Transient surface permeabilities

Fig. 45 Comparison of the transient concentration profiles during methanol uptake by the MOF-type crystal as recorded by interference microscopy (symbols) with the corresponding profiles recalculated from the measured diffusivities with surface permeabilities (full line in Fig. 44) which lead to the best fit to the experimental points...

Soo and Radke (11) confirmed that the transient permeability reduction observed by McAuliffe (9) mainly arises from the retention of drops in pores, which they termed as straining capture of the oil droplets. They also observed that droplets smaller than pore throats were captured in crevices or pockets and sometimes on the surface of the porous medium. They concluded, on the basis of their experiments in sand packs and visual glass micromodel observations, that stable OAV emulsions do not flow in the porous medium as a continuum viscous liquid, nor do they flow by squeezing through pore constrictions, but rather by the capture of the oil droplets with subsequent permeability reduction. They used deep-bed filtration principles (i2, 13) to model this phenomenon, which is discussed in detail later in this chapter. 

The permeability trends found in this study for the 1000 pm thick membranes under flowing 1000 ppm H2S/H2 were compared to transient permeability results reported previously . Nearly identical trends in permeability were found for the three Pd-Cu alloys by these two measurement methods reinforcing the suggested correlation between the alloy crystal structure and H2S tolerance. The results of these steady-state permeability experiments indicated that when the Pd-Cu alloys had an fee structure, H2S had little impact on flux but when the structure was bcc, H2S had a moderate to severe impact. However, both of these methods had limitations that probably impacted the results. The transient or batch method was limited by finite H2S availability, competition for the available H2S by other metal surfaces, relatively short test durations and a limited data set. The steady-state 1000 pm thick membrane method was potentially limited by membrane thickness which could have... 

Each team performed an uncoupled transient flow simulation. The spatial distribution and evolution of the simulated head are qualitatively similar to the coupled H-M simulations above a certain depth, which varies with time from near surface to 900-m depth. Below this depth, due to the very low rock permeability there is hardly any hydraulic response to glaciation. Residual excess pressures similar to the coupled H-M model results are predicted 9500 years after glacier is gone. In this case, however, the values (350-400 m of pressure head) are almost twice as high. 

The pure gas (99.99% or higher purity) transport properties were tested with an instrument (GKSS, Geesthacht, Germany) with eonstant permeate volume described elsewhere [14], The short response time of the instrument allows one to record transient permeation behaviours of less than 1 second. The pure gas permeability is the amount of gas permeating in the unit time, multiplied by the thickness of the membrane and normalized for the membrane surface and the pressure gradient. The recorded pressure vs. time plots were used to derive diffusion coefficients from the initial transient permeation, and permeability at the steady state. The time-lag 9 is the intercept of the linear part of the pressure vs. time curve with the axis of time. In a homogeneous membrane in which the solubility of a gas obeys Henry s law, its diffusion coefficient D can be calculated from the ratio ... 

The surface-pressure history during polymer injection is predicted by Eq. 1 in the following manner. A skin/permeability quotient that is consistent with brine injectivity at 481 and 301 B/D as well as a pressure-transient test is calculated for use in the model. The shut-in-pressure difference between the injection well, P, and the observation well, Pg, is calculated to obtain the hydrostatic-pressure contribution. A viscosity/velocity functionality is assumed, and the hourly flow rates and BHP s during polymer injection are entered into the flow model to calculate injection pressure vs. time. This injection-pressure history is compared with the actual pilot data. As previously mentioned, the predicted injection pressure is inversely proportional to the area under the viscosity/velocity curve, starting at high velocities close to the wellbore and moving... 

A modelling approach that could fulfil this need is based on the stationary-state approximation to coupled fluid flow and water-rock interaction (Lichtner 1985, 1988). This model represents the chemical evolution of an open, flow-through system as a sequence of relatively long-lived stationary states of the system, which are linked in time by short-lived transients. The basis for the model is the observation that within a representative elemental volume of a rock-water system, the aqueous concentration of any particular species is generally much less than its concentration in minerals. Long periods of time are therefore necessary to dissolve, or precipitate, minerals such that the spatial distribution of mineral abundances, surface area, porosity and permeability is altered significantly. Each time interval represents a stationary state of the system, in which fluid composition, reaction rates and the distribution of primary and alteration minerals vary only as a function of position in the flow path, not of time. 

---