Gas hydrates (naturally occurring ice-like combinations of gas and water have the potential to provide an immense source of natural gas from the world?s oceans and polar regions. However, the technical challenges of economically producing gas from hydrates are substantial. Proposed recovery methods such as dissociating or melting gas hydrates by heating and/or depressurizing the reservoir are currently being tested. One such test was conducted recently in northern Canada by the partners in the Mallik 2002 Research Well Program.
Before the test, the Mallik 5L-38 well was cored, and gas hydrates and associated sediments were recovered for study. An extensive geophysics research program was also launched, which included seismic and advanced well log studies. A 5-day thermal test was then conducted within a 13-m interval of highly concentrated gas hydrates. The well was cased and a pipe circulating hot water was landed to the base of the perforated test section. The heating of the formation caused the hydrate to dissociate into water and natural gas, with the dissociated gas expanding into the wellbore. The amount of gas produced at the surface was monitored using gauges.
Resistivity logs were used to answer the important question of how deeply the hydrate dissociation extended into the formation. The open hole logging suite run prior to the thermal test included array induction, array laterolog, nuclear magnetic resonance, and 1.1-GHz electromagnetic propagation logs, as well as resistivity images. The reservoir saturation tool was run both before and after the thermal test to monitor formation changes. A cased hole formation resistivity log was run after the test.
The radius of dissociation as a function of depth was determined by means of iterative forward modeling of cased hole formation resistivity response. Pretest baseline resistivity values in each formation layer (Rt were established from the deep induction and laterolog curves. The resistivity in the region of hydrate dissociation near the wellbore (Rxo was determined from electromagnetic propagation and reservoir saturation tool measurements. The dissociation radius was systematically increased in each layer, and cased hole formation resistivity tool response was iteratively modeled with a finite element code. A solution for the dissociation radius in each layer was obtained when the modeled log overlaid the field log.
Pretest production computer simulations predicted that dissociation should take place at a constant radius. However, the post-test resistivity modeling showed that this was not the case. The resistivity-derived dissociation radius was greatest near the outlet of the circulating pipe, where the highest temperatures were recorded. The radius was smallest near the center of the test interval, where a conglomerate section with lower porosities and permeabilities appears to have inhibited dissociation. The free gas volume calculated using the resistivity-derived dissociation radii yielded a value within 20% of the surface gauge measurements. These results indicate that most of the produced gas was actually measured, and that resistivity modeling and inversion show promise for future use in gas hydrate monitoring.