Hydraulic versus pneumatic measurements of fractured sandstone permeability

Transcription

1 Journal of Petroleum Science and Engineering 36 (2002) Hydraulic versus pneumatic measurements of fractured sandstone permeability Salima Baraka-Lokmane* Department of Geology and Geophysics, Grant Institute, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, UK Received 22 November 2001; accepted 30 September 2002 Abstract This paper investigates the issue of ascertaining whether gas can replace water for determining the flow parameters in fractured porous media. This is accomplished by the determination of the hydraulic parameters using brine, the pneumatic parameters using air, and the study of the correlation between these two parameters. The measurements are obtained for fractured sandstone cores from the middle Stubensandstein unit in the Southwest German Trias. In most cases, the intrinsic liquid permeability is lower than the intrinsic gas permeability. Intrinsic gas permeability (k g ) ranged from 32 to 159 md, while intrinsic liquid permeability (k l ) ranged from 12 to 47 md. The ratio of intrinsic gas permeability to intrinsic liquid permeability (k g /k l ) shows two subgroups: (1) ratios ranging from 1 to 2 (62.5% of samples) and (2) ratios ranging from 4 to 5 (37.5% of samples). The reduction in the intrinsic liquid permeability is governed by three phenomena: physicochemically, by the migration of the clay particles which clog the pores, mechanically, by the breakdown of original fabrics caused by the passage of wetting fronts across relatively delicate clay mineral complexes, and experimentally, by the undersaturation of samples during liquid permeability measurements. This study concludes that gas permeability is more accurate than liquid permeability because it measures more closely intrinsic permeability especially for clay-rich rocks. In addition, because gas experiments can be conducted much faster than liquid flow experiments, gas is a desirable replacement fluid. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Fractured sandstone; Liquid permeability; Gas permeability; Laboratory study 1. Introduction * Fax: address: S.Baraka-Lokmane@glg.ed.ac.uk (S. Baraka-Lokmane). In aquifers, where fractures make a significant contribution to the transmissivity, the estimated average permeability is some complex function of fracture characteristics, e.g. fracture connectivity, fracture aperture and spacing distributions (Bloomfield and Williams, 1995) as well as the matrix properties of the aquifer material itself. In the field, it is difficult to characterise fractures and to quantify their contribution to the permeability of the medium because the fracture geometry is generally complex and unknown, especially in 3D. Laboratory permeability studies can contribute significantly to the quantification of aquifer heterogeneity and to a better understanding of the relationship between flow properties and fracture geometry by means of permeability measurements and determination of the /02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S (02)

2 184 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) fracture geometry using resin impregnation techniques (Baraka-Lokmane, 2002). They can also help to analyse the interaction between fracture and matrix. Much work has been carried out to study the correlation between liquid and gas permeabilities in sandstone (Lovelock, 1977; Sampath and Keighin, 1982; Rasmussen et al., 1993; Bloomfield and Williams, 1995; Jaritz, 1998), but the comparison between liquid and gas permeability has not yet been determined for a fractured sandstone. In this study, the intrinsic permeability is determined using air and brine as the fluid medium. The measurements are obtained for 0.1 m diameter fractured sandstone cores. The permeabilities derived using these methods (k l for liquid, k g for gas) were then compared. This study confirms the importance of difference between gas and liquid permeabilities and explains its causes in the case of fractured and porous sedimentary rock. 2. Characterization of samples Eight core samples of 0.1 m in diameter chosen for the present study were obtained from the middle Stubensandstein unit of the Middle Keuper succession in the Southwest German Trias. The Stubensandstein is characterised by a mixed mineralogy and high matrix porosity. The mineralogical study showed that the samples present well to poorly sorted grains. Quartz is the dominant framework of the sandstone samples (70 80%), with some feldspar (10 25%), and rare micas. The remaining fraction is cement, the most common being clay minerals including kaolinite, smectite and illite, carbonate (calcite or dolomite) and quartz (in the form of overgrowths). Figs. 1 and 2 show examples of the coexistence of several types of cements. The dominant cements of the different samples are listed in Table 1. The samples can be subdivided into three groups with respect to the dominant cementing agent: samples where carbonate is the dominant cement (samples 4 and 5); sample where kaolinite is the dominant cement (sample 8); samples where the dominant cements are clay minerals and carbonate (samples 1, 2, 3, 6 and 7). The samples contain fractures in the form of opening-mode joints, with no appreciable shear off- Fig. 1. SEM photo of cement containing kaolinite (K) and carbonate minerals (C).

3 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) Fig. 2. SEM photo of cement containing smectite (S), kaolinite (K) and illite minerals (I). set and no cataclastic deformation bands. The type, nature and orientation of these fractures all have an effect on the permeability of the sample. Poor choice of sample could mean that the permeability of the matrix would dominate the permeability of the whole sample. The samples were selected so that they contained visible, open, unhealed and unfilled fractures that could be cored roughly parallel to the axis of the Table 1 Cement types and the dominant cement for the eight selected samples Sample number Type of cement Dominant cement 1 kaolinite, carbonate, quartz kaolinite, carbonate 2 kaolinite, carbonate kaolinite, carbonate 3 carbonate, smectite, illite, carbonate, smectite kaolinite, quartz 4 carbonate, kaolinite carbonate 5 carbonate, kaolinite carbonate 6 carbonate, quartz, smectite carbonate, smectite 7 carbonate, quartz, smectite carbonate, smectite 8 kaolinite, quartz, carbonate kaolinite coring instrument. The sampling was carried out with a special electric bore core drill with a diamonddrilling bit (Hilti, DD-160E). 3. Determination of the hydraulic and pneumatic parameters The air and brine permeability tests were conducted with eight samples. The permeability of clay sandstones decreases rapidly and significantly when the salt water present in the sandstone is replaced by fresh water. The sensitivity of sandstone to fresh water is primarily due to the blocking of pore passage by dispersed particles (Khilar et al., 1983), and the permeability reduction due salinity changes occurs regardless of the type of clay (Mungan, 1965). It has been found that the permeability of the core samples begins to decrease at a specific salt concentration. In the literature, this salt concentration has been termed the critical salt concentration (CSC) (Rahman et al., 1995). The CSC is defined as the minimum salinity required protecting the formation from swelling clay damage and mobilisation of clay fines. When the salt concentration decreases below the critical value, the

4 186 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) double layer expands and the force of repulsion between the clay particles increases, weakening the bonds between them. The clay particles can then separate more readily under the hydrodynamic shear stress created by fluid flow, migrate and bridge at pore constrictions, leading to pore plugging (Rahman et al., 1995). Much work has been carried out to study the formation damage in sandstone (Mungan, 1965; Khilar et al., 1983; Khilar and Fogler, 1984; Mohan et al., 1993; Rahman et al., 1995; Civan, 2000; De Lima and Niwas, 2000). The experiments conducted by Mohan et al. (1993) and Mohan and Fogler (1997) have shown that the critical salt concentration (CSC) of sodium chloride needed to prevent damage is between 4 and 5 M for Stevens sandstone and 0.07 M for Berea sandstone. According to the studies reported above, the brine used in the experiments was 9% NaCl brine, with a density of g/cm 3 and a viscosity at 20 jc of 1.31 cp Determination of the gas permeability The gas permeability tests were measured using air and were chosen for eight samples. The intrinsic matrix permeability was measured from small bore cores with volumes equal to approximately 50 cm 3 (4 cm in length and 4 cm in diameter), which did not contain visible fractures. The effective parameters of the matrix/fracture system were estimated with a special tri-axial test cell (Fig. 3). In contrast to the conventional procedures for the determination of the liquid permeability, because of the increase in the loss of the potential in the supply, the gas pressure was measured directly between the sample input and output with two potential probes, which were placed at each end of the sample. The dried samples are sealed with a flexible latex membrane to avoid leakage. The dry air flows from the bottom (sample input) to the top. Because of gas compressibility, flow through the sample is allowed to stabilize before measuring permeability. The pressure difference can be regulated in a short time using the pressure difference adjuster, where the permeability is higher than 1 md, steady state is achieved within 1 min. The apparatus allows the mean sample pressure to be increased without increasing the pressure gradients. Thus, the rate of flow remains low avoiding turbulent conditions. Two parameters are measured: pressure difference (Dp) and the flow rate of air ( Q po ) using a flow meter. The flow rate of air ranged from 0.3 to 3 cm 3 / s and the pressure difference ranged from to Pa. Fig. 3. Equipment for the measurement of gas permeability ( p 1 [Pa]: input pressure, p 2 [Pa]: output pressure, p z [Pa]: confining pressure, Dp [Pa]: pressure difference, Q po [m 3 /s]: flow rate of air.

5 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) Fig. 4. Klinkenberg correction of the apparent permeability (k a ) for sample 3. The differential gas pressure is made as small as possible to provide a suitable rate of air flow. The permeability is determined by means of a multiplepoint gas flow test. The permeability of air, k a,as calculated from Darcy s law (Klinkenberg, 1941) is plotted against the inverse of the mean absolute air pressure on the sample, 1/p (Fig. 4). The permeability is evaluated by extrapolation k ¼ limk a as p! l Compressibility of the gas, the effect of Klinkenberg slip flow as well as the effect of the turbulence was taken into consideration (Bloomfield and Williams, 1995). To avoid turbulent conditions, low gas pressure gradients were used. To insure that turbulent flow did not occur, the flow rates for each sample were measured using a range of pressure differences. The resulting gas flow rate versus gas pressure for sample 5 is shown in Fig. 5; the linear function demonstrates laminar flow. The calculated Reynolds numbers ranged from to confirming that the permeability tests were conducted under laminar flow conditions, and that Darcy s law is valid over the range of pressure gradients used. The gas permeability values (k g ) for the eight core samples varied between 30 and 180 md, while matrix gas permeability values (k gm ) varied between 1 and 4 md (Table 2) Determination of the liquid permeability The intrinsic permeability was measured using 9% NaCl brine (saturated under vacuum) and small bore cores of a volume of 50 cm 3 (4 cm in length and 4 cm Fig. 5. Correlation between hydraulic gradient and gas flow rate for sample 5.

6 188 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) Table 2 The hydraulic, pneumatic and physical parameters of the eight selected samples Sample number k g (md) k l (md) k g /k l Porosity, / (%) Saturation, Sr (%) k gm (md) k lm (md) k gm /k lm Characteristics more than one fracture. large fracture aperture. presence of clay nodules in the matrix one fracture. large fracture aperture more than one fracture more than one fracture one fracture more than one fracture more than one fracture one fracture. large fracture aperture in diameter) without visible fractures, taken from each of the previously described eight samples. Determination of permeability using brine was conducted with the equipment shown in Fig. 6. The sample was sealed in a latex membrane to avoid leakage and a confining pressure of at least Pa was applied. The brine present in the storage container is injected into a saturated sample and the volumes of the brine flow (input and output) were measured with burettes. The effluent brine flowed into a closed storage container where a constant pneumatic pressure was used to maintain a saturation pressure. To insure that the samples were completely saturated, the measurements of liquid permeability were carried out at high saturation pressures ranging from to Pa; any residual air in the core would be and rendered negligible. Pressure differences (Dp) across the samples were ranged from to Pa and flow rates ( Q) from about 0.02 to 0.1 cm 3 /s. The intrinsic permeability value (k l ) are listed in Table 2 along with the matrix permeabilities (k lm ). The intrinsic permeability values (k l ) measured for the eight samples vary between 14 and 50 md. Intrinsic matrix permeability values (k lm ) vary between 7 and 26 md. They are given in Table 2. Fig. 6. Equipment for the measurement of liquid permeability ( p 1 [Pa]: input pressure, p 2 [Pa]: output pressure, p z [Pa]: confining pressure, Q 1 [m 3 /s]: water input volumetric rate of flow, Q 2 [m 3 /s]: water output volumetric rate of flow).

7 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) Fig. 7. Correlation between hydraulic gradient and brine flow rate for sample 3. To check whether the assumption that laminar flow was valid in the presence of fractures, the flow rate was measured at different pressure differences for each sample. The calculated Reynolds numbers p range from to ðre ¼ qu ffiffi k =lþ. This is significantly less than the critical Reynolds number of 1, which indicates incipient turbulent flow. Fig. 7 shows a linear dependency between brine flow rate and the hydraulic gradient for sample 3, demonstrating that Darcy s law is valid over the range of pressure gradients studied. Therefore, the permeability tests were conducted at laminar flow condition. 4. Correlation between liquid and gas permeability The liquid and gas permeability tests were conducted under laminar flow conditions. The intrinsic gas permeability (k g ) ranged from 31.9 to 159 md, while intrinsic liquid permeability (k l ) ranged from 12.3 to 46.9 md. The degrees of correlation between the two permeability measurements are analysed on the basis of k g /k l ratios, Table 1. Two tendencies in the distribution of k g /k l were observed. (1) Ratios ranging from 1 to 2 (samples 3, 4, 5, 6, and 7), and (2) ratios ranging from 4 to 5 (samples 1, 2, and 8). Scatter between liquid gas permeabilities are also reported in the literature (Lovelock, 1977; Sampath and Keighin, 1982; Rasmussen et al., 1993; Bloomfield and Williams, 1995; Jaritz, 1998) with a tendency for liquid permeability to be lower than gas permeability. The ratio of gas permeability to liquid permeability for sandstone cores ranges from 1 to 30 (Table 3). The sandstone used in this study is from the same outcrop used by Jaritz (1998). The ratio between intrinsic gas and liquid permeabilities is between 1 and 3, for the sandstone cores and between 1 and 5 for the fractured sandstone cores (reported in Table 2). There are several reasons for the scatter of liquid gas permeabilities Migration of clay particles Three types of clays are present in the cement of the measured samples: kaolinite, illite and smectite. Strongly hydrophilic minerals, particularly swelling clays can be important in controlling the liquid permeability (Mungan, 1965; Sampath and Keighin, 1982; Khilar et al., 1983; Khilar and Fogler, 1984; Bitton and Gerba, 1984; Appelo and Postma, 1993; Mohan et al., 1993; Rahman et al., 1995; Bloomfield and Williams, 1995; De Lima and Niwas, 2000; Civan, 2000). Smectites can swell with changing ionic conditions and eventually disperse and migrate with the Table 3 Scatter of liquid gas permeability data from the literature k g /k l Nature of samples Number of samples Authors 1! 5 sandstone 1155 Lovelock, ! 10 sandstone 10 Sampath and Keighin, ! 4 fractured tuff 105 Rasmussen et al., ! 30 sandstone 55 Bloomfield and Williams, ! 3 sandstone 15 Jaritz, 1998

8 190 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) flowing fluid. Swelling alone reduces the effective area for flow and causes reduction in permeability. Kaolinites and illites are non-swelling clays that tend to detach from the rock surface and migrate when the colloidal conditions are conducive for release. The migrating particles can get trapped in pore throats, thus causing a reduction in permeability (Mohan et al., 1993). This is generally observed for the samples with a large fracture apertures (greater than 40 Am) (Baraka- Lokmane, 2002), where the values of liquid permeability are lower than the values of the gas permeability; this is explained by the presence of the clay minerals within the fracture aperture, which are responsible for the reduction of the water flow rate. The presence of clay minerals within the fracture aperture was verified by independent methods (resin casting and magnetic resonance imaging techniques) (Baraka-Lokmane, 1999, 2002; Baraka-Lokmane et al., 2001) Breakdown of original fabrics The passage of wetting fronts across relatively delicate clay mineral complexes can breakdown the original fabrics and may physically alter the geometry of some of the flow channels, causing irreversible changes in the permeability of the medium and therefore modifying liquid permeabilities in a complex manner (Lovelock, 1977; Bloomfield and Williams, 1995) Undersaturation of samples Undersaturation of samples during liquid permeability measurement may cause a systematic underestimation of liquid permeability. It results liquid permeability values consistently lower than the gas permeability values. It may not be possible to fill all of the pore space with brine during the liquid permeability measurement, particularly in finer grained sandstones (Lovelock, 1977). To determine the reasons for the scatter of liquid and gas permeability data, physical parameters such as: matrix porosity (/), and the degree of saturation (Sr), matrix gas (k gm ) and liquid permeability (k lm ) tests were conducted for the eight samples whose characteristics are presented in Table 1. The matrix porosity measured with the help of a Helium Pycnometer, as well as the matrix gas and liquid permeability were measured from small bore cores with volumes equal to approximately 50 cm 3 (4 cm in length and 4 cm in diameter), which did not contain visible fractures. During the liquid permeability measurements, the saturation degree Sr is calculated as follows: Sr = 1 + po/p (Sro 1), where p is the mean absolute air pressure on the sample during the liquid permeability test, po is the atmospheric pressure, and Sro is the degree of saturation at atmospheric pressure. The saturation degrees were measured from the previously described fractured bore cores. 5. Discussion The ratio between matrix gas permeability (k gm ) and liquid permeability (k lm ) for sample 2 is equal to 2.46 (Table 2). The high value of the ratio is due to the presence of clay nodules in the matrix (Baraka-Lokmane, 1999). In addition, the scatter between the liquid and the gas permeability is particularly high for the samples where a large fracture aperture is observed (43.6 Am for sample 1 and 43.6 Am for sample 8) (Baraka-Lokmane, 2002). In these samples, the two first processes cited above (migration of clay particles and the breakdown of original fabrics caused by the passage of wetting fronts across relatively delicate clay minerals complexes) are probably due to the presence of clay minerals within the relatively large fracture aperture. The mineralogical study shows that the clay minerals are the dominant cement of these samples. For sample 8, the scatter between matrix gas and liquid permeability is also explained by undersaturation of this sample during liquid permeability measurements; indeed, the degree of saturation is lower than 80% (Baraka-Lokmane, 1999). This phenomenon occurs for finer grained sandstone, which is the case for this sample. For sample 2, there is no difference between matrix and the sample gas and liquid permeability ratio. The fracture of this sample does not run right through the core sample (Baraka-Lokmane, 2002); thus, the contribution of hydraulic conductivity of the fracture to the total hydraulic conductivity of the sample is negligible. The gas permeability is about 4.5 times higher than the liquid permeability; therefore, the clay minerals

9 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) present (as cement in the matrix) seem to be responsible for the scatter between gas and liquid permeability values. This study has shown that the gas permeability is carried out much faster than liquid permeability technique and is less prone to experimental problems and errors. Gas permeability technique measures more closely the intrinsic permeability, therefore it is to be recommended especially for clayrich fractured porous rocks. 6. Conclusions Both liquid and gas permeability tests were performed under conditions that insure laminar flow. Intrinsic gas permeability (k g ) ranged from 32 to 159 md, while intrinsic liquid permeability (k l ) ranged from 12 to 47 md. There is a general tendency for intrinsic liquid permeability to be lower than intrinsic gas permeability. For fractured sandstone cores, there are two tendencies in the distribution of the ratio of intrinsic gas permeability to intrinsic liquid permeability (k g /k l ): (1) ratios ranging from 1 to 2 (62.5% of the samples), and (2) ratios ranging from 4 to 5 (37.5% of the samples). Three samples (1, 2 and 8) differed from the remaining five samples. These three samples exhibit ratios of intrinsic gas permeability to intrinsic liquid permeability ranging between 4 and 5. The factors reducing the liquid permeability are governed by three phenomena: (1) physicochemically, by the migration of the clay particles which clog the pores, (2) mechanically, by the breakdown of original fabrics caused by the passage of wetting fronts across relatively delicate clay mineral complexes, and (3) experimentally, by undersaturation of samples during water permeability measurements. It has been recognised that liquid permeability tests are very time consuming and because gas permeability measures more closely intrinsic permeability especially for clay-rich rocks, gas is a desirable replacement fluid. List of symbols k Intrinsic rock permeability, L 2 k a Apparent permeability using air pressure gradient, L 2 k g Intrinsic gas permeability, L 2 k gm Intrinsic matrix gas permeability, L 2 k l Intrinsic liquid permeability, L 2 k lm Intrinsic matrix liquid permeability, L 2 p Pressure, Pa p Mean absolute air pressure on the sample, Pa p 1 Input pressure or the pressure at x 1,Pa p 2 Output pressure or the pressure at x 2,Pa po Atmospheric pressure ( Pa) p z Confining pressure, Pa Q Liquid volumetric rate of flow, L 3 /T Q g Gas volumetric rate of flow, L 3 /T Q po Flow rate of air, L 3 /T Q 1 Liquid input volumetric rate of flow, L 3 /T Q 2 Liquid output volumetric rate of flow, L 3 /T Re Reynolds number Sr Saturation degree Sro Saturation degree at atmospheric pressure u Darcy s velocity or filtration velocity, L/T Dp Pressure difference, Pa l Dynamic viscosity of fluid, M/LT q Density of fluid, M/L 3 / Matrix porosity Acknowledgements This work was conducted as part of the project Festgesteins-Aquiferanalog: Experimente und Modellierung, funded by Deutsche Forschungsgemeinschaft (DFG), Germany. The author thanks Prof. Dr. Georg Teutsch and Dr. Bryne Ngwenya for their constructive comments on drafts of the manuscript, as well as the managing editor and the two anonymous reviewers for their constructive critical review of the manuscript. References Appelo, C.A.J., Postma, D., Geochemistry, Groundwater and Pollution. A.A. Balkema, Rotterdam, The Netherlands. 536 pp. Baraka-Lokmane, S., Determination of hydraulic conductivities from discrete geometrical characterisation of fractured sandstone Cores, PhD Dissertation. Department of Applied Geology, University of Tuebingen, Germany. 119 pp. Baraka-Lokmane, S., A new resin impregnation technique for characterising fracture geometry in sandstone cores. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 39, S15 S23.

10 192 S. Baraka-Lokmane / Journal of Petroleum Science and Engineering 36 (2002) Baraka-Lokmane, S., Teutsch, G., Main, I.G., Influence of open and sealed fractures on fluid flow and water saturation in sandstone cores using magnetic resonance imaging. Geophys. J. Int. 147, Bitton, G., Gerba, C.P., Groundwater Pollution Microbiology. Wiley, New York. Bloomfield, J.P., Williams, A.T., An empirical liquid permeability gas permeability correlation for use in aquifer properties studies. Q. J. Eng. Geol. 28, Civan, F., Reservoir Formation Damage. Butterworth-Heinemann Publ. 740 pp. De Lima, O.A.L., Niwas, S., Estimation of hydraulic parameters of shaly sandstone aquifers from geoelectrical measurements. J. Hydrol. 235, Jaritz, R., Quantifizierung der Heterogenität einer Sandsteinmatrix am Beispiel des Stubensandsteins. PhD Dissertation. Department of Applied Geology, University of Tübingen, Germany. 106 pp. Khilar, K.C., Fogler, H.S., The existence of a critical salt concentration for particle release. J. Colloid Interface Sci. 101, Khilar, K.C., Fogler, H.S., Ahluwalia, J.S., Sandstone water sensitivity: existence of a critical rate of salinity decrease for particle capture. Chem. Eng. Sci. 38 (5), Klinkenberg, L.J., The permeability of porous media to liquids and gases. Drill. Prod. Pract., Lovelock, P.E.R., Aquifer properties of Permo-Triassic sandstones in the United Kingdom. Bull. Geol. Surv. G. B. 56, Mohan, K.K., Fogler, H.S., Colloidally induced smecticic fines migration: existence of microquakes. AIChE J. 43 (3), Mohan, K.K., Vaidya, R.N., Reed, M.G., Fogler, H.S., Water sensitivity of sandstones containing swelling and non-swelling clays. Colloids and surfaces A. Physicochem. Eng. Aspects 73, Mungan, N., Permeability reduction through changes in ph and salinity. J. Pet. Technol., (Volume, December). Rahman, S.S., Rahman, M.M., Khan, F.A., Response of lowpermeability, illitic sandstone to drilling and completion fluids. J. Pet. Sci. Eng. 12, Rasmussen, T.C., Evans, D.D., Sheets, P.J., Blanford, J.H., Permeability of apache leap tuff: borehole and core measurements using water and air. Water Resour. Res. 29 (7), Sampath, K., Keighin, C.W., Factors affecting gas slippage in tight sandstones of cretaceous age in the Uinta basin. SPE, J.P. Tech.,