Click
to article in PDF format.
Subsurface Fluid Compartments: Report*
By
D. E. Powley1
Search and Discovery Article #60006 (2006)
Posted March 14, 2006
*Adapted from Amoco Geological Research Report, December 30, 1984
1Amoco Production Company, retired, Tulsa, Oklahoma 74136
|
Purpose, Summary, and Introduction The purpose of this report is to summarize progress made over the last few years in understanding the movement of subsurface fluids as interpreted from subsurface fluid pressures. During that time it was recognized that, in most of the deep basins in the world, there are at least two superimposed hydraulic systems. The shallowest hydraulic system generally extends from the surface down to about 9000-12,000 ft greatest historical depth of burial and typically exhibits normal pressures. The deeper hydraulic system generally consists of individual compartments which are sealed off from each other and from the overlying system. The pressures within the compartments in the deep hydraulic system are usually markedly overpressured or underpressured relative to the pressures in the overlying hydraulic system. In some basins there is a deeper, near normally pressured section. This report deals with the geological and exploration implications of fluid compartments. The study was conducted on an informal when-time-is-available basis and draws from experience derived from several limited-objective technical-service-type studies conducted for various Company locations. Also, literature and personal-communication data were collected on about 200 of the world’s nearly 500 basins. Intensive collection and review of data were concentrated on about 70 basins, over half of which are in North America. This report does not attempt to discuss each of those basins; it deals with the summation of observations and conclusions drawn from all of the basins studied. A seminar or slides on the data assembled pertinent to specific basins can be supplied if needed.
Conclusions1. There are so many basins with a layer of fluid compartments that the formation and preservation of compartments appear to be parts of normal basin development. 2. Seals bounding subsurface fluid compartments may trap or localize the entrapment of oil and gas. 3. Recognition and mapping of subsurface compartments mainly on the basis of pressure data are generally relatively easy and can become parts of the normal procedures used in development and assessment of exploration plays.
Prior StudiesThere
have been three periods of innovative interpretations of subsurface
pressures. The first was an amazingly perceptive study of the
relationship of oil and gas pools to water density and water pressures
in sand reservoirs in Rocky Mountains basins conducted in 1924 and 1925
by Clare Coffin, a geologist then with an Amoco predecessor company
(Coffin, 1925). He introduced the concepts and mathematics of tracing
water flow through subsurface formations from entry at high elevation
outcrops to exits at lower elevation outcrops. He correctly pointed out
that, while fresh water entering high outcrops may, on the basis of
elevations of water inlet outcrops and water outlet outcrops alone,
appear to have sufficient head to push heavier salty water from deep
within basins, it actually selects the path of least resistance over the
heavier brines and thus moves within individual permeable rock layers
around the basins in the shallow basin-periphery The
next period of interest in pressure indicators of underground fluid flow
conditions was in the mid 1950’s when King Hubbert (Shell) and William
Russell (Texas A&M University), working independently, presented
mathematical explanations for the tilted oil/water interfaces observed
in a few pools. The explanations involved rapid movement of the water in
basal contact with the base of oil columns. Russell (1956) recognized
that tilted contacts may be produced by stratigraphic variations;
however, that cautionary note was lost in an almost competitive
outpouring of literature and talks on tilted oil/water interfaces by
those and other authors. Hubbert (1940, 1954), to a greater extent than
others, displayed the misconception that each aquifer is independent of
all other aquifers and the resultant misconception that aquifers in
basins may be likened to U tubes in which fresh water enters at high
elevation outcrops and proceeds directly across basins displacing heavy
deep-basin brines rather than flowing over or around the brines. Hubbert
became a distinguished lecturer for the American Association of
Petroleum Geologists on interpretive tracing of water flows by the use
of pressure data. He labeled his interpretation technique
“hydrodynamics”. His lectures included a movie of an experiment using a
large glass U tube with small kinks to represent an aquifer in a basin
with anticlines. He flowed water and oil into the apparatus and showed
that, by adjustment of flow When the pace of drilling revived after the 1957-1965 imports-induced domestic drilling slump, many wells were taken to greater than usual depths. Many of those deeper wells encountered higher-than-anticipated subsurface pressures. In many basins it was recognized that the top of the high pressures does not follow traditional stratigraphic layers, so many geologists’ interests shifted to searches for unusual local factors which might control fluid pressures. There was an ensuing flood of published papers and Company reports which attempted to relate the high pressures to various rock and mineral properties. Colin Barker (1972) and John Bradley (1973, 1975, 1976) broke the stream of rock and mineral properties-pressures relationship papers by resurrecting an old concept (Versluys, 1932; Illing 1938) that the expansion of water when heated exceeded the expansion of rock pores and they pointed out that thermal expansion and thermal contraction of confined pore waters resultant from changes in subsurface temperatures during progressive basin filling and subsequent progressive erosion satisfactorily accounted for most of the abnormal pressures being encountered in wells worldwide. Rock and mineral properties induced pressure effects, such as compaction, clay minerals changes and osmotic membranes, were left with a minimally important, contributory role in generation of pore pressures. Studies of pressures were generally terminated at this point.
During the
late 1960’s and 1970’s, simultaneously with the interest in the origin
of abnormal pressures, geochemists were collecting data and developing
concepts which placed the depth of thermochemical generation of
petroleum in the general 10,000 to 16,000 ft range in low-in-kerogen-associated-sulfur
Figures 1-16
TextInasmuch as there may be a wide range of reader prior knowledge regarding subsurface fluid pressures, this section is included to accommodate those with limited experience. More experienced readers may find it convenient to proceed directly from here to the next section. Readers desiring more illustrations of the basic concepts of subsurface pressures are referred to Amoco Geological Research Report M76-G-16 (Powley, 1976). Pressure is the force per unit area which fluids (liquids and gases) exert on the surface of any solid which they contact. Pressure exists at every point in a fluid at rest. The magnitude of the pressure is proportional to the depth below the surface; i.e., the pressure is the same at all points at the same level within a single fluid at rest. Also, the pressure at any depth is proportional to the density of the fluid (Figure 1). The pressure in a fluid at rest is independent of the shape of the containing vessel and is the same whether the vessel contains a fluid only or contains a fluid and a quantity of solids in grain-to-grain contact; i.e., not a suspension. Thus, in the earth, the pressure in a subsurface fluid is independent of the shape and size of the rock pores but is dependent upon the density of the fluid and upon the depth below its surface (Figure 2). In the earth, the datum water surface usually cannot be seen. However, pressure calculations commonly indicate that the rock pores are fluid-filled and interconnected from the top of the free water in the soil down to at least intermediate depths. Inasmuch as the soil water surface is usually only a few inches to a few feet below the topographic surface, it has become common practice to consider the free water surface and the topographic surface to be the same. In marine areas, the free water surface is considered to be mean sea level. The pressure previously discussed is that caused by the weight of a free-standing fluid column without any external pressure being applied. If any external pressure is applied to any confined fluid .at rest, the pressure at every point within the fluid is increased by the amount of the external pressure. This statement is known as Pascal’s Principle, after the French philosopher who first clearly expressed it. An example of a confined fluid is the fluid below a piston in a closed cylinder. The pressure in the fluid increases as external pressure is applied and returns to normal when the pressure is removed. Within the confined fluid, the rate of increase in pressure downward is the same with or without an external pressure (Figure 3). In
geology, the counterpart to the piston and cylinder walls previously
shown is any combination of rock layers and interfaces which completely
enclose a body of fluid-bearing rock in a low-permeability envelope. The
low-permeability envelope is usually referred to as a seal. A seal is
usually thin with respect to both thickness and lateral extent of the
enclosed rock body. An abnormally pressured rock body is like a huge
bottle. It has a thin, essentially impermeable outer seal and an
internal volume which exhibits effective hydraulic communication. The
interval rate of increase in pressure with increasing depth within the
internal volume is in direct accordance with the density of the internal
fluids (Figure 4). The fluid pressures in the internal volume may be
greater than, equal to, or less than the pressures in the fluids in the
The Keyes Field in northwestern Oklahoma is illustrative of the case in which the rock matrix at the base of each of the two top seals bears the entire weight of the overburden, so the fluid pressures start from zero at these levels. All of the fluid pressures are markedly less than the normal +/- 45 psi/l00 ft from the surface, so the pressures in the Keyes Field, except those in the shallow beds above the uppermost seal, are termed underpressures (Figure 6). The Carpathian Basin in Hungary is an example of a rock load,being partially borne by the fluids below a seal. The fluid pressures are normal from the surface down to the base of the Pliocene clastics but are greater than normal below a thick series of lava flows which separate the Pliocene clastics from the Miocene clastics. Wells drilled on the northern shelf penetrate the normally-pressured Pliocene clastics, the seal, and the subseal high-pressured Miocene formations, whereas wells drilled in the southern basin usually penetrate only normally-pressured Pliocene clastics. The interval rate of pressure increase with depth in the Miocene section is the same as the rate of change in the normally pressured Pliocene section (Figure 7). This figure and the previous figure illustrating the Keyes Field demonstrate the field applicability of Pascal’s Principle. Pressures which are less than can be attributed to a free-standing water column to the surface were termed underpressures during the discussion of the Keyes Field. Likewise, pressures which are greater than can be attributed to a free-standing water column to the surface are termed overpressures. Underpressures and overpressures together comprise the well-known classification, abnormal pressures (Figure 8). Overpressures develop when there is an excess of pore fluids over available pore space. The state of volume imbalance may be due to pore space shrinkage or to pore fluid expansion (Figure 9). Most of the origins suggested in the geological literature involve pore space reduction by extensive mechanical collapse of the rock matrix under conditions of increased depth of burial. The rock collapse theories may be applicable in regions of very incompetent rock, like the shallow gumbo shales in the North Sea Basin, but are not acceptable in regions of competent rock like the Uinta, Green River, and Anadarko basins, the Grand Banks, the basal Jurassic sands in Mississippi, and the clastics in much of the Gulf Coast Basin. For instance, laboratory “rock squeezing” experiments on overpressured Gulf Coast shales indicate that those shales have many of the strength characteristics of limestone; hardly a low strength material (Figure 10). The first widely acceptable origin for most of the abnormal pressures encountered in wells was suggested by Barker in 1972, and expanded upon by Bradley in 1973, 1975, and 1976. It involves creation of overpressures by thermal expansion of pore fluids as pore fluids become warmer under conditions of deeper burial and creation of underpressures by thermal contraction of pore fluids resultant mainly from removal of cover by erosion. The Barker-Bradley proposal is based on considering a sealed rock body to be an essentially constant-volume vessel similar to a high-pressure boiler. As the temperature of a constrained volume of fresh water is increased, the pressure will increase at about 125 psi/oF of temperature change. Pressure in a saturated brine will change approximately 110 psi/oF of temperature change under the same conditions. If these conditions are applied to a geological model involving 1000 ft of increase in depth of burial, moderately salty water in a confined rock body and a geothermal gradient of 1.75°F/100 ft, the increase in pressure due to thermal expansion will be 2048 psi. If the rate of subsurface pressure buildup is in the order of 2000 psi per each 1000 ft of additional burial, the pressures would quickly become excessive and the confined fluids would burst through their seal by natural hydraulic fracturing. The internal pressure would then be able to blow down to some lower pressure. The fact that overpressures and underpressures are so common indicates that the fractures self-seal in some manner when the pressures in sealed-off rock bodies change sufficiently. Thus, it seems likely that the pressures in a sealed-off rock body undergoing continuous temperature increase through progressive burial are in a continuous cycle of buildup to fracture pressure, then fracture of the seal, followed by pressure drop, fracture healing and then buildup of pressures again (Figure 11). It is
likely that pressure release by fracturing is localized at the
shallowest depth of burial of overpressured rock masses. For
illustration purposes, consider an overpressured rock mass at 15,000 ft
depth under a 200-ft-thick seal (Figures 12,
13, and 14). The fluid
pressure above the seal is 6882 psi, and 11,310 psi below the seal. Let
there be a local upbulge of the top of the overpressured mass to a depth
of 10,000 ft. The pressure differential across the seal remains the same
(Pascal’s Principle). The overpressured fluids could vent themselves by
natural hydraulic fracturing when the pressure is great enough to
overcome the horizontal rock stress plus the tensile strength of the
rock, plus overcome the fluid pressure in the formation being invaded.
The horizontal stress ratio used was taken from Anderson, et al. (1973),
and for the tensile strength figure it is assumed that the seal is
indurated shale. The pressure below the seal at its shallowest depth
(9000 psi at 10,000 ft) is greater than the pressure required to induce
natural hydraulic fractures (8629 psi at 9800 ft), so fracturing would
occur. However, at the regional depth of 15,000 ft, the pressure below
the seal (11,325 psi) does not exceed the pressure required to induce
natural hydraulic fracturing (12,904 psi) at 14,800 ft, so fracturing
would not occur at that depth. This model demonstrates that venting of
overpressures preferentially occurs at the point of shallowest depth of
burial of an overpressured rock mass. Likewise, any hydraulic fracture
venting of fluids from a normally pressured rock mass into an underlying
underpressured rock mass will preferentially occur at the point of
shallowest depth of burial of the underpressured Large reductions in pressure may be accomplished by the release of very small quantities of water. To depressure a 1000-ft-thick rock body with 10% porosity by 1000 psi requires the release of only 9.6/in3 water/in2 top surface. A square mile of rock, 1000 ft thick, 10% porosity, would have to yield only 6200 barrels of water to be depressured by 1000 psi (Figure 15). Most of
the discussions to this point have dealt with pressures in individual
wells or in local areas. When interpretations are extended from well to
well over large areas, it is generally easier to deal in terms of
potentiometric surfaces than with pressures in psi and pressure-depth
ratios. A potentiometric surface is the elevation of the upper free
surface of a fluid, generally at rest. In subsurface geology, the
potentiometric surface for an aquifer is the elevation to which a
free-standing fluid column of some specific density would rise if free
to do so in a well penetrating only that aquifer. In normally pressured
Calculations of potentiometric surfaces derived from elevations and
measured pressures data are highly vulnerable to error if the fluid
Fortunately, selection of an appropriate water density for use in
potentiometric surface calculations is usually quite easy. If carefully
measured pressures in normally pressured formations at several depths in
wells in a single basin are plotted against depth, the resultant profile
in each basin usually portrays a strictly linear relationship. The
pressure-depth ratio may be different from basin to basin, but within
any single basin, the ratio usually is remarkably constant. This is
apparently due to some equalization process in nature such that water
salinity, water compressibility, water temperature, and work being done
on the system jointly reach a vertical and horizontal equilibrium
density balance in all normally pressured formations within each basin.
For instance, a pressure-potentiometric head conversion factor of 45 psi/100
ft from the surface is applicable in normally pressured During
the 1950’s, an interpretive technique known as “hydrodynamic analysis”
was widely applied, mainly by the Petroleum Research Corporation, a
Denver based consulting company, in the Rocky Mountains states and in
western Canada. The technique was based on the assumption that there is
a component of fluid flow from all sites with high potentiometric
surfaces to all sites with lower potentiometric surfaces within the same
rock layer. The technique was developed into elegant mathematical
derivations of the velocity of steady rate flow between pressure control
sites. The velocity figures were commonly developed further to establish
minimum rates of dip which would be required to retain oil and gas on
structures. The principal shortcomings in the hydrodynamics rationale
was the failure to recognize that there is a close vertical and
horizontal interdependence of aquifers (as outlined in the foregoing
paragraph) and the failure to recognize that water bodies laterally
sealed off from each other in the same rock layer could have substantial
differences in potentiometric surfaces and thus lateral differences in
potentiometric surfaces could signify either flow or no flow. Also,
nearly all of the early practitioners of the hydrodynamics art used
fresh water Sources of Pressure DataTextThe pressures in fluids in subsurface formations are generally determined by measurements within wellbores which have penetrated those formations. Well log data may also be combined with empirically derived relationships to derive reasonably reliable indicated subsurface pressures. The art of measuring shut-in pressures in wells is in good order, and there are a variety of reliable pressure measurement tools available. Inasmuch as those measurements are regularly conducted under the supervision of Company drilling engineers, many of whom are available for consultation by readers of this report, there seems to be no necessity for compiling a “how to” section in this report.
Overpressures have been known and studied in the Gulf Coast Basin for
many years. Most of the techniques to safely drill and complete wells in
overpressured formations now in use worldwide were developed in the Gulf
Coast. One of the most significant techniques is the use of well logs to
identify and quantify overpressures. The techniques now in use are
modified from those introduced in a paper presented by Hottman and
Johnson in 1965. They reported the coincidence of high fluid pressures
in sands and lower-than-normal electrical resistivities and acoustic
The
reader should follow through the example calculations on Figures
18, 19,
20, 21,
22, and 23
at this point to insure that the technique is understood. This example
calculation is somewhat misleading inasmuch as the accuracy obtained is
better than that which can be routinely derived from average quality
well logs. The importance of the foregoing well log interpretation
technique is that it is possible to construct pressure-depth profiles
for overpressured sections without requiring downhole pressure
measurements. Geologists are now able to know more about the pressures
in overpressured Several
authors have noted that abnormally high pressures are frequently
accompanied by higher-than-normal geothermal gradients. Interval
geothermal gradients in overpressured Hottman
and Johnson (1965) contended that porosity in shale is abnormally high
relative to its depth if the fluid pressure is abnormally high. That
statement led to a flood of measurements of porosity and density of Gulf
Coast shales. Amoco measured dry bulk Sonic
well logs and field seismic data may indicate overpressures by a
reduction in interval
Geology of Abnormal PressuresFigures 25-26Return to top.
TextIn most
deep basins in the world there is a layered arrangement of at least two
superimposed hydraulic systems (Figure 25). The shallowest hydraulic
system generally extends from the surface down to about 9000 ft greatest
historical depth of burial in normal geothermal gradient basins and to
slightly greater depths in cool basins. There are a few remarkable
deviations, like the central North Sea Basin, the South Papua Basin, and
the Canadian Arctic Basin, where the base of the shallow system has
apparently never been buried more than a few thousand feet. The shallow
hydraulic systems are basin wide in extent and typically exhibit normal
pressures. The deeper hydraulic system usually is not basin wide in
extent. It generally consists of a layer of individual compartments
which are sealed off from each other and from the overlying system. In
some basins, mainly onshore, there is a deeper, near normally pressured
section (Figure 26). The compartmented layer, known as the Elisian
Regime in eastern European literature, is generally in the sequence of
Recognition of the layered arrangement of hydraulic systems is generally
quite easy. Only a few widely spaced, well documented deep wells with
several tests run over perforated intervals are generally necessary to
outline the overall arrangement of hydraulic systems in each basin.
However, in some young, foreign basins and in the Copper River Basin in
Alaska, fluidized rock material, mainly shale, and high pressure water
with minor hydrocarbons are being locally ejected upward from subsurface
overpressured compartments, through overlying normally pressured The individual compartments in the compartmented layer may be very extensive, as in some of the Rocky Mountains basins, or may be only a few miles across, as in the Gulf Coast Basin. The pressures within the compartments are usually markedly overpressured or underpressured relative to the pressures in both the shallower and deeper hydraulic systems. The compartmented hydraulic systems in geologically young basins are almost universally overpressured and are underpressured in most old basins. Thus, it appears that the compartments have an amazing longevity as they undergo a continuum from overpressures through normal appearing pressures to underpressures as their host basins progress from deposition, to quiescence, to basin uplift and erosion. In those basins with three layers of hydraulic systems, the boundary between the middle compartmented layer and the underlying layer usually follows a single stratigraphic horizon. For instance, the basal boundary of the compartmented section in the central Powder River Basin appears everywhere to be within the thin Cretaceous Fuson shale. However, the top of the compartmented layer is in many basins more complicated. It (1) tends to follow an irregular sands-over-massive-shale boundary in the Gulf Coast and Niger Delta basins, (2) it follows thin evaporites in many onshore European and southwestern U.S. basins, and (3) it follows horizontal or gently dipping planes which cut indiscriminately across structures, facies, formations, and geological time horizons in the northern Cook Inlet Basin, in the Alberta Basin, in the Anadarko Basin, and in many Rocky Mountains basins (Figure 26). Those top surfaces which do not follow a specific stratigraphic horizon are generally restricted to clastics-dominated sections. The planar-topped, compartmented sections are almost universally in basins which are older than the basins in which the compartmented sections exhibit much top surface irregularity. Thus it appears that there is some process in nature whereby the top surfaces of compartments in clastics-dominated sections can smooth themselves over time. The leveling process must be quite rapid because the tops of the two principal pressure compartments in the central North Sea Basin are horizontal over distances in excess of 100 miles despite the recent salt-induced structure development in the area. Planar seals may occur within, as well as on the top of, the compartmented layer. For instance, the shallowest seal in the Mill Creek Graben in southern Oklahoma is everywhere within the thin Marmaton shale; the next deeper seal is horizontal (-10,400 to -11,500 ft elevation), cuts through many Paleozoic formations across the graben and even extends, at the same elevation, across the adjacent Ardmore Basin. No deeper seals have been encountered in the graben or in the Ardmore Basin. Earlier in this report it was pointed out that the individual compartments in the compartmented layer are like huge bottles with thin bounding seals and huge fluid-communicating internal volumes. Seals are particularly annoying to work with because they do not have unique 1ithologic properties other than extremely low permeability. In the absence of unique lithologic properties, recognition must be accomplished from indirect evidence, such as well log indicators, measured pressures in local reservoirs encased in seal rock, and often only from the requirement that they must be there separating reservoirs which, from measured pressure data, are obviously hydraulically separated from each other. The transition of pressures across the thickness of top seals is linear wherever data have been obtained (Figure 24). No data have been accumulated to determine the patterns of pressures within lateral seals. In some areas, seals may be recognized by calcite and/or silica mineralization, probably resultant from dissolved minerals being precipitated as water seeps through the seals. The mineral infill of porosity and fractures may be so readily recognizable that it becomes an identifier of present or past seals. For instance, calcite infill is so ubiquitous in seals in southwestern Louisiana that it has been given the name “AI’s Cap,” named for Al Boatman, a local geologist, who first publicly drew attention to the phenomenon there. Silica infill may be recognizable on the basis of drastically reduced rates of drilling penetration across a seal. For instance, it took 24 hours to cut a 60-ft core in a silica-enriched seal in chalk in the Shell-Esso 30/6-2 well in the North Sea. Chalk normally cores very rapidly. Top seals in clastics-dominated sections range in thickness from 150 ft to over 2000 ft; however, the majority are uniformly near 600 ft. Seals in carbonate-evaporite sections are generally somewhat thinner; in fact some salt and anhydrite beds as thin as 10 ft form effective seals. An example of the latter is the Devonian Davidson Evaporite which, except for a small area in central Saskatchewan, is about 20 ft thick but forms a regional pressure seal over almost the entire extent of the Williston Basin. Lateral
seals appear to be generally vertical or very nearly vertical. They
range in thickness from less than one eighth of a mile (within the
distance between wells on 10-acre spacing) to about six miles, with the
majority being about one eighth of a mile in width. They tend to be
quite straight, suggesting that they may tend to follow fault trends.
There have not been any satisfactory suggested geochemical mechanisms
which could create impermeable walls over thousands of feet of vertical
extent through The
The
fractures in the internal volume are, in a few areas, open enough to
permit commercial-rate extraction of oil and gas even in the absence of
significant matrix porosity and permeability. However, the distribution
of open fractures is generally not uniform enough to allow field
development without a substantial proportion of dry holes unless the
fracture porosity is augmented with matrix porosity and permeability
within the internal volume
Basin FluidsFigures 27-34
TextThe
four most important recent developments in basin fluids concepts are (1)
the recognition that there probably is much less lateral movement of
pore fluids than was envisioned in the heyday of hydrodynamics, (2) the
recognition that vertical migration of pore fluids is more prevalent
than earlier recognized, (3) the recognition of the ubiquity of pressure
compartments and their effects on the movement of all pore fluids and
(4) underground hydraulic fracturing of
Combination of the foregoing concepts provides a speculative indication
of how petroleum starts its path from its source The foregoing is applicable where the shallowest depth of burial is due to an upbulge of the top of an overpressured section. However, in regions of great topographic relief, the shallowest depth of burial may be due to a local very low surface elevation. In that situation, hydraulic fracture breakout also occurs at the location of the shallowest depth of burial; however, there is no buried local upbulge to pre-collect oil or gas there. The Transylvanian Basin may be an example of that situation. The top seal of the overpressured section is horizontal and Romanian geologists have reported the ascent of hot, medium-salinity water with minor oil and gas in a few localities beneath major river valleys, but they have not reported similar ascending water plumes below the adjacent high plateaus. Around 1970 Bobby Newton, then the Region Geologist in New Orleans, attempted to categorize the localities of large oil and gas pools in southern Louisiana relative to their pressure environments. His system categorized pressures by relations to stratigraphy; i.e., pressure boundaries rising across stratigraphy, parallel to stratigraphy, or dropping across stratigraphy. Newton’s descriptive categories, which required precise correlation of beds, were difficult to recognize and difficult to work with, particularly during the wildcatting and early field development stage, so the system was not adopted; however, his diagrams indicate that all of the major pools studied are in very close proximity to local points of shallowest depths of burial of the overpressured hydraulic system. Figures 31, 32, 33, and 34 are from a seminar prepared by Newton in support of his descriptive category system. Those readers who remember the Newton seminar may note that wildcat wells would be located at the same sites if the Newton descriptive system is used or if the proximity-to-the-shallowest-depth-of-burial genetic concept is adhered to. Some oil and gas will escape entrapment in proximity to the regions of fracture breakout and may move far into the shallow hydraulic system. It may become trapped in shallow formations and, with luck, will escape degradation by water-borne bacteria. Conventional updip migration and trap concepts explain the occurrence of much of the “escaped” oil and gas. Oil and gas has also accumulated in abundance within the internal volumes of abnormally pressured compartments. Accumulations may be in traps within the internal volume or may be trapped within or against the bounding seals. Many of the pools located within the internal volumes of abnormally pressured compartments have tended to be rather small in the Gulf Coast Basin, probably because the individual compartments are small and because the reservoir sands are thin and discontinuous. All of the internal sands are oil-filled in the Altamont-Bluebell overpressured compartment in the Uinta Basin; all of the internal sands are gas filled in the Blanco underpressured compartment in the San Juan Basin; and all of the sands in the deep Wind River Basin overpressured compartment appear to be gas filled. Over one billion barrels of oil have been produced from Paleozoic reservoirs within the internal volume and within the top seal of the underpressured compartment in the Seminole Sag, a small graben adjacent to the Arkoma Basin in southern Oklahoma. The trapped-against seals accumulations may be at the highest internal elevation regions where the oil and gas awaits release by hydraulic fracturing or may be in regionally permeable beds where they are cut by bounding seals. The oil and gas pool in the fractured, overpressured Monterey shale reservoir on the Lost Hills Anticline in the San Joaquin Basin appears to be an example of oil and gas in a pending-fracture-release pool. The large, underpressured Viking oil and gas pools trending from the Oyen-Sedalia area through the Provost, Killam, Bruce, Beaverhill Lake, Fort Saskatchewan, Fairydell-Bon Accord, Westlock and Judy Creek fields in the Alberta Basin provide an example of entrapment against a bounding seal. Those fields are in a 350-miles-long megatrap where the southwestward dipping Viking sand is regionally cut by a horizontal seal at about sea level elevation (see Figure 50). An important extra benefit from petroleum remaining within abnormally pressured compartments, particularly in deeply eroded regions, is that oil pools are protected from contact with bacteria-bearing meteoric waters. For instance, the regional pressure seal in the Castile-Salado evaporites has protected the shallow, underpressured giant oil pools in West Texas from being bacterially degraded. The only bacterially degraded oil in the entire West Texas-southeastern New Mexico area is in above-the-seal beds, mainly near Santa Rosa, New Mexico. The foregoing generalization applies to conditions prior to the intrusion of man. Injection of bacteria bearing water during waterflooding has resulted in local bacterial degradation of oil in some pools. Many oil and gas pools within abnormally pressured compartments exhibit isolated ponds of water well above the pools’ water tables, probably because there was not enough vigor to fluid movements within sealed-off compartments to sweep all of the water out of the oil and gas pools. An example is provided by the Recluse - Bell Creek underpressured compartment in the northern Powder River Basin. In that compartment, the pool to pool interval pressure/depth ratio in the Muddy Formation is 33 psi/100 ft; i.e., the pool to pool pressure transmitting medium has the density of oil, rather than water. Thus a very large, continuous oil pool is indicated. However, there are large ponds of water which apparently have been prevented, by tight areas and local shaleouts in the Muddy sand, from moving downdip to the main water body. Thus the entire area is a huge oil pool with internal local ponds of water. Another example, of more direct interest to Amoco, is provided by the underpressured Wolfcampian Bough “C” limestone in eastern New Mexico. During the late 1950’s and early 1960’s, six widely separated oil pools were discovered. The pool to pool interval pressure/depth ratio was 33.6 psi/l00 ft; a density figure compatible with the oil in the six pools. Thus a very large, continuous oil pool was indicated. Amoco drilled an inter-pool wildcat, the State “DO” No.1, in a slight structural depression. The well yielded 350 ft of oil and gas cut mud and 5500 ft of slightly oil cut salty water on a drillstem test of the Bough “C”. Pipe was run and the well yielded 39 barrels of oil and 1728 barrels of salty water per day through perforations. The well was sold to an independent operator who placed a large pump on the well. After ten years, the water pond in the structural sag had been pumped out and the well had also produced 300,000 barrels of oil. The Bough “C” is now oil productive over almost its entire 500 square mile extent; however, there still are a few internal water ponds. Of more current interest to Amoco, both the North Poui oil and gas pool offshore Trinidad and the Saaja layered gas pool in Sharjah have some well log indications of water ponds; however, it appears that bottom water has not yet been encountered in either of those overpressured pools. Oil and gas may have been generated below the compartmented layer in some basins. For instance, the sub-Fuson pays in the central Powder River Basin probably were generated below and have remained below the compartmented layer. The oi1 and gas appears to have been unaffected by the overlying compartments, except that the deep section may have been effectively shielded from surface influences like meteoric water drives. Despite the obvious advantage of not being subject to strong water flushes, the hydraulic layer below the compartmented layer has not been found productive in many basins. Some of the low productivity is probably due to limited deep drilling.
Ruptured CompartmentsFigures 35-37Return to top.TextUp to now we have been dealing with compartments in which the bounding seals have been continuously intact since their formation or have undergone brief episodes of hydraulic fracturing and subsequent healing. There are several recognized compartments in which the bounding seals have been permanently ruptured by erosion or faulting or were breached by hydraulic fracturing without subsequent healing. The remaining seal segments apparently are still as impervious to gas, oil, and water as they were when the seals were complete; therefore recognition of seal segments is very important in petroleum exploration.
Pressures within a newly ruptured compartment will progressively change
toward equilibrium with the pressures in the external water through
fluid leakage into or out of the compartment at the point of rupture.
When pressure equilibrium is reached at the elevation of the rupture,
there is no pressure differential to move fluids farther. If the rupture
is large, or if the adjacent
Compartments in which the bounding seals were breached by hydraulic
fracturing at their shallowest depths of burial without subsequent
healing may still contain huge amounts of oil and gas. The giant Milk
River gas field in Alberta may be of this type. It fills an
underpressured compartment with an internal gas pressure of 625 psi at
its updip terminus. The adjacent external water pressure is also 625 psi.
Figure 37 portrays the pressures at Milk River if PGl, the pressure in
the gas, equals PWl, the pressure in the adjacent updip water. Note that
the compartment is normally pressured at its updip leak point but,
because gas is less dense than the external water, the gas pool is
underpressured relative to the external water in its full downdip
extent. The better known, underpressured, giant Medrano oil pool on the
Cement Anticline in Oklahoma is of the same type. The Medrano pool is
ruptured underground at its updip terminus. It has continued to leak oil
as erosion has progressively removed cover and thereby reduced the
external normal pressures. The leakage plume over this field has been
extensively used by promoters of geological and geophysical techniques
which directly sense the hydrocarbon plume or sense the chemical changes
in A compartment may be breached by erosion, generally at the pre-breaching site of the shallowest depth of burial of the upper seal. When this occurs, any oil or gas awaiting hydraulic fracture breakout would suddenly be exposed to the atmosphere. The giant Athabasca tar sands deposit in Alberta probably had this history. The deposit is at the northeastern updip terminus of the sub-Viking pressure compartment which extends over most of the Alberta Basin. Inasmuch as the water-bearing formations contain salty water all the way up to the outcrops, the compartment, now underpressured except at the rupture area, may have been overpressured until erosional breaching. The
giant Oklahoma City Field apparently had a similar early history. That
field is located at the updip terminus of the lower (sub-Meramec) tier
of compartments in the Anadarko Basin compartmented layer. The
compartment was breached by early Pennsylvanian erosion. A thin, but
extensive tarry layer at the unconformity attests to the pre-erosion
presence of a large oil pool. The unconformity was reburied by thousands
of feet of Pennsylvanian and younger Rupture
of a seal downdip from the updip terminus of a dipping compartment will
lead to pressure equalization at elevation of the point of rupture but,
if the rupture is small or if the adjacent The evidence for ruptured seals in the tight gas sand areas is not unassailable. It is possible that some of those compartments have not been ruptured; rather they are fully sealed but are in the “midlife identity crisis” period when compartments are passing from early basin-life overpressures to late basin-life underpressures. A large rupture in a compartment seal may lead to a normally pressured water column within a compartment, not only downdip from the point of rupture, but also updip to the base of any oil or gas column trapped against the remaining updip seal segment. Thus, any wells drilled into the water-bearing sector of the compartment would not yield an abnormal pressure indicator of the presence of a compartment. It would be easy to overlook the petroleum trapping potential of the unruptured updip segment of the compartment seal. There is no current geological or geophysical method known to the author for recognition of such seal segments except by inference. For instance, a trend of pools at the same elevation, an unusually straight line trend of pools, superimposed “stratigraphic trap” pools, or even the apparent abutment of different salinity waters in apparently continuously permeable beds may spark intuitive interest and lead to recognition of the fundamental trapping mechanism while there is still time remaining and acreage available to wring a reward from the interpretation. The terms “point of pressure equalization” and “seal rupture,” used in the preceding paragraphs, may be misleading because they may create the impression that the internal-external pressure equalization path is necessarily quite short. In a few cases, the path from the internal volume of a compartment to the external normal pressure control is a very long distance, particularly if the equalization path extends from one compartment into, across and out of an adjacent compartment. For instance, the internal pressure at the base of the gas-filled Blanco pressure compartment in the central San Juan Basin appears to be controlled by the elevation of the Paleozoic rock outcrops in the Grand Canyon, slightly over 200 miles away. The intervening path is interpreted, on the basis of coincident elevations of potentiometric surfaces, to be through the underpressured Paleozoic formations in the Paradox Basin.
Mapping CompartmentsFigures 38-51
TextThe
most fundamental elements of the petroleum geology of abnormal pressures
and of the geology of compartments are the geology and the geometry of
seals. In general, for petroleum exploration purposes, it is unimportant
whether the pressures in a compartment are markedly abnormal or only
slightly abnormal, whether the compartment encloses a thousand square
miles or is only half that size, or whether a compartment contains
Paleozoic and Mesozoic For
mapping purposes, seals may be considered to be subsurface layers or
surfaces which are recognizable on both regional and local scales, may
be correlated from place to place and may be mapped like other
subsurface layers or surfaces. Top seals and bottom seals are like
“thick” unconformities; i.e., they may cut across or may parallel
depositional layers and are identifiable mainly on the basis of the
differential properties of the shallower and deeper sections. Vertical
seals are like “thick” faults; i.e., they cut across depositional layers
and are identifiable mainly on the basis of the differential properties
of the abutting sections. In the case of seals, the differential
properties referred to are the fluid pressure regimes in the adjacent
When
commencing a study of subsurface pressures in a previously unstudied
basin, an investigator should first determine if, and approximately
where, abnormal pressures have been encountered in wells within the area
of study. Most government field-development regulatory bodies in the
United States and Canada require sworn-to public disclosure of the
discovery shut-in pressures in all productive pools; so this data source
is generally the best place to start. In most domestic basins that data
source is sufficient to roughly outline the main pressure compartments,
if present. In those Company locations which have old potentiometric
surface maps on file, those maps should be examined for bands of
over-steep dip; i.e., very high rates of change (Figure 38), reversed
dip, or bands of no dip in an otherwise dipping potentiometric surface.
Even if very inappropriate fluid The
next step is to construct work maps and probably also construct
supporting cross sections using only very reliable (preferably Amerada
or Kuster gauges) pressure data from vertical wellbores. The outlines of
any large compartments probably will become quite clear. Additional data
will likely be required along the boundaries of the compartments, but
there is generally little to be gained at this stage from an exhaustive
gathering of test data from wells centrally located within large
compartments. Having assembled a body of measured pressures data, a map
of the potentiometric surfaces should be constructed. It is important to
use a pressure-water head conversion factor which fits the The work maps must now be fleshed-out with more data along either side of each seal. In some basins, like the downdip Gulf Coast Basin and the western Sacramento Basin, the vertical seals tend to coincide with major faults so regional structure should be considered in selecting the mapped locations of seals. In assembling data from wells, an investigator should be wary of pressures measured in formations which are, or were, productive or are, or were, water disposal zones in nearby fields or are productive of water in nearby cities; the pressures may have been significantly altered by fluid withdrawal or by fluid injection. The next step is only a slight variation of the procedure well known to most experienced subsurface geologists; i.e., examine every indicated updip interruption in carrier bed continuity to determine if a stratigraphic trap type or fault type play may be made. Figures 39, 40, 41, 42, 43, and 44 portray the suggested steps using the Anadarko Basin as the illustrative area of study. Figures 39, 40, and 41 are pressure-elevation profiles of discovery pressures in individual pools, using data derived mainly from state government regulatory sources. Note the generally clean separation of pressure profiles, hence leading to an early recognition of the reality and approximate locations of major pressure compartments. Figure 42 shows the approximate outline of the individual compartments in the combined Morrow and Springer formations. The structure map of the Morrow Formation (Figure 43) is referred to next to determine where interruptions in updip carrier bed continuity are indicated and thus where infill data are required. Combination of Figures 42 and 43 (Figure 44) portrays several updip compartment boundaries and corners, which are prospective. Many of those sites have been tested; some have been productive for years, and one is currently (1984) being developed; however, a few prospective updip corners are still untested and are currently being studied further by the Denver Region. The
Anadarko example may be misleadingly simple, inasmuch as the pressure
control from fields is adequate to outline most of the pressure
compartments. The more usual situation is that there are only a few
pools in each compartment (Figure 45); so the investigator is faced with
a large fleshing-out job using data from wireline tests, drillstem
tests, echometer readings, and even Pressure-depth profiles using only reliable pre-drawdown shut-in pressures in several formations in individual fields or wells may facilitate recognition of the vertical arrangement of pressure compartments. The investigator should correct all within-pay pressures for buoyancy to pressures at or below the bottom water surface. Figure 46 illustrates the error which could be introduced by using a within-pay pressure, particularly if that pressure was measured high up in a long oil or gas column. Figure 47 portrays pressures measured at several depths in an individual field. The profile indicates one abnormally pressured compartment and the approximate location of the top seal. That information is sufficient to get an investigation underway. Figures 48 and 49 portray the follow-up steps; i.e., gather more data and then construct maps. The next step will be to determine if the lateral seal between the two compartments crosses the plunging, northwest-trending anticlines in the area. There may be down-plunge plays yet to be made there. Another “getting started” technique .is to construct regional cross sections using only very reliable data. Figure 50 is a portion of the regional cross section which led to .the.author’s investigation of the pressure compartments in the Alberta Basin. Note that the hydraulic interruption in the Viking sand near Killam is not readily apparent using reservoir pressures alone; the potentiometeric surface is required for recognition (Figures 50 and 51).
Future WorkThis report is intended to provide a technical and conceptual background for using pressure data in developing and modifying exploration plays. The application techniques are sufficiently developed that the Regions may take over most pressure related applications and Geological Research may shift farther into a support-when-needed role on the subject. There is need for further development of geophysical techniques for identification and quantification of abnormal pressures. Also, further understanding of seals might be worthwhile, but it is not clear that new work by the Research Center is required now. If we wait until the Regions have worked with seals, research could be in response to real needs, not just to anticipated needs.
ReferencesAlliquander, O., 1973, High pressures, temperatures plague deep drilling in Hungary: Oil and Gas Jour., v. 71, no. 21 (May 21), p. 97-100. Anderson, R.A., Ingram, D.S., and Zanier, A.M., 1973, Determining fracture pressure gradients from well logs: Jour. Petrol. Tech., v. 25, p. 1259-1268. Barker, C., 1972, Aquathermal pressuring, role of temperature in development of abnormal pressure zones: AAPG Bulletin, v. 56, p. 2068-2071. Bradley, J.S., 1973, Abnormal formation pressure: Amoco Geological Research Report F73-G-6, 33 p., 25 figures. Bradley, J.S., 1975, Abnormal formation pressure. AAPG Bulletin, v. 59, p. 957-973. Bradley, J.S., 1976, Abnormal formation pressure: Reply: AAPG Bulletin, v. 60, p. 1127-1128. Coffin, R.C., 1925. Notes on the circulation of water in the sands of structural basins as related to the occurrence of oil and gas in the Rocky Mountain region. Preliminary Report to Midwest Oil Company, March 14: 55 p., 2 plates.
Handin, J., and Hager, R.V., Jr., 1958, Experimental
deformation of sedimentary Hottman, C.E., and Johnson, R.K., 1965, Estimation of formation pressures from log-derived properties. Jour. Pet. Tech., v. 17, p. 717-722. Hubbert, M.K., 1940, The theory of ground water motion: Jour. Geo1., v. 48, p. 785-944. Hubbert, M. K., 1954, Entrapment of petroleum under hydrodynamic conditions: AAPG Bulletin, v. 37, p. 1954-2026. Illing, V.C., 1938, The origin of pressures in oil-pools: Science of Petroleum, Oxford Univ. Press, v. 6, p. 224-229. Narr, W., and Currie, J.B., 1982, Origin of fracture porosity - example from Altamont Field, Utah: AAPG Bulletin, v. 66, p. 1231-1247. Powley, D.E., 1976, Pressures, normal and abnormal: Amoco Geological Research Report M76-G-16, 11 p., 56 slides, 1 tape recording. Powley, D.E., 1982, The relationship of shale compaction to oil and gas pools in the Gulf Coast Basin: Amoco Geological Research Report F82-G-23, 13 p., 60 figures, Appendix 500 figures. Rogers, L., 1966, Shale-density log helps detect overpressure: Oil and Gas Journal, v. 64, no. 37, p. 126-130. Ronai, A., 1978, Hydrogeology of great sedimentary basins: Proceedings of the Budapest Conference, 1967: International Association of Hydrologic Sciences Publication no. 120, 829 p. Russell, W.L., 1956, Tilted fluid contacts in Mid-Continent region: AAPG Bulletin, v. 40, p. 2644-2668. Vers1uys, J., 1932, Factors involved in segregation of oil and gas from subterranean water: AAPG Bulletin, v. 16, p. 924-942.
AppendixAir standard and water standard (psi/100 ft)
Density vs. temperature and pressure for water and NaCl solutions Density of average natural gas versus depth Crude oil density versus temperature and pressure Determination of subsurface density and pressure gradient from stock-tank API gravity and GOR Depth correction to find true vertical depth of nonvertical boreholes |