, all those calculations can be refined with the formation data and a better estimate can be produced. It is not as exact an estimate as further wells, but it can be pretty accurate.