REVIEW OF LITERATUR.....
homeworkhelp
review_17.pdf
SPE 163292
Permeable Tar Mat Formation Within the Context of Novel Asphaltene Science Hadrien Dumont, Vinay Mishra, Julian Y. Zuo, Oliver C. Mullins (Schlumberger)
Copyright 2012, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Kuwait International Petroleum Conference and Exhibition held in Kuwait City, Kuwait, 10-12 December 2012. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of th e paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessar ily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohi bited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
ABSTRACT
Tar mats at the oil-water contact (OWC tar mats) in oilfield reservoirs can have enormous, pernicious effects on production
due to possibly preventing of any natural water drive and precluding any effectiveness of water injectors into aquifers. In
spite of this potentially huge impact, tar mat formation is only now being resolved and integrated within advanced asphaltene
science. Herein, we describe a very different type of tar mat which we refer to as a “rapid-destabilization tar mat”; it is the
asphaltenes that undergo rapid destabilization. To our knowledge, this is the first paper to describe such rapid-destabilization
tar mats at least in this context. Rapid-destabilization tar mats can be formed at the crest of the reservoir, generally not at the
OWC and can introduce their own set of problems in production. Most importantly, rapid-destabilization tar mats can be
porous and permeable, unlike the OWC tar mats. The rapid-destabilization tar mat can undergo plastic flow under standard
production conditions rather unlike the OWC tar mat. As its name implies, the rapid-destabilization tar mat can form in very
young reservoirs in which thermodynamic disequilibrium in the oil column prevails, while the OWC tar mats generally take
longer (geologic) time to form and are often associated with thermodynamically equilibrated oil columns. Here, we describe
extensive data sets on rapid-destabilization tar mats in two adjacent reservoirs. The surprising properties of these rapid-
destabilization tar mats are redundantly confirmed in many different ways. All components of the processes forming rapid-
destabilization tar mats are shown to be consistent with powerful new developments in asphaltene science, specifically with
the development of the first equation of state for asphaltene gradients, the Flory-Huggins-Zuo Equation, which has been
enabled by the resolution of asphaltene nanostructures in crude oil codified in the Yen-Mullins Model. Rapid-destabilization
tar mats represent one extreme while the OWC tar mats represent the polar opposite extreme. In the future, occurrences of tar
in reservoirs can be better understood within the context of these two end members tar mats. In addition, two reservoirs in the
same minibasin show the same behavior. This important observation allows fluid analysis in wells in one reservoir to indicate
likely issues in other reservoirs in the same basin.
INTRODUCTION
Tar mats are of critical importance in the oilfield; however, mechanisms of tar mat formation have not been well understood.
The confusion extends into terminology where terms bitumen and tar are sometimes meant to imply distinct yet ill-defined
provenances. In large part, this is due to the prior lack of detailed understanding of asphaltenes. Indeed, it is only recently that
the nanoscience of asphaltenes in crude oil has been largely resolved and codified in the “Yen-Mullins Model”.[1,2] In turn
this has led to the Industry’s first predictive equation for asphaltene gradients, the Flory-Huggins-Zuo Equation of State (FHZ
EoS).[3,4] With this new understanding, asphaltenes are now treated within standard methods of physical chemistry much as
gas-liquid equilibria of crude oils (GOR, bubble point, etc.) have long been treated with standard physical chemistry models
such as the cubic equation of state. The cubic EoS has been extended to try to treat the solid asphaltenes because there had
been no alternative. However, the cubic EoS was never designed to treat colloidal solids such as asphaltenes; consequently,
these cubic EoS methods to treat asphaltenes fail.[1-4] Just as gradients (e.g. GOR gradients) and phase behavior for gas
liquid equilibria are treated by the cubic EoS, we now have an equation, the FHZ EoS that treats both asphaltene gradients in
reservoir fluids as well as phase behavior of asphaltenes. Thus, asphaltenes whether in carbonaceous deposits or dissolved (or
colloidally suspended) in crude oil are now treated within one scientific framework. Consequently, tar mats are now
understood within the framework of reservoir fluids and corresponding geologic processes.
2 SPE
By definition, asphaltenes are destabilized by light alkane addition to a crude oil. (Asphaltenes are defined as being the
component of crude oil, or carbonaceous material that is insoluble in n-heptane and soluble in toluene.) As has long been
known, the lightest alkane, methane, causes just such asphaltene precipitation. Moreover, light ends are often added to the
reservoir late in the charge process. In a normal burial sequence, the lightest charge is expelled at the latest times from the
kerogen with longer catagenesis times, and higher catagenesis temperatures. This is similar to refining where higher
temperatures and longer times yields cracking to lighter hydrocarbons. Second, in colder reservoirs biogenic methane in
substantial quantities can enter the reservoir, again, after the oil charge, the food for the microbes. The light alkane or gas
enters the reservoir in high permeability streaks and goes right to the top of the fluid column without fluid mixing except in
the vicinity of the charge path or plane,[5] a schematic of this process is shown in Fig. 1.[6] In this process, the gas migrates
to the top of the oil column, then diffuses down from the top of the oil column, asphaltenes are destabilized at the gas-oil
contact and increasingly at lower points with time as the solution gas gradually increases with methane diffusion.
Figure. 1. Charge history mechanism determines the hydrocarbon distribution in the Stainforth model.[5] Initially, the heaviest charge from the kerogen charges the reservoir. Later in the charging process, lighter hydrocarbons are expelled from kerogen. These light hydrocarbons migrate to the top of the oil column through high permeability streaks and do not mix with in situ reservoir fluids in this process. Given sufficient time, equilibration can take place. Light ends can destabilize asphaltenes; this destabilization process is seen to occur at the top of the oil column, not at its base where tar mats are frequently found.
The former lack of understanding of asphaltene nanoscience created considerable confusion as to how and where asphaltene
destabilization events take place. Specifically, it has long been assumed that asphaltene instability occurred where asphaltene
deposits are now found. Along these lines, in the laboratory, light alkane addition to crude oil destabilizes the asphaltenes
then and there. Thus, regional tar mats have been thought to form at the OWC due to gas entry at the oil-water contact
(OWC). This concept violates well known processes in reservoirs (cf. Fig. 1) and recent work has clarified actual reservoir
processes leading to tar mat formation. The reservoir is not like a distillation tower with bubble plates ensuring equal gas
entry everywhere at the OWC. This is in contrast to the laboratory where fluid mixing is simple and is part of asphaltene
flocculation procedures. The crucial point is that asphaltenes can migrate great distances in reservoirs even when partially
destabilized. Previously it had been thought that with asphaltene instability, there can only be floc formation, and all agree
that flocs cannot migrate through reservoirs. We now understand that there are two nanocolloidal asphaltene species, and that
instability can result in formation of the larger nanocolloidal asphaltene species; this nanocolloidal particle can migrate
through reservoirs.[7] This will be discussed in greater detail in the Asphaltene Nanoscience section.
163292
SPE 163292 3
In addition, tar mats can be as thick as 60 meters. Those who propose that the tar mat forms at the OWC because asphaltene
instability is mediated by a process involving water never explain how it is that the tens of meters of tar do not seal off the oil
from the water. (Many high-rise buildings have a thin layer, ≤ 1cm, of roofing tar sealing off water entry into the building,
tens of meters of tar are not needed for this purpose!) Once this sealing process takes place, the tar mat cannot grow further
IF tar mat growth depends on water. For example, if 3 meters of OWC tar mat creates this seal, then there should be no
further growth of the tar mat. In fact, this conventional explanation is incorrect; generally, tar mat growth is not mediated by
water. Frequently, tar mat growth occurs via asphaltene instability at the top of the oil column with asphaltene transport
through the reservoir to the flank. The thickness of tar mats in this process is rather unconstrained and is consistent with
thicknesses exceeding 30 meters that are frequently observed. The key improvement in understanding is the mechanism of
transport of unstable asphaltene through the reservoir. This is where new asphaltene science, the Yen-Mullins model, has had
a significant impact. Instability of asphaltene nanoaggregates can yield asphaltene clusters which are only 5 nm in size so
easily migrate through porous media, yet are large enough to accumulate at the base of the oil column.[1-4]
Case Studies. Gas addition to reservoirs mostly occurs through spatially restricted high permeability channels without
mixing with existing fluids in the reservoir. The added gas quickly finds its way to the top of the reservoir or at least to a
local high point in the reservoir. With new gas addition at the top of the reservoir, the gas diffuses down and expels
asphaltene from the oil in this process. Figure 2 shows a reservoir caught in the middle of such a process. One can visually
see lack of asphaltenes at the top of the reservoir, that is where asphaltene instability occurs.[8]
Figure. 2. A black oil reservoir had a substantial late gas charge which gave rise to a huge color gradient (see actual dead oil bottles on the right), and a huge gradient in solution gas.[8] The late gas charge quickly went to the top of the black oil column without mixing into this reservoir crude oil. Subsequently, the gas diffused down as indicated by white arrows in the cartoon on the right. Substantial solution gas near the gas-oil contact caused the asphaltenes to be expelled – thus the oil has very little color at the top of the column. The gas has not had time to diffuse to the base of the column, thus, there, the solution gas is low, and the oil maintains a high concentration of asphaltenes. In the figure, the calculated color (asphaltene) gradient is shown from the FHZ EoS coupled with a methane diffusion term and agrees with the color gradient evident in the samples.[8] Figure 2 shows a reservoir where substantial gas has diffused only partially into the oil column; thus expelling asphaltene
only towards the top of the column. It would take geologic time for this diffusion process to transport methane to the base of
the column. If the late gas charge has sufficient time to diffuse to the base of the oil column then the asphaltenes can become
very concentrated at the base and form a tar mat. Figure 3 shows just such a reservoir. In this case, large volumes of gas
entered the oil column expelling most of the asphaltenes. The asphaltenes migrated ahead of the gas front, most likely by
convective waves. The asphaltene rich fluid went to the base of the column, and was trapped there by cement underneath.
The gas caught up to the asphaltenes creating a condensate (high GOR, little asphaltene) and a tar mat underneath. This tar
mat is on cement so has nothing to do with water. The conventional explanation that the tar mat forms at the OWC because
that is the location of the asphaltene instability obviously does not apply here and is generally incorrect. For oil reservoirs at
or very close to surface, biodegradation and even evaporative and water washing processes can yield very viscous oil, such as
the Athabasca tar sands, but that is distinct from a reservoir with a tar mat.
4 SPE
Figure 3. A reservoir with a late gas charge. The gas went to the top of the reservoir and diffused down to the base of the oil column (depicted by long white arrows in the left cartoon).[9] The expelled asphaltenes stayed ahead of the diffusive gas front, presumably by convective currents. The tar mat is visible in core sections, the fluorescence image of the core (right image in each of the six core panels) shows strong fluorescence of the condensate immediately above the tar mat with no fluorescence (bottom two core panels with visible image, left; fluorescence, right). This tar mat rests on cement, water had nothing to do with this tar mat formation.[9]
Equilibration of reservoir fluids is a geologically slow process.[10] Recently, a reservoir was reported to have two separate
gas caps that could each be seen in seismic data. Figure 4 shows an image of this reservoir.[11] Logging data from the two
wells established that the two separate gas-oil contacts (GOCs) differ by 20 meters in the true vertical depth (TVD).[11]
Either the reservoir is compartmentalized, each compartment with its own gas cap, or there is lateral disequilibrium in
solution gas in a single reservoir with two gas caps.
Figure 4. (Left) A reservoir with two separate gas caps (brown) over oil (green); the two gas-oil contacts differed by 20 meters true vertical depth.[11] (Right) The continuous asphaltene gradient measured in three wells indicated connectivity which was proven in production. There is lateral disequilibrium of solution gas; the diffusive processes to cause equilibration in the reservoir are very slow even compared to geologic time.[10]
The downhole fluid analysis (DFA) data clearly established that there is a continuous color gradient across the reservoir (in
three wells including the two depicted in the Fig. 4).[11] This color gradient matched the FHZ EoS analysis and indicates that
the reservoir is connected. The heavy ends are equilibrated across the field – this requires connectivity. Production proved the
reservoir is connected.[11] Thus, the reservoir fluid is out of equilibrium for gas-liquid properties such as GOR. When gas
charges into the reservoir it can get stuck in local reservoir highs. To equilibrate the two gas caps, gas from the lower GOC
gas cap would have to dissolve in the oil, diffuse across the reservoir and release into the higher GOC gas cap. This process is
extremely slow, consequently the reservoir gas-liquid properties are out of equilibrium. In particular, if there is a late gas
phase charge into the reservoir, it is likely that gas would collect in local highs and would remain out of equilibrium for a
long time. In contrast, the asphaltenes only partition to the liquid phase so can equilibrate provided that the reservoir has good
connectivity. This is indeed what happened in the oilfield shown in Fig. 4.[11]
163292
SPE 163292 5
Another reservoir experienced only a small light end influx into a black oil reservoir. With just a slight destabilization event,
but with time for equilibration to take place, black oil can remain in the crest, whereas heavy oil and a tar mat can reside in
the low points or the flank or rim of the field. Figure 5 shows a cartoon of a very large Jurassic anticline field with a four way
dip closure. There was a slight asphaltene instability leading to asphaltene migration and accumulation of the asphaltenes in
the rim of the field. Most likely, this migration took place in convective waves of asphaltene rich oil (thus high density).
Diffusion would require on the order of one trillion years (much longer than the age of the universe), so diffusion alone
cannot account for asphaltene migration in this reservoir!
Figure 5. A large, Jurassic anticline oilfield in Saudi Arabia has black oil in most of the field, has a mobile heavy oil rim which is underlain by a tar mat.[12,13] The mobile heavy oil rim exhibits a gigantic asphaltene gradient (6x) as shown. The asphaltene gradient fits the gravity term of the FHZ EoS, with one tightly constrained adjustable parameter, the asphaltene cluster size, thus the asphaltenes are equilibrated throughout this huge volume. The fitted data (above) gives an asphaltene cluster size of 5.2 nm, very close to the published nominal size of 5.0 nm (cf. Fig. 6). This Jurassic field of Fig. 5 had experienced some asphaltene instability, but not too much as substantial asphaltene remains
in the crude oil, and the GOR is low – both conditions in contrast to conditions depicted in Fig. 3. The destabilized asphaltene
accumulated in the flank. The asphaltenes in the mobile heavy oil section equilibrated laterally over the entire tens of
kilometers circumference of the field and in height of 50 meters as shown in Fig. 5. Equilibration ultimately does have a
diffusive component; the simple diffusion relation is Dt = x 2 where D is the diffusion constant (which is very small for
asphaltene clusters), t is time, and x is mean distance of displacement. This large field has a large value of x >>10 kilometers;
consequently, a very long time is needed to reach equilibration. As noted above, convection must also play a large role in
equilibrating this field. The field is Jurassic, and being equilibrated, this field identifies what a “long time” is for such
reservoir processes, it is ~150 million years.[13]
Where the asphaltene content exceeded 35%, this is tar mat. For asphaltene concentration above 4% and below 35%, this is
mobile heavy oil (viscosity < 1000 centipoise). We also note that unlike the mobile heavy oil, the asphaltene content in the
tar mat in this field is not even slightly equilibrated. The asphaltene content in the tar mat ranges from ~35% to ~60%, with
large variations up and down in concentration within a few meters within individual wells.[12,13] We propose that the tar
mat represents a phase transition; asphaltene is not soluble in crude oil in all proportions.[13] As the asphaltene continues to
enter low points in the reservoir by accumulation of 5 nm asphaltene clusters, the crude oil can become supersaturated in
asphaltene content. The asphaltenes then plate out on grain surfaces. As this process continues, the pore throats become
occluded, and no further fluid exchange can take place. That is, tar mats are not equilibrated in two meters (vertical) whereas
the heavy oil is equilibrated over many tens of kilometers (lateral) because the carbonaceous grain coating in the tar mat
precludes any mass exchange necessary for equilibration.[12,13] Precepts in this explanation are under study.
What had not been appreciated is that destabilized asphaltenes can migrate in reservoirs, but not when precipitated as flocs.
Asphaltenes in crude oils can exist as three distinct species as shown in Fig. 6.[1] All of these species are in the nanometer
size range so are tiny with respect to pore throats in rock in conventional reservoirs. When asphaltenes are slightly
destabilized for example by slow gas addition to a crude oil, asphaltene nanoaggregates form clusters (containing ~8
nanoaggregates). Clusters are relatively large compared to the other asphaltene species and consequently, they accumulate
6 SPE
towards the base of oil column by gravity, much more so than asphaltene nanoaggregates. By this means, asphaltenes that are
destabilized can migrate in reservoirs. The migration process most likely involves convective waves of asphaltene rich (and
cluster rich) fluids.
Asphaltene Nanoscience and Equation of State
Figure 6. The Yen-Mullins Model. Typical structures for asphaltene molecules, nanoaggregates (of molecules) and clusters (of nanoaggregates). At low concentrations as in condensates, asphaltenes are dispersed as a true molecular solution (left); for black oils, asphaltenes are dispersed as nanoaggregates of ~6 molecules (center); for heavy oils, asphaltenes are dispersed as clusters of ~8 nanoaggregates (right).[1,2]
With the size known, the effect of gravity is determined. For the asphaltene equation of state, the gravity term is given by
Archimedes buoyancy in the Boltzmann distribution. That is, the asphaltene particles are negatively buoyant in the crude oil
as described by Archimedes buoyancy. Combining the gravity term, with a chemical solubility term and an entropy term we
have the equation of state for asphaltene gradients, the Flory-Huggins-Zuo (FHZ) Equation of state. The Flory-Huggins
theory has long been used to describe polymer solubility, here we use this theory but we also include a gravity term to treat
asphaltene gradients.
12
21
11 expexpexp
1222
1
2
1
2
hh
a aa
haha a
a
a
vv v
RT
hhgv
RT
v
h
h
hOD
hOD
1.
where OD(hi) is the optical density (color) measured by DFA [6] of the fluids at height hi in the oil column, a(hi) is the
asphaltene fraction at that height, a is the molar volume of the relevant asphaltene species (cf. Figure 6), is the molar
volume of the oil, R is the ideal gas constant, T is temperature, a is the solubility parameter of the asphaltene, is the
solubility parameter of the oil, g is earth’s gravitational acceleration, a is the asphaltene density (~1.2g/cc), is the oil
density. The solubility parameter of the asphaltene can be obtained from literature values, and, in an oil column, the solubility
parameter variation of the oil is primarily due to GOR variations. In the FHZ EoS, the first exponential factor is the solubility
term, the second is the gravity term and the third is the Flory-Huggins entropy term. For low GOR oils, the gravity term
dominates. For moderate GOR oils (700 scf/bbl), typically both the solubility term and the gravity term contribute to the
asphaltene gradient. With this foundation, the understanding of many reservoirs is dramatically improved.
In this study, we examine two Pliocene reservoirs each with multiple horizons with considerably smaller dimensions than the
reservoir of Figure 5. In the case herein, there is massive, recent gas addition to a black oil with dramatic but unusual effects
manifested in many ways. In particular, this reservoir is evidently grossly out of equilibrium, but not in a systematic way as
depicted in Fig. 2, but rather in a seemingly stochastic (random) disequilibrium variation of fluids properties. With the
reservoir rock being <5 million years, the black oil charge being younger than that and the gas charge more recent still, we
call this time frame roughly 1 million years; this defines what is very rapid regarding reservoir fluids. This case study and the
Saudi Aramco study (Fig. 5) bracket short and long times for reservoir fluids process, 1 million years to 150 million years.
RESULTS AND DISCUSSION
Two fields that share the same minibasin are probed herein; both fields exhibit similar characteristics of importance to this
paper. We will focus primarily on one reservoir with many details presented. These fields and the contained oil and gas are
quite young. Any fluid process that has occurred could take no more than a few million years. Frequently, reservoir processes
take longer than that, and often such young reservoirs have processes that are still ongoing, such as gas charging;
consequently fluid equilibration is not expected. Two horizons in each field are of interest. These reservoirs have substantial
structure containing many lobes. The wireline pressure survey along with some PVT data is shown in Fig. 7 for one reservoir.
163292
SPE 163292 7
x x
x
x
x
x
x x
Well 1
Well 2
Well 3
X,100
X,200
X,300
DEPTH (TVD, feet)
Figure 7. Pressure survey and PVT data for one of the fields. The pressures in the one sand in three wells are essentially on the same trend. The peculiar observation is that the GOR of one intermediate sampling point is significantly different and smaller than all other samples. It is possible the fluids are grossly out of thermodynamic equilibrium (but in pressure equilibrium) in a connected baffled reservoir.
Well Logging. The pressure data for the three wells are on the same trend. Prior to production, aligned pressure
measurements are not a strong indicator of connectivity; nevertheless, these pressure measurements are consistent with
connectivity. The GOR from the lab PVT reports is also shown in Fig. 7. The GOR from one sample (one well) is
substantially smaller than the others. Normally this could indicate compartmentalization. Here, there is a second explanation.
There could be ongoing massive gas influx into this reservoir, with gas pockets getting trapped in connected, lobate systems.
In this Pliocene reservoir with (likely) current gas charged, insufficient time has passed for equilibration. After production,
this reservoir was shut in and all measured pressures returned to virgin pressure indicating 1) excellent connectivity and 2)
strong aquifer support. This observation is consistent with connected but disequilibrium fluids as the possible origin of the
unusual GOR measurements in Fig. 7. Other unusual observations are consistent with this interpretation.
Figure 8. Log of an interval in one primary sand. A whole core was taken confirming the excellent porosity and permeability obtained in wireline logging. Nevertheless, large sections of the producing interval exhibited no fluorescence as indicated by two brown rectangles in the figure labeled “Non-fluorescing core”; this is traced to a tar coating in these core sections.
Whole Core. In one of the wells, whole core was obtained for much of one of the target sands. Lab data confirmed excellent
porosity and permeability of the target sand. However, a surprising observation from the whole core is that large sections of
the oil bearing sand of the whole core did not fluoresce as indicated in Fig. 8, in spite of the crude oil being highly
8 SPE
fluorescent. The log data exhibits density–neutron crossover that is associated with gas (shaded yellow in Fig. 8). The gas is
towards and at the top of this interval which is what would be expected if this interval is vertically connected (but perhaps
around baffles). In addition, the reservoir pressure greatly exceeds saturation pressure of this oil. The observations of gas and
an extended section of tar are related as will be discussed.
Figure 9. Whole core (6 feet section in two contiguous 3 feet sections) from the well in Fig. 8 at the transition from fluorescent to nonfluorescent sands under ultraviolet illumination. The core sections are displayed under visible illumination (left in each 3 foot panel) and under UV illumination (right in each 3 foot panel). Visible light shows increasing optical absorption in the nonfluorescent section. Laboratory analysis confirmed that tar in the core produced little fluorescence and the dark color of the core. n.b. The core section with tar is permeable and porous. The transition from fluorescent to nonfluorescent sand core is shown in Fig. 9. The nonfluorescent section is due to a coating
of tar in the core. This tar is not typical: it is not at the low point in the reservoir, it covers roughly ½ of the producing interval
(cf. Fig. 8) and most importantly the tar zone is porous and permeable. Moreover, this reservoir exhibits excellent
connectivity and natural aquifer support; shut-in resulted in virgin pressure. These properties are quite distinct from tar mats
routinely encountered in oilfields. Typically, tar mats are found at the base of the oil column, they cover a small overall
section of the producing interval, they are not permeable, and have not been considered porous. Tar mats at the OWC often
preclude natural aquifer support and preclude utility of water injectors.
To confirm that the tar mat herein is in fact permeable, it is desirable to perform a downhole flow test. That is, core properties
can be altered upon depressurization with subsequent lab measurements. To test the tar zone permeability, a one foot interval
in the tar zone was perforated and produced by straddling the perforation with the MDT Dual Packer. Figure 10 shows the
section of core from the well depth that was perforated.
Figure 10. Whole core including the one foot interval that was perforated and sampled using the MDT Dual Packer. The entire permeable section of this core depicted here is coated with tar. (Shale at the base of this core section is not permeable.) The sampled crude oil is less than 1% asphaltene while the tar is 35% asphaltene confirming the coexistence in the reservoir or two immiscible hydrocarbon phases.
163292
SPE 163292 9
The tar zone allowed flow of a nice light crude oil in the wireline sampling test confirming that the tar zone is permeable.
The produced oil contains less than 1% asphaltene while the residual tar was found to have ~ 35% asphaltene. Consequently,
it is evident that there are two immiscible hydrocarbon phases present in the formation at the same depth, a light oil and a tar.
Indeed, four hydrocarbon phases have been experimentally determined to coexist in asphaltenic materials, a gas, two
immiscible liquids and a solid;[14] this reservoir is not violating any thermodynamic principles. Nevertheless, the unusual
nature of this tar mat must be explained.
Figure 11 shows a detailed analysis of the phase behavior of crude oils produced (Left) from the tar zone and (Right) from a region not near the tar zone. The complex phase behaviors of these two crude oils are similar. The tar mat does not signify a large change in properties of the local crude oil. The asphaltene onset pressure is near or at reservoir pressure which is expected for a late gas charge into crude oil that resulted in tar.
Given that the tar represents a phase transition of heavy ends precipitated out of the oil, it is important to che ck the phase
behavior of crude oils, produced from the tar zone and produced from a point not so close to the tar. Figure 1 1 shows that the
overall complex phase envelops of these crude oils are similar, meaning that the tar mat is not associated with some dramatic
change of the corresponding crude oil. Nevertheless, there are some differences noted in bubble point and details of the
asphaltene onset. As expected for reservoirs with late gas charge and tar, the asphaltene onset pressure is near or at reservoir
pressure. Asphaltene is not a homogenous chemical substance. Some fractions of asphaltenes are more stable in solutions,
others less stable. When gas destabilizes asphaltene sufficiently, some fraction of asphaltene can for tar. Another fraction can
remain in the liquid phase, but very unstable such that any pressure reduction causes some of this fraction to precipitate as
shown in Fig. 11. The bubble points of the crude oils are not near reservoir pressure even though density neutron cross was
observed (cf. Fig. 9) in upper sections of the sand. This is another indicator that the reservoir fluids are very much out of
equilibrium.
Geoscenario. The explanation consistent with a broad array of observations is the following: the reservoir rock is of Pliocene
age. More recently, a black oil charged into the reservoir. More recently still, likely ongoing, the reservoir experienced a
massive gas influx. Roughly, this corresponds to processes occurring in the last 1 million years. The oil in proximity to the
gas experienced a large, rapid increase in GOR causing rapid destabilization of the asphaltenes. This destabilization was so
complete that the asphaltenes could fall only small distances and only vertically in the oil column before they stuck to
available surfaces, the grain surfaces of the rock. This rapid destabilization did not allow time for the asphaltenes to migrate
to the low points in the reservoir, the OWC. That migration process, for example that did occur in a large field in Saudi
Arabia, is primarily lateral.[12,13] This rapid destabilization event did not even allow time for the asphaltenes to fall
vertically all the way to a shale break at the base of the sand. That process would yield a relatively thin tar mat of no
permeability. Instead, the rapid destabilization of asphaltenes caused the asphaltenes to “paint” the rock surface over an
extended vertical interval; this occurring after the asphaltenes fell a short distance in the sand. An extended vertical tar mat
interval is consistent with only a thin layer of tar on the rock; there was not enough asphaltene in the oil to fill ½ the
producing interval with space-filling tar (cf. Fig. 8). This geoscenario is consistent with many observations: 1) the remaining
oil has contains very little asphaltenes, 2) reservoir pressure is the asphaltene onset pressure, 3) an extended vertical interval
of tar is permeable, 4) much tar is found up-structure and near gas bearing zones, 5) the evident lack of equilibrated GOR
indicates these processes are very recent; the large GOR variations indicate massive gas influx recently occurred 6) the
asphaltene destabilization was so dramatic that evidently even some resins of lower viscosity phase separated. The last point
has significant implications for production as discussed below in the Production Section.
10 SPE
This rapid reservoir process yielding a disequilibrated fluid column is essentially at one extreme with a very rapid time frame.
That is, the GOR and thus methane are not equilibrated, and the asphaltenes are plated out locally on reservoir rock up-
structure. This defines “very young” and is roughly 1 million years old. The massive Saudi Arabian reservoir with
equilibrated asphaltenes over a huge length scale is at the other extreme of time, essentially defining a “very old”. That is, the
asphaltenes are equilibrated over great distance in spite of their tiny diffusion constant. This Saudi Arabian reservoir defines
what is very slow - and is roughly 150 million years old. These two case studies, the one herein and the one from Saudi
Arabia [12,13] bracket reservoir fluid processes in time scale. Other reservoirs should be in between these two in terms of
time frame and thus in terms of observables that affect production, such as GOR distributions, asphaltene distributions, tar
location etc.
An important component of tar mats is their structure. Some tar mats appear to consist of both an immo vable carbonaceous
phase and a heavy oil phase.[13] This is expected from slow dynamics of asphaltene sedimentation. Tar mats produced in
rapid asphaltene destabilization can have fundamentally different properties. Asphaltenes from rapid destabilization can have
lower asphaltene content and higher mobility. That is, strong asphaltene destabilization that causes fast asphaltene deposition
also causes deposition of some heavy resin components that are more mobile than a higher purity asphaltene deposit. There
are important consequences of some mobility of tar, even if permeable.
A confusion can occur in evaluating OWC tar mats vs rapid destabilization tar mats. In both cases, the tar mats have >35%
asphaltene content (cf. Fig. 10, and Ref. [12, 13]). In both cases, thin sections exhibit porosity and exhibit a carbonaceous
coat on the grain surface. However, there are distinct differences. The OWC tar mats can go as high as 60% asphaltene. And
the OWC tar mats are not permeable, while the rapid destabilization tar mat is permeable. In both cases, there are two
immiscible hydrocarbon phases present. In the rapid destabilization tar mat, in addition to the tar, there is a light oil. In the
OWC tar mat, there is a heavy oil of ~35% asphaltene plus a carbonaceous coat of extremely high asphaltene content (≥60%
asphaltene). The OWC carbonaceous coat seals off pore throats trapping heavy oil, and precluding the ability to acquire pure
samples of the carbonaceous coating. The rapid destabilization tar mats are porous and allow easy isolation of the tar mat
from the light oil. The huge difference is that the rapid destabilization tar mat is not ultra-high viscosity and can flow (like a
heavy oil) while the 60% carbonaceous coat of the OWC tar mat is ultra-high in viscosity and cannot flow under any
conditions.
A thin section of the tar mat is shown in Figure 12.
Figure 12. Thin section of the tar mat. Black is the tar. The blue is epoxy that displaced movable fluids prior to preparation of the thin section, and white is the sand grain. This image is consistent with significant porosity in the tar zone.
Production. A well test following long term production was performed after perforation of an interval containing a
permeable tar. (This well test is different than the test presented in Fig. 10 where a one foot interval was flowed.) In this test,
significant and relatively low viscosity tar was obtained in tubulars during this test. Figure 13 shows viscous heavy ends
remaining in the tubulars after the well test. Obviously, flow of such a material is of major concern in production. As shown
herein, understanding the distribution of reservoir fluids and their organic solids alike within a single framework helps to
identify corresponding production issues that are significant.
163292
SPE 163292 11
Figure 13. Residual tar in tubulars after a well test. This tar is thought to arise from mobility of an existing tar mat. This material differs from typical asphaltene deposits in flow assurance that have a physical consistency and appearance similar to coal.
Figure 14. A cumulative-oil dependent skin in production is observed and is attributed to asphaltene concerns, both mobile tar and asphaltene onset with pressure reduction. Xylene treatment of the producing well significantly improves performance. With repeated xylene washes, the rate of skin deterioration can be reduced.
Consistent with this well test result, a skin that is dependent on cumulative produced oil has been observed in the formation
when analyzing extensive production data as shown in Fig. 14. Asphaltene concerns, including both mobile tar as well as
asphaltene deposition with pressure drop are considered responsible for this increasing skin. Xylene treatments are effective
in mitigating these production problems. In particular, repeated xylene treatments reduce the rate of skin deterioration.
Understanding these complex production issues at the outset is desirable in order to optimize production by dynamic
intervention.
12 SPE
Figure 15. Barrels of oil per day and the GOR of the produced oil over a multi-year period. It is uncommon to have such large, seemingly random variations in GOR. The existence of pockets of connected, disequilibrium fluids in this young reservoir is consistent with these observations. Another observation that is not common is the large variation of the GOR of the produced oil. Figure 15 shows that the GOR
varies up and down by a factor of three in production from one well in one reservoir. Again, note that the reservoir appears to
be connected with a strong aquifer drive. The large, nonmonotonic variations of GOR coupled with many observations
discussed above suggest that there are pockets of significantly different fluids in the reservoir that have not had time to
equilibrate. Baffles but not barriers might be separating different fluids. Using literature diffusion constant (D where t=D/x 2 )
of methane through hydrocarbon filled porous rock of ~10 -5
cm 2 /sec, [15,16] one obtains that it takes a million years (t) to go
a distance (x) of 200 meters. It is plausible that reservoir fluid variations exist at that length scale being separated by baffles,
with current gas charging, and with no time to equilibrate.
Different Field, Same Observations. Another field in the same basin exhibits very similar behavior to the field discussed
above. There is tar deposition in a well near the crest of the field.
Figure 16. Core and logs from a well near the crest in another field in the same basin. Very similar observations are made to the case study above; there is tar deposition up-structure that is porous and permeable. All production issues discussed above apply to this field.
163292
SPE 163292 13
A question arises as to how common the above case study is. Figure 16 shows that another field in the same basin but a
significant distance away exhibits the same ‘unusual’ phenomena. There is tar deposition up-structure that is porous and
permeable. Other production issues such as a skin dependent on cumulative production and variable GOR in production are
also observed. It is evident that the phenomena discussed in this paper apply to a class of reservoirs, those with rapid and
recent gas charge into black oil.
CONCLUSIONS A Pliocene reservoir study is presented with a variety of putatively unusual observations: there is tar deposition up-structure
that is porous and permeable. There are large, nonmonotonic variations in GOR obtained in wireline logging and with
production data over years. There is mobile tar as shown in photographs, yet the produced crude oil is rather light. A
consistent geoscenario is for a rapid and recent gas charge into black oil, the time frame being roughly one million years.
This short time is not sufficient for equilibration of reservoir fluids even though the reservoir exhibits excellent connectivity
and pressure build-up behavior in shut-in. The asphaltenes were knocked out of solution so rapidly and strongly that they did
not have time to descend in the reservoir to the OWC; rather, they made it only part way down the individual sand lobes
before sticking to and ‘painting’ the rock surface – thereby leaving permeability. This deposition process naturally leads to
somewhat higher mobility tar than typically found in OWC tar mats, enabling limited but important mobility of this tar.
Consequently, a production dependent skin develops and requires intervention via xylene treatment. These rapid ~1 million
year old processes are in distinct contrast to equilibrated asphaltenes in a giant, Jurassic Saudi Arabia field, thus old in a
geologic sense. Consequently, short and long time scales are establish as ~1 million years to ~150 million years for reservoir
fluid processes of interest to major production concerns. Many other reservoirs are intermediate in this time scale. These new
methods, particularly employing new asphaltene science and downhole fluid analysis technology are enabling significant
increases in efficiency as increasingly difficult reservoirs are exploited. Moreover, as shown herein, neighboring reservoirs
within a basin can exhibit very similar fluid variations and production concerns.
ACKNOWLEDGEMENTS The authors are deeply indebted to the technologists in the operating company. These technologists recognized the origins of surprising reservoir complexities and performed definitive yet uncommon tests to
validate these physical origins. Simply stated, their skill and clarity of thought is inspiring. We are also indebted to these
technologists and the operating company for permitting this publication.
REFERENCES
[1] Mullins, O.C., “The Asphaltenes”, Annual Review of Analytical Chemistry, 2011 Vol. 4, page 393-418
[2] Mullins, O.C.; Sabbah, H.; Eyssautier, J.; Pomerantz, A.E.; Barré, L.; Andrews, A.B.; Ruiz-Morales, Y.; Mostowfi, F.;
McFarlane, R.; Goual, L.; Lepkowicz, R.; Cooper, T.; Orbulescu, J.; Leblanc, R.M.; Edwards, J.; Zare, R.N.; Advances in
Asphaltene Science and the Yen-Mullins Model, Energy & Fuels, 26, 3986–4003, (2012)
[3] Freed, D.E., Mullins, O.C., Zuo, Y.J.: “Theoretical Treatment of Asphaltene Gradients in the Presence of GOR
Gradients”, Energy & Fuels, 24, 3942-3949, (2010)
[4] Zuo, J.Y.; Mullins, O.C.; Freed, D.; Elshahawi, H.; Dong, C.; Seifert, D.J.; Advances in the Flory-Huggins-Zuo Equation
of State for Asphaltene Gradients and Formation Evaluation, submitted, Energy & Fuels
[5] Stainforth, J.G., “New Insights into Reservoir Filling and Mixing Processes” in Cubit J. M., England, W.A., Larter, S.
(Eds.) Understanding Petroleum Reservoirs: toward and Integrated Reservoir Engineering and Geochemical Approach,
Geological Society, London, Special Publication, (2004)
[6] Mullins, O.C., The Physics of Reservoir Fluids; Discovery through Downhole Fluid Analysis, Schlumberger Press,
Houston, (2008)
[7] V. Mishra, N. Hammou, C. Skinner, D. MacDonald, E. Lehne, J.L. Wu, J.Y. Zuo, C. Dong, O.C. Mullins, Downhole
Fluid Analysis & Asphaltene Nanoscience coupled with VIT for Risk Reduction in Black Oil Production, Accepted, SPE
ATCE, (2012)
[8] Zuo, J.Y., Elshahawi, H., Dong, C., Latifzai, A.S., Zhang, D., Mullins, O.C., DFA Assessment of Connectivity for Active
Gas Charging Reservoirs Using DFA Asphaltene Gradients, SPE 145438, ATCE, (2011)
[9] Elshahawi, H., Latifzai, A.S., Dong, C., Zuo, J.Y., Mullins, O.C., Understanding Reservoir Architecture Using Downhole
Fluid Analysis and Asphaltene Science, Presented, Colorado Springs, SPWLA, Ann., Symp., (2011)
14 SPE
[10] Pfeiffer, T.; Reza, Z.; Schechter, D.S.; McCain, W.D.; Mullins, O.C.; Determination of Fluid Composition Equilibrium
under Consideration of Asphaltenes – a Substantially Superior Way to Assess Reservoir Connectivity than Formation
Pressure Surveys, SPE #145609 ATCE, (2011)
[11] Gisolf, A., Dubost, F.X., Zuo, J., Williams, S., Kristoffersen, J., Achourov, V., Bisarah, A., Mullins, O.C., SPE 121275,
SPE Europe/EAGE Ann. Conf. Ex., Amsterdam, The Netherlands, 8-11 June, (2009)
[12] Seifert, D.J., Zeybek, M., Dong, C., Zuo, J.Y., Mullins, O.C., Black Oil, Heavy Oil and Tar in One Oil Column
Understood by Simple Asphaltene Nanoscience, SPE ADIPEC 158838, Abu Dhabi (2012)
[13] Seifert, D.J., Qureshi, A., Zeybek, M., Zuo, J.Y., Pomerantz, A.E., Mullins, O.C., Mobile Heavy Oil and Tar Mat
Characterization Within a Single Oil Column Utilizing Novel Asphaltene Science, SPE KIPCE 163291, Kuwait International
Petroleum Conference and Exhibition, Kuwait City, Kuwait, Dec 10-12, (2012)
[14] Shaw, J.M., Zou, X.; Phase behavior of heavy oils, Ch. 19 in Asphaltenes, Heavy Oils and Petroleomics, Mullins, O.C.
Sheu, E.Y., Hamami, A., Marshall, A.G., Editors; Springer, New York, (2007)
[15] Chen, L.L., Katz, D.L., Tek, M.R., Binary gas diffusion of methane-nitrogen through porous solids, AICHE, 23, 336-
341, (1977)
[16] Ghorayeb, K., Firoozabadi, A., Modeling multicomponent diffusion and convection in porous media, SPE Journal, 5,
158-171, (2000)
163292
