Scribd Logo
Upload

Welcome to Scribd!

  • Spe 38682

    Uploaded by

    Fareed
    5 views16 pages

    Original Title

    SPE-38682

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    5 views16 pages

    Spe 38682

    Uploaded by

    Fareed

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 16
    Q SPE 38682 Society of Petroleum Engineers A Comprehensive Dynamic Theory of Hydrocarbon Migration and Trapping Fareed |. Siddiqui, SPE, Mobil Technology Company and Larry W. Lake, SPE, The University of Texas, Austin Copyright 187, Sooo Fate Eraners. ne “Sipser nas pepe to patric he 157 SPE 72 hal Tet Cato Team tiotSe ro Ton 58 Oct 197. anton toe Sac arcaum Engnears, te oaon sr eb Pana peseed [Fe une we nent to pussy rot by Eee Cea oe Sony {erconnarom poner stra Pe wen sao tt Seco Pesinen Crores ‘oben Porn erecuce Mp eeriand on bea of x ne han 30 ‘rte "havetore wey tt be coped Te soaiec ust conan “cnapca Seana Rebun TH /sseoaeae USK eolat ese en EPS Abstract Understanding hydrocarbon migration and trapping is important since it can mean the difference between success and failure in exploration projects. The current understanding is that a capillary pressure change between a seal and a cartier bed (or reservoir) is the main factor responsible for the trapping. The current theory uses capillary pressure gradients under static (no flow) conditions to define the maximum amount of hydrocarbon that can be trapped under a particular seal It assumes that at the very low flow rates encountered in secondary migration viscous pressure drops are negligible. Using numerical simulation and theoretical analysis, we show that, even at very low flow rates, viscous pressure drops are not negligible and that pressure gradients within phases can be substantially different from the static gradients. We present a theory that includes the effect of both the capillary and viscous forces. An innovative way of including the effect of capillary pressures in the method of characteristics is used to solve the migration and trapping problem. Migration and trapping are explained as a result of reflection and refraction of non-linear saturation waves from the heterogeneity boundaries. When viscous forces are included the seals can trap substantially more hydrocarbons than those predicted by the current theory. It is possible to classify seals into static and dynamic seals based on their capillary pressure curves and on the petrophysical properties of the carrier bed. In both cases, wwe are able to associate a time scale to the accumulation and indicate explanations for several other features commonly observed in secondary migration. The results from the proposed theory are confirmed using numerical simulations. 395 Introduction Hyrocarbons (oil and gas) are formed by the decomposition ‘of organic solids deposited in fine grained sediments, mostly shales. With subsequent burial, the pressure and temperature in these rocks increase and some of the bonds in the kerogen are broken to produce oil or gas. After their production, these hydrocarbons must be transported and concentrated into more porous and permeable regions to form hydrocarbon reservoirs. ‘The movement of hydrocarbons just after their formation in the source rocks until they reach the more permeable rocks is called primary migration (Fig. 1). Primary migration finishes when hydrocarbons are expelled from the fine grained rocks into the large permeability rocks called carrier beds. ‘The subsequent movement of hydrocarbons after they emerge from the source rock is called secondary migration (Fig. 1). ‘There are different ways in which secondary migration can ‘occur: in solution, as micelles or as a separate hydrocarbon Phase. Most of the evidence points to separate phase migration.’ Even in the case when the migration occurs in ‘ther forms, the hydrocarbons must come out of solution to form traps. Thus, the final stage of secondary migration will always be a separate hydrocarbon phase migration. This paper assumes that all the migration occurs in a separate hydrocarbon phase, either an oil or a gas phase, depending on the temperature and pressure conditions and the composition of the fluids. ‘As hydrocarbons enter the large pores of a carrier bed they may coalesce to form larger globules. These large globules will move up by buoyancy. Hydrocarbons move in these carrier beds until they reach locations where further ‘movement is partially or totally stopped. The obstacles to the farther movement of hydrocarbons are called seals. The region beneath the seal that contains the trapped hydrocarbons at high concentrations is called a hydrocarbon trap or a hydrocarbon reservoir (Fig. 1) No seal is perfect. They all fail under certain conditions, allowing the hydrocarbons to leak from the trap. Leakage is in effect a continuation of secondary migration although it is sometimes also referred as tertiary migration. After leaking from a trap, the hydrocarbons may trap under another seal or ‘may ultimately seep tothe surface. ‘A thorough understanding of secondary migration is 2 Fareed |. Siddiqui and Larry W. Lake SPE 3682 important as it is the necessary final step before the entrapment of hydrocarbons. It is useful in explaining the distribution of hydrocarbons in a basin, and can be used to predict the nature of petroleum in an undeilled prospect ** Background. ‘The process of hydrocarbon trapping during secondary migration is the same as that which traps hydrocarbons at a heterogencity boundary during water fooding, Describing it involves the study of multiphase flow across a heterogeneity with flow perpendicular to the heterogeneity boundary. In reservoir engineering the description of flow across a heterogeneity is important in both secondary and enhanced oil recovery (EOR) processes since the heterogeneities can impede the flow of a particular phase and thus affect recovery ‘The process of hydrocarbon trapping during secondary migration is the same as that which traps hydrocarbons at a heterogeneity boundary during water flooding. Describing it involves the study of multiphase flow across a heterogencity with flow perpendicular to the heterogeneity boundary. In reservoir engineering the description of flow across a heterogeneity is important in both secondary and enhanced oil recovery (EOR) processes since the heterogeneities can lmpede the low of a particular phase and thus affect recovery. Under isothermal conditions capillary, viscous and buoyancy forces are the most important in multiphase flow in permeable media. The current theory of migration and trapping completely ignores viscous forces and only partly includes capillary forces. Buoyancy is the main driving force and the displacement pressure, because of the seal's capillary Pressure, is the only retarding force.” ‘An earlier paper presented a capillary pressure-free theory ‘of secondary migration and trapping.” It showed the importance of viscous forces and that they alone can cause trapping to occur. This work is an extension of the previous ‘work where we add the effects of capillary pressure to present complete, unified theory of secondary migration and trapping. We first present the general equations for the wo- phase flow. Then we derive the current static theory by neglecting viscous forces. We then briefly present the capillary pressure-free theory and the method of characteristic solution technique.*? Pressure profiles show thatthe effect of both viscous and capillary forces can not be neglected in general. We then extend the method of characteristic solution of the capillary pressure-free theory to include the effects of capillary pressure and use it to present the new unified dynamic theory. The theory is compared to numerical simulations, tothe current theory and tothe capillary pressure- free theory. The effect of finite injection time from the source rock and the effects of wettability are also studied. Although wwe have used oil in the following discussion, the theory is valid for any hydrocarbon phase. General Equations of Two Phase Flow. Following assumptions are made in the general flow equations to arrive at the mass conservation equations. These assumptions are 396 restitive, but they preserve the most important variables in the migration and trapping problem |. One dimensional flow. Most of the migration occurs initially venically upward and then along the top ofthe carrier bed, both of which can be considered a one dimensional flow. 2. Isothermal flow. The distances of secondary migration can often be large and may include significant vertical migration that may entail significant temperature change. This may be important if the secondary migration is occurring in solution because it accounts for exsolving of hydrocarbons.’ For the case of separate phase hydrocarbon migration, the only effect is on the density, viscosity and interfacial tension, but these effects are considered secondary. Therefore, energy balances are not required and mass and momentum conservation ‘equations are enough. Other assurmptions are as follows 3. No adsorption or chemical reactions 4. Incompresible fluids and rock 5. Two phase flow of oil and water 6. No partitioning between the phases Invoking the above assumptions, the conservation of mass equation for each phase becomes Be, Ae 4 a a (la) Bw , aw “a & uD where u and u, are the oil and water fluxes respectively, and x and ¢ are space and temporal variables respectively. x is positive in the vertical upward direction, If only oil and water are occupying the pore space then @ Sy + Sy ‘Thus S, and Sy are dependent and only one saturation is needed for the formulation of fluid flow equations. We use ‘So. as the independent variable. ‘The oil and water fluxes are defined by Darcy’s law wo - Ma [ Bes p,esina] 09 4 : - thee [a : px sina] 6) we me La Where & is the single phase permeability, i, is the relative permeability, 11 isthe viscosity, is the density, o and w are the subscripts for oil and water respectively and a is the dip angle measured from the horizontal (Fig. 1). P, and Py are the pressures in the oil and water continuous phases, respectively, The capillary pressure at any point is defined as the difference in pressure between the oil and water continuous phases, i, SPE 38682 ‘A Comprehensive Dynamic Theory of Hydrocarbon Migration and Trapping 3 Pe=Po-Py (a) ‘Adding Eq, (1a) and (1b), and using Eq, (2) gives My tay = (0) (9) where 1, the total flux, is space invariant. Equation (5) is @ result of the incompressible flow assumption. For the sake of simplicity a constant flow rate, 1, is assumed. Solving for and eliminating pressure gradients between Eqs. (3)-(5), gives the oil lux in terms of the total flux and capillary pressures, Aro. + Fe 4, aaa Hira are =} Cy Where Ayo" hro/#o is the oil relative mobility and Ar hry / Hy isthe water relative mobility Detring f= 2 — an exif (6 gt 16 = fi + Pry dP~ Retry FE Fin the above equation is the fractional flow of oil when capillary pressure and gravity are neglected. Substituting Eq. (7) into Eg. (1a) gives an equation that describes one dimensional, two phase, unsteady state flow including the effects of capillary pressure and gravity. oSo+8 22-0 aa wae The above equation is a non-linear parebolic_ partial differential equation that is difficult to solve analytically. In addition to the above equation, appropriate initial and boundary conditions, relative permeability and capillary pressure functions are required to completely specify the problem The following forms of relative permeability functions are used. [sa1* Phapydre- fh 8) rw = Kew (1-S)" @) where £2, and 42, are the oil and water end-point relative permeabilities, and ny and nz are the saturation exponents, respectively. Sis the normalized oil saturation defined as kro kf s- —Se-Se_ (19 T= Sor - Swr where Sq, and Sypare oil and water residual saturations, respectively.” We have used k%y= 0.2, kf)= 0.8, n)=4 and ny = 2. In the rest of the paper we have used S,, instead of S to define the normalized saturations. Capillary Pressure Model. Capillary pressure was defined in Eq, (4) as the difference in pressure between the oil and water 397 continuous phases. Capillary pressure is a strong function of the phase saturations, mean pore size and pore size distribution in the permeable medium. Leverett'' proposed a non-dimensional form of capillary pressure relationship which when modified for wettability becomes Pe = Io) an ves We use the Brooks-Corey form of the Leverett j-function.”” Ho) = (1 - So) (12) where Sis the normalized oil saturation. Equations (11) and (12) show that a fixed pressure (c'cas6)/J&/¢ ) is required to start the injection of a non-wetting phase into a medium completely saturated with the wetting phase (S,~0). The value of this pressure is @ function of the interfacial tension (7), the wettability of the permeable medium () and the ‘mean pore diameter (f/d). This pressure is called the displacement pressure, Py. nye in Eq. (12) is a positive ‘number that characterizes the distribution of pore throat sizes in the permeable medium. An np, of zero represents a permeable media in which all of the pore throats are of the same size. For such a permeable medium the capillary pressure assumes a constant value equal to Py Present Understanding of Migration and Trapping. The current theory of trapping is usually derived by makin force balance between the buoyancy and capillary forces.” However, here we derive it from the general theory of multiphase flow. We do this to show the static nature of the present theory. The static theory arises from setting. 1, =u, =0 in Eq. (1). Thus Pe 4 14 a7 ee (Ih) ‘The present theory can be explained using the schematic of Fig. 1. In Fig. 1 the permeability of layer 2 (the seal) is smaller than the permeability of layer 1 (the reservoir). The displacement pressures for each of these layers is given by Eq. I. According to Eq. 11 the displacement pressure for the seal, Pgs, is greater than the displacement pressure for the reservoir, Pg. When a hydrocarbon slug reaches the reservoir-seal boundary it can not enter the seal until the hydrocarbon phase pressure increases to balance the displacement pressure of the seal Pys. This results in the trapping of hydrocarbons below the reservoir-seal boundary. With a continuous supply of hydrocarbons from the source rock the length of trapped hydrocarbon column increases. If ‘Pz = Pag. at the bottom of the hydrocarbon column in the reservoir (x=0), then the hydrocarbon column required to 4 Fareed | Siddiqui id Larry W. Lake SPE 38682 attain Pas at the reservoir-seal boundary can be found by integrating Eq. (14). This hydrocarbon column length is denoted a5 hygye- Thus Ps oat Tah, = TA pe de foraconstant Ap this becomes Pa 0 Ione = TFB. 15 ane ~ “BR 3) Froanic's the maximum amount of hydrocarbons than can be trapped under 2 specific seal, any further supply of hydrocarbons leak into the seal. Deviations from the state theory are explained by using ground water flow, which is approximated by a constant pressure head against a static hydrocarbon column.'* The effect of relative permeabilites and saturation dependence of capillary pressure is completely neglected. Insufficiency of the Static Theory. The only published data in which actual hydrocarbon column thicknesses are compared to those predicted by Eq. (15) was published by Berg.’ Table 1 is a reproduction from Berg,* Of the four reservoirs, three showed hydrocarbon columns greater than those predicted by Eq. 15. Even in the fourth reservoir, of 2 Milbur, Texas, field, the smaller of the hydrocarbon columns should be taken because the leakage will occur from the ‘weakest point in the seal, i. the point with the largest pore diameter. Table 1 suggests that the static theory consistently underestimates. the maximum length of the hydrocarbon column tapped under a seal. Berg suggested that the discrepancy could be resolved if ground water flow was assumed. Ground water flow has been used very often as a fudge factor to explain away any deviance from the predicted values. But many times it is very difficult to prove its presence, Berg himself noted that “ground water flow has not been demonstrated” in any of the cases he studied. The theory's underestimation of the hydrocarbon column points 0 the fact that some of the retarding forces are neglected in its formulation. The static theory completely neglects the effects of viscosity and relative permeability, which are very imporant in multiphase flow, The relative permeability effects are a function of saturation of the phases which are a function of time. The additional pressure drops because of viscous forces may be able to explain the longer- than-static hydrocarbon columns Because of the current theory's inherent static nature, hydrocarbon influx rates have no effect on the location ofthe trapping or on the capacity ofa seal. We can isolate the effets of viscous forces by neglecting the capillary pressure completely. In our earlier work we presented a capillary pressure-iee theory of hydrocarbon migration and trapping that shows that viscous forces alone can cause trapping of hydrocarbons. We present a brief review of the method of 398 Solution here as we will extend the same method of solution for the unified dynamic theory. The theory describes the migration and trapping of hydrocarbons as an interaction between waves moving through a hetcrogencous permeable ‘medium. Waves are saturation or composition changes which, under the conditions used here, move with constant velocity Capillary | Pressure-Free Theory of Hydrocarbon Migration.* When capillary pressure is neglected in Eq. (7) the oil (or water) flux is uniquely determined as a function of oil saturation through the relative permeability relations. WolSo)= fu, + FkAy APE (16) With this flux function, the oil conservation equation, Ea. 8, reduces to a hyperbolic quasi-linear partial differential tquation that, can be solved using the method of characteristics.*'*" Be 1 die Beg aay 9 So & The method of characteristics provides graphical solutions that give fundamental insights that are difficult to obtain by ‘other techniques. According to the method of character the specific velocity of a constant saturation, S.,, is equal to the slope of the flux-saturation curve at that saturation, thus* a 1 | duol So) vee |* tolSe) (184) “ (3, al Bo lL » ‘The relationship between saturation and flux is non-linear, causing du,/dS, to be a strongly non-linear function of S, ‘The non-linearity allows continuous waves to develop discontinuities or shock waves because of the overtaking of slower saturations by faster ones. The specific velocity of these shock waves (shocks) follow from the conservation of mass considerations." J Au Fug = uc toc =1 Ate = Ete ue age Yet 91s, 855-87 where (ug, $3) and (us. Sc) are the points upstream and downstream of the shock, respectively. Equation (186) suggests that vsjock 8 the slope of the chord joining the poims (uj, $6) and (x, 8°) ona uy vs. Sy pl Shocks are characteristic features of ‘di free hyperbolic conservation equations, When dissipation effects (uch as diffusion, capillary pressure, compressibility) are present, the shocks are smeared or spread out around the shock front position, but the positon of the shock is unaltered ‘We will make use of this fact to develop the dynamic theory later in defining the leaking and trapped fluxes. Equations (18a) and (b) show that the shape of flux-saturation (1, ~S,) ‘SPE 38682 ‘A Comprehensive Dynamic Theory of Hydrocarbon Migration and Trapping 5 curve is the main factor in defining the character of the solutions. Fiux-Saturation Relationship. Figure 2 shows 1, -5, curves for layers of two different permeabilities. The high permeability layer is labeled as the reservoir while the low permeability layer as the seal. The relative permeability dependence isthe same in both layers. Figure 2 is a schematic showing the u, ~S, curves for 1, =0 (no regional flow). The case with regional flow is discussed in cetal in reference 7. ‘The case of u,=0 defines pure countercurrent flow in which the oil and water fluxes are equal in magnitude but ‘opposite in direction at every point in the permeable medium. uy goes to zero at both S, =O and Sq ~/ and has a maximum at some intermediate saturation. The oil flux is zero at Sy =0 because the relative permeability to oil is zero. At S, =/ the relative permeability to oil is maximum but the relative permeability to water is zero. As no water can flow to balance the oil, the oil flux is also forced to be zero. This makes these ccurves different from the S-shaped curves of the Buckley- Leverett problem." The curves also have negative slopes showing the possibility of reverse flow or waves moving away from the seal for certain saturations. Similarly, in contrast to cocurrent flow where the maximum flux occurs at the extremes of saturation, the maximum in countercurrent flow ‘occurs at some intermediate S, where both of the phases are mobile, The maximum in the flux curve defines the maximum. possible flux st any point in the permeable medium (Clomar)-Momaxf0r the seal is called the seal capacity. The shape of the curve is also sensitive to wettability. An increase in wettability to a certain phase shifts the peak hydrocarbon. flux to occur at higher saturation of that phase. The effect of wettability is discussed briefly later in this paper and in detail in references 6 and 7. Method of Solution for Capillary Pressure-Free Theory. ‘We assume that there is a source rock supplying hydrocarbons into the carrier bed at a constant rate, 1, . Both the reservoir and the seal are completely saturated with water initially (Sq1=0). The trapping behavior of the seal will depend on Whether the influx fiom the source rock is less than or greater than the seal capacity. The rule for the construction of the solution isto take the maximum possible lux that satisfies the conditions of continuity of flux across the heterogeneity boundary. The solution for the case when the influx is less than the capacity of the seal is shown on a u, ~S,curve in Fig. 2 and on time-distance (x-/) diagram in Fig. 3. The initial and boundary conditions in the reservoir result in a shock wave whose velocity is given by the slope of the chord joining points J7 and Jon the tu, —S, curve for the reservoir. When this wave reaches the reservoir-seal boundary at x (Fig. 3), continuity of flux implies thatthe saturation atthe reservoir- seal boundary changes from Sy down-stream to Sz up- stream (Figs. 3(a) and (b). The solution for x> x, is governed by the flux function of the seal in Fig. 2. The shock velocity, given by the slope of the chord f-/2 in Fig. 2, is less than the shock velocity in the reservoir. No accumulation results in the reservoir; all the hydrocarbons coming into the reservoir pass into the seal and proceed in it with a velocity defined by its properties. Figures 4 and 5 show the solution when the influx from the source rock is greater than the seal capacity. The initial shock wave in the reservoir is the same as in the previous case. When this wave reaches the boundary, all of the ofl carried by this wave cannot be transmitted into the seal because of its lower capacity. Continuity of flux at the imerface can only be satisfied if the flux at the interface is reduced to a value equal to the capacity of seal, ugsmpar This reduction results in a reflected shock wave whose velocity, as shown in Fig. 4, isthe slope of the chord between J/ and 77. The rule used for the construction of the solution isto allow the maximum flux, that satisfies the condition of continuity of | ‘lux, to pass through to the seal. This reflection will start the accumulation of oil in the reservoir. The flow in the seal will follow the curve in Fig. 4 from (Somae:¥%as max) 10 1, which will result in a shock wave (from 1 to J2) followed by @ spreading waves (from tps max'0 J2XFigs. 5(a) and (b)). The saturation at the interface between reservoir and the seal will Jump from S; t0 Somax~ The spreading waves are shown as shaded regions on the x-! diagrams Siddiqui’ and Siddiqui and Lake® discuss several cases with the effect of groundwater flow (1, = 0), the effect of hydrocarbon and water partitioning into each other, the effect of wettability and the effec of finite injection time. Siddiqui’ shows that when capillary pressure is neglected the traps formed are transient. After the stopping of hydrocarbon influx from the source rock, all of the dynamically trapped hydrocarbons leak off. However, the time taken to leak these dynamically trapped hydrocarbons may be geologically significant. The capillary pressure-free theory shows that viscous forces alone can cause the trapping of hydrocarbons at a heterogeneity boundary. However, capillary forces must be important and can not be ignored for a general theory. Next wwe study the pressure behavior of oil and water to study the combined effect of capillary and viscous forces. Pressure Profiles. Using Darcy's law (Eqs. (3) and (6)), with 14,= 0, the equations defining the oil, water and capillary pressure gradients become He Bo Bo bw 6 Faroed |. Siddiqui and Larry W. Lake SPE 38682 6 pe twltrg Be Bey Ge = —___©0_&_ - pyg (196) a hy + Ky HO By oP. te Hay He ce tay He 4 Hoy + doe (19 a k kw ra CIES scons? Dynamic Efecr The first terms in the above equations denote the dynamic ‘or viscous effects while the second terms are the static effects The static effects are independent of the movement of fluids They are also independent of time. The dynamic effects are a function of both the flux rate ofthe phases and time. Viscous effects are a result of the drag caused by the movement of fluid elements against each other and the walls of the pores. They are a function of the viscosity and saturation of the fluids. The dynamic effect is a combination of both the viscous and capillary effects, The contribution of these dynamic effects changes as the saturation changes. According to Eq. (19c) in the regions where cither hydrocarbon or water relative permeabilities are close to zero, the dynamic effect may not be negligible even for small flux rates (uo) ‘According to Eq. (19), in the regions where viscous forces are negligible both oil and water pressures will follow their respective pressure gradients. However, when viscous forces are not negligible both oil and water pressures gradients will be different from their respective static gradients. An extreme case of this occurs in the regions where either the oil or water saturation is close to zero. In the regions where the water saturation is small k;y =0. In these regions the viscous forces are dominant. Equation (19) gives fee. B= - oe #e pg - He So a Poe os, If'in addition 3, /d =0, then Fe 2, ani Gennes and ‘Thus, in tis region both oil and water pressures gradients wil approximate the oil static gradient. Figure 6 is a typical pressure profile for a reservoir-seal system where viscous forces are not negligible. itis based on a large number of numerical simulations run at low influx rates (107 m/s-i0" mi). It shows that for most of the oil column the capillary forces are insignificant, since the oil and water pressure gradients are both equal to the oil pressure gradient (Region Ill). It i only in the region close to the reservoir-seal boundary (Region 1V) and around the saturation front (Region Po 400 1) where capilary forces are large and viscous forces can be neglected, In these regions the dynamic term in Eq. 19 can be roped and the pressure behavior can be described by the static pressure gradients. This pressure behavior provides an entra pressure drop, shown as the dynamic effect in Fig. 6 ‘that causes the pressure in the oil phase atthe boundary to be less than the static case. Thus, in the dynamic case the leakage ‘of hydrocarbons can not be explicitly related to the height of the hydrocarbon column but is @ complex interplay of capillary and viscous forces that determine the pressure Profiles in the hydrocarbon region. Hydrocarbons will Teak into the seal when the oil pressure below the reservoir-seal boundary becomes large enough to overcome the displacement pressure of the seal. If P defines the pressure jjust below the reservoir seal boundary then the hydrocarbons will start to leak into the seal when > = Pys . Unified Dynamic Theory of Secondary Migration and ‘Trapping. Now we present a unified dynamic theory of hydrocarbon migration and trapping that retains the ease of the method of characteristics while includes the effects of capillary pressure, It combines the capillary pressure-free flux-saturation curves and the capillary pressure-saturation ‘curves to explain migration and trapping. The theory explains how the amount of hydrocarbons trapped, and the type of trap ccan be affected by the influx rate from the source rock and its relationship to the displacement pressure of the seal. The time- distance diagrams and the flux diagrams of the method of characteristics give powerful insight into the complicated process of hydrocarbon migration and trapping under dynamic conditions. Boundary Conditions at the Reservoir-Seal Boundary. As discussed in the capillary pressure-fee theory section, the fluxes of both the phases (hydrocarbon and water), upstream and downstream of the reservoir-seal boundary should be the same. In addition to the individual phase fluxes, the pressures in both phases across the reservoir-seal boundary must be equal also, Therefore, @ re @) Where - indicates a point just below the boundary (upstream) and + indicates 2 point just above the boundary (downstream). The rule we use for the construction of the solution is to take the maximum possible flux that will simultaneously satisfy the conditions of continuity of flux and continuity of pressures across the reservoir-seal boundary. ‘The capillary pressure vs. saturation ( P, ~ 5) curves can be used to define the saturations allowed just upstream and downstream of the reservoir-seal boundary. Figure 7(b) shows ‘SPE 38682 ‘A Comprehensive Dynamic Theory of Hydrocarbon Migraton and Trapping 7 the capillary pressure-saturation curves for a reservoir and two different seals. In Fig. 7, S* represents the saturation in the reservoir corresponding to the displacement pressure of the seal (Pys). For oil saturations less than S* in the reservoir, continuity of capillary pressure dictates thatthe corresponding, cil saturation just above the boundary in the seal should be zero. This implies that no oil can flow into the seal as long as ‘the saturation atthe top ofthe reservoir is less than S*. When the saturation at the top of the reservoir becomes equal to S*, hydrocarbons start to flow into the seal. After this time the saturation at the top of the reservoir will remain at S* Any further hydrocarbons reaching the boundary will leak into the seal Dynamic vs. Static Seal. Figure 7 shows how the capillary pressure-free flux-saturation (u,-S,) curves can be used ‘with the capillary pressure-saturation curves to define the trapping characteristics ofa given seal. uy in Fig. 7 isthe cil flux from the source rock into the carrier bed (reservoir) and is the saturation corresponding to this flux on the higher oil saturation flank of the 1, -S, curve. It is also the sninimum saturation for which uaz can produce a reflected wave into the reservoir. Figure 7 shows two types of seals for a given reservoir, depending on whether S* corresponding to the seal is greater or smaller than Sy If S* for the seal is greater than Sg, then the seal is, termed as a namic seal (shown by Sj in Fig. 7(0). In a and ,/a is the only adjustable parameter. Thus &,/& at the boundary adjusts such thatthe flux at the top of the reservoir decreases to tg, (shown by point 77). Figure 14(a) shows the solution on a time-distance lagram and Fig. 14(b) shows the saturation profile at time 1 .As inthe previous cases the injected Mux, way results in a shock, /JI. When this shock reaches the boundary the hydrocarbons stat to trap below the boundary. The saturation at the top of the boundary increases until it becomes equal to 'S} at time, f,. At this time the saturation in the hydrocarbon column varies continuously from 5} at the top to Spat the bottom of the hydrocarbon column. Numerical simulations show that the hydrocarbon column length at ¢ is equal to ‘uae This is expected because of the capillary dominance required to satisfy the condition of continuity of fuxes and pressures simultaneously. ‘The time taken to trap a column of length Aygye under a static seal is much less than the time taken to trap a column of the same length for the no capillary pressure case. This happens because the hydrocarbon column in a static seal is more diffused and thus requires less volume of hydrocarbons to fill the same height. Comparison with the No Capillary Pressure Case. ‘Compare Figs. 13 and 14 to the saturation velocities and the time-distance diagrams for the P, = Ocase (Figs. 2 and 3). ‘The wave velocities in the two cases for the upward moving shocks are the same, The only difference is that capillary pressure delays the leaking of hydrocarbons into the seal until continuity of capillary pressures at the boundary is satisfied This results in an accumulation in the high permeability layer equal 10 Asanee Validation From Simulation. For the same reservoir of permeability $ Darcy, if the low permeability layer is of permeability 1.5 Darcy then from the P. ~ S, relationship for the two layers, S* = 0.7. If myx 10-9 mis, then $470.85. As S*] L ct nj = Exponent for water phase relative permeability mn Exponent for oil phase relative permeability pe = Capillary pressure exponent P) = Pressure [=] FIL? -P Copillary pressure [=] F/? Pq = Displacement pressure [=] F/L? r Radius ofthe capillary throats (=] L Sr Residual saturation of phase j ' = Time [=}t u Volumetric flux [ ¥ = Specific wave velocities x = Distance, positive in the upward direction{-] L @ = Inclination angle of the medium from the horizontal va 4 Difference operator é Porosity 4 Mobility [=] L'vm “a Viscosity [=] m/L-t @ - Density @ = Interfacial tension [=] mit” Subscripts D_— = Dimensionless HC = Hydrocarbon 1 = Initial J injected ° Oil R= Reservoir = Residual or relative Ss Seal t= Total w= Water Superseripts +) = Downstream 2 Fareed | Siddiqui and Larry W. Lake SPE 38682 = = Upstream 0 = End point value References. LTissot, B. P. and Welte, D. H., 1984, Petroleum Formation and ‘Occurrence: Springer-Veriag, Berlin 2Roberts, W. H. Ill and Cordell, R.T., 1987, Problems of Petroleum Migration: Introduction: in W. H. Robert, 111 and R. J. Corde, (Eds), Problems of Petroleum Migration, American Association ‘of Petroleum Geologists Studies in Geology, No. 10: American Association of Petroleum Geologiss, Tulsa, Oklahoma, p. 1 3.Hamms, J.C, 1966, Stratigraphic Traps in a Valley Fill, Westem ‘Nebraska: AAPG Bull, v. $0, p. 2119-2149 4Stone, D. Sand Hoeger, RL, 1973, Importance of Hydrodynamic Factor in Formation of Lower Cretaceous Combination Traps, Big-Muddy-South Glen rock Arca, ‘Wyoming: AAPG Bull, v. $7, p. 1714-1733, S.Berg. B. B; 1975, Capillary Pressure in Stratigraphic Traps: ‘AAPG Bull, v. $9, n0.6, 939 GSiddigui F. I. and Lake, L. W., 192, A Dynamic Theory of ‘Hydrocarbon Migration: Mathematical Geology, v.24, n0. 3, p. 305.327 TSiddiqui, F. 1, 1996, A Dynamic Theory of Hydrocarbon Migration and Trapping, Pb. D. Dissertation: The University of Texas, Austin Table 1. Actual hydrocarbon column lengths vs. Prass(mnodtied from Ref. 5) BLake, L. W..1989, Enhanced Oil Recovery: Prentice-Hall Inc. Englewood Clifls, NJ 9.Wood, J. R., and Hewett, T. A., 1984, Reservoir Diagenesis and Convective Fluid Flow, Part 1. Concepts and Principles: in D. A, McDonald and R. C. Surdam (Eds.), Clastic Diageness: Memoir 37: American Association of Petroleum Geologists, Tulsa, Oklahoma 10.Corey, A. T., 1986, Mechanics of Immiscible Fluids in Porous Media: Water Resources Publications, Littleton, CO U1.Leverett, M. C.,1941, Capillary Behavior in Porous Solids: Trans. AIME, v. 142, p, 152 12Schowalier, T. T., 1979, Mechanics of Secondary Hydrocarbon Migration end Entrapment, AAPG Bul v.63, n0. 5, p.723 13.England, W. A. , Mackenzi, A. S., Quigley, T. M, 1987, The Movement and Entrapment of Petroleum Fluids in the Subsurface: Geol. Soc. London, v. 144, p. 327 14.Buckley, S. E, and Leverett, MC, 1942, Mechanisms of Fluid Displacement in Sands: Trans. AIME, v. 146, p. 107-116 15.Rhee, Hyun-Ku, Aris, Rand Amundson, N. R., 1986, First Order Partial Differential Equations, Vol I: Prentice-Hall, Englewood Cliffs, N. J I6.Saad, N., 1989, Field Scale Simulation of Chemical Flooding, Ph.D. Dissertation: The University of Texas, Austin Table 2. Time to drain 1 m of dynamically trapped hydrocarbons Calculated State Fra | Columns, tiEa. 15) ot__|cas_| ow Gas Mibur, Texas eo1075 osorsa |- Lane, Nebraska | 49, 140190 Main Pass Block 35, [87 [20 | 290r25 |10 otshore Louisiana Paduca, NewMorico | 120e |- | 100__|- ‘Seat Permeability | Leakage Rate | S* | Timetoteak 1m (darcy) (ris (years) 04 | s05x1o% | ose 96 o4 terxi® | 098 16200 ocor_| 110x102 | 0000 | 2.66% 10% ‘SPE 38682 Figure |, Schemati showing diferent components of «hydrocarbon migration system ina malt-layer carr bed, aac fom Suc Rock Tet ® Oi Sain, 5, ° Figure 3, Process of oil migration with influx less than the capacity ofthe seal (a) on une-astancedagra,(b) saturation profile ‘A Comprehensive Dynamic Theory of Hydrocarbon Migration and Trapping 3 oo 02 04 o8 08 19 Normatized Oil Saturation, $y Figure 2. Of and wave velo wth inux rate less than the fx eapacty ofthe seal, oo 02 Sm oa oe ry 10 Normalized Oi! Saturation, S Figure 4. Oilfux and wave velocities with int rate greater than the fux eapacty ofthe seal ih hac fam Soe Rack Disa fom Soc Rak © Figure S. Process fol migration with influx greater han the ‘capacity ofthe eal (a on ume-distance diagram. (6) saturaion profi “4 Fareed |. Siddiqui and Larry W. Lake 1 | Figure 6, Schematic showing the combined effect of viscous and capillary forees onthe presse profes. o O11 Fas, Sy Repo. oo on aa Shoe Son Normalized Oil Suuration Figure 7. Classification of seals onthe basis of fax mateo the source rock nd he displacement pressure ofthe sel. Nommalized Oi Saturation, S, Figure 8. Efecto capillary pressure on cimensionles cil fax and wave velocties fora dynamic sel, aoe inc om Sone Rack x ‘SPE 36682 WoC Weer Cena: Figure 9, Process foil accumulation fora dynamic sel () on tie-

    You might also like