|
Black_Oil |
Top Previous Next |
|
{ BLACK_OIL.PDE *************************************************************** This example considers the transport of oil and water in soil.
The model is given in Gelinas, et al, "Adaptive Forward-Inverse Modeling of Reservoir Fluids Away from Wellbores", (Lawrence Livermore National Laboratory report UCRL-ID-126377) and in Saad & Zhang, " Adaptive Mesh for Two-Phase Flow in Porous Media" (in Recent Advances in Problems of Flow and Transport in Porous Media, Crolet and El Hatri, eds., Kluwer Academic Publishers, Boston, 1998).
The saturation of water is represented by S, with the saturation of oil defined as 1-S. The relative permeabilities of water and oil are assumed to be S^2 and (1-S)^2, respecitvely. The total mobility M is defined as M = S^2/muw + (1-S)^2/muo, where muw and muo are the viscosities of water and oil.
The total velocity, V, and the fractional flux, f, are defined as V = - K*M grad(P) f = [S^2/muw]/M, where K represents the saturation-independent permeability coefficient, and P is the pressure, assuming capillary to be zero and oil and water pressures equal.
If the porosity Phi is taken as constant and gravity effects are negligible, the PDE's governing the system reduce to Phi*dt(S) + div(V*f) = 0 div(V) = 0.
Here we study the flow through a 30-meter box with an inlet pipe in the upper left and an outlet pipe in the lower right. The box is initially filled with oil, and water is pumped into the inlet pipe at a constant pressure. Time is measured in seconds.
-- Submitted by Said Doss, Lawrence Livermore National Laboratory.
******************************************************* }
TITLE 'Black Oil Model'
SELECT smoothinit { Smooth the initial conditions a little, to minimize the time wasted tracking the initial discontinuity } prefer_speed { nonlinearity is not strong - we can get away with shortcuts }
VARIABLES s, p { Saturation and Pressure }
DEFINITIONS muo = 4.e-3 { oil viscosity } muw = 1.e-3 { water viscosity } K = 1.e-12 { Saturation-independent permeability coefficient } Pin = 1.5e6 { Inlet pressure } Pout = 1.e6 { Outlet pressure } M = S^2/muw + (1-S)^2/muo { Total mobility } f = S^2/muw/M { Fractional flux } krw = S^2/muw { Relative permeability of water } phi =.206 { porosity }
xmax = 30 { Box dimensions } ymax = xmax out_ctr = 8 tfrac = 2*out_ctr diam = 2 in_ctr = ymax-out_ctr rad = diam/5
epsvisc = 1.e-6 { A little artificial diffusion helps smooth the solution }
sint = integral(s) { the total extraction integral }
hour = 60*60 day = hour*24 { seconds per day }
INITIAL VALUES s = 0 { start with all oil } p = Pin + (Pout-Pin)*x/xmax { start with a rough approximation to the pressure }
EQUATIONS s: phi*dt(s) - div(K*krw*grad(p)) - epsvisc*div(grad(s)) = 0 p: div(K*M*grad(p)) = 0
BOUNDARIES REGION 1 { fillet the input pipe, and define no-flow boundaries of the box } start(-2*rad,in_ctr-diam) natural(p)=0 natural(s) = 0 line to (0,in_ctr-diam) fillet(rad) line to (0,0) to (xmax,0) to (xmax,out_ctr-diam) fillet(rad) line to (xmax+2*rad,out_ctr-diam)
{ set constant outlet pressure, and "tautological" saturation flux } value(p) = Pout natural(s) = -K*krw*dx(p) line to (xmax+2*rad,out_ctr+diam)
{ reset no-flow box boundaries } natural(p)=0 natural(s)=0 line to (xmax,out_ctr+diam) fillet(rad) line to (xmax,ymax) to (0,ymax) to (0,in_ctr+diam) fillet(rad) line to (-2*rad,in_ctr+diam)
{ set constant inlet pressure and saturation } value(p) = Pin value(s) = 1 line to close
TIME 0 to 120*day by 10
MONITORS for cycle=5 contour(s) as "Saturation" range(0,1) contour(s) zoom(xmax-tfrac+2*rad,0, tfrac,tfrac) as "Outflow Saturation" range(0,1) contour(p) as "Pressure" vector(-K*M*grad(p)) norm as "Flow Velocity"
PLOTS for t = day by day to 20*day by 10*day to 120*day grid(x,y) contour(s) as "Saturation" range(0,1) painted surface(s) as "Saturation" range(0,1) painted viewpoint(60,-120,30) contour(s) zoom(xmax-tfrac+2*rad,0, tfrac,tfrac) as "Outflow Saturation" range(0,1) painted contour(p) as "Pressure" painted vector(-K*M*grad(p)) norm as "Flow Velocity" contour(K*M*magnitude(grad(p))) norm as "Flow Speed" painted
HISTORIES history(sint) at (0,0) as "Extraction"
END |