Accurate Measurement of Steam Flow Properties

Kewen Li,R. Horne

Abstract:The accurate estimation of steam flow properties is important for geothermal reservoir engineering and the numerical simulation of reservoir performance. It has been found that the steam phase relative permeabilities at some water saturations may be significantly greater than unity. This unexpected phenomenon has been interpreted physically and reproduced in nitrogenbrine flow. A new method for measuring and correcting gas-liquid relative permeabilities has been developed. The unusual values of gas phase or steam phase relative permeabilities were found to be attributable to the gas slip (Klinkenberg) effect in two-phase flow. The effect of water saturation on gas slippage has been investigated; experimental results show that water saturation has a significant influence on the gas slippage. The steam slip factor was measured using a new apparatus and used to correct the previously measured steam phase relative permeability values. Introduction Steam-water relative permeabilities are fundamental properties in reservoir engineering and numerical simulation for geothermal reservoirs. There are few measurements of steam-water relative permeabilities reported in the literature, and published values are frequently inconsistent. The main difficulties in making steam-water relative permeability measurements arise from the following four aspects: 1) mass transfer between phases makes it difficult to measure or calculate fractional flow; 2) extremely low intrinsic permeabilities of geothermal rocks; 3) difficulties in measuring heat loss during experiments; 4) unclear mechanisms of gas-liquid flow. For example, the effect of gas slippage (Klinkenberg effect) on gas relative permeabilities in steam-water flow is not clear; very few studies related to the gas slip effect in steam-water flow have been published. In this study, we focused on the precise measurement of the slip effect in steam-water flow. Satik and Horne (1998) observed an unusual phenomenon in that steam-water relative permeabilities (krg), at some water saturations sometimes appear to be greater than 1.0 (see Figure 1). A similar phenomenon was also found in a recent experiment conducted by Mahiya (1999). What causes the steam phase relative permeabilities greater than 1.0? Steam is subject to a prominent flow characteristic of gas molecules: gas slippage. Gas permeability kg at a low mean pressure is greater than the absolute liquid permeability (k) due to the gas slip effect. If the gas slippage is large enough, the gas permeability at low water saturation may still be greater than the absolute permeability. From the definition of relative permeability, kg/k, the gas relative permeability may be greater than one. Hence the phenomenon of gas slippage may be the reason. Steam-water relative permeability experiments have shown values of krg greater than one. Kewen Li and Roland N. Horne 2 The gas slippage theory established by Klinkenberg (1941) in single-phase gas flow is expressed as follows: ) 1 1 ( m g g p b k k + = ∞ , (1) where kg is the gas permeability at a mean pressure, pm, and kg∞ the intrinsic permeability of gas at an infinite pressure, respectively; b is the gas slip factor. The mean pressure is defined by pm = (pi + po) / 2; pi and po are inlet and outlet pressures of the cores, respectively. Jones and Owens (1980) measured the gas permeabilities on more than 100 tight gas sand samples with permeabilities ranging from 0.0001 to 1.0 md and their data yielded a relationship between b and kg∞ for single-phase gas flow: 33 . 0 86 . 0 ∞ = g k b , (2) Sampath and Keighin (1982) thought that a correlation, if it exists, should be between b and kg∞/φ rather than between b and kg∞; they suggested a correlation between the gas slip factor and the intrinsic gas permeabilities based on their experimental results using nitrogen: 53 . 0 ) / ( 0955 . 0 − ∞ = φ g k b , (3) where φ is the porosity of the core sample. Sampath and Keighin (1982) found that the intrinsic gas permeabilities decrease with increasing partial water saturations. Counsil (1979) discussed gas slip effect and showed that the effect of gas slippage was small when b = 0.2 atm and pm= 10 atm. So the gas slip effect was not considered in the analysis of the gas flow data. Counsil (1979) stated that “For the case of steam-water relative permeab ilities, slip could be reduced by running experiments at very high pressures, and therefore very high temperatures”. There are certain difficulties in running experiments at very high pressures and temperatures for measuring steam-water relative permeabilities, for example, when the X-ray CT method is used to measure the water saturation in a core (Ambusso, et al., 1996). On the other hand, increasing experimental temperature will increase the value of b significantly as reported by Wel et al. (1986). Increasing b will increase the influence of gas slippage on steam flow properties. Geothermal rock often has a very low intrinsic permeability in an order of 10 -6 md. The value of b computed using Eq. 2 is about 82.1 atm for this permeability. The mean pressure should be equal to about 1000 atm so as to neglect the gas slip effect in single-phase gas flow. In two-phase flow, the mean pressure necessary to neglect the gas slip effect is greater than that in single-phase gas flow. Obviously, it is not easy to operate experiments at pressures over 1000 atm. It can be seen from this simple calculation that the effect of gas slippage on steam flow properties may be very significant in geothermal reservoirs. Herkelrath, et al. (1983) recognized the importance of gas slippage and incorporated it into a single-phase steam flow model in order to model steam pressure-transient experimental data. In their study, the gas slip factor measured using nitrogen gas was assumed to apply to the steam flow. Kewen Li and Roland N. Horne 3 Hence it is seen that it is necessary to measure the slip factor of steam flow in a porous medium in order to obtain reliable relative permeability data for steam. When the data of steam slip factor are available, it will be possible to determine how important the steam slip effect is in steam-water two-phase flow. In this work, an apparatus was developed for measuring gas-liquid relative permeability using an on-line weighing method with a balance. A new procedure for measuring and correcting gas relative permeabilities was proposed. The slip factor of steam was measured using another new apparatus and used to correct the steam phase relative permeabilities. The effects of water saturation on the gas slip factor and the intrinsic gas permeabilities were also investigated. Theory Assuming that Darcy’s equation is valid for steam-water two-phase flow and that capillary pressures are neglected, the effective permeability of steam phase is calculated using the following equation: m o s s w s p p p A L q S k ∆ = μ ) ( , (4) where ks(Sw) is the effective steam phase permeability at a water saturation of Sw; qs and μs are the flow rate and viscosity of steam phase; ∆p is the differential pressure across the core sample; A and L are the cross-section area and the length of the core sample. The relative permeability of steam phase is usually calculated without the correction of steam slip effect as follows: k S k S k w s w rs ) ( ) ( = , (5) where krs(Sw) is the relative permeability of steam phase at water saturation of Sw; k is the absolute permeability of the rock sample. With consideration of the steam slip effect, the intrinsic effective permeability of the steam phase is calculated using the following formula: ) / 1 ( ) ( ) ( m S w s w s p b S k S k