Summary We developed a new method involving crushed core samples to measure porosityand fluid content accurately in the black, organic-rich Devonian shales in the Appalachian basin. In the four wells studied, average core bulk volume of gasand porosity ranged to 4% and 8%, respectively, values substantially higherthan previously reported. previously reported. Introduction The Devonian shale in the Appalachian basin is considered a major potentialgas source. Because of analytical difficulties, core potential gas source. Because of analytical difficulties, core analysis in shales for reservoirproperties is not done often. Previous reports 1-3 show typical core porositiesof 2% to 3% in Previous reports 1-3 show typical core porosities of 2% to 3% inthe Devonian shale and free-gas content less than 0.1 % to 0.2 % by bulkvolume. Estimates of gas in place associated with these low free-gas contentsare an order of magnitude lower than is customarily produced in Devonian shalewells with long-term production history. produced in Devonian shale wells withlong-term production history. The goal of this study, sponsored by the Gas Research Inst. (GRI), was to improve our understanding of formation evaluationand reservoir description in shales. Four wells have been air-drilled recentlyin Kentucky and West Virginia from which extensive formation evaluation data, including 519 ft of conventional cores in the Devonian shale, have beencollected (Fig. !). Core recovery by formation is shown for each well in Table1 (the wells were drilled in the sequence shown). These wells were designatedcomprehensive study wells (CSW's) because of the extensive GRI- fundedevaluation-data program. One purpose of the CSW program was to develop newformation evaluation methods to identify gas-producing intervals, for bothdeliverability and reserves, with logging tools. Significant progress has beenmade. A key element of these new is to progress has been made. A key element ofthese new is to establish the reservoir rock and fluid properties throughcareful core analysis and then to develop the log interpretation methodsnecessary to match core analysis and production results. In the four CSW'S, cores range from shaly siltstones and laminated siltstones and shales to siltyshales. Mean grain size ranges from 0.01 to 0.06 mm. Mineralogy and petrographyshow the shale to be predominantly quartz and illite, with other mineralsincluding pyrite and kerogen (which can range up to 30 vol %). In CSW No. 2, the first well cored in the CSW program, very little kerogen and oil werepresent. Core analyses were similar to those in other low-permeabilitygas-productive reservoirs: water content was determined by weight loss afterextended drying in a convection oven at 230F, which removes both free andclaybound water. We measured grain volume of the dry core with helium and bulkvolume by immersing the core in mercury. PV is the difference between bulk andgrain volume. Free-gas content is the difference between PV and water content. These measurements were made at 1-ft intervals from the cored interval. Becausethese cores, and those from CSW Nos. 4A, 5, and 1A, were drilled in air-mistand the cores were preserved at the wellsite, the fluid saturations in the coreshould represent those in the reservoir after appropriate correction forcompaction to reservoir stress. When core analyses were begun on the next well, CSW No. 4A, different methods were required because the cores containsignificant amounts of oil and kerogen. Initially, core pieces were extractedin boiling toluene with the Dean Stark method for up to 2 weeks; the cores weredried in a convection oven at 230F for about 2 weeks; and PV's were determinedwith helium. This resulted in grain densities and porosities that were much toolow compared with log analyses. A new method of core analysis was developedthat involves crushing the core to speed up the processes of extraction, drying, and grain volume measurement. This new method results in much higherporosity and gas content values in the shale samples from CSW No. 4A, whereboth methods were used. The following sections describe the new core analysismethods. Porosity results are shown by a comparison of the new methods to Porosity results are shown by a comparison of the new methods to conventionalanalyses in normal-permeability rocks and in cores from the kerogen-free CSWNo. 2 to verify the new methods. Porosity and fluid content are shown fromcompanion sets of 95 crushed samples run by two separate laboratories toillustrate repeatability. Core analysis results are compared with a loganalysis in one of the study wells. Core analysis results for all four studywells are summarized. Finally, examples of other applications of analyses ofthe companion crushed sample sets are shown. P. 1184