Professional Documents
Culture Documents
Geologic Problems Related To Characterization of Clastic Reservoirs For EOR
Geologic Problems Related To Characterization of Clastic Reservoirs For EOR
Characterization of Clastic
Reservoirs for EOR
M. Szpakiewicz, Natl. Inst. for Petroleum and Energy Research
K. McGee, Natl. Inst. for Petroleum and Energy Research
B. Sharma, SPE, Natl. Inst. for Petroleum and Energy Research
Summary. Geologic problems encountered in four U.S. DOE-sponsored EOR projects in clastic reservoirs were analyzed. The
reservoir heterogeneities were grouped into four categories: depositional, diagenetic, structural, and formation-fluid
composition/distribution. Each category had a variable effect on the performance of the EOR projects.
Requirements for developing one static geologic model for each category of heterogeneity are proposed. The interrelationship of
geologic factors and the effect of heterogeneities resulting from the origin and timing of geologic events are described.
Besides heterogeneities associated with the depositional environment, which has a significant effect on EOR performance,
diagenetic effects resulting from geochemical alterations and different types and polymorphs of clays are also important but are
often neglected. The structural component of the model needs to be improved through the application of new methodology or
improvements in existing technology (e.g., remote sensing, high-resolution seismic). A realistic prediction of fluid movement in
clastic reservoirs subjected to EOR requires the integration of the four models that define all the heterogeneities.
Introduction
The purpose of this continuing research is to develop a generic ap- oil and to define the most cost-effective recovery method, reser-
proach to reservoir characterization for EOR rather than to pro- voir parameters altered during primary, secondary, and/or tertiary
vide a technique for constructing individual reservoir models. recovery must be idel\tified and quantified. In practice, however,
Results of this work will bring awareness to many problems that comprehensive geologic reservoir characterization is not always in-
recur in EOR projects within clastic reservoirs and that indicate cluded in designing EOR projects.
the need for further refinement. Geologic data from four DOE-sponsored EOR pilot projects were
Small-scale heterogeneities, not always critical to primary and analyzed to determine the effect of various geologic heterogenei-
secondary recovery, may affect sweep and displacement efficien- ties on EOR performance. Results of the four case history analy-
cies in EOR operations significantly. In addition to the log and core ses as well as our experience with similar projects indicate that the
data available, new methodology is needed to improve reservoir various heterogeneities assigned to four categories can be studied
characterization in the interwell areas. The importance of recon- best through development of four static geologic models, one cor-
structing the paleohistory of formations is emphasized to provide responding to each category. The requirements for optimum geo-
information on the variability of diagenetic, structural, and forma- logic modeling of clastic reservoirs for EOR are proposed
tion-fluid heterogeneities. subsequent to the review of the geologic problems encountered in
Improvement in reservoir characterization for EOR requires an the EOR pilot projects. Geologic factors that affect fluid movement
understanding of the origin, timing, trend, and magnitude of geo- in reservoirs are defined, and attention is given to certain aspects
logic events that affected the reservoir anatomy (internal structure of reservoir characterization-geochronology, geochemistry, and
and composition of the rock). A biased or partial approach to reser- man-induced alterations. A final hydrodynamic model derived from
voir characterization may result in complete failure or suboptimal the integration of four partial geologic models will provide the means
results. In 1979, Dickey 1 described several EOR projects that for a more accurate prediction of fluid movement in the reservoir.
failed because the actual geology had not been considered. In 1986,
Dickey 2 further stated that the most common cause of EOR proj- Pilot Area Studies
ect failure was heterogeneity of the reservoir.
Reservoir characterization for EOR purposes should include data The historical background, performance, and geologic problems
from all available sources (analogous reservoirs, outcrops, aqui- encountered in four EOR pilot projects (Big Muddy, WY, North
fers , etc.). The multimodel system developed from these data would Burbank, OK, Bell Creek, MT, and EI Dorado, KS) reported in
include a subsequent ranking and integration of the various factors the literature were studied. Information from these studies contribut-
responsible for specific flow behavior. ed to the development of the proposed modeling for improving
The basic framework for a clastic reservoir description is derived characterization of clastic reservoirs for EOR.
from the geometry and anatomy of the oil-producing formation at
the time of deposition. The span of time from deposition to the pres- Big Muddy Field. Production in Big Muddy field is primarily from
ent introduces a variety of changes in internal structure and com- the Second Wall Creek sandstone member of the Upper Cretaceous
position in the formation. Frontier formation. 3.4 There are three Wall Creek sandstones in
Reconstruction of postdepositional conditions should include (1) the field. The upper and lower sandstones are unproductive and
interpretations of maximum depth of burial, residence time at differ- are separated from the middle Second Wall Creek sandstone by 20
ent depths, and tectonic history; (2) the extent of paleoexposure, and 30 ft [6.1 and 9.1 m], respectively. 3-5 Structural entrapment
erosion, weathering, and reburial conditions; and (3) knowledge is within an east-plunging anticline 6 characterized by more than
of the time of hydrocarbon entry into the reservoir. During the 300 ft [91.4 m] of closure above the oil/water contact (OWC).5
evolutionary history of reservoir rock, mineral assemblages con- The reservoir characteristics for the Second Wall Creek sand in
tinue to equilibrate chemically with formation fluids, and intermedi- Big Muddy field are summarized in Table 1. Several interpreta-
ate byproducts are produced that contribute to the overall tions ofthe depositional environment have been given (deltaic, bar-
heterogeneity of the rocks. To determine the amount of residual rier bar, and blanket sand).
Two EOR pilot projects have been implemented in Big Muddy North Burbank Field. The oil reservoir at North Burbank field,
field. Conoco began a I-acre [O.40-ha], five-spot, low-tension sur- OK, is a large sandstone body that was deposited in a fluvial envi-
factant polymer flood in 1970. 3-9 In 1978, Conoco and the DOE ronment on the north-central Oklahoma platform along the western
initiated a 9O-acre [3.6-ha] expansion pilot (nine lO-acre [4-ha] five- shore of the Cherokee Sea during the Pennsylvanian Age. This mas-
spots) to test surfactant flooding in a low-permeability, high-fracture- sive sand, known as the Burbank, was created by the superimposed
density, freshwater reservoir. The natural fracture system and geo- effect of a series of channels cut laterally into each other, which
logic problems such as low-pressure parting within the reservoir created the effect of a wide, massive body with a few internal dis-
were studied. 7 The daily tertiary oil producing rates were some continuities. The individual fluvial channels are about 1,000 ft
220 bbl [3.5 m 3 ]lower than that predicted, possibly because of [304.8 m] wide. Examinations of the core samples suggest that ma-
highly conductive joints and because of the reduced injection and rine incursions have disrupted the continuity of the Burbank fluvi-
production rates mandated by low-pressure parting in the for- al channel system periodically. 14 The Burbank structure is an
mation. 9 undulating monocline dipping at a rate of 35 ft/mile [6.63 m/km]
in a westerly direction. Several domes and synclines are present
Big Muddy Field Geologic Problems. Many geologic problems in the field,
experienced during the first pilot project were circumvented be- Reservoir and fluid properties that are important in understand-
fore the second pilot project began. The reservoir fractured below ing the geologic heterogeneities in the North Burbank field are sum-
the hydrostatic head. 9 One-half of the fluid production was from marized in Table 2. Petrographic analysis of thin-sections from Well
formation water influx as a result of a highly jointed system within 27-W26 indicate that the sandstone typically contains 40 to 50%
a low-matrix-permeability reservoir. Core analyses indicated that quartz, with substantial amounts of cementing materials consisting
the reservoir was of uniform thickness, 7,10.11 yet existing interpre- of silica, dolomite, ankerite, or clays. Lesser amounts of feldspars,
tations of the depositional environment are contradictory. Geolog- mica, chert, and metamorphic rock fragments were also observed,
ic and engineering data provided evidence of an east/west fracture None of the clays in the North Burbank reservoir are water-sensitive.
pattern. The general fracture orientation was determined; however, The North Burbank reservoir is an attractive target for tertiary
the density, distribution, and size of the fractures were not defined oil recovery because more than 360 x 10 6 STB [57,2 x 10 6 stock-
sufficiently to document their influence on fluid migration within tank m 3 ] of sweet 39° API [0,829-g/cm 3 ] oil will remain unrecov-
the reservoir. 5.11 This definition becomes critical when fracture ered when waterflooding reaches its economic limit. 15 Because of
orientation is oblique to the natural reservoir gradient direction. 3. 7 the severe vertical heterogeneities and consequent poor sweep ef-
Tracer surveys and temperature logs indicated that injected fluids ficiency, oil recovery by waterflooding in certain parts of the North
were migrating up-section, possibly along the annulus of old aban- Burbank field, like the Tract 49 area, has been about half that of
doned wells or within the prevailing fracture system. other. parts of the unit. In Tract 97 (site of a surfactant/polymer
Diagenesis in the form of clay alteration, patchy carbonate ce- pilot), oil recovery by waterflooding was excellent, mainly because
ments, and quartz overgrowths was observed in previously studied the staggered linedrive configuration swept the reservoir fairly uni-
equivalent outcrops of the Second Wall Creek sandstone in the formly.
Casper Arch area. 12 Diagenesis was not considered detrimental in Cumulative tertiary oil production in North Burbank is less than
the subsurface; however, little information was available to sup- that predicted. In the surfactant pilot project, for example, projected
port that contention. Pressure-pulse data indicated that there were ultimate oil recovery is estimated at 300,000 bbl [47,7 x 10 3 m 3 ],
no permeability barriers across several previously defined fault about half of that predicted. 16 Although oil recovery has been less
traces. Further study indicated that many of the "faults" were than that predicted, the pilot project was considered technically suc-
mapped from poor-quality well logs and that what was interpreted cessful because a significant amount of tertiary oil was recovered.
as fault gaps were probably changes in log response resulting from
localized changes in stratigraphy. 5 Consequently, many of the North Burbank Geologic Problems. The Burbank sand is a flu-
original faults were omitted from EOR structure maps. Several ar- vial deposit that characteristically exhibits rapid vertical and horizon-
guments support the concept for a highly tectonic reservoir, how- tal lithologic changes with corresponding changes in reservoir flow
ever, including the proximity to highly faulted reservoirs near the properties (see Table 2). The severe permeability stratification of
Laramie Uplift, and the complexity of photo lineaments observed the reservoir rocks at North Burbank, primarily related to the deposi-
on Landsat imagery throughout the region. 5. 12 tional environment, created numerous problems in the two chemical
Knowledge of the formation anatomy, water chemistry, and ef- EOR pilots.
fects of mineralization during chemical flooding is needed to im- When a highly stratified reservoir like North Burbank is subjected
prove reservoir characterization for predicting EOR performance' to surfactant/polymer flooding, the high-permeability zones will
in Big Muddy. The key to improvement first requires an accurate be exposed to more surfactant and should have lower ultimate residu-
and detailed interpretation of the depositional environment because al oil saturation (ROS), In a reservoir like North Burbank, how-
it furnishes the "map" from which diagenesis can be formulated. ever, the situation is complicated because cross flow could lead to
partial resaturation of high-permeability zones by oil displaced by Correlation and mapping of the various permeability zones from
more slowly moving fluids in adjacent lower-permeability zones. available information on depositional history and core, log, and high-
There is also the possibility of formation brine crossflow from a resolution seismic data and integration of this information with reser-
low-permeability layer into a preflushed zone containing surfac- voir engineering data (tracer test, pressure-transient list) would be
tant and/or polymer, causing degradation of their chemical activity. critical in assessing the magnitude of the crossflow problems like-
Studies of models have indicated that such crossflow could occur ly to be encountered in different parts of the reservoir.
and may have some effect on performance ofthe surfactant/polymer A primary cause for the less-than-expected oil recove7 at North
method. I? Burbank was high consumption of injected sulfonate. 16. I Loss of
Besides heterogeneities imparted to the rocks because of the flu- sulfonate was attributed to adsorption to the oil-wet pore surfaces
vial nature of the deposits, a strong east/west-trending, vertical-joint- and mixing of micellar fluids with in-place water that has a high
ing, or fracturing system causes preferential movement of fluid in sal inity and hardness.
the east/west direction. Owing to this system of fractures, the ef- In oil-wet rocks, the displacement of one nonwetting phase by
fective permeability in the east/west direction is five times as great a second non wetting phase is inefficient; therefore, the preflush at
as that in the north/south direction. I? Results of pressure-falloff North Burbank might have displaced only part of the formation
tests on several wells in Tract 97 and neighboring areas indicate water, creating an unsuitable salinity and hardness environment for
vertical fractures in these wells ranging in length horizontally from the micellar fluid. 16 It has been reported that compaction, authi-
165 to 623 ft [50.29 to 189.9 m] and flow capacities ranging from genesis, replacement, and dissolution have altered the texture and
665 to 2,600 md-ft [203 to 792 md·m]. composition of the Burbank sandstone significantly. 22
Pressure interference test data from the area also suggest a roughly Further research on the diagenetic effect and the wettability char-
east/west preference in the directional permeability of the North acteristics is needed. Careful analysis of the formation fluid would
Burbank reservoir. The radioactive tracer study performed during help in designing an alternative preflush to remove excess salinity
preflush injection showed that several wells were receiving little and hardness from the reservoir. Further research in developing
or no fluid from certain quadrants of their drainage areas. The flow a suitable surfactant system may also prove beneficial under the
of surfactants and polymer would also be restricted in these quad- prevailing salinity and hardness environment in the reservoir.
rants, and oil recovery from these areas would be low. The radi- Natural and man-induced fractures also presented problems.
oactive and subsequent chemical tracer studies also pinpointed the Natural fractures are well developed in outcrops of Pennsylvanian
location of flow channels in a number of wells in the area studied. limestone and sandstone and are visible on the ground and on areal
The channeling probably resulted from the directional fracture sys- photographs. I The average fracture azimuth is 70° (N700E) and
tem in the pilot area. 14 This fracture system is dynamic in that the is constant, with only a few fractures less than 65° or more than
fractures open and extend to great lengths with increasing injec- 75°. Information on fracture distribution, length, and infilling (detri-
tion rates but close when fluid injection is shut off. The low for- tal or geochemical) could be obtained from correlative studies of
mation parting pressure also prevented high injection rates. 18 cores, outcrops, and areal photographs.
The oil-wet nature of reservoir rocks resulting from a coating Another problem was the low parting pressure of the reservoir
of the iron-rich clay mineral chamosite on 70% of the pore sur- rocks, which opened the fractures whenever injection pressure ex-
faces causes adsorption of sulfonate surfactant and leads to ineffi- ceeded the critical pressure in the reservoir. When this occurs, large
cient displacement of oil. The relative displacement index measured volumes of fluid may circulate readily without moving much oil.
on a Burbank sandstone core from Tract 97 has an average value Although several fracture treatments increased fluid production,
of -0.45,14 indicating the highly oil-wet nature of the sand. it was speculated that they also induced interwell channeling. The
Formation-water analysis data from two samples 19.20 indicate channels were created from extension of fractures at producing wells
that the formation water from the producing sandstone at North Bur- when the micellar fluids were injected. This channeling could have
bank has very high salinity and high hardness (high Ca + + and prevented the polymers from following the micellar fluids at the
Mg + + content). The salinity and the chemical composition of proper sequence in all parts of the reservoir. 16
brine were different in the two samples. The sulfate content reported Further research on estimating and controlling fracturing pres-
in one analysis was low (58 ppm.) The high hardness of formation sures and improving workover procedures is needed to improve
water caused precipitation of calcite and sulfonate salts, thereby sweep efficiency in this project area.
reducing the permeability of the medium.
The reservoir description of North Burbank was incomplete when Bell Creek Field. Production in Bell Creek field is from the "Zone
the surfactant/polymer pilot was begun and presented a more favora- 2" barrier bar sand of the Lower Cretaceous Muddy formation.
ble picture than was actually found. 18 Permeability stratification Bell Creek field is a combination of six separate waterflood units,
adversely affected the performance of the chemical flood. each representing a separate barrier bar sandstone reservoir. 23-25
These stratigraphic traps are defined by updip pinchouts to the east Bell Creek Geologic Problems. Unexpected geologic heterogenei-
(into lagoonal facies) and by multiple OWC's along the western ties that adversely affected the performance of the first pilot were
downdip direction.23.24.26 Each waterflood unit has at least two related primarily to the. depositional environment of the Zone 2
sandstone intervals separated by a shale barrier. 25 The lower sand, sand. 27 Authigenic clay infilling had altered the primary porosity
commonly called the Zone 2 sand, is the most favorable reservoir and permeability in the reservoir. 23 Fine-grained sands, interbed-
host rock in Bell Creek field. 24 ,27 The flow capacity of the over- ded siltstones, and shales formed a permeability barrier in the south-
lying Zone I sand is one-eighth that of the Zone 2 sand, and vari- western part of the pilot area that adversely affected the sweep
ous studies have indicated little or no communication between the efficienc~ by diverting the flow of the injected fluid out of the pilot
two sands. 23,27 The Zone 2 reservoir is a very-fine- to fine-grained area. 27 ,2 Reservoir continuity was assumed on the basis of a par-
quartzose sandstone that is clean, moderately well sorted, and semi- tial geologic assessment, and drilling was continued before suspected
consolidated. Clay and silt fractions range from 2 to 8 wt%. 26,28 heterogeneity problems were confirmed by a complete analysis of
Because of the excellent reservoir qualities conducive to micel- pressure-pulse test data. 29 Brine resistivity values measured from
lar flooding, two tertiary-mode EOR pilot projects were implement- swab-water samples and log calculations indicated two chemical
ed to test the Zone 2 reservoir in the northernmost waterflood types of formation water. This analysis suggested that the water-
segment area known as Unit A.24 Although a small gas cap was flood was not successful in displacing oil and brine in the pilot area
present at the time of discovery, the Zone 2 reservoir in the Unit adequately. 24 .
A area was depleted at a faster rate than the other w.aterflood seg- A three-layer black-oil model failed to incorporate the complex
ment areas. 23 The reservoir characteristics for the Zone 2 sand in heterogeneity withil} the pilot area adequately. The acceptance of
Unit A are summarized in Table 3. inadequate waterflood history-matching results and the over-
The first tertiary project in Bell Creek field was started in 1976 optimistic determination of the ROS resulted in failure to scale up
with the development of a 160-acre [64.8-ha] micellar/polymer pi- from laboratory to field applications and prediction. 27 ,29 Success-
lot demonstration project. 25,27 Technically, the pilot was margi- ful prediction and characterization of heterogeneity are directly cor-
nally successful, recovering 27,000 bbl [4.29 x 10 3 m 3] of oil, or
related to the selection process of reservoir zonation. Reservoir
14 % of the oil in place (DIP) after waterflooding. Through DOE's
zonation can be improved following a complete waterflood tracer
provisions of the Tertiary Oil Incentive Program, a second, 179-acre
[72-ha] micellar/polymer pilot was initiated in 1981. 25 Unlike the study. A full understanding of the tracer survey results was not pos-
first pilot, the second pilot was successfully located in the optimum sible because of the interaction of the preflush, micellar, and poly-
portion of the reservoir. mer chemistry in the reservoir. 24 Inconsistent results were derived
The reservoir quality in the second pilot area was better than origi- from the tracer survey and pressure-pulse test. 29
nally expected. The postwaterflood DIP was 2.773 x 10 6 bbl In a heterogeneous reservoir such as the Bell Creek barrier bar,
[4.441 x 10 5 m 3] and the tertiary recovery was 39.6% of the the selection of an optimum pilot site that represents the perform-
DIP. 25 ance behavior of the entire reservoir remains a challenge.
Reservoir heterogeneity in Bell Creek field must be sufficiently north pilot. the test area for Shell Oil Co.·s "high-water-content
quantified before successful waterflood history matching or EOR process," was within crevasse splay and distributary channel sand
performance prediction can be expected. This requires not only a facies. 37
close cooperative effort by geologists, petrophysicists, and engi- Perhaps the greatest single problem resulted from a major frac-
neers, but also a reservoir simulator designed to include the com- ture artificially induced in a south pilot area injector well, which
plexity of the observed heterogeneities as well as the prediction of introduced complications that prevented an adequate waterflood his-
potential geochemical reactions and their products. Unfortunately, tory match, and the reservoir could not be suitably simulated. 37
the literature does not provide comprehensive information on frac- A reservoir pressure gradient in a northeasterly direction in the pi-
ture studies. diagenesis. or phase behavior relationships-the fun- lot area caused fluid movement at the linear rate of 29.6 ft/D [9.0
damental elements for heterogeneity analysis in Bell Creek. Results m/d).37,40 This prompted the invasion of the preflush fluids by
of tracer surveys and pressure-pulse data should be considered be- resident brine, resulting in the adsorption and neutralization of the
fore new wells are drilled and expensive chemical fluids are injected. surfactants d~spite an initial observation of a mobilized oil bank
that represented 43 % of the residual OIP after waterflood. 37
EI Dorado Field. EI Dorado field, Butler County, KS, produces The gypsum in the reservoir was responsible for depleting the
oil and gas from several carbonate and clastic reservoirs ranging alkaline silicate preflush before it moved through the reservoir. 37
from the Ordovician Stapleton limestone at a depth of 2.700 ft The rapid increase in salinity at the onset of tertiary oil production
[823 m) to the Lower Permian Admire sand at a depth of 600 ft suggests that the mobile oil moves through parts of the reservoir
[183 m).30-32 Structural entrapment is within an elongated asym- not swept by the preflush. 37
metrical anticline 40 sq miles [103.60 km 2) in area. directly over- The rapid changes from reservoir (distributary channel and
lying the basement granite Nemaha uplift. 31-33 Reservoir rock and crevasse splay) to nonreservoir (interdistributary channel) facies
shales thin radially toward the apex of the anticline. 34.35 EI Dora- that occur between the 6.4-acre [2.6-ha) spacing in the pilot areas
do field was developed rapidly to furnish oil for the Allies in World clearly indicate that the Admire sand is a very heterogeneous reser-
War I. Oil production peaked in 1918 31 and water injection was voir. 36 The 40-acre [16-ha) well spacing in EI Dorado field does
terminated in Feb. 1971. when essentially all wells in the 650-ft not allow accurate reconstruction of facies and nonfacies bounda-
[198-mJ Admire sand were plugged before tertiary production was ries unless improvements in reservoir characterization or additional
initiated. 36 infill drilling are made. The success of an EOR project in EI Dora-
The reservoir characteristics of the Admire sand arc summar- do field will ultimately require significant improvements in facies
ized in Table 4. A cost-shared tertiary micellar/polymer demon- analysis through an understanding of depositional-environment and
stration project by the DOE and Cities Service Co. was initiated diagenesis models.
in 1974 for the Admire sand. 37 -43 The project included two pilot
areas, and the objective was to compare two separately designed Summary of Geologic Problems
tertiary oil recovery methods. 36-38 The pilot areas included 51.2 The geologic problems encountered in the four EOR pilots may
acres [20-haJ with a total of 61 wells based on 6.4-acre [2.6-ha) be grouped into four categories: depositional, diagenetic, structur-
spacing and a five-spot pattern. 37 The project was terminated in al, and interstitial fluid. Heterogeneities related to each category
1982 as a marginal technical success and an economic failure. 44 had a variable effect on EOR performance in the four reservoirs.
The low cumulative tertiary oil recovery of only 26.734 bbl The results of Natl. Ins!. for Petroleum and Energy Research
[4.2 x 10 3 m 3) was a result of chemical and geologic problems. 45 studies of these EOR pilots are summarized in Table 5. The data
The Admire sand is a sublitharenite ( < 75 % quartz, < 25 % rock base for this paper was taken from the literature; thus all evalua-
fragments. and < 10% feldspar) deposited in a complex heteroge- tions and conclusions on the case studies in Table 5 are subjective.
neous distributary channel system within a prograding delta sys- The effect of each category of reservoir heterogeneity has been
tem and proximal beach sands. 37 Nine reservoir and nonreservoir evaluated and ranked (high, moderate, or low) in terms of its in-
facies have been recognized within the Admire sand. 37 Distribu- fluence in EOR. The potential contribution from additional research
tary channel sands constitute most of the reservoir facies; however, in each category of reservoir heterogeneity was also evaluated and
the highest measured permeabilities are within the less-extensive ranked (maximum, moderate, or minimum) for each reservoir
crevasse splay sand facies. 37 Reservoir qualit is a function of the studied.
1
total sand thickness and its distribution. 3 Primary porosity This study indicates that the depositional environment defines the
predominates. and secondary porosity occurs as the result of clay original framework of reservoirs and is responsible for a category
removal, alteration of calcite cements, leaching of plagioclase, of heterogeneity that has affected all four reservoirs, and the poten-
deformation of ductile grains, and quartz overgrowths. 37.46 tial contribution from further research is estimated to be maximum
for Big Muddy, North Burbank, and EI Dorado, and moderate for
EI Dorado Field Geologic Problems. Efforts to simulate this reser- Bell Creek. A more detailed understanding of log-to-core relation-
voir were hampered by several factors. A primary geologic prob- ships and interpretations is needed. Combining the depositional
lem was the failure to spot-locate the two pilot areas within a similar model with engineering test analyses and photogeologic interpre-
facies for comparing the two EOR processes. The south pilot, the tations from Landsat, Spot-I, SLAR, and other remote sensing data
test area for Union Oil Co.'s "soluble oil process," was in inter- would improve EOR simulation and tertiary prediction/perform-
distributary bay shales, crevasse splays. and beach sands. 37 The ance results with less risk. The postdepositional alterations (i.e.,
SPE Formation Evaluation. December 1987 453
PROCESSES DURATION
diagenesis and tectonics) are not adequately defined in all four cases. voir characteristics in the interwell region. Studies of the four
The diagenetic effect in three of the four EOR case studies was categories of heterogeneity can proceed from the construction of
moderate except at North Burbank, where it significantly affected one partial model for each category. This is discussed more fully
fluid displacement in the chemical EOR pilot. in the next section.
The structural component of the geologic heterogeneities signif-
icantly influenced the fluid flow behavior in Big Muddy and Nqrth
Burbank reservoirs and was moderately influential in Bell Creek Proposed Requirements for Optimum
and El Dorado reservoir pilot areas. Considering that El Dorado Characterization of Clastic Reservoirs
field directly overlies the basement Nemaha uplift, one would ex- General Requirements for Characterization and Modeling. The
pect this anticlinal structure to be fractured. With the exception of general requirements for characterizing clastic reservoirs for EOR
Big Muddy, where definite conclusions could not be made because are described, with emphasis on certain aspects of reservoir charac-
of paucity of data, the effects of formation fluids on EOR perform- terization that are often neglected and underestimated. The proposed
ance were moderate in Bell Creek and El Dorado fields and severe requirements address heterogeneity problems encountered not only
in North Burbank fields. These facts illustrate the need for research in the four pilot studies, but also in other similar projects where
and development of methods to improve quantification of reser- EOR has been implemented.
Sedimentary Postsedimentary
Lithostratification Compaction pattern resulting from burial and
Geometry, continuity of strata, and boundaries pressure regime'
Bedding and internal sedimentary structures Deformation of layers resulting from dewatering
Texture and related original porosity and clay collapsing
Detrital mineralogy Distortion of continuity and geometry of shale
Chemistry of synsedimentary fluids drapes
Redox potential of depositional environment Geochemical alteration of original detrital grains
Early biogenic and geochemical processes and Fracture pattern related to the thickness of rock
products layers
Contemporaneous tectonics Fracture pattern related to the type of lithologies
Detrital filling of fractures related to alternating
lithologies
.. Hydrostatic and under- or overpressured regimes in open or closed systems.
Geochemical Tectonic
Lithification and resulting brittleness Subsidence, uplifting
Dissolution/precipitation/mineral transformation Folding, faulting, overthrusting, tilting, flexuring
Eogenetic stage Fracture generations resulting from the history
Mesogenetic stage of shear stresses
Telogenetic stage, resulting from chemical Fracture pattern resulting from the structuring
equilibration of rock minerals with of the basement
migrating fluids Fracture pattern related to the position on folds
Geochemical filling and reopening of fractures Fracture pattern related to the brittleness of
Late biogenic alterations and products layers
Man-induced geochemical alterations since Man-induced reopening of natural joints
primary production
Final properties and distribution of formation
fluids
zation or overgrowths, leaching or precipitation of cement, relations. Major rock alterations that are depth-related are shown
transformation of de!rital clay minerals, and precipitation of authi- in Fig. 5. 58
genic clays and their further transformations. Changes in porosity The structural varieties (polymorphs) of final clays in shaly layers
and permeability are clearly the result of reconstruction of grains, and clay/sand mixtures may react quite differently under induced
pores, and infillings. conditions. For example, swelling properties are characteristic for
Diagenetic modeling incorporates both the above geochemical halloysite, which is structurally the same as nonswelling kaolinite
alterations induced into the formation during its postdepositional but contains a layer of water in its crystalline structure. Glauconite
history and compaction resulting from physical reduction of thick- (a variet?, of nonswelling illite) may contain up to 55 % of swelling
ness and porosity of the original sediment. The proposed elements layers,6 and may be quite common in certain marine clastics.
of the diagenetic model are shown in Table 8. Some sandstone ~ers may contain up to 40% of glauconite ag-
Effects of Paleohistory. The time-dependent depth history of a gregates (pellets) dispersed within the rock matrix and particu-
British Carboniferous formation in two locations situated only 18.6 larly concentrated in trough-bedding laminae (Fig. 6). These 0.04-
miles [30 km] apart reveals drastically different evolutionary stages, to 0.08-in. [1- to 2-mm]-diameter glauconitic pellets consist ofmyri-
as described by Dickson 59 (Fig. 3). ad tiny (2-JLm) plates. If these pellets break up during EOR, pores
An example of severe diagenetic alterations connected with paleo- will certainly be plugged. Behavior of glauconite under induced
historic evolution comes from the Piceance basin in Colorado at thermal or chemical conditions is not known.
the site of the Multiwell Experiment (MWX) being performed by Vermiculite, another type of clay, has a very high ion exchange
the DOE (Fig. 4). There, the Upper Cretaceous Mesaverde for- capacity and when heated to about 572°F [300°C] dehydrates quick-
mation was buried to a depth of about 13,000 ft [4,000 m], thus Iy, flakes, and expands 18 to 25 times its original volume. 63 ,65 The
probably being subjected to a zone of low-grade regional ther- process terminates at about 932°F [500°C]. Owing to its unusual
mometamorphism before being uplifted to its current depth of 7,000 property, vermiculite may cause plugging during an in-situ com-
ft [2,134 m].60 bustion process, because the temperatures of the burning front and
Interpretation of the diagenetic sequence in matrix and fractures the combustion zone range from 600 to 1,200°F [316 to 648°C].
for all cores in the MWX was performed and included the effects Expansiveness resulting from dehydration is also characteristic for
of compaction and geochemical alterations. In certain intervals, the dickite and muscovite. 63
porosity is nearly all secondary and was formed by dissolution of Thermal effects of dehydration and pyrolysis on different clay
feldspars, lithics. and calcite cement. However, nearly all the in- minerals have been studied extensively by the ceramics and build-
tergranular pore space is filled by authigenic clays. Despite the dis- ing industries, but little attention has been given to these altera-
solution of detrital grains and cement, open space within the matrix tions in reservoirs subjected to thermal EOR.
and fractures is rare, with permeabilities in the microdarcy Clearly, there are many reasons to study the origin and history
range. 61 •62 of clays and their proportions and structural polymorphic varieties
Information obtained from the study of analogous deposystems, both in clay/sand mixtures and in shale layers. drapes, or clasts.
outcrops, and aquifers is useful if carefully selected. However, reser- Simple identification of major groups of clay minerals may be in-
voirs of similar age and origin that were subjected to different di- adequate for reservoir characterization.
agenetic environments during their evolutionary history may possess Man-Induced Geochemical Alterations. Man-induced geochemical
quite different properties. Comparison of properties of analogous alterations affect current and future projects in tne same reservoir.
deposystems belonging to different age groups and/or different geo- The magnitude and rate of these alterations, particularly those re-
logic settings may be even more misleading. sulting from dissolution/precipitation processes, are important be-
Effects of Clays. The mineralogy of detrital or authigenic clays cause they are responsible for significant changes in reservoir
resulting in plugging of pores, pore throats, and joints is usually properties throughout the course of field development. The pre-
ill-defined in many EOR projects. The transformation of clay diction of these changes should be included in the diagenetic (geo-
through geologic time is related to changes in temperature. pres- chemical) model and be an integral part of reservoir characterization
sure, interstitial water ionic composition. ion activity, and pH/Eh (Table 8).
456 SPE Formation Evaluation, December 1987
MILLIONS OF YEARS BEFORE PRESENT
'0
A. SEDIMENTATION
OepthofRolllq
Fig. 4-Geologic history of the Piceance Creek basin at the MWX site. After Lorenz. 60
Depressurizing and degassing of producing reservoirs disturb ing steamdrive and in-situ combustion. As the silica solution spreads
long-term established equilibrium or semiequilibrium between fluids from injection wells, a decrease in temperature may cause precipi-
and rocks. Reservoir porosity and permeability obtained during early tation at a certain distance and plug fluid pathways (Fig. 7).
field development may not always be applicable to EOR. The kinetics of chemical processes under natural and induced con-
Waterflooding introduces new elements to the environment, and ditions in reservoirs has recently been studied intensively. 44.66-71
the chemical composition of injected water is controlled to avoid One experimental study 71 revealed dramatic effects of silica dis-
problems of precipitation and scaling. However, this may trigger solution/precipitation during steam injection at different flow rates,
slight changes in redox potential that may cause oxidation of form- temperatures, and pH levels. Another experiment with Heart of
erly reduced species, such as iron or manganese, and produce par- Texas sand documented extensive dissolution of quartz grains even
ticles in suspension. These suspended solids, mobile silt, and clay after 24 hours of steam treatment. 71
particles can block pore throats and prevent fluid flow. Diagenetic modeling involves a variety of natural and man-made
Extensive alterations may be induced during tertiary stimulation alterations that influence hydraulic properties of a reservoir under
when reservoirs are subjected to changes in the temperature and/or consideration for EOR. However, few of these problems were satis-
chemical regime. The rate of precipitation/dissolution of any sub- factorily studied before reservoir simulation was performed.
stance depends on concentrations of reacting ions, total salinity, Structural Model. The triplex porosity (primary, secondary, and
temperature, partial pressure of gases, lithology, and fluid flow ve- fractures) and triplex directional permeability (resulting from the
locity.65.66 The dissolution rate is fastest for strongly undersatu-
combined effect of matrix, fracture, and bedding-plane permeabil-
rated solutions. 66
ities) are commonly observed in reservoirs. 72 If the distribution
The injection of CO 2 to the system changes the trend and the
magnitude of dissolution, particularly with respect to calcium, mag- of directional fracture permeabilities is such as that presented by
nesium, and iron carbonates precipitated as cements. Excess C02 Bergosh (Table 9) for the Monterey formation, however, the con-
forms a highly aggressive system in reservoirs in which carbonates tribution to the overall permeability from the other sources would
and significant amounts of ubiquitous iron can be mobilized. Con- probably be negligible. In such a situation, more attention should
versely, the liberation of CO 2 from a calcite-saturated solution may be paid to structural modeling rather than to other partial models.
induce carbonate precipitation. Although the physical and mathematical approach to modeling
The solubility of silica is low at 77°F [25°C] but increases rapidly and simulation of naturally fractured reservoirs has been recently
at higher temperatures,45 and siliceous minerals may corrode dur- presented,73-75 several geologic aspects that may be important
Mlcrobiol
CARBONATE Initlol pore wot.r,
molt'll,
FezO, "HzO
AI20, nH 20 CH.
SATURATED I Carbonot. ~ment;
".
m,chanlcal Elimination CalCite,
r'ductlon of Little 510 2 ferm.ntatlOn dolomite, Redl,trlbuted
primary Chanol un,tabl. 'ld,nte corbonat.
pOro'lty compaction M
M ,ilicat" In
E
S
by comPoct Ion
SEMI-MAT\.flE
Smect.te
+1/S
initial Dor.
wot,r, CO,. HzO -20 % poro.lty
E
S -{}
65
n
0
G
mot"'' chemtCQl
porO'lty l/S+ K-f,ld,por. th.rmOI
::i~~o~ottr CARBONATE
0 Kaolinite
pr'Clpltates as 2
r.duction 80 0/ o .llit. colcit., ellfftlnotfon
~H~'~'-fr
\;
E koolinit.,
80 M;;I~';-- N MATURE 'A'
v,rmlcullte ';I FIRST CLAY AGGRESSIVE
f"... - 9roln.d
.....
DEHYDRATION
90 ';~~'-- E :~ondory
ro,.t ChlOrite cloy. OIL 2 .t.PI? Co, + H,o E
oX
U 95 T leocI\ln9
~
MATuRE 'B' Fin ••' Addition, Oil FLUSHED OUT? 3 Ci
cloy 'rom furth.r F.rroon
m S,coftdary
porolity
Improved
clo,
fractiQnl
GAS clo)' carbonates,
late illite,.
"'
0
destruction
r.ductlon cr),.tolllnlt)' Silica, carbonotes ? Quartz
over9rowth,
'.t
4
120 6
160
)
SECOND CLAy STRUCTURALLY ChlOrite
DEHYDRATION HELD wATER r,plOC8$
Ii SUPER MATURE
ani, '1191'1t Ilhf' .... pnen91f, Oolomit., GAS? lIoolln,t,
poro •• ty Iron ? Flnol reduction of
220 -----yNciPtENT-METAMORPHISM
carbonat"
-Closed system
7
Fig. 5-Summary of depth-related diagenetic stages and Important reactions in mud-rocks that may contribute solutes to sand-
stones by the associated trends in pore-water evolution and expUlsion. After Ref. 58.
-f
~~OISS~OLU_TIOO~t~=-~wm~1l
PROGRESSING FRONT OF STEAM
9. The distribution of heterogeneities, their characteristics, and 13. Painter, T.R .. Borah, M.T .• and Ferrell, H.H.: "Big Muddy Field Low
interrelationships in pilot studies should be supplemented by geo- Tension Flood Demonstration Project." Fifth Annual Report. April
logic data from analogous reservoirs, outcrops, and aquifers that 1982-March 1983, Contract No. DOE/SF/01424-47. U.S. DOE (Aug.
1983)
have similar genetic and evolutionary history.
14. Trantham. J.C .• Threlkeld. C.B., and Patterson, H.L.: "Reservoir
10. The distribution, proportion, and properties of different types Description for a Surfactant/Polymer Pilot in a Fractured. Oil-Wet
and polymorphs of authigenic clays need to be quantified. Reservoir-North Burbank Unit Tract 97." JPT(Sept. 1980) 1647-55.
11. In addition to the geometry and the morphology of fractures, 15. Trantham, J. C.: "Prospects of Commercialization, Surfactant/Polymer
structural modeling should also incorporate information on the Flooding, North Burbank Unit. Osage County. OK." JPT(May 1983)
degree of their openness and the type of detrital or geochemical 872-80.
infilling. 16. "An Evaluation of the North Burbank Unit Tertiary Recovery Pilot
12. A dynamic hydraulic model derived from the integration of Test." Keplinger and Assocs. Inc., Contract No. DE-ACI9-80/BC/
four partial geologic models will provide a more accurate predic- 10033-2 (Aug. 1982).
17. "North Burbank Unit Tertiary Recovery Pilot Test, Final Report," Phil-
tion of fluid movement in the reservoir. The dynamic hydraulic
lips Petroleum Co .. Contract No. DOE/ET/13067-6O (June 1980).
model will be unique because of the triplex nature of the effective 18. Trantham. J.C. and Moffitt, P.D.: "North Burbank Unit 1.440-acre
permeability distribution in each reservoir. Polymer Flood Project Design." paper SPE 10717 presented at the 1982
SPEIDOE Enhanced Oil Recovery Symposium, Tulsa, April 4-7.
Acknowledgments 19. Boneau, D.F. and Clampitt. R.L.: "A Surfactant System for the Oil-
This work was sponsored by the U.S. DOE under Cooperative Wet Sandstone of the North Burbank Unit." JPT(May 1977) 501-06.
20. Riggs, C.H.: "Waterflooding in the Burbank Oil Field. Osage County.
Agreement DE-FC22-83FE60149. We thank K. Westhusing, T.
Oklahoma." publication of U.S. Dept. of the Interior. Washington. DC
Geffen, M. Tham, and M. Fowler for their critical review of the (May 1954) 19.
manuscript. We are also thankful to B. Linville and I. Palmer for 21. Lorenz. P.B.. Trantham. J.C., and Zornes. D.R.: "A Postflood Eval-
their valuable editorial comments and suggestions for improvement. uation of the North Burbank Surfactant/Polymer Pilot." paper SPE
12695 presented at the 1984 SPE/DOE Enhanced Oil Recovery Sym-
References posium. Tulsa. April 15-18.
I. Dickey, P.A.: Petroleum Development Geology. PennWeli Publishing 22. Hufford. W.R. and Tieh. T.T.: "Diagenesis of Burbank Sandstone.
Co .. Tulsa (1979) 350. North Burbank Field. Osage County, Oklahoma." paper presented at
2. Brown. D.: "Geologists Eye Developing Future." MPG Explorer. the 1984 AAPG Annual Convention. San Antonio. May 20-23.
(Jan. 1986) 20-21. 23. Burt. R.A., Haddenhorst, F.A .• and Hartford. J.C.: "Review of Bell
3. "Big Muddy-Can a Chemical Flood Breathe New Life into a Tired Creek Waterflood Performance-Powder River County. Montana."
Old Giant?" Drilling-DeW (June 1978) 36-37. paper SPE 5670 presented at the 1975 SPE Annual Technical Confer-
4. Ogren. DJ. et al.: "Big Muddy Field Low Tension Flood Demon- ence and Exhibition. Dallas (Sept. 28-0ct. I)
stration Project." Second Annual Report. April1979-March 1980, Con- 24. Vargo. J.l.: "Site Selection, Reservoir Definition and Estimation of
tract No. DOE/SF/01424-26. U.S. DOE (Aug. 1980) 70. Tertiary Target Oil for the Bell Creek Unit A Micellar-Polymer Proj-
5. Troiani. L.R. et al.: "Big Muddy Field Low Tension Flood Demon- ect," paper SPE 7072 presented at the 1978 SPE Symposium on Im-
stration Project." First Annual Report. April I 978-March 1979. Con-. proved Methods for Oil Recovery. Tulsa. April 16-19.
tract No. DOE/SF/01424-13. U.S. DOE (Aug. 1979) 81. 25. Hartshorne, J.M., and Nikonchik, 1.S.: "MiceliariPolymer Flood Suc-
6. Haller. R.W.: "Big Muddy Water Flood. "JPT (Nov. 1955) 9-12. cess in Bell Creek Field." paper SPE 13122 presented at the 1984 SPE
7. Gilliland. H.E. and Conley. F.R.: "Pilot Flood Mobilizes Residual Oil." Annual Technical Conference and Exhibition. Houston. Sept. 16-19.
Oil & Gas J. (Jan. 19. 1976) 43-48. 26. Berg. R.R. and Davies. D.K.: "Origin of Lower Cretaceous Muddy
8. Davis. J.G .• Ferrell. H.H .• and Stewart. W.C.: "Big Muddy Field Low Sandstone at Bell Creek Field. Montana," MPG Bulletin (1968)
Tension Flood Demonstration Project." Third Annual Report. April 1888-98.
1980-March 1981. Contract No. DOE/SF/01424-39, U.S. DOE (Nov. 27. "Bell Creek Field Micellar-Polymer Pilot Demonstration." final report.
1981) 85. Gary Energy Corp., June 1976-March 1982, Contract No.
9. Ferrell. H.H .• Gregory, M.D .• and Borah. M.T.: "Progress Report: DOE/SF/01802-61, U.S. DOE (Sept. 1982).
Big Muddy Field Low-Tension Flood Demonstration Project With Em- 28. McGregor. A.A. and Biggs, C.A.: "Bell Creek Field, Montana: A Rich
phasis on Injectivity and Mobility." paper SPE 12682 presented at the Stratigraphic Trap." MPG Bulletin (1968) 1869-87.
1984 SPE/DOE Enhanced Oil Recovery Symposium. April 15-18. 29. "An Evaluation of Bell Creek Field Micellar-Polymer Pilot," Keplinger
10. Gilliland. H.E. and Conley. F.R.: "A Pilot Test of Surfactant Flood- and Assocs. Inc .• Contract No. DOE/BC/l0033-5. U.S. DOE (Dec.
ing in the Big Muddy Field." paper SPE 5891 presented at the 1976 1982).
SPE Rocky Mountain Regional Meeting. Casper. WY. May II and 12. 30. Fath. A.E.: "Geology of the EI Dorado Oil and Gas Field. Butler
II. Marrs. R.W. and Raines. G.R.: "Tectonic Framework of Powder River County. Kansas." Bull.• Geo!. Survey of Kansas. 7, (1921).
Basin. Wyoming and Montana. Interpreted from Landsat Imagery." 31. Reeves. 1. R.: "EI Dorado Oil Fields. Butler County. Kansas." Struc-
MPG Bulletin (1984) 1718-31. ture of Typical American Oil Fields. MPG Bulletin II (1929) 160-67.
12. Towse. D.: "Frontier Formation, Southwest Powder River Basin. 32. Uhrlaub; R.: "Butler County. Kansas-Geological Notes." MPG Bulle-
Wyoming." MPG Bulletin (1952) 1962-2010. tin (1921) 421-24.