APPEA Journal 2013—31 A. Al Hinai 1 , R. Rezaee 1 , A. Saeedi 1 and R. Lenormand 2 . 1 Department of Petroleum Engineering ARRC Building, Curtin University 26 Dick Perry Avenue Kensington WA 6151 2 Cydarex 31 Avenue Gabriel Péri Rueil-Malmaison 92500, France Adnan.al-hinai@postgrad.curtin.edu.au ABSTRACT For shale gas reservoirs, permeability is one of the most important—and difficult—parameters to determine. Typical shale matrix permeabilities are in the range of 10 microdar- cy–100 nanodarcy, and are heavily dependent on the presence of natural fractures for gas transmissibility. Permeability is a parameter used to measure the ability of a rock to convey fluid. It is directly related to porosity and depends on the pore geometry features, such as tortuosity, pore shape and pore connectivity. Consequently, rocks with similar porosity can exhibit different permeability. Generally, permeability is measured in laboratories using core plugs. In some cases, however, it is difficult to obtain suitable core plugs. In these instances, other approaches can be used to predict permeability, which are chiefly based on mathematical and theoretical models. The approach followed in this peer-reviewed paper is to correlate permeability with capillary pressure data from mercury injection measurements. The theoretical and empirical equations, introduced in the literature for various conventional and unconventional reser- voir rocks, have been used to predict permeability. Estimated gas shale permeabilities are then compared with results from transient and steady state methods on small pieces of rocks embedded in a resin disk. The study also attempts to establish a suitable equation that is applicable to gas shale formations and to investigating the relationship between permeability and porosity. KEYWORDS Permeability, shale, mercury injection capillary pressure (MICP), porosity, pore size distribution, steady state, unsteady state. INTRODUCTION Typical shale gas matrix permeabilities are in the range of 10 microdarcy—100 nanodarcy, and are heavily dependent on the presence of natural fractures for gas transmissibility. Permeability is a parameter used to define, or measure, the ability of a rock to convey fluid. It is directly related to porosity and depends on the pore geometry (Fredrich et al, 1993). Consequently, rocks with similar porosity can exhibit different permeability (Costa, 2006). In reservoir rocks, the pores are connected by pore throats and each pore is accompanied by a certain range of throat sizes. A common technique in evaluating a number of petrophysical properties of a reservoir rocks is by measuring its capillary pres- sure. Capillary pressure and permeability are measured in labo- ratories using core plugs. In some cases, however, it is difficult to obtain suitable core plugs. In these instances, other approaches can be used to predict permeability and capillary pressure to pro- vide an insight into the petrophysical properties; such approaches are based on empirical and theoretical models. Core analysis, under ambient or reservoir conditions, is a com- mon method for direct measurement of permeability. Because of their high costs, only a limited number of core analyses are done for any particular field, although cuttings are available in almost all wells. The mercury injection technique may be used with well cuttings, or chips (Jennings, 1987). As the reservoir prop- erties—such as porosity and permeability—are controlled by the size and arrangement of pores and throats, the mercury injection method is commonly employed to characterise pore size distribu- tion and permeability in porous media (Swanson, 1981b; Katz and Thompson, 1986; Pittman, 1992; Kale et al, 2010; Kamath et al, 1998; Shouxiang et al., 1991; Owolabi and Watson, 1993; Purcell, 1949). As most studies have been for sandstones, there is a lack of comprehensive studies for shale. To bridge the information gap, this peer-reviewed paper will determine the applicability of the various models to estimate permeability from mercury injection measurements for shale gas samples. The authors assessed eight samples from one well, at various depths, in the Perth Basin. Predicted mercury injection capillary pressure (MICP) permeabilities are compared with those measured using transient and steady state techniques on small pieces of rock embedded in a resin disk. Models evaluated in this study include the Kozeny-Carman (Wylllie and Gregory,1955) and Swanson (1981), Winland (Kolodzie, 1980), Jorgensen (1988), Pape et al (1999), Rezaee et al (2006), Katz-Thompson (1986), Pittman (1992) and Dastidar et al (2007) methods. The key objectives of the study are to compare the results of the MICP permeability prediction methods versus laboratory mea- sured permeabilities, and to develop an improved relationship between permeability and pore throat size. Reservoir description Te Perth Basin is a 100,000 km 2 area covering the Western Australian margin between Augusta and Geraldton. Te North- hampton block is north of the basin, and the north-south trend- ing of Darling Fault is east. Te samples in this study are from the Carynginia Formation and are classifed as claystone. X-ray difraction (XRD) analy- sis indicates a composition of 63% non-clays and 37% clays. Figure 1 summarises the average percentage of the minerals present in the samples. Mercury porosimetry Te technique involves the intrusion of mercury (non-wet- ting liquid) at a high pressure into a porous material, through the use of a special assembly called a penetrometer. Te pressure is PERMEABILITY PREDICTION FROM MERCURY INJECTION CAPILLARY PRESSURE: AN EXAMPLE FROM THE PERTH BASIN, WESTERN AUSTRALIA Lead author Adnan Al Hinai