- Email: [email protected]
Chapter 14 Primary Hydrocarbon Migration
Chapter 14 PRIMARY HYDROCARBON MIGRATION
Hydrocarbons generated in fine-grained sedimentary rocks are probably disseminated at first, but eventually they must move from their host rocks into more permeable. and porous sedimentary rocks t o form an accumulation. The movement of hydrocarbons from nonreservoir rocks to reservoir rocks is called primary migration, and is distinguished from their concentration and accumulation within the reservoir rocks known as secondary migration (Levorsen, 1967). A discussion of the primary migration of hydrocarbons includes three different kinds of problems: (1)The form in which they migrate, such as molecular solution, micellar solution and separate hydrocarbon phase. (2) The cause of migration. (3) The water source and the cause of its movement. Whether the hydrocarbons being discussed are mainly gas or mainly liquid will affect the proportions of them that are moving in solution and in separation, because the solubility of gaseous hydrocarbons is generally higher. The migration mechanism for hydrocarbons in separate phase may differ from that for hydrocarbons dissolved in water. If water movement is important in moving hydrocarbons, the source of the water and the cause of its movement must be examined carefully. The molecular solubility of liquid hydrocarbons in water at relatively high temperatures was recently discussed by Price (1976), who showed that the solubility increased with increasing temperature (Fig. 14-1). The solubility of the Farmer's oil at 160°C is approximately 150 ppm, and the curve shows that it tends t o increase with further increase in temperature. However, these temperature values are much higher than the known temperature range of 60" to 150°C for active oil generation. Dickey (1975) on the other hand, suggested that the flowing stream would have to contain at least 10,000 ppm of hydrocarbons at the time of primary migration. Therefore, it may be very difficult t o believe that most oil migrates as a molecular solution in water. In the case of migration of gas, the situation can be completely different. According t o Dodson and Standing's chart (1944, see Fig. 14-2), the solubility of natural gas in water ranges from 4 cu f t per barrel of water at 500 psi, up to 30 cu f t at 10,000 psi. In other words, most of the gas might migrate in solution in water at the primary migration stage. Another solution mechanism, called micellar solution, was proposed by
27 6
-E
l
FARMERS WHOLE OIL
ALASKA REEDY CREEK
L. A. STATE
TEMPERATURE IN OC
Fig. 14-1. Solubilities of two whole oils (Wyoming Farmers and Louisiana State) and four topped oils (Amoseas Lake, Reedy Creek, Alaska and Union Moonie) as functions of temperature in water. Topping temperature is 2OO0C (392OF). (From Price, 1976.)
Baker (1962). He suggested that hydrocarbon solubility is substantially high if the water contains micelles formed by soaps of organic acids. However, there are several reasons why Baker’s proposal is not plausible as the princi-
-
EXPERIMENTAL D A T A
---- E X T R A P O L A T E D
DATA
Fig. 14-2. Solubility of natural gas in formation water in cu ft/bbl. (From Dodson and Standing, 1944.)
277 pal mechanism of hydrocarbon migration in the subsurface, First, there is no good evidence that such solubilizing micelles exist in substantial quantity in shales. And even if they d o exist in shales, they would not be easily moved because they are not small. Then, the micelles would increase the solubility of the heavier hydrocarbons in water only to a few parts per million - nowhere near the 10,000 ppm o r more that now appears to be necessary (Dickey, 1975). Another difficult point in believing micellar solution to be important in primary migration is that the process of unloading the hydrocarbons carried by the fluid (water, micelles and hydrocarbons) at the final trapping position in the reservoir cannot be thoroughly explained. The preceding discussion may lead us t o conclude that the larger proportion of liquid hydrocarbons must migrate in a separate phase, although the rest can migrate in solution in water. In the case of gas migration, the proportion moving in solution in water can be relatively large, because of its greater solubility in water. Although the form in which hydrocarbons moved at the time of primary migration, and the mechanism of that migration, are not completely understood, the movement of water in fine-grained source rocks must be one of the most important factors. The amounts of organic matter and generated hydrocarbons in the source rocks are quite small in comparison with the amount of water. The movement of the large quantity of water must have influenced and may have controlled the direction and effectiveness of hydrocarbon migration. If the water concerned is meteoric water, the direction in which it moves is controlled by excess pressures generated by the difference in elevations of the water-intake areas of the aquifers. If, however, the moving fluid originated in the sediments, the loading of the sediment layers would be the principal cause of the excess fluid pressure that determines the direction of fluid movement. Compaction fluid movement This problem was discussed in Chapter 8 ; the fluids move from an area of more loading (thicker deposition) to one of less. The volume of horizontal fluid movement relative to vertical increases as the permeability and thickness of interbedded permeable rocks increase, and as the rate of thickness change of newly deposited sediments increases. If the shales are thick and homogeneous, most fluid will move vertically. The presence of some contiguous or lenticular sandstones in a thick shale sequence may not drastically change this basic direction of fluid flow. Price (1976) recently proposed the importance of growth faults through massive .and undercompacted shales in the Gulf Coast as the main fluidmigration pathways. He suggested that hydrocarbons generated in the deep and hot undercompacted shale section have migrated upward along these
278 FLUID PRESSURE 1000 p r i 0 2 4 6
Fig. 14-3. Fluid-pressure profile in the Beaufort Basin.
faults, in the form of molecular solution in water. The fluid-flow model discussed in Chapter 8 , however, indicates that the principal direction of flow through massive shales is vertically upward, whether the shales are faulted or not; in other words, such upward movement of generated hydrocarbons through these shales is always possible. Its importance in the total petroleum accumulations in this area, however, may not be so great because the total volume of vertical fluid flow through these undercompacted shales may not have been large. The model discussed in Chapter 8 is applicable if the sediments reached the compactionequilibrium condition after each increment of instantaneous loading. If, however, some shales were t o stay (slightly) undercompacted while other shales attained almost compaction equilibrium, significant pressure differences and barriers within the shale zones would be developed. This type of facies was named “mixed compaction facies” (Evans et al., 1975). It occurs in an intermediate depth range below the normal compaction facies. Examples of the calculated fluid-pressure profiles of the mixed compaction facies in the Beaufort Basin, Canada, are shown in Fig. 14-3.Fluid moves from a higher excess-pressure point to a lower, and the inferred directions of fluid flow are shown by arrows. Similar pressure or compaction patterns were reported in other sedimentary basins (see Chapter 5 ) . Once an interbedded sequence has reached such an intermediate compaction stage, essentially all the compaction fluids may have to move laterally
279 Ol
O l
$65
5%: 2Ccz psE Ccw
9500 FEET
-
10052
10000 12500 FEET
=+
f%
Eu L h O K
0%
z
0
c 20052
20000
2
u
i
0 4
-
c U
>
YI
30000
30000
0
5000 10000 CUMULATIVE WATER VOLUME
CU FT I S 0 FT
15000 CU FT is(1 FT CUMULATIVE WATER VOLUME
Fig. 14-4.Cumulative water-loss volumes from shales in the Gulf Coast (combined vertical and horizontal migration model). (From Magara, 1976.)
through the interbedded sandstones. There is some vertical fluid flow in the shales, too, but the flow is only local. In summary, the development of the mixed compaction facies could facilitate lateral fluid flow from syncline areas, and this flow would take place after the sandstoneshale sequence has reached an intermediate depth range where petroleum may have been generated by the thermal process. The discussion in Chapter 8 introduced the method of calculating the horizontal and vertical fluid volumes that have moved from a given block of rock. In the subsurface, however, fluids expelled from the other blocks below and beside a particular block will also influence the fluid-flow condition within that block. In other words, the cumulative effect of fluid migration will be three-dimensional. Although estimating such a fluid-migration condition is extremely complicated, it may be worthwhile in that the migration of hydrocarbons may be affected by the cumulative fluid migration after the hydrocarbons have been generated. Magara (1976)recently estimated cumulative compaction fluid volumes, using Dickinson's (1953)porosity-depth curve and a simplified Gulf Coast model. In this model, the upper geological sequence is composed of sandstoneshale interbeds in which fluids have moved horizontally, and the lower sequence consists of massive and homogeneous shales where compaction fluid 'has moved vertically upward. The horizontal migration distance in the upper sequence is assumed to be 10 miles, and the total thickness of the sedimentary column 33,000 f t (10km). Fig. 14-4shows the cumulative vol-
280 umes of fluid loss since burial t o 2000 f t from a shale column whose base area is 1 sq f t (the respective depths to the boundaries of the upper and lower sequences are assumed t o be 9500 f t and 12,500 ft). It is interesting t o note that the cumulative fluid-volume plot based on the model that simulates the Gulf Coast sedimentary basin resembles the oilproduction frequency plot for the same area (Burst, 1969). This similarity suggests that fluid movement due to sediment compaction is one of the controlling factors in hydrocarbon occurrence in that area. A fact that could affect the importance of mechanical shale compaction and fluid expulsion in petroleum migration is that the rate of compaction decreases continuously as the shales become more deeply buried. In other words, by the time the source rocks had reached deep burial where the ternperature was high enough to generate hydrocarbons, the movement of cornpaction fluid might have become too slow and insignificant. If the fluids expand at such depths, the expansion might facilitate latestage fluid movement. Subsurface temperature increase with burial depth might cause such fluid expansion in most sedimentary basins. Aquathermal fluid movement Fig. 14-5 is a pressure-temperature diagram for water with selected isodensity lines, adapted from Barker (1972). The vertical scale is pressure in psi, and the horizontal scales are temperature in both Centigrade and Fahrenheit. Density values in g/cc (and specific volume values in cc/g) of water are shown along the isodensity lines. The original data for constructing this diagram were obtained by Kennedy and Holser (1966). The three geothermalgradient lines of 25"CFm (1.37"F/100 ft), 18"C/km (1"F/100 f t ) and 36"C/ km (2"F/100 ft) for hydrostatically pressured water (the system is not closed) are superimposed; the lines intercept water isodensity lines whose values decrease as the pressure (or burial depth) increases. A hydrostaticpressure gradient of 0.47 psi/ft was used. This progression t o lower densities and higher specific volumes means that a given weight of water expands with burial: the reason is that the increase of pressure associated with the 0.47 psi/ft hydrostatic gradient is inadequate t o hold the water volume constant. The amount of expansion can be derived easily from the specific volume values (cc/g), shown in brackets. When the geothermal gradient is 25"C/km (1.37"F/100 ft), for example, the specific volume increases from 1 cc/g at 0 psi pressure to 1.10 cc/g at 11,600 psi, which corresponds to a burial depth of about 25,000 ft. Thus, a 10%water expansion results from about 25,000 f t of burial; this is a significant amount. Continuous expansion of water for the three geothermal gradients is depicted in Fig. 14-6, where specific volume of water (cc/g) is shown on the vertical scale and depth (ft) on the horizontal scale. At 20,000 ft, for example, about 3% expansion has occurred for the geothermal gradient of 1"F/
281
TEMPERATURE b 32
100
O C
ZOO
300 TEMPERATURE
400
4
SO0
O F
Fig. 14-5. F'ressure-temperature-density (or specific volume) diagram for water. Three geothermal lines of 25O, 18O and 36OC/km for hydrostatically pressured fluid are superimposed on a basic diagram derived from Barker (1972).
100 ft, about 7% expansion for 1.37°F/100 ft, and 15%for 2"F/100 ft. Fig. 14-6 shows that rates of increase in specific volume, or rates of expansion, increase with burial depth. This fact is interesting because the amount of water expelled by compaction decreases with burial depth, but the subsurface temperature tends to expand water volume. This expansion could facilitate fluid migration at depth and hence could favour hydrocarbon migration. Expansion of rock grains also may be considered in the discussion of fluid migration. The grain expansion would create more intergrain spaces, thus more spaces for water. Its effect, however, is much less significant: the thermal expansion of quartz, for example, is only about & that of water (see Skinner, 1966). Thermal expansion data for dry clay matrix are not readily available; the value for quartz may be the closest approximation. In other
282
2 0 )OO
00
DEPTH [ F T
3 c I00
I
Fig. 14-6. Specific volume (of waterkdepth relationships in normally pressured zones for three geothermal gradients of 25', 18' and 36'C/km. (From Magara, 1974b.)
words, if the ratio of volume of water to that of rock grains is more than about 1:15 (porosity is more than about 6%),the effect of water expansion overrides that of grain expansion, resulting in water movement. In the Gulf Coast, a shale porosity of 6% would not be attained above 24,000 f t (Dickinson, 1953). Note that the above-mentioned aquathermal model is valid when pore water is not completely isolated. Such a relatively open system is developed in the normal and mixed compaction facies, which usually occur in the shallow t o intermediate depth range in many sedimentary basins. If the pore fluids are more isolated, as in the case of undercompacted facies, the fluid cannot expand freely and the fluid pressure will increase (Magara, 1974b). The directions of fluid migration due to the aquathermal effect are from a hot place to a cold, from a deep section t o a shallow, and from a basin's centre t o its edges. These directions are essentially the same as those of fluid movement caused by sediment compaction. Therefore, the significance of the aquathermal effect in the subsurface may simply be t o increase the effectiveness of compaction fluid flow at deep burial.
283 B
A SHALE
C
FLUID PRESSURE WATER SALINITY
POROSITY
IN SHALE
I N SHALE
SHALE
SHALE
+
DEPTH
DEPTH
4
~
SHALE
SHALE
R
\+
E
T I O ON F OSMOTIC FLUID FLOW
C
7
LHYDROSTATIC PRESSURE
Fig. 14-7. Schematic diagram showing shale porosity, fluid pressure and pore-water salinity distributions in interbedded s a n d s h a l e sequences (From Magara, 1974a.)
Now let us assume a geological model at intermediate depths in which sandstones are interbedded with shales. A shale porosity profile such as is shown in Fig. 14-7A may be developed. If the interbedded sandstones are permeable, the maximum fluid expulsion or the maximum shale porosity reduction will occur in the shales directly above and below the sandstones. The porosity in the middle of a shale bed may remain relatively high. The corresponding fluid-pressure plot is shown in Fig. 14-7B,in which arrows depict the inferred directions of compaction fluid flow. If water expands from the thermal effect, water will move within the shale bed from the centre to the upper and lower edges, because more expansion can be expected at the point of higher porosity (more water). The directions of the small-scale fluid migration due to the aquathermal effect are essentially the same as those of compaction fluid migration. Osmotic fluid movement In many sedimentary basins the salinity of the formation water increases with depth or compaction. These salinity values are usually higher than that of sea water (about 35,000ppm). In the undercompacted zones, the salinity is lower than those of normal and mixed compaction zones. The principal cause of these salinity variations in sedimentary rocks may be ion filtration by shales (see Chapter 10). Ion filtration by clays or shales has also been documented by laboratory methods (McKelvey and Milne, 1962;Engelhardt and Gaida, 1963),which showed that clays and shales filter salt from a solution. Therefore, the fluids
284 moving through the shales must be fresher than the original solution that saturated the shales. As mentioned in Chapter 10, Hedberg (1967) studied pore-water chlorinities and porosities in shales, using cores from several areas in the world. Fig. 10-10 shows the chloride content (ppm) versus porosity plots from the Burgan field in Kuwait and several oil fields in Texas. The relation between the chlorinity * and porosity in the Burgan data may be approximated by a hyperbola: the chlorinity increases as the porosity decreases. The data from the three Texas fields are too scattered and insufficient to prove or disprove the hyperbola relation. It is, however, interesting that most of the plotted data from Texas fall within the extension of the general Burgan trend. Combining the the concept of ion filtration and the shale porosity profile as shown in Fig. 14-7A enables a possible water-salinity profile for the shales to be drawn (Fig. 14-7C). Salinity is the reciprocal of shale porosity; i.e., it increases as the porosity decreases. Salinity, therefore, would increase from the centre to the edges of each shale bed. Because osmosis tends to move water from a fresher to a more concentrated side, the fluid-flow direction due to osmosis can be inferred as shown by the arrows in Fig. 14-7C. The osmotic-pressure difference due to salinity change is probably not very large as compared with that due to compaction. According to the chart shown by Jones (1967), the osmotic-pressure difference caused by a salinity difference of 50,000 mg/l is only about 600 psi (see Fig. 4-18 in Chapter 4). Because osmotic fluid flow is in the same direction as compaction fluid flow, however, osmotic flow could facilitate the primary hydrocarbon migration from shales to permeable sandstones. This combined fluid flow due to compaction and osmosis may continue until the shales reach equilibrium, at which time no fluids can be expelled from them by compaction, and salinity also may reach equilibrium. If any of the freshening mechanisms alter the salinity distribution at the later stages, the osmotic fluid flow may be changed also. The most important point in this combined mechanism of flow, however, is that the salinity contrast resulting from ion filtration seems to start at relatively early stages of compaction, and the resultant osmotic-pressure difference seems to support fluid flow from the shales at the early stages of water expulsion. Another important effect of the combined fluid flow on hydrocarbon migration is that the fluids moved both by compaction and by osmosis are relatively fresh, so that hydrocarbons would be more soluble in them than in more saline water. High hydrocarbon solubility will favour hydrocarbon migration.
* Salinity (NaCl) may be calculated by multiplying the chlorinity by 1.65.
285 Fluid movement due to clay dehydration Powers (1967)showed that alteration of montmorillonite to illite in the Gulf Coast area begins at a depth of about 6000 ft and continues at an increasing rate to a depth, usually about 9000-12,000 ft, where there is no montmorillonite left. This alteration offers a mechanism for desorbing the last few layers of adsorbed or bound water in clay and transferring it into interparticle locations as free water. If the last few layers of bound water have a greater density than free water, this released water tends to increase its volume as it is desorbed from between unit layers. If water expansion is restricted, the pore-water pressure will increase to abnormally high levels. According to Burst (1969),clay dehydration depends mainly on subsurface temperature, the average dehydration temperature being 221°F in the Gulf Coast. Certain chemical conditions for potassium fixation also are required for this conversion. Phase change and possible expansion of bound water at the time of dehydration may, as proposed by Burst (1969),be important agents for flushing hydrocarbons, at least from clay-interlayer locations t o interparticle locations (shale pore space). Martin (1962) summarized data on adsorbed water density in montmorillonite analysed by several different investigators. This summary is shown in Fig. 14-8,which plots the calculated and measured water density versus amount of water in the clay (g HzO/g clay) Fig. 14-8appears to support Powers’ (1967)and Burst’s (1969)proposals of the higher-than-normal (greater than 1 g/cc) density of the adsorbed water. However, the validity of
0
&WIT
X
MaeKENZlE
o^
0
MOONEY E l AL
2 13-
A
NOURISH
B . a
t E 12-
& ARENS
ANDERSON & LOW
E
a
8
8 1 1 -
>
t
5
0 10-
Fig. 14-8. Adsorbed water density on Na-montmorillonite. (From Martin, 1962, for which see also the references in this figure.)
286 the higher-than-normal density shown in Fig. 14-8 is not very great, because most of these higher values were derived from calculations rather than direct measurements. Martin (1962) stated that “the only unambiguous adsorbed waterdensity data are (those) of Anderson and Low (1959)”, which show values less than 1g/cc. Therefore, from the data existing at present, it is difficult t o prove or disprove the water expansion and flushing effect associated with clay dehydration. However, we may be able t o say that clay dehydration could be an additional source of liquid water at relatively deep burial where hydrocarbons may have been generated. Van Olphen (1963) demonstrated that at 25°C the pressure needed to remove the last interlayer of water is 65,000-70,000 psi, and that needed for the second-to-last water interlayer is 30,000 psi. These values are considerably higher than the pressure at depths less than 20,000 ft. Overburden pressure alone, then, may not suffice to release at least the last two layers of bound water. This is the main reason why Burst and Powers developed their concepts of the temperaturedependent water-release mechanism associated with clay-mineral conversion. If, however, the interlayer water is released by clay dehydration in response to temperature, and subsequently remains in the pore spaces as free water, the same overburden pressure could be enough to push it out of the shales, provided drainage is available. This type of water movement is essentially the same as that caused by sediment compaction discussed above (Magara, 1975a). The validity of the average dehydration and mineralconversion temperature of 221°F proposed by Burst must also be examined. Schmidt (1973) studied the proportions of expandable clay (mostly montmorillonite) and nonexpandable clay in a well drilled in the Gulf Coast (Fig. 14-9). This figure shows that the rate of mineral conversion increases at about 10,500 ft, which corresponds t o a subsurface temperature of about 200°F (Fig. 14-10). However, the geothermal gradient in this well also increases at that depth (10,500 ft), which is the top of the undercompacted (abnormally pressured) section. Because water has a thermal conductivity significantly lower than that of most rock matrix, the undercompacted section, which contains an excessive amount of water, tends to have a thermal conductivity lower than that in the normally compacted section (Lewis and Rose, 1970). If heat flows upward at a given rate, the geothermal gradient in the undercompacted section would become greater than that in the normal compaction section. An important point shown in Fig. 14-9 is that clay-mineral conversion is not a drastic event. The conversion temperature of 221”F, as suggested by Burst, may not be required. Rather, conversion seems t o begin almost immediately after deposition, and continues to depth. The higher the geothermal gradient, the faster the rate of conversion. Because in essentially all the world’s sedimentary basins temperature tends
287
0
LESS THAN 2 . 0 MICRON SIZE C L A Y
0
L E S S THAN 0.15 MICRON SIZE C L A Y
I-
Y Y)
8-
I II -
-
Y 0
a
12-
14 I 100
40
60
80
20
% EXPANDABLE CLAY
Fig. 14-9. Plot of per cent expandable clay versus depth showing accelerated increase in diagenesis of montmorillonite to mixed-layer mon tmorilIoniteillite. (From Schmidt, 1973.)
.
TEMPERATURES FROM BOTTOM HOLE PRESSURt SURVEYS I N SHUT-IN PRODUCING WELLS
'1 \:
aTfMPERATURES OF SHALLOW GROUND WATERS MEASURED B Y THE ffiGS
1
\
5 1 $
E
8
9-
D 10 - TOP-HIGH -PRESSURE 11 - ZONE 12
-
13 14
HOLE TEMPERATURE TREND
.* . ..
b*
I
I
I
I
I
to increase with depth, the bound water will be released in any case. Claymineral conversion could create an additional source of liquid water at depth. Its significance in primary migration, however, cannot be understood clearly, because whether such conversion causes fluid expansion and migration is not known. Other possible causes of primary migration There are several other possible causes of primary migration, such as capillary pressure, buoyancy, diffusion, generation of hydrocarbons - especially gas, etc. These causes are mostly unassociated with the movement of water. Although there is no solid reason to deny their importance, I personally feel that water movement of some kind must be important at the primary-migration stage, and that therefore a mechanism unrelated to water movement may be of secondary importance. As pointed out previously, we are dealing with a large amount of water and a relatively small amount of hydrocarbons in the sediments, which have a very fine network at the time of primary migration.
Form of hydrocarbons at primary migration If all the hydrocarbons are in molecular solution in water at the primarymigration stage, estimating the volume and direction of sediment-source water as discussed is of prime importance in understanding hydrocarbon migration. The water volume may be tied directly t o the amount of hydrocarbons. However, the solubility of liquid hydrocarbons in water is relatively low even at elevated temperatures (Fig. 14-1).Micellar solution as proposed by Baker (1962)cannot be very important in the subsurface for the several reasons mentioned above. According to Dickey’s (1975)estimate, there must be at least 10,000 ppm hydrocarbons in the flowing stream at the time of primary migration. Another approach to estimating the required concentration of oil in the flowing stream is as follows: Tissot and Pelet (1971)analysed the amounts of hydrocarbons, resins and asphaltenes in shales adjacent to a reservoir in Algeria. Fig. 14-11shows the results of their analyses in mg/g organic carbon. Although the amounts of resins and asphaltenes in the shale remain relatively constant, the amount of hydrocarbons decreases toward the reservoir, suggesting primary hydrocarbon migration. The difference in hydrocarbon contents at the 14-m point and at the near-reservoir point is about 40 mg/g organic carbon. If the level of total hydrocarbon generation per gram of organic carbon is constant throughout the shale section, this 40 mg represents the lowest possible amount of hydrocarbons expelled per gram of organic carbon from the shale closest to the reservoir. If the shale has a density of 2 g/cc and 1 weight per
289 mdg ORGANIC CARBON
20
40
M
HYDROCARBONS
80
1W
120
ALGERIA
RESERVOIR
Fig. 14-11. Plot of amounts of hydrocarbons, resins and asphaltenes versus organic carbon (g) of Devonian shales adjacent to reservoir in Algeria. Original data derived from Tissot and Pelet (1971).
cent of organic carbon, 1 cc of this shale lost 0.8 mg of hydrocarbons *. If the porosity difference between these two points is lo%, which seems to be the largest porosity difference possible under these conditions, the amount of hydrocarbons in the flowing stream can be estimated to be about 8000 PPm. As mentioned above, this estimate is based on the lowest possible estimate of hydrocarbons in the compaction fluid; the true value could be higher. In any case, this figure is at least one order higher than the highest molecularsolubility figure in the temperature range for oil generation, and is surprisingly close to the >10,000 ppm given by Dickey (1975).Note that the density and porosity data for the shales studied by Tissot and Pelet (1971) are not readily available, so that they have had to be assumed for this estimate. Vyshemirsky et al. (1973)experimented with squeezing the mixture of clay, liquid hydrocarbons and water up to 300 atoms. They found that the
* The figures used for this estimate would be the lowest possible values to produce the lowest possible hydrocarbon yield.
290 amount of hydrocarbons squeezed with the water was more than could be accounted for by the solution mechanism alone. From the above estimate and other observations, it is clear that the greater proportion of liquid hydrocarbons must move in a separate phase. Gas, however, can migrate in aqueous solution because of its higher solubility. The question then arises: Why is the movement of water important if most of the liquid hydrocarbons move in their separate phases? The next section will suggest an answer. Migration of oil in oil phase
A comprehensive discussion of the mechanisms associated with oil-phase migration was published in 1954 by Hobson. Recently Dickey (1975) rediscussed this possibility from a slightly different angle. In compacted shales, the larger proportion of water is electrically charged at the clay surfaces, and has a relatively high viscosity (Fig. 8-6), which means that some water is semisolid. The amount of liquid (or free) water in the compacted shales is probably not great. In these circumstances, if the shales compact further, the oil as well as the liquid water will migrate provided the oil saturation in the liquid phase is higher than the critical value for oil migration. Basing his argument on the concept of relative permeability in sandstone, Dickey (1975) stated that, “oil, will move along with the water only if it occupies about 20 per cent or more of the pore volume” (p. 341). If the sandstone is partly water-wet and partly oil-wet, the critical residual-oil saturation can be as low as 10%.Dickey also suggested that the residual-oil satbecause a uration in shales may be less than 10% and possibly as low as 1%, considerable fraction of the internal surfaces of shales can be oil-wet (p. 342). A schematic diagram of the relative-permeability and oil- (or water-) saturation relationship is shown in Fig. 14-12. The critical residual-oil saturation is marked by an X and for shales may be assumed t o be a value between 10 and 1%.For oil t o migrate along with water, this critical oil saturation must be exceeded; that is, if the oil saturation is at X ’ in Fig. 14-12, there will be some oil migration. If, for example, the oil saturation in the total water (solid and liquid) is 100 ppm (0.01 wt%), and if only 1%of the water is in liquid phase (and 99% is solid), the oil saturation in the liquid water will be 10,000 ppm (1 wt%). Assuming that the density of oil is 0.8 g/cc and that of water 1g/cc, this figure will correspond to about 1.2 ~01%.This is the concept suggested by Dickey. If some of the liquid water is expelled, possibily with a small amount of oil, then the oil saturation of the liquid phase in the shale pores will increase, ensuring more oil migration. However, as the liquid water is further expelled as compaction proceeds, the permeability will be reduced to
291
X'Y
x
0
OIL SATURATION
0
WATER SATURATION
100%
Fig. 14-12. Schematic diagram showing relative permeabilitywater (or oil) saturation relationship for sandstone.
an extremely low level and the movement of the total fluids (water and oil)
eventually may become difficult. Possible changes in oil saturation with the gradual removal of liquid water are depicted in Fig. 14-13. The two diagonal lines refer to the original oil concentrations of 1 0 ppm and 100 ppm, respectively, when the liquid water occupied 10% of the bulk shale volume. If the liquid is expelled from these shales until 0.01% liquid water remains (there would still be a lot of solid water left at this stage), the respective oil saturations in the liquid phase will become 10,000 ppm ( 1 wt%) and 100,000 ppm (10 wt%) *. The corresponding volume percentages are about 1.2 and 12%, respectively. At this stage, oil may move along with the water (Fig. 14-12). The boundary between the solid and liquid water phases in shales would not be clear-cut; the change is probably quite gradual. In other words, it would be difficult to define how much solid water and how much liquid water are in a shale at any given compaction stage. However, for further discussions in this chapter, it would be wise to get some approximate figures on the amounts of solid water in shales. Let us assume that there is an illite clay sample whose bulk density is 2 g/cc. The specific surface area of illite clay is about 100 m2/g (Grim, 1953, p. 311). If illite has one solid-water
* Movement of oil up to the state of 0.01%liquid water is ignored in this case.
292
""
.
10
100
1.m
1o.m
1M.MO
Wl
01
1
1
10
W1
01
1
1
10
A
=
lOppm OIL CONCENTRATION WHEN LIOUID WATER OCCUPIES 10% OF BULK VOLUME
6
=
IWppm OIL CONCENTRATION WHEN LIQUID WATER OCCUPIES 10% OF BULK VOLUME
1.m.MO ppm
loow% I W VOL %
CONCENTRATION OF OIL IN LIQUID PHASE
Fig. 14-13.Chart showing increasing tendency of concentration of oil in liquid phase a liquid water per cent in shale decreases, assuming no oil migration.
layer whose thickness is about 2.5 A (or 0.26 nanometer), then 1 cc of this illite contains about 0.06 cc of solid water. In other words, about 5 vol% of illite clay would be solid water. Shales usually contain other nonclay minerals (quartz, feldspar, etc.), so the actual per cent of solid water in an illitic shale may be slightly lower than this estimate. Now let us estimate the amount of solid water in montmorillonite clay, Theoretically speaking, montmorillonite has a specific surface area of about 800 m2/g (Grim, 1953, p. 311); therefore, if only one water layer is considered to be relatively solid, the amount of solid water might be about 40%. Because montmorillonite should have more than one water layer, this estimate probably is too low. However, we also have t o allow for the effect of nonclay minerals in actual shales, which will reduce the proportion of solid water in bulk volume as discussed above. The other factor we might have to consider is that numerous water layers attached t o montmorillonite clay surfaces will reduce the bulk density of the sample significantly, so that to assume a bulk density of 2 g/cc may not be warranted; it could be lower. This factor in turn would reduce the surface area within 1 cc of the montmorillonite clay sample, and hence the percentage of solid or structured water. Powers' (1967) estimate is that about 60% of montmorillonite clay is (structured) water. In the light of the above considerations, let us assume that 4 6 5 0 % of montmorillonitic shale is relatively solid or structured water. This reasoning suggests that montmorillonitic shale contains approximately 8 t o 10 times as much solid water as illitic shale.
293
W
<
I
..
0
I
0
I
4.000
I
I
8.000
I
I
12.000
I
I
16.000
1
20,OOo
I
24.000 FEET
DEPTH
Fig. 14-14. Shale porosity-depth relationship in Gulf Coast by Dickinson (1953). 5% porosity line represents possible solid or structured water per cent in illitic shale and 4050%porosity zone indicates such in montmorillonitic shale.
In Fig. 14-14,the lines marking the 5% solid water for illitic shale and 40-5076 solid water for montmorillonitic shale are added to Dickinson’s (1953)shale porosity4epth relationship in the Gulf Coast. The 5%porosity is not reached above 24,000 ft; the porosity at 24,000 f t is about 9%. In
other words, for the sake of discussion, if all the Gulf Coast sediments are assumed to have been illitic at the time of deposition, and there was no conversion of minerals from montmorillonite to illite, the amount of liquid water in sediments was relatively large, ranging from about 75% at the surface to 4% at 24,000 ft. The 10% liquid-water level would have been attained at about 12,000 f t because the shale porosity at this depth is about 15% (155 = lo%, Fig. 14-14).If the concentration of oil is assumed to have been 100 ppm at 12,000 f t , the compaction from 12,000 f t to 24,000 f t would have increased the concentration only to about 250 ppm or 0.025 wt% (Fig. 1413). This level of oil concentration would not suffice for any oil migration in the oil-phase. Note that the above-mentioned estimate is based on the assumption that all the clays were illite at the time of deposition: such is not the case in the Gulf Coast and many other sedimentary basins. In the next schematic model, let us consider the situation where all the
294 clays at the time of deposition were montmorillonite. The 40-5096 porosity level of the shale can be attained at relatively shallow depths, such as those of 500 to 1000 f t (Fig. 14-14). In other words, the critical situation at which the amount of liquid water becomes extremely small, facilitating the possible oil-phase migration, would be reached at a very shallow depth - at which stage there may not have been enough oil generated t o enable any effective oil migration. If the solid water is not effectively removed by overburden pressure alone, as suggested by Van Olphen (1963), the compaction of this montmorillonitic shale may have t o terminate entirely. In the actual subsurface, however, it does not, because heat helps release continuously some of the relatively solid water, and some of the relatively less-bounded water may be expelled hydraulically if the threshold pressure is exceeded. The common observation of gradual porosity decline in normally compacted shales suggests that water has been expelled one way or another. In order t o keep this relatively small amount of liquid water available in shales over periods of geological time, the liquid water generated must be expelled effectively. In other words, the generation of liquid water and its expulsion must occur hand-in-hand. Good drainage is a necessary condition. If the liquid water generated cannot go out and stays in the shale pores, the oil concentration in the liquid phase will become less. This situation may be observed in the deep, undercompacted shales of the Gulf Coast, within which most of the solid water in the montmorillonite has already been released by relatively high temperatures (Burst, 1969), but the liquid water generated seems not t o have been expelled through lack of good permeable zones. Fluids will still be moving through these shales at an extremely slow rate, but effective oil migration in the oil phase is not likely because the oil saturation is so low. However, some oil may move in solution in water. Comparison of the illite and montmorillonite models described above suggests that the presence of montmorillonite at the time of sediment deposition, and its conversion by heat, can be beneficial in primary oil migration in the oil phase. However, this migration mechanism would not require any critical temperature such as 221"F, because it seems t o be a long and continuous process. It may be concluded that, t o have effective primary migration of oil in the oil phase, most of the liquid water available in the shales must be expelled effectively t o maintain a relatively high oil saturation in the liquid phase. The longer the sediments can maintain this effective drainage situation, the greater the chances of effective primary migration of oil, other geological and geochemical conditions being equal. This may explain why most oil pools have been found in relatively low-pressured zones (Timko and Fertl, 1971), where drainage conditions are generally excellent. On the basis of the above discussion, I propose a model for primary migra-
29 5
COMPACTION
-
Fig. 14-15. Hypothetical relationships of relative permeability, absolute permeability, and fluid movement versus degree of compaction of shale.
tion in the oil phase, as shown in Fig. 14-15.The top diagram shows a schematic of relative permeability versus degree of compaction in a shale. As the shale compacts, the relative permeability to water decreases and that to oil increases. Although the relative permeability to oil increases drastically with compaction, the absolute permeability of the shale will continually decrease
296
as the shale loses more liquid water and becomes more compacted (middle diagram of Fig. 14-15).Oil migration in the oil phase will reach its maximum at an intermediate compaction stage, then decline as the absolute permeability of the shale decreases (bottom diagram of Fig. 14-15).If this peak oilmigration stage is not very far from the peak oil-generation stage, we may be able t o expect significant oil accumulation. An important conclusion derived from the concepts discussed above is that effective drainage of fluids is essential to effective oil migration in the oil phase. The effectiveness of the drainage can be worked out from the calculated cumulative fluid-loss volumes or calculated pressure plots, as discussed in Chapters 3, 5,6 and 8. References Anderson, D.M. and Low, P.F., 1958. Density of water adsorbed by lithium-, sodium-, and potassium-bentonite. Soil Sci. SOC.A m . Proc., 22: 97-103. Baker, E.G., 1962. Distribution of hydrocarbons in petroleum. Bull. A m . Assoc. Pet. Geol., 46: 76-84. Barker, C., 1972.Aquathermal pressuring - role of temperature in development of abnormal-pressure zones. Bull. A m . Assoc. Pet. Geol., 56: 2068-2071. Burst, J.F., 1969. Diagenesis of Gulf Coast clayey sediments and its possible relation to petroleum migration. Bull. A m . Assoc. Pet. Geol., 53: 73-93. Dickey, P.A., 1975. Possible primary migration of oil from source rock in oil phase. Bull. A m . Assoc. Pet. Geol., 59: 337-345. Dickinson, G., 1953. Geological aspects of abnormal reservoir pressures in Gulf Coast Louisiana. Bull. A m . Assoc. Pet. Geol., 37: 410-432. Dodson, C.R. and Standing, M.B., 1944. Pressureuolume-temperature and solubility relations for natural gas-water mixtures, In: Drilling and Production Practice. Amer. Petrol. Inst., pp. 173-178. Engelhardt, W.V. and Gaida, K.H., 1963.Concentration changes of pore solutions during compaction of clay sediments. J. Sediment. Petrol., 33: 919-930. Evans, C.R., McIvor, D.K. and Magara K., 1975. Organic matter, compaction history and hydrocarbon occurrence - Mackenzie Delta, Canada. Proc. 9 t h World Pet. Congr., 3: 147-157. (Panel Discussion) Graton, L.C. and Fraser, H.J., 1935. Systematic packing of spheres with particular relation to porosity and permeability. J. Geol. 43: 785-909. Grim, R.E., 1953.Clay Mineralogy. McGraw-Hill, New York, N.Y., 384 pp. Hedberg, W.H., 1967. Pore-Water Chlorinities of Subsurface Shales. Univ. Microfilms, Ann Arbor, Mich. (Thesis, Univ. Wisconsin). Hobson, D.G., 1954. Some Fundamentals of Petroleum Geology. Oxford Univ. Press, London, 139 pp. Jones, P.H., 1967. Hydrology of Neogene deposits in the northern Gulf of Mexico Basin. Proc. 1st Symp. Abnormal Subsurface Pressure. Louisiana State Univ., Baton Rouge, La., pp. 91-207. Kennedy, G.C. and Holser, W.T., 1966. Pressure-volume-temperature and phase relations of water and carbon dioxide: Section 16 in Handbook of Physical Constants (revised ed.). Geol. SOC.A m . Mem., 97: 371-383. Levorsen, A.I., 1967. Geology of Petroleum. Freeman, San Francisco, Calif., 2nd ed., 724 pp.
297 Lewis, C.R. and Rose, S.C., 1970. A theory relating high temperatures and overpressures. J. Pet. Technol. 22: 11-16. Magara, K., 1972. Compaction and fluid migration in Cretaceous shales of western Canada. Geol. Surv. Can. Pap., 72-18: 81 pp. Magara, K., 1974a. Compaction, ion-filtration, and osmosis in shales and their significance in primary migration. Bull. A m . Assoc. Pet. Geol., 58: 283-290. Magara, K., 1974b.Aquathermal fluid migration. Bull. A m . Assoc. Pet. Geol., 58: 25132516. Magara, K., 1975a. Reevaluation of montmorillonite dehydration as cause of abnormal pressure and hydrocarbon migration. Bull. A m . Assoc. Pet. Geol., 59: 292-302. Magara, K.,1975b. Importance of aquathermal pressuring effect in Gulf Coast, Bull. A m . Assoc. Pet. Geol., 59: 2037-2045. Magara, K., 1976. Water expulsion from elastic sediments during compaction - directions and volumes. Bull. A m . Assoc. Pet. Geol., 60:543-553. Martin, R.T., 1962. Adsorbed water on clay: a review. Clays Clay Miner., 9 (ROC. 9th Natl. Conf. Claysand Clay Minerals, 1960, Pergamon, New York, N.Y., pp. 28-270). McKelvey, J.G. and Milne, I.H., 1962.The flow of salt solutions through compacted clay. Clays Clay Miner., 9. (Proc. 9th Natl. Conf. Clays and Clay Minerals, Pergamon, New York, N.Y.,Earth Sci. Ser. Mongr., 11: 248-259). Powers, M.C., 1967. Fluid-release mechanisms in compacting marine mudrocks and their importance in oil exploration. Bull. A m . Assoc. Pet. Geol., 51: 1240-1254. Price, L.C., 1976. Aqueous solubility of petroleum as applied to its origin and primary migration. Bull. A m . Assoc. Pet. Geol., 60: 213-244. Salathiel, R.A., 1973. Oil recovery by surface film drainage in mixed-wettability rocks. J. Pet. Technol., 25: 1216-1224. Schmidt, G.W., 1973.Interstitial water composition and geochemistry of deep Gulf Coast shales and sandstones. Bull. A m . Assoc. Pet. Geol., 57: 321-337. Skinner, B.J., 1966. Thermal expansion: Section 6 in Handbook of Physical Constants (revised ed.). Geol. SOC.A m . Mem., 97:75-96. Timko, D.J. and Fertl, W.H., 1971.Relationship between hydrocarbon accumulation and geopressure and its economic significance. J. Pet. Technol., 22: 923-930. Tissot, B. and Pelet, R., 1971. Nouvelles donndes sur les mdcanismes de genese et de migration du petrole: simulation mathhmatique et application B la prospection. Proc. 8 t h World Pet. Congr., pp. 35-46. Van Olphen, H., 1963. Compaction of clay sediments in the range of molecular particle distances. Clay Clay Miner. 11. (Proc. 11th Natl. Conf. Clays and Clay Miner. 1962) Macmillan, New York, N.Y., pp. 178-187. Vyshemirsky, V.S., Trofimuk, A.A., Eontorovich, A.E. and Neruchev, S.G., 1973. Pitumoids fractionation in the process of migration. In: B. Tissot and F. Rienner (Editors), Advances in Organic Geochemistry. Editions Technip., Paris, pp. 359-365.
Hydrocarbons generated in fine-grained sedimentary rocks are probably disseminated at first, but eventually they must move from their host rocks into more permeable. and porous sedimentary rocks t o form an accumulation. The movement of hydrocarbons from nonreservoir rocks to reservoir rocks is called primary migration, and is distinguished from their concentration and accumulation within the reservoir rocks known as secondary migration (Levorsen, 1967). A discussion of the primary migration of hydrocarbons includes three different kinds of problems: (1)The form in which they migrate, such as molecular solution, micellar solution and separate hydrocarbon phase. (2) The cause of migration. (3) The water source and the cause of its movement. Whether the hydrocarbons being discussed are mainly gas or mainly liquid will affect the proportions of them that are moving in solution and in separation, because the solubility of gaseous hydrocarbons is generally higher. The migration mechanism for hydrocarbons in separate phase may differ from that for hydrocarbons dissolved in water. If water movement is important in moving hydrocarbons, the source of the water and the cause of its movement must be examined carefully. The molecular solubility of liquid hydrocarbons in water at relatively high temperatures was recently discussed by Price (1976), who showed that the solubility increased with increasing temperature (Fig. 14-1). The solubility of the Farmer's oil at 160°C is approximately 150 ppm, and the curve shows that it tends t o increase with further increase in temperature. However, these temperature values are much higher than the known temperature range of 60" to 150°C for active oil generation. Dickey (1975) on the other hand, suggested that the flowing stream would have to contain at least 10,000 ppm of hydrocarbons at the time of primary migration. Therefore, it may be very difficult t o believe that most oil migrates as a molecular solution in water. In the case of migration of gas, the situation can be completely different. According t o Dodson and Standing's chart (1944, see Fig. 14-2), the solubility of natural gas in water ranges from 4 cu f t per barrel of water at 500 psi, up to 30 cu f t at 10,000 psi. In other words, most of the gas might migrate in solution in water at the primary migration stage. Another solution mechanism, called micellar solution, was proposed by
27 6
-E
l
FARMERS WHOLE OIL
ALASKA REEDY CREEK
L. A. STATE
TEMPERATURE IN OC
Fig. 14-1. Solubilities of two whole oils (Wyoming Farmers and Louisiana State) and four topped oils (Amoseas Lake, Reedy Creek, Alaska and Union Moonie) as functions of temperature in water. Topping temperature is 2OO0C (392OF). (From Price, 1976.)
Baker (1962). He suggested that hydrocarbon solubility is substantially high if the water contains micelles formed by soaps of organic acids. However, there are several reasons why Baker’s proposal is not plausible as the princi-
-
EXPERIMENTAL D A T A
---- E X T R A P O L A T E D
DATA
Fig. 14-2. Solubility of natural gas in formation water in cu ft/bbl. (From Dodson and Standing, 1944.)
277 pal mechanism of hydrocarbon migration in the subsurface, First, there is no good evidence that such solubilizing micelles exist in substantial quantity in shales. And even if they d o exist in shales, they would not be easily moved because they are not small. Then, the micelles would increase the solubility of the heavier hydrocarbons in water only to a few parts per million - nowhere near the 10,000 ppm o r more that now appears to be necessary (Dickey, 1975). Another difficult point in believing micellar solution to be important in primary migration is that the process of unloading the hydrocarbons carried by the fluid (water, micelles and hydrocarbons) at the final trapping position in the reservoir cannot be thoroughly explained. The preceding discussion may lead us t o conclude that the larger proportion of liquid hydrocarbons must migrate in a separate phase, although the rest can migrate in solution in water. In the case of gas migration, the proportion moving in solution in water can be relatively large, because of its greater solubility in water. Although the form in which hydrocarbons moved at the time of primary migration, and the mechanism of that migration, are not completely understood, the movement of water in fine-grained source rocks must be one of the most important factors. The amounts of organic matter and generated hydrocarbons in the source rocks are quite small in comparison with the amount of water. The movement of the large quantity of water must have influenced and may have controlled the direction and effectiveness of hydrocarbon migration. If the water concerned is meteoric water, the direction in which it moves is controlled by excess pressures generated by the difference in elevations of the water-intake areas of the aquifers. If, however, the moving fluid originated in the sediments, the loading of the sediment layers would be the principal cause of the excess fluid pressure that determines the direction of fluid movement. Compaction fluid movement This problem was discussed in Chapter 8 ; the fluids move from an area of more loading (thicker deposition) to one of less. The volume of horizontal fluid movement relative to vertical increases as the permeability and thickness of interbedded permeable rocks increase, and as the rate of thickness change of newly deposited sediments increases. If the shales are thick and homogeneous, most fluid will move vertically. The presence of some contiguous or lenticular sandstones in a thick shale sequence may not drastically change this basic direction of fluid flow. Price (1976) recently proposed the importance of growth faults through massive .and undercompacted shales in the Gulf Coast as the main fluidmigration pathways. He suggested that hydrocarbons generated in the deep and hot undercompacted shale section have migrated upward along these
278 FLUID PRESSURE 1000 p r i 0 2 4 6
Fig. 14-3. Fluid-pressure profile in the Beaufort Basin.
faults, in the form of molecular solution in water. The fluid-flow model discussed in Chapter 8 , however, indicates that the principal direction of flow through massive shales is vertically upward, whether the shales are faulted or not; in other words, such upward movement of generated hydrocarbons through these shales is always possible. Its importance in the total petroleum accumulations in this area, however, may not be so great because the total volume of vertical fluid flow through these undercompacted shales may not have been large. The model discussed in Chapter 8 is applicable if the sediments reached the compactionequilibrium condition after each increment of instantaneous loading. If, however, some shales were t o stay (slightly) undercompacted while other shales attained almost compaction equilibrium, significant pressure differences and barriers within the shale zones would be developed. This type of facies was named “mixed compaction facies” (Evans et al., 1975). It occurs in an intermediate depth range below the normal compaction facies. Examples of the calculated fluid-pressure profiles of the mixed compaction facies in the Beaufort Basin, Canada, are shown in Fig. 14-3.Fluid moves from a higher excess-pressure point to a lower, and the inferred directions of fluid flow are shown by arrows. Similar pressure or compaction patterns were reported in other sedimentary basins (see Chapter 5 ) . Once an interbedded sequence has reached such an intermediate compaction stage, essentially all the compaction fluids may have to move laterally
279 Ol
O l
$65
5%: 2Ccz psE Ccw
9500 FEET
-
10052
10000 12500 FEET
=+
f%
Eu L h O K
0%
z
0
c 20052
20000
2
u
i
0 4
-
c U
>
YI
30000
30000
0
5000 10000 CUMULATIVE WATER VOLUME
CU FT I S 0 FT
15000 CU FT is(1 FT CUMULATIVE WATER VOLUME
Fig. 14-4.Cumulative water-loss volumes from shales in the Gulf Coast (combined vertical and horizontal migration model). (From Magara, 1976.)
through the interbedded sandstones. There is some vertical fluid flow in the shales, too, but the flow is only local. In summary, the development of the mixed compaction facies could facilitate lateral fluid flow from syncline areas, and this flow would take place after the sandstoneshale sequence has reached an intermediate depth range where petroleum may have been generated by the thermal process. The discussion in Chapter 8 introduced the method of calculating the horizontal and vertical fluid volumes that have moved from a given block of rock. In the subsurface, however, fluids expelled from the other blocks below and beside a particular block will also influence the fluid-flow condition within that block. In other words, the cumulative effect of fluid migration will be three-dimensional. Although estimating such a fluid-migration condition is extremely complicated, it may be worthwhile in that the migration of hydrocarbons may be affected by the cumulative fluid migration after the hydrocarbons have been generated. Magara (1976)recently estimated cumulative compaction fluid volumes, using Dickinson's (1953)porosity-depth curve and a simplified Gulf Coast model. In this model, the upper geological sequence is composed of sandstoneshale interbeds in which fluids have moved horizontally, and the lower sequence consists of massive and homogeneous shales where compaction fluid 'has moved vertically upward. The horizontal migration distance in the upper sequence is assumed to be 10 miles, and the total thickness of the sedimentary column 33,000 f t (10km). Fig. 14-4shows the cumulative vol-
280 umes of fluid loss since burial t o 2000 f t from a shale column whose base area is 1 sq f t (the respective depths to the boundaries of the upper and lower sequences are assumed t o be 9500 f t and 12,500 ft). It is interesting t o note that the cumulative fluid-volume plot based on the model that simulates the Gulf Coast sedimentary basin resembles the oilproduction frequency plot for the same area (Burst, 1969). This similarity suggests that fluid movement due to sediment compaction is one of the controlling factors in hydrocarbon occurrence in that area. A fact that could affect the importance of mechanical shale compaction and fluid expulsion in petroleum migration is that the rate of compaction decreases continuously as the shales become more deeply buried. In other words, by the time the source rocks had reached deep burial where the ternperature was high enough to generate hydrocarbons, the movement of cornpaction fluid might have become too slow and insignificant. If the fluids expand at such depths, the expansion might facilitate latestage fluid movement. Subsurface temperature increase with burial depth might cause such fluid expansion in most sedimentary basins. Aquathermal fluid movement Fig. 14-5 is a pressure-temperature diagram for water with selected isodensity lines, adapted from Barker (1972). The vertical scale is pressure in psi, and the horizontal scales are temperature in both Centigrade and Fahrenheit. Density values in g/cc (and specific volume values in cc/g) of water are shown along the isodensity lines. The original data for constructing this diagram were obtained by Kennedy and Holser (1966). The three geothermalgradient lines of 25"CFm (1.37"F/100 ft), 18"C/km (1"F/100 f t ) and 36"C/ km (2"F/100 ft) for hydrostatically pressured water (the system is not closed) are superimposed; the lines intercept water isodensity lines whose values decrease as the pressure (or burial depth) increases. A hydrostaticpressure gradient of 0.47 psi/ft was used. This progression t o lower densities and higher specific volumes means that a given weight of water expands with burial: the reason is that the increase of pressure associated with the 0.47 psi/ft hydrostatic gradient is inadequate t o hold the water volume constant. The amount of expansion can be derived easily from the specific volume values (cc/g), shown in brackets. When the geothermal gradient is 25"C/km (1.37"F/100 ft), for example, the specific volume increases from 1 cc/g at 0 psi pressure to 1.10 cc/g at 11,600 psi, which corresponds to a burial depth of about 25,000 ft. Thus, a 10%water expansion results from about 25,000 f t of burial; this is a significant amount. Continuous expansion of water for the three geothermal gradients is depicted in Fig. 14-6, where specific volume of water (cc/g) is shown on the vertical scale and depth (ft) on the horizontal scale. At 20,000 ft, for example, about 3% expansion has occurred for the geothermal gradient of 1"F/
281
TEMPERATURE b 32
100
O C
ZOO
300 TEMPERATURE
400
4
SO0
O F
Fig. 14-5. F'ressure-temperature-density (or specific volume) diagram for water. Three geothermal lines of 25O, 18O and 36OC/km for hydrostatically pressured fluid are superimposed on a basic diagram derived from Barker (1972).
100 ft, about 7% expansion for 1.37°F/100 ft, and 15%for 2"F/100 ft. Fig. 14-6 shows that rates of increase in specific volume, or rates of expansion, increase with burial depth. This fact is interesting because the amount of water expelled by compaction decreases with burial depth, but the subsurface temperature tends to expand water volume. This expansion could facilitate fluid migration at depth and hence could favour hydrocarbon migration. Expansion of rock grains also may be considered in the discussion of fluid migration. The grain expansion would create more intergrain spaces, thus more spaces for water. Its effect, however, is much less significant: the thermal expansion of quartz, for example, is only about & that of water (see Skinner, 1966). Thermal expansion data for dry clay matrix are not readily available; the value for quartz may be the closest approximation. In other
282
2 0 )OO
00
DEPTH [ F T
3 c I00
I
Fig. 14-6. Specific volume (of waterkdepth relationships in normally pressured zones for three geothermal gradients of 25', 18' and 36'C/km. (From Magara, 1974b.)
words, if the ratio of volume of water to that of rock grains is more than about 1:15 (porosity is more than about 6%),the effect of water expansion overrides that of grain expansion, resulting in water movement. In the Gulf Coast, a shale porosity of 6% would not be attained above 24,000 f t (Dickinson, 1953). Note that the above-mentioned aquathermal model is valid when pore water is not completely isolated. Such a relatively open system is developed in the normal and mixed compaction facies, which usually occur in the shallow t o intermediate depth range in many sedimentary basins. If the pore fluids are more isolated, as in the case of undercompacted facies, the fluid cannot expand freely and the fluid pressure will increase (Magara, 1974b). The directions of fluid migration due to the aquathermal effect are from a hot place to a cold, from a deep section t o a shallow, and from a basin's centre t o its edges. These directions are essentially the same as those of fluid movement caused by sediment compaction. Therefore, the significance of the aquathermal effect in the subsurface may simply be t o increase the effectiveness of compaction fluid flow at deep burial.
283 B
A SHALE
C
FLUID PRESSURE WATER SALINITY
POROSITY
IN SHALE
I N SHALE
SHALE
SHALE
+
DEPTH
DEPTH
4
~
SHALE
SHALE
R
\+
E
T I O ON F OSMOTIC FLUID FLOW
C
7
LHYDROSTATIC PRESSURE
Fig. 14-7. Schematic diagram showing shale porosity, fluid pressure and pore-water salinity distributions in interbedded s a n d s h a l e sequences (From Magara, 1974a.)
Now let us assume a geological model at intermediate depths in which sandstones are interbedded with shales. A shale porosity profile such as is shown in Fig. 14-7A may be developed. If the interbedded sandstones are permeable, the maximum fluid expulsion or the maximum shale porosity reduction will occur in the shales directly above and below the sandstones. The porosity in the middle of a shale bed may remain relatively high. The corresponding fluid-pressure plot is shown in Fig. 14-7B,in which arrows depict the inferred directions of compaction fluid flow. If water expands from the thermal effect, water will move within the shale bed from the centre to the upper and lower edges, because more expansion can be expected at the point of higher porosity (more water). The directions of the small-scale fluid migration due to the aquathermal effect are essentially the same as those of compaction fluid migration. Osmotic fluid movement In many sedimentary basins the salinity of the formation water increases with depth or compaction. These salinity values are usually higher than that of sea water (about 35,000ppm). In the undercompacted zones, the salinity is lower than those of normal and mixed compaction zones. The principal cause of these salinity variations in sedimentary rocks may be ion filtration by shales (see Chapter 10). Ion filtration by clays or shales has also been documented by laboratory methods (McKelvey and Milne, 1962;Engelhardt and Gaida, 1963),which showed that clays and shales filter salt from a solution. Therefore, the fluids
284 moving through the shales must be fresher than the original solution that saturated the shales. As mentioned in Chapter 10, Hedberg (1967) studied pore-water chlorinities and porosities in shales, using cores from several areas in the world. Fig. 10-10 shows the chloride content (ppm) versus porosity plots from the Burgan field in Kuwait and several oil fields in Texas. The relation between the chlorinity * and porosity in the Burgan data may be approximated by a hyperbola: the chlorinity increases as the porosity decreases. The data from the three Texas fields are too scattered and insufficient to prove or disprove the hyperbola relation. It is, however, interesting that most of the plotted data from Texas fall within the extension of the general Burgan trend. Combining the the concept of ion filtration and the shale porosity profile as shown in Fig. 14-7A enables a possible water-salinity profile for the shales to be drawn (Fig. 14-7C). Salinity is the reciprocal of shale porosity; i.e., it increases as the porosity decreases. Salinity, therefore, would increase from the centre to the edges of each shale bed. Because osmosis tends to move water from a fresher to a more concentrated side, the fluid-flow direction due to osmosis can be inferred as shown by the arrows in Fig. 14-7C. The osmotic-pressure difference due to salinity change is probably not very large as compared with that due to compaction. According to the chart shown by Jones (1967), the osmotic-pressure difference caused by a salinity difference of 50,000 mg/l is only about 600 psi (see Fig. 4-18 in Chapter 4). Because osmotic fluid flow is in the same direction as compaction fluid flow, however, osmotic flow could facilitate the primary hydrocarbon migration from shales to permeable sandstones. This combined fluid flow due to compaction and osmosis may continue until the shales reach equilibrium, at which time no fluids can be expelled from them by compaction, and salinity also may reach equilibrium. If any of the freshening mechanisms alter the salinity distribution at the later stages, the osmotic fluid flow may be changed also. The most important point in this combined mechanism of flow, however, is that the salinity contrast resulting from ion filtration seems to start at relatively early stages of compaction, and the resultant osmotic-pressure difference seems to support fluid flow from the shales at the early stages of water expulsion. Another important effect of the combined fluid flow on hydrocarbon migration is that the fluids moved both by compaction and by osmosis are relatively fresh, so that hydrocarbons would be more soluble in them than in more saline water. High hydrocarbon solubility will favour hydrocarbon migration.
* Salinity (NaCl) may be calculated by multiplying the chlorinity by 1.65.
285 Fluid movement due to clay dehydration Powers (1967)showed that alteration of montmorillonite to illite in the Gulf Coast area begins at a depth of about 6000 ft and continues at an increasing rate to a depth, usually about 9000-12,000 ft, where there is no montmorillonite left. This alteration offers a mechanism for desorbing the last few layers of adsorbed or bound water in clay and transferring it into interparticle locations as free water. If the last few layers of bound water have a greater density than free water, this released water tends to increase its volume as it is desorbed from between unit layers. If water expansion is restricted, the pore-water pressure will increase to abnormally high levels. According to Burst (1969),clay dehydration depends mainly on subsurface temperature, the average dehydration temperature being 221°F in the Gulf Coast. Certain chemical conditions for potassium fixation also are required for this conversion. Phase change and possible expansion of bound water at the time of dehydration may, as proposed by Burst (1969),be important agents for flushing hydrocarbons, at least from clay-interlayer locations t o interparticle locations (shale pore space). Martin (1962) summarized data on adsorbed water density in montmorillonite analysed by several different investigators. This summary is shown in Fig. 14-8,which plots the calculated and measured water density versus amount of water in the clay (g HzO/g clay) Fig. 14-8appears to support Powers’ (1967)and Burst’s (1969)proposals of the higher-than-normal (greater than 1 g/cc) density of the adsorbed water. However, the validity of
0
&WIT
X
MaeKENZlE
o^
0
MOONEY E l AL
2 13-
A
NOURISH
B . a
t E 12-
& ARENS
ANDERSON & LOW
E
a
8
8 1 1 -
>
t
5
0 10-
Fig. 14-8. Adsorbed water density on Na-montmorillonite. (From Martin, 1962, for which see also the references in this figure.)
286 the higher-than-normal density shown in Fig. 14-8 is not very great, because most of these higher values were derived from calculations rather than direct measurements. Martin (1962) stated that “the only unambiguous adsorbed waterdensity data are (those) of Anderson and Low (1959)”, which show values less than 1g/cc. Therefore, from the data existing at present, it is difficult t o prove or disprove the water expansion and flushing effect associated with clay dehydration. However, we may be able t o say that clay dehydration could be an additional source of liquid water at relatively deep burial where hydrocarbons may have been generated. Van Olphen (1963) demonstrated that at 25°C the pressure needed to remove the last interlayer of water is 65,000-70,000 psi, and that needed for the second-to-last water interlayer is 30,000 psi. These values are considerably higher than the pressure at depths less than 20,000 ft. Overburden pressure alone, then, may not suffice to release at least the last two layers of bound water. This is the main reason why Burst and Powers developed their concepts of the temperaturedependent water-release mechanism associated with clay-mineral conversion. If, however, the interlayer water is released by clay dehydration in response to temperature, and subsequently remains in the pore spaces as free water, the same overburden pressure could be enough to push it out of the shales, provided drainage is available. This type of water movement is essentially the same as that caused by sediment compaction discussed above (Magara, 1975a). The validity of the average dehydration and mineralconversion temperature of 221°F proposed by Burst must also be examined. Schmidt (1973) studied the proportions of expandable clay (mostly montmorillonite) and nonexpandable clay in a well drilled in the Gulf Coast (Fig. 14-9). This figure shows that the rate of mineral conversion increases at about 10,500 ft, which corresponds t o a subsurface temperature of about 200°F (Fig. 14-10). However, the geothermal gradient in this well also increases at that depth (10,500 ft), which is the top of the undercompacted (abnormally pressured) section. Because water has a thermal conductivity significantly lower than that of most rock matrix, the undercompacted section, which contains an excessive amount of water, tends to have a thermal conductivity lower than that in the normally compacted section (Lewis and Rose, 1970). If heat flows upward at a given rate, the geothermal gradient in the undercompacted section would become greater than that in the normal compaction section. An important point shown in Fig. 14-9 is that clay-mineral conversion is not a drastic event. The conversion temperature of 221”F, as suggested by Burst, may not be required. Rather, conversion seems t o begin almost immediately after deposition, and continues to depth. The higher the geothermal gradient, the faster the rate of conversion. Because in essentially all the world’s sedimentary basins temperature tends
287
0
LESS THAN 2 . 0 MICRON SIZE C L A Y
0
L E S S THAN 0.15 MICRON SIZE C L A Y
I-
Y Y)
8-
I II -
-
Y 0
a
12-
14 I 100
40
60
80
20
% EXPANDABLE CLAY
Fig. 14-9. Plot of per cent expandable clay versus depth showing accelerated increase in diagenesis of montmorillonite to mixed-layer mon tmorilIoniteillite. (From Schmidt, 1973.)
.
TEMPERATURES FROM BOTTOM HOLE PRESSURt SURVEYS I N SHUT-IN PRODUCING WELLS
'1 \:
aTfMPERATURES OF SHALLOW GROUND WATERS MEASURED B Y THE ffiGS
1
\
5 1 $
E
8
9-
D 10 - TOP-HIGH -PRESSURE 11 - ZONE 12
-
13 14
HOLE TEMPERATURE TREND
.* . ..
b*
I
I
I
I
I
to increase with depth, the bound water will be released in any case. Claymineral conversion could create an additional source of liquid water at depth. Its significance in primary migration, however, cannot be understood clearly, because whether such conversion causes fluid expansion and migration is not known. Other possible causes of primary migration There are several other possible causes of primary migration, such as capillary pressure, buoyancy, diffusion, generation of hydrocarbons - especially gas, etc. These causes are mostly unassociated with the movement of water. Although there is no solid reason to deny their importance, I personally feel that water movement of some kind must be important at the primary-migration stage, and that therefore a mechanism unrelated to water movement may be of secondary importance. As pointed out previously, we are dealing with a large amount of water and a relatively small amount of hydrocarbons in the sediments, which have a very fine network at the time of primary migration.
Form of hydrocarbons at primary migration If all the hydrocarbons are in molecular solution in water at the primarymigration stage, estimating the volume and direction of sediment-source water as discussed is of prime importance in understanding hydrocarbon migration. The water volume may be tied directly t o the amount of hydrocarbons. However, the solubility of liquid hydrocarbons in water is relatively low even at elevated temperatures (Fig. 14-1).Micellar solution as proposed by Baker (1962)cannot be very important in the subsurface for the several reasons mentioned above. According to Dickey’s (1975)estimate, there must be at least 10,000 ppm hydrocarbons in the flowing stream at the time of primary migration. Another approach to estimating the required concentration of oil in the flowing stream is as follows: Tissot and Pelet (1971)analysed the amounts of hydrocarbons, resins and asphaltenes in shales adjacent to a reservoir in Algeria. Fig. 14-11shows the results of their analyses in mg/g organic carbon. Although the amounts of resins and asphaltenes in the shale remain relatively constant, the amount of hydrocarbons decreases toward the reservoir, suggesting primary hydrocarbon migration. The difference in hydrocarbon contents at the 14-m point and at the near-reservoir point is about 40 mg/g organic carbon. If the level of total hydrocarbon generation per gram of organic carbon is constant throughout the shale section, this 40 mg represents the lowest possible amount of hydrocarbons expelled per gram of organic carbon from the shale closest to the reservoir. If the shale has a density of 2 g/cc and 1 weight per
289 mdg ORGANIC CARBON
20
40
M
HYDROCARBONS
80
1W
120
ALGERIA
RESERVOIR
Fig. 14-11. Plot of amounts of hydrocarbons, resins and asphaltenes versus organic carbon (g) of Devonian shales adjacent to reservoir in Algeria. Original data derived from Tissot and Pelet (1971).
cent of organic carbon, 1 cc of this shale lost 0.8 mg of hydrocarbons *. If the porosity difference between these two points is lo%, which seems to be the largest porosity difference possible under these conditions, the amount of hydrocarbons in the flowing stream can be estimated to be about 8000 PPm. As mentioned above, this estimate is based on the lowest possible estimate of hydrocarbons in the compaction fluid; the true value could be higher. In any case, this figure is at least one order higher than the highest molecularsolubility figure in the temperature range for oil generation, and is surprisingly close to the >10,000 ppm given by Dickey (1975).Note that the density and porosity data for the shales studied by Tissot and Pelet (1971) are not readily available, so that they have had to be assumed for this estimate. Vyshemirsky et al. (1973)experimented with squeezing the mixture of clay, liquid hydrocarbons and water up to 300 atoms. They found that the
* The figures used for this estimate would be the lowest possible values to produce the lowest possible hydrocarbon yield.
290 amount of hydrocarbons squeezed with the water was more than could be accounted for by the solution mechanism alone. From the above estimate and other observations, it is clear that the greater proportion of liquid hydrocarbons must move in a separate phase. Gas, however, can migrate in aqueous solution because of its higher solubility. The question then arises: Why is the movement of water important if most of the liquid hydrocarbons move in their separate phases? The next section will suggest an answer. Migration of oil in oil phase
A comprehensive discussion of the mechanisms associated with oil-phase migration was published in 1954 by Hobson. Recently Dickey (1975) rediscussed this possibility from a slightly different angle. In compacted shales, the larger proportion of water is electrically charged at the clay surfaces, and has a relatively high viscosity (Fig. 8-6), which means that some water is semisolid. The amount of liquid (or free) water in the compacted shales is probably not great. In these circumstances, if the shales compact further, the oil as well as the liquid water will migrate provided the oil saturation in the liquid phase is higher than the critical value for oil migration. Basing his argument on the concept of relative permeability in sandstone, Dickey (1975) stated that, “oil, will move along with the water only if it occupies about 20 per cent or more of the pore volume” (p. 341). If the sandstone is partly water-wet and partly oil-wet, the critical residual-oil saturation can be as low as 10%.Dickey also suggested that the residual-oil satbecause a uration in shales may be less than 10% and possibly as low as 1%, considerable fraction of the internal surfaces of shales can be oil-wet (p. 342). A schematic diagram of the relative-permeability and oil- (or water-) saturation relationship is shown in Fig. 14-12. The critical residual-oil saturation is marked by an X and for shales may be assumed t o be a value between 10 and 1%.For oil t o migrate along with water, this critical oil saturation must be exceeded; that is, if the oil saturation is at X ’ in Fig. 14-12, there will be some oil migration. If, for example, the oil saturation in the total water (solid and liquid) is 100 ppm (0.01 wt%), and if only 1%of the water is in liquid phase (and 99% is solid), the oil saturation in the liquid water will be 10,000 ppm (1 wt%). Assuming that the density of oil is 0.8 g/cc and that of water 1g/cc, this figure will correspond to about 1.2 ~01%.This is the concept suggested by Dickey. If some of the liquid water is expelled, possibily with a small amount of oil, then the oil saturation of the liquid phase in the shale pores will increase, ensuring more oil migration. However, as the liquid water is further expelled as compaction proceeds, the permeability will be reduced to
291
X'Y
x
0
OIL SATURATION
0
WATER SATURATION
100%
Fig. 14-12. Schematic diagram showing relative permeabilitywater (or oil) saturation relationship for sandstone.
an extremely low level and the movement of the total fluids (water and oil)
eventually may become difficult. Possible changes in oil saturation with the gradual removal of liquid water are depicted in Fig. 14-13. The two diagonal lines refer to the original oil concentrations of 1 0 ppm and 100 ppm, respectively, when the liquid water occupied 10% of the bulk shale volume. If the liquid is expelled from these shales until 0.01% liquid water remains (there would still be a lot of solid water left at this stage), the respective oil saturations in the liquid phase will become 10,000 ppm ( 1 wt%) and 100,000 ppm (10 wt%) *. The corresponding volume percentages are about 1.2 and 12%, respectively. At this stage, oil may move along with the water (Fig. 14-12). The boundary between the solid and liquid water phases in shales would not be clear-cut; the change is probably quite gradual. In other words, it would be difficult to define how much solid water and how much liquid water are in a shale at any given compaction stage. However, for further discussions in this chapter, it would be wise to get some approximate figures on the amounts of solid water in shales. Let us assume that there is an illite clay sample whose bulk density is 2 g/cc. The specific surface area of illite clay is about 100 m2/g (Grim, 1953, p. 311). If illite has one solid-water
* Movement of oil up to the state of 0.01%liquid water is ignored in this case.
292
""
.
10
100
1.m
1o.m
1M.MO
Wl
01
1
1
10
W1
01
1
1
10
A
=
lOppm OIL CONCENTRATION WHEN LIOUID WATER OCCUPIES 10% OF BULK VOLUME
6
=
IWppm OIL CONCENTRATION WHEN LIQUID WATER OCCUPIES 10% OF BULK VOLUME
1.m.MO ppm
loow% I W VOL %
CONCENTRATION OF OIL IN LIQUID PHASE
Fig. 14-13.Chart showing increasing tendency of concentration of oil in liquid phase a liquid water per cent in shale decreases, assuming no oil migration.
layer whose thickness is about 2.5 A (or 0.26 nanometer), then 1 cc of this illite contains about 0.06 cc of solid water. In other words, about 5 vol% of illite clay would be solid water. Shales usually contain other nonclay minerals (quartz, feldspar, etc.), so the actual per cent of solid water in an illitic shale may be slightly lower than this estimate. Now let us estimate the amount of solid water in montmorillonite clay, Theoretically speaking, montmorillonite has a specific surface area of about 800 m2/g (Grim, 1953, p. 311); therefore, if only one water layer is considered to be relatively solid, the amount of solid water might be about 40%. Because montmorillonite should have more than one water layer, this estimate probably is too low. However, we also have t o allow for the effect of nonclay minerals in actual shales, which will reduce the proportion of solid water in bulk volume as discussed above. The other factor we might have to consider is that numerous water layers attached t o montmorillonite clay surfaces will reduce the bulk density of the sample significantly, so that to assume a bulk density of 2 g/cc may not be warranted; it could be lower. This factor in turn would reduce the surface area within 1 cc of the montmorillonite clay sample, and hence the percentage of solid or structured water. Powers' (1967) estimate is that about 60% of montmorillonite clay is (structured) water. In the light of the above considerations, let us assume that 4 6 5 0 % of montmorillonitic shale is relatively solid or structured water. This reasoning suggests that montmorillonitic shale contains approximately 8 t o 10 times as much solid water as illitic shale.
293
W
<
I
..
0
I
0
I
4.000
I
I
8.000
I
I
12.000
I
I
16.000
1
20,OOo
I
24.000 FEET
DEPTH
Fig. 14-14. Shale porosity-depth relationship in Gulf Coast by Dickinson (1953). 5% porosity line represents possible solid or structured water per cent in illitic shale and 4050%porosity zone indicates such in montmorillonitic shale.
In Fig. 14-14,the lines marking the 5% solid water for illitic shale and 40-5076 solid water for montmorillonitic shale are added to Dickinson’s (1953)shale porosity4epth relationship in the Gulf Coast. The 5%porosity is not reached above 24,000 ft; the porosity at 24,000 f t is about 9%. In
other words, for the sake of discussion, if all the Gulf Coast sediments are assumed to have been illitic at the time of deposition, and there was no conversion of minerals from montmorillonite to illite, the amount of liquid water in sediments was relatively large, ranging from about 75% at the surface to 4% at 24,000 ft. The 10% liquid-water level would have been attained at about 12,000 f t because the shale porosity at this depth is about 15% (155 = lo%, Fig. 14-14).If the concentration of oil is assumed to have been 100 ppm at 12,000 f t , the compaction from 12,000 f t to 24,000 f t would have increased the concentration only to about 250 ppm or 0.025 wt% (Fig. 1413). This level of oil concentration would not suffice for any oil migration in the oil-phase. Note that the above-mentioned estimate is based on the assumption that all the clays were illite at the time of deposition: such is not the case in the Gulf Coast and many other sedimentary basins. In the next schematic model, let us consider the situation where all the
294 clays at the time of deposition were montmorillonite. The 40-5096 porosity level of the shale can be attained at relatively shallow depths, such as those of 500 to 1000 f t (Fig. 14-14). In other words, the critical situation at which the amount of liquid water becomes extremely small, facilitating the possible oil-phase migration, would be reached at a very shallow depth - at which stage there may not have been enough oil generated t o enable any effective oil migration. If the solid water is not effectively removed by overburden pressure alone, as suggested by Van Olphen (1963), the compaction of this montmorillonitic shale may have t o terminate entirely. In the actual subsurface, however, it does not, because heat helps release continuously some of the relatively solid water, and some of the relatively less-bounded water may be expelled hydraulically if the threshold pressure is exceeded. The common observation of gradual porosity decline in normally compacted shales suggests that water has been expelled one way or another. In order t o keep this relatively small amount of liquid water available in shales over periods of geological time, the liquid water generated must be expelled effectively. In other words, the generation of liquid water and its expulsion must occur hand-in-hand. Good drainage is a necessary condition. If the liquid water generated cannot go out and stays in the shale pores, the oil concentration in the liquid phase will become less. This situation may be observed in the deep, undercompacted shales of the Gulf Coast, within which most of the solid water in the montmorillonite has already been released by relatively high temperatures (Burst, 1969), but the liquid water generated seems not t o have been expelled through lack of good permeable zones. Fluids will still be moving through these shales at an extremely slow rate, but effective oil migration in the oil phase is not likely because the oil saturation is so low. However, some oil may move in solution in water. Comparison of the illite and montmorillonite models described above suggests that the presence of montmorillonite at the time of sediment deposition, and its conversion by heat, can be beneficial in primary oil migration in the oil phase. However, this migration mechanism would not require any critical temperature such as 221"F, because it seems t o be a long and continuous process. It may be concluded that, t o have effective primary migration of oil in the oil phase, most of the liquid water available in the shales must be expelled effectively t o maintain a relatively high oil saturation in the liquid phase. The longer the sediments can maintain this effective drainage situation, the greater the chances of effective primary migration of oil, other geological and geochemical conditions being equal. This may explain why most oil pools have been found in relatively low-pressured zones (Timko and Fertl, 1971), where drainage conditions are generally excellent. On the basis of the above discussion, I propose a model for primary migra-
29 5
COMPACTION
-
Fig. 14-15. Hypothetical relationships of relative permeability, absolute permeability, and fluid movement versus degree of compaction of shale.
tion in the oil phase, as shown in Fig. 14-15.The top diagram shows a schematic of relative permeability versus degree of compaction in a shale. As the shale compacts, the relative permeability to water decreases and that to oil increases. Although the relative permeability to oil increases drastically with compaction, the absolute permeability of the shale will continually decrease
296
as the shale loses more liquid water and becomes more compacted (middle diagram of Fig. 14-15).Oil migration in the oil phase will reach its maximum at an intermediate compaction stage, then decline as the absolute permeability of the shale decreases (bottom diagram of Fig. 14-15).If this peak oilmigration stage is not very far from the peak oil-generation stage, we may be able t o expect significant oil accumulation. An important conclusion derived from the concepts discussed above is that effective drainage of fluids is essential to effective oil migration in the oil phase. The effectiveness of the drainage can be worked out from the calculated cumulative fluid-loss volumes or calculated pressure plots, as discussed in Chapters 3, 5,6 and 8. References Anderson, D.M. and Low, P.F., 1958. Density of water adsorbed by lithium-, sodium-, and potassium-bentonite. Soil Sci. SOC.A m . Proc., 22: 97-103. Baker, E.G., 1962. Distribution of hydrocarbons in petroleum. Bull. A m . Assoc. Pet. Geol., 46: 76-84. Barker, C., 1972.Aquathermal pressuring - role of temperature in development of abnormal-pressure zones. Bull. A m . Assoc. Pet. Geol., 56: 2068-2071. Burst, J.F., 1969. Diagenesis of Gulf Coast clayey sediments and its possible relation to petroleum migration. Bull. A m . Assoc. Pet. Geol., 53: 73-93. Dickey, P.A., 1975. Possible primary migration of oil from source rock in oil phase. Bull. A m . Assoc. Pet. Geol., 59: 337-345. Dickinson, G., 1953. Geological aspects of abnormal reservoir pressures in Gulf Coast Louisiana. Bull. A m . Assoc. Pet. Geol., 37: 410-432. Dodson, C.R. and Standing, M.B., 1944. Pressureuolume-temperature and solubility relations for natural gas-water mixtures, In: Drilling and Production Practice. Amer. Petrol. Inst., pp. 173-178. Engelhardt, W.V. and Gaida, K.H., 1963.Concentration changes of pore solutions during compaction of clay sediments. J. Sediment. Petrol., 33: 919-930. Evans, C.R., McIvor, D.K. and Magara K., 1975. Organic matter, compaction history and hydrocarbon occurrence - Mackenzie Delta, Canada. Proc. 9 t h World Pet. Congr., 3: 147-157. (Panel Discussion) Graton, L.C. and Fraser, H.J., 1935. Systematic packing of spheres with particular relation to porosity and permeability. J. Geol. 43: 785-909. Grim, R.E., 1953.Clay Mineralogy. McGraw-Hill, New York, N.Y., 384 pp. Hedberg, W.H., 1967. Pore-Water Chlorinities of Subsurface Shales. Univ. Microfilms, Ann Arbor, Mich. (Thesis, Univ. Wisconsin). Hobson, D.G., 1954. Some Fundamentals of Petroleum Geology. Oxford Univ. Press, London, 139 pp. Jones, P.H., 1967. Hydrology of Neogene deposits in the northern Gulf of Mexico Basin. Proc. 1st Symp. Abnormal Subsurface Pressure. Louisiana State Univ., Baton Rouge, La., pp. 91-207. Kennedy, G.C. and Holser, W.T., 1966. Pressure-volume-temperature and phase relations of water and carbon dioxide: Section 16 in Handbook of Physical Constants (revised ed.). Geol. SOC.A m . Mem., 97: 371-383. Levorsen, A.I., 1967. Geology of Petroleum. Freeman, San Francisco, Calif., 2nd ed., 724 pp.
297 Lewis, C.R. and Rose, S.C., 1970. A theory relating high temperatures and overpressures. J. Pet. Technol. 22: 11-16. Magara, K., 1972. Compaction and fluid migration in Cretaceous shales of western Canada. Geol. Surv. Can. Pap., 72-18: 81 pp. Magara, K., 1974a. Compaction, ion-filtration, and osmosis in shales and their significance in primary migration. Bull. A m . Assoc. Pet. Geol., 58: 283-290. Magara, K., 1974b.Aquathermal fluid migration. Bull. A m . Assoc. Pet. Geol., 58: 25132516. Magara, K., 1975a. Reevaluation of montmorillonite dehydration as cause of abnormal pressure and hydrocarbon migration. Bull. A m . Assoc. Pet. Geol., 59: 292-302. Magara, K.,1975b. Importance of aquathermal pressuring effect in Gulf Coast, Bull. A m . Assoc. Pet. Geol., 59: 2037-2045. Magara, K., 1976. Water expulsion from elastic sediments during compaction - directions and volumes. Bull. A m . Assoc. Pet. Geol., 60:543-553. Martin, R.T., 1962. Adsorbed water on clay: a review. Clays Clay Miner., 9 (ROC. 9th Natl. Conf. Claysand Clay Minerals, 1960, Pergamon, New York, N.Y., pp. 28-270). McKelvey, J.G. and Milne, I.H., 1962.The flow of salt solutions through compacted clay. Clays Clay Miner., 9. (Proc. 9th Natl. Conf. Clays and Clay Minerals, Pergamon, New York, N.Y.,Earth Sci. Ser. Mongr., 11: 248-259). Powers, M.C., 1967. Fluid-release mechanisms in compacting marine mudrocks and their importance in oil exploration. Bull. A m . Assoc. Pet. Geol., 51: 1240-1254. Price, L.C., 1976. Aqueous solubility of petroleum as applied to its origin and primary migration. Bull. A m . Assoc. Pet. Geol., 60: 213-244. Salathiel, R.A., 1973. Oil recovery by surface film drainage in mixed-wettability rocks. J. Pet. Technol., 25: 1216-1224. Schmidt, G.W., 1973.Interstitial water composition and geochemistry of deep Gulf Coast shales and sandstones. Bull. A m . Assoc. Pet. Geol., 57: 321-337. Skinner, B.J., 1966. Thermal expansion: Section 6 in Handbook of Physical Constants (revised ed.). Geol. SOC.A m . Mem., 97:75-96. Timko, D.J. and Fertl, W.H., 1971.Relationship between hydrocarbon accumulation and geopressure and its economic significance. J. Pet. Technol., 22: 923-930. Tissot, B. and Pelet, R., 1971. Nouvelles donndes sur les mdcanismes de genese et de migration du petrole: simulation mathhmatique et application B la prospection. Proc. 8 t h World Pet. Congr., pp. 35-46. Van Olphen, H., 1963. Compaction of clay sediments in the range of molecular particle distances. Clay Clay Miner. 11. (Proc. 11th Natl. Conf. Clays and Clay Miner. 1962) Macmillan, New York, N.Y., pp. 178-187. Vyshemirsky, V.S., Trofimuk, A.A., Eontorovich, A.E. and Neruchev, S.G., 1973. Pitumoids fractionation in the process of migration. In: B. Tissot and F. Rienner (Editors), Advances in Organic Geochemistry. Editions Technip., Paris, pp. 359-365.