SPE 82252
The Impact of Hidden Emulsions on Oil Producing Wells
- Stimulation Concept and Field Results
R. Espinoza, SPE, Preussag Energie GmbH Venezuela, W. Kleinitz, SPE, Preussag Energie GmbH Germany
Copyright 2003, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE European Formation Damage
Conference to be held in The Hague, The Netherlands 13-14 May 2003.
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 the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. 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 where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
Ninety-eight per cent of the total cumulative production from
the Cabimas Field has originated from Miocene reservoirs.
The difference in density between Miocene oil and water is
very small. In addition, this crude oil has a tendency to wet the
surfaces of the reservoir rock. Under these conditions, a water-
in-oil emulsion, and thus higher viscosities are favoured; this
in turn impairs the well production capacity. Three wells were
selected (R-A1, R-A2, and R-A3) for the treatment of hidden
emulsions with the use of a wetting agent in an organic
solvent system.
Another factor that was thought to impair the well
production capacity was plugging of the Miocene slotted-liner
gravel-packed completions by organic solids which are
present. For considering this problem separately, two further
wells (R-B1 and R-B2) were treated only with an organic
solvent system.
Treatment preparation and pump operation (between 250
and 400 bbl/well) for all five wells were completed in a single
week. All of these field operations were simple and were
carried out with no operational or safety problems.
Thus, the use of the wetting agent proved to be a
successful method for breaking hidden emulsions. Two of the
wells treated in this manner, R-A2 and R-A3, respond with an
increase in the fluid production rate. An exception was well R-
A1; this negative result was due to the existence of a thief
zone that fails to receive enough treatment fluid. Wells treated
with organic solvents only did not show any change in
production rate.
Introduction
During the last 80 years, more than 360 MMBls have been
produced in the Cabimas field. Ninety-eight per cent of the
total oil has accrued from Miocene Reservoirs. More than 900
wells have been drilled to this target, which has proved to be
highly prolific and successful: “La Rosa formation” [1]. This
formation is subdivided into three units: “La Rosa Superior”,
“La Rosa Intermedio”, and “La Rosa Inferior”, which is also
called “Santa Bárbara” Member and is the most
important unit.
In the central, northern, and north-eastern areas of the
Cabimas field, the Icotea formation (Oligocene Age) is also
present. Where this formation is in direct contact with the
Santa Bárbara Member, they are considered together as a
single reservoir unit. In this area, the production from most of
the Miocene wells has originated jointly from both of these
two members (Santa Bárbara and Icotea). On the average,
these formations have the following properties: Age: Miocene;
formation: Sta. Barbara/Icotea; depth: 700 - 2600 ft; thickness:
80 - 120 ft; porosity: 20 - 25 per cent; permeability: 100 -
1000 md; pressure: 100 - 500 lb/in2. The main reservoir drive
mechanisms have changed during time (as a function of the
pressure depletion) from liquid expansion (undersaturated oil)
to the current solution gas expansion. Regarding pressure
maintenance, different trials have been done in the past
(firefloods, waterflood, steam soak, steam drive and gas
injection). However, they have failed for technical and/or
economical reasons. Currently no secondary pressure support
is used on Miocene reservoirs [2].
In the early seventies, SHELL performed a sidetrack
campaign in more than 60 Miocene wells. Of these, 58 were
completed with slotted liners and gravel-packed. These wells
have been the best producers in Miocene reservoirs. However,
this completion technique is currently not
feasible economically.
The crude oil from the Miocene is highly aromatic and
therefore has a tendency to wet the surfaces of the reservoir
rock. Since the difference in density between the oil and water
is so small, water-in-oil emulsions, and therefore higher
viscosities, are favoured. The mobilisation of ultra-fine
particles results in additional stabilisation of the emulsion and
thus causes a noticeable impairment in flow behaviour. Hence,
washing of the Miocene slotted-liner gravel-packed
completions and avoiding the production of emulsion are the
main targets of this study.
Some of these wells have shown indications of plugging
by emulsions incremented and supported by ultra-fine
particles in the gravel-pack area. Furthermore, high-density oil
deposits contribute to the plugging of the areas which
surround the gravel pack. Hence, washing the gravel pack
2 SPE 82252

system and avoiding the production of emulsion constitute the of this kind stabilise the water-in-oil emulsions drastically
basis of these stimulations. (figure 1) [5]. This tendency is supported by wetting of the
particles by the aromatic crude oil.
Properties of the Reservoir Components. Figure 1 Sand Particles in Water Phase (Cabimas)
Crude Oil.
Because of the simultaneous production from different
reservoirs, the API gravity of Cabimas crude oil from the Effect of Hidden Emulsions.
Miocene varies from 13º to 26º, with an average of 21º. This
oil is also characterised by an aromatic content (SARA) of 50
to 60 per cent. As indicated by GC analyses, the hydrocarbon
distribution in the crude oil corresponds to a high degree of
biodegradation. Only small quantities of n-alkanes above C-20
are still present. The biomarkers phytane and pristane indicate
that the oil has been generated under marine conditions.
The high aromatic content in the Cabimas oil is also the
cause of the measured densities from 0.96 to 0.98 g/cm³,
among other effects. Aromatic compounds generally exhibit a
strong affinity or solubility toward water; that is, aromatic
crude oil is capable of binding water. (The solubility at 20 °C
of water in n-hexane is 135 ppm, whereas for water in
benzene, the value is of 520 ppm.) This property is favourable
for wetting of the reservoir rock [3]. The enveloping phase is
removed from the otherwise water-wet sandstone, and the
negative charges of the quartz are compensated by the
opposite polarity of aromatic compounds which contain
nitrogen. Chemical bonds of this kind can be highly stable. Fluid properties (such as viscosity and ºAPI) in an oil-water
system can change dramatically in correspondence with the
Reservoir Water. amount of water present. Because of the small difference in
There is hardly any difference in the reservoir water produced density and the predominantly aromatic character, there is a
from the individual formations, Miocene and Eocene. The pronounced tendency to form emulsions. Even during
TDS concentration is generally about 7500 mg/l, a value laboratory investigations, these emulsions can be very difficult
which is considered low. Because of the low salt content, the to break (even in the centrifuge and with demusifiers) [6].
density (20 °C) is close to 1 g/cm³ (1.0075 g/cm³). In relation However, dehydration (ASTM 4006) in the laboratory is time-
to the crude oil, the density differences are very small for the consuming and is applied only for LACT unit measurements.
Cabimas oil/water system ( = 0.0475 g/ml). For all other measurements, the centrifuge method is applied.
However, because of the small difference in density and In figure 2, the difference between results of the centrifuge
the predominantly aromatic character of the oil, there is a and dehydration methods is shown for the determination of the
pronounced tendency to form emulsions. The results of BSW water content in different wells. In considering this figure, it
as well as Dean Stark analyses (ASTM 4006) of the water must be emphasised that only 15 per cent of the total water can
present in the produced crude oil indicate the occurrence of so- be removed by centrifugation of the oil, whereas 75 per cent of
called hidden emulsions, especially in the Miocene. A water the total water (based on the xylene method) can be removed
determination by the distillation method with the use of xylene by dilution in Diesel oil and addition of demulsifier. The
(Dean Stark) yields decidedly higher water contents, whereas remaining water in the system can cause significant errors in
a complete separation in the centrifuge is not possible, even the determination of fluid properties such as viscosity
after addition of a selected demulsifier. The deviation of the and ºAPI.
results between the two methods varies between 10 and 30
per cent.
In accordance to the literature [4, 5], the viscosity of a water-
in-oil emulsion increases with increasing content of the
aqueous phase in the system. For production, this implies that
a higher viscosity aggravates problems with the flow
behaviour in the near-wellbore area.

Solids in the Production Fluid.

In the Miocene and Eocene formations at Cabimas, there is a
tendency toward sand entrainment because of the stratigraphic
sequence. As indicated by microscopic examinations, the grain
size of the fines is in the range from 2 to 4 µm. Quartz is the
major component in each case, and is followed by minerals
such as albite, orthoclase, muscovite, and kaolinite. Particles
SPE 82252 3

20 basis of previous intensive tests on demulsifiers from various

suppliers, the Baker/Petrolite demulsifier DC-64F PDV was
Water Content by Centrifuge method, %

Centrifuge (50% Oil + 50% Diesel + Demulsifier)

16
Centrifuge (100% Oil)
selected for the treatment of Cabimas crude oil. For the
purpose of stimulation, the function of this product is to
accelerate demulsification in the zone immediately
12
surrounding the well and thus to decrease the viscosity in
this zone.
8

Wetting Agents.
4 In the Cabimas wells with a higher water cut in production
(> 10 %), a solid phase (as fines) was present within the water
0 phase; these fines increased the stability of the existing
0 5 10 15 20
Water Content by Xylene method, %
25
emulsions (figure 1). For dispersing the agglomerated sand
phase, the use of a wetting agent (Baker Petrolite Sepapar P)
has proved to be effective in the past [7]. This product is a non-
Figure 2 Effect of Hidden Emulsions ionic surfactant on a nonylphenyl ethoxylate basis. The
conditions for oil solubility (nonylphenyl residue) and water
solubility (11 – 15 ethoxylate groups) have been optimised to
Stimulation Fluid Selection.
permit the broadest possible distribution in both phases. With
The Cabimas Miocene wells were classified into two groups:
the combination of wetting agent and demulsifier in a ratio of
the first group were low-water producers (water cut £ 10 %),
1:1, the agglomerated “fines“ can be destabilised and thus
R-B1 and R-B2. The second group were the high-water
washed out by entrainment, as demonstrated by microscopic
producers (water cut > 10 %), R-A1, R-A2 and R-A3. In order
examination, among other methods.
to dissolve and remove the oil-specific deposits, the injection
The commercial product is a pure chemical and must be
of an organic-based fluid is necessary for all wells. Access to
carefully diluted in advance for field applications (content:
the continuous (oil) phase in the reservoir is thus facilitated.
500 - 1000 ppm in water). The possibility of forming liquid
For this specific task, two fluids were selected: gas-oil and an
crystals is thus avoided. The addition of polyglycol can
aromatic solvent.
increase the solubility of the product in water. Because of the
high dispersivity, organic deposits as well as agglomerated
Gas-oil.
solids are undermined and penetrated; they are then distributed
The use of gas-oil (Diesel) is a logical choice for Venezuela,
in the transport fluid. The surfaces thus remain water-wet. The
not only because of the economic advantage. Gas-oil is highly
cleaning action of this wetting agent has proved to be highly
soluble, or miscible, in crude oil and also has a very low
effective over a long period of time.
viscosity. At 40 °C, the viscosity of gas-oil is only 2.7 mPas.
Favourable mixing conditions are thus ensured.
Stimulation
Wells Selection.
Aromatic solvent.
From the 58 wells completed with slotted liners and gravel
Disolin 88 is an aromatised gasoline cut which also contains
packs, five were selected on the basis of their recent decline in
about 12 per cent of polyglycol. The function of Disolin 88 is
production rate: R-A1-A3, R-B1, and R-B2. These five wells
to dissolve and remove oil-specific deposits which may
are located in different Miocene reservoirs simultaneously; the
possibly be present in the zone immediately surrounding the
average °API value therefore varies in production. The
well and in the gravel after a long period of production.
completed reservoirs are listed on a well-by-well basis in
Furthermore, Disolin 88 increases the aromatic character of
table 1. The following characteristics of these wells deserve
crude oil at reduced viscosity.
special attention:
As indicated by previous investigations, gas-oil and
All five wells were completed in the La Rosa (~23ºAPI)
Disolin 88 are miscible in all proportions. The contents of the
formation, which is the most prolific zone in the Cabimas
two base fluids can be determined and easily quantified by gas
Miocene. All five wells were completed in Lower Lagunillas
chromatography. For economic reasons, a ratio of about 30 per
Inferior (LLi)(~12ºAPI). However, this member cannot be
cent Disolin 88 and about 70 per cent gas-oil was chosen for
producing. This situation applies especially to wells R-A2 and
the stimulation.
R-B1, where the average API gravity is not affected by the
The biomarkers pristane and phytane were detected by GC
presence of heavy oils. In two wells, R-A1 and R-B1, TD was
analysis. These biomarkers are not present in detectable
reached some 10 to 20 ft below the Eocene unconformity.
amounts in Cabimas crude oil because of the high level of
Hence, production from upper Eocene members is possible to
biodegradation. Thus, by identifying these biomarkers in the
a certain extent. In two wells, R-A2 and R-B1, the Icotea
produced oil after treatment, the amount of injected fluid can
formation is present and could be producing together with the
be correlated.
La Rosa formation. In two Wells, R-A1 and R-B2, Lower
Lagunillas Superior (LLs) (~12ºAPI) is producing together
Demulsifier.
with the La Rosa formation. Thus, the average API gravity in
Demulsification of the water-in-oil system in the Miocene
these wells is affected (15.5ºAPI in R-A1 and 16.5ºAPI
reservoir is promoted by the addition of a demulsifier. On the
in R-B2).
4 SPE 82252
Table 1 Completed Members/Formations by Well
Completed Formations Produced
Well LLs LLi La Rosa Icotea E. Unc.
(~12ºAPI) (~12ºAPI) (~23ºAPI) (~23ºAPI)
Avg. ºAPI
R-A1 X X X X 15.5
R-A2 X X X 21.0
R-B1 X X X X 22.0
R-B2 X X X 16.5
R-A3 X X 15.5
With this well selection, it was ensured that all Miocene
formations (in all possible combinations) were included in
the study.
Stimulation Operation
The selected Miocene wells were treated by two
different methods:
o Wells with an increased water cut at present: In the
three wells selected for this treatment, the water cut
ranged from 5 to 12 per cent (R-A1, R-A2, R-A3).
Basis fluids: gas-oil (64 %), aromatic solvent: Disolin
88 (36 %); chemical additive wetting agent,
Sepapar-P (0.05 %), Demulsifier, DC-64F: 0.05 %
(contents by volume)
o Wells with a low water cut at present: In the two
wells selected for this treatment, the water cut had
remained below 2 per cent during the past 15 months
(R-B1, R-B2). Basis fluids: gas-oil, Disolin 88;
chemical additive: demulsifier, D-64F; 0.05 %.
The length of the slotted liners was taken as basis for
calculating the total volume required. Hence, 0.25 m3/m (0.48
bbl/ft) was used. The basis fluid consisting of gas-oil and
aromatic solvent was prepared by pumping Disolin 88 into
the gas-oil.
An emulsion was formed during the preparation of the
basis solvent, Diesel/Disolin 88; as the components were
mixed, the glycol contained in Disolin 88 favoured an oil-in-
water emulsion. A highly viscous emulsion was formed during
the addition of the wetting agent / water phase to the Disolin
88 / gas-oil phase; this emulsion depends on the concentration
ratios. The addition of water, that is, dilution of the emulsion,
results in a significant decrease in viscosity. Even after 48
hours at about 40 °C, the emulsion remained stable.
Important: For the preparation of mixtures containing the
wetting agent (Sepapar-P), this product was first dissolved
with the use of 10 bbl of water per gallon of Sepapar-P with
controlled agitation (foaming). This mixture was then added to
the organic basis fluid; emulsions are formed during this
process. Circulation for 20 minutes with an external
centrifugal pump was sufficient to ensure adequate
homogenisation; on the other hand, the emulsification was
also stabilised by this process.
Injection.
The five selected wells are characterised by the following
special features: Miocene formation, slanted wells completed
with “slotted liner“, gravel-sand packed in the under-reamed
zone, disconnection of the sucker-rod pump in the final depth
zone, decrease in production during recent months,
accessibility of the well for pump and chemical
transport vehicles.
The selected wells were originally completed in 1929-1930
(more than 70 years ago!). Some precautions were then taken
to protect the well integrity: Similar treatments usually involve
injection through the annulus; this process is easier, since there
is no need to pull the pump. However, this practice requires
perfect well integrity, a condition that may not prevail in
Miocene wells, since most of them have holes in shallow
zones as a result of corrosion. For these reasons, it was
decided to inject through the tubing, in order to protect the
casing and also avoid injecting into a shallow thief zone. As a
precaution against possible casing leakage, the wellhead
pressure should not exceed 300 lb/in2 (1.5 bar) during the
treatment. The injection rate must therefore be adjusted to
match the pressure conditions.
For ensuring that the washing fluid acts over the entire
length of the liner, production from the well was continued
until one hour before treatment, in order to lower the liquid
level. The pump was then pulled from the pump seat, and
pumping of chemicals started immediately.
No injection problems occurred in any cases of chemical
injection. The average flow rate was about 1.5 bbl/min (about
240 l/min). It was observed that a higher injection pressure
(150-300 lb/in2) was necessary for the wells treated with
Sepapar-P because of the emulsion formed in the treatment
fluid during injection. In order to promote the penetration into
the reservoir rock, an after-flush volume of 50 bbl gas-oil was
injected into each well through tubing and casing after about
14 days of initial treatment [8].
Because of the favourable pressure conditions, pumping
was performed with the operator’s own vehicles. In the course
of this measure, the pressure in the well was relieved, in order
to allow movement of the liquid level in the casing. After the
injection of the gas-oil slug, the casing was closed. As a result,
the flowing gas slowly drives the solvent into the formation
and thus penetrates the liner zone. As a follow-up to the after-
flush, this procedure was repeated once or twice for each well.
The total shut-in time was 24-29 days.
In order to have a representative evaluation, it was decided
to monitor each well until they have produced back three times
the total volume injected. The total evaluation time was
around 2 months.
Result Discussion
For each well, the starting dates and production rates were
different. For this reason, evaluating and comparing well
properties vs. time or cumulative production could be
confusing. Then, for better understanding, a dimensionless
production index concept was defined in order to evaluate the
different properties [9]. The index equation is:
Cum. production volume . By using this index,
Index =
Total injection volume
regardless rate or starting date, for all wells, index=0 is the
starting point; and index=1 is when the total produced volume
is equal to the total initially injected. From this point
(index=1) it was expected to have the fluid properties (API,
BS&W, Viscosity) stabilization.
SPE 82252
For analyses purposes, the results were separated in two
groups showing different results: Wells treated without the
wetting agent (Sepapar-P): these are the wells that produce
less than 10% water cut (R-B1 and R-B2); Wells treated with
the wetting agent (Sepapar-P): these are the wells that produce
more than 10% water cut (R-A1, R-A2 and R-A3).
During back production samples were taken from the
wellhead for analysing API gravity, BS&W and the chemical
composition by gas chromatography. For calibration oil has
been used from the different treated wells and the base fluid
(Disolin 88 and gas-oil). The quantitative amount of the
treatment fraction was calculated by individually selecting
characteristic peaks from Disolin 88 and gas-oil. Then, the two
different values were plotted. By these methods it was
observed that the calculated treatment fraction by Disolin 88
peaks shown a tendency to be lower than the calculated by the
gas-oil. This is due to the higher volatile components in the
Disolin 88, which can evaporate during sampling.
Wells Treated without Wetting Agent.
Since the wells R-B1 and R-B2 produced less than 10%
BS&W, it was decided not to use the wetting agent (Sepapar-
P) on these wells. The preparation process was easy because
no emulsions were created during circulation on the Frac-
Tank. During injection only 100 psi wellhead pressure
were detected.
The fluid behaviour for these two wells was similar to the
others treated without Sepapar-P: after the injection, the fluid
level was at the wellhead. After that, during the shut-in period
(soaking time), the level was lowered constantly below the
liner top. Again, fluid levels information was not available.
For this reason, during the production time, all acoustic shots
were also interpreted as the top of the liner. Thus, real
production fluid level was not available, and data from pump
efficiency was used to identify those cases were the fluid level
was at the pump. For the fluid treatment the API gravity was
35º (API before treatment R-B1: 22,1;R-B2: 16,5).
As long as the injected fluid was 13 to 18º API higher than
the produced, during the first production test it was expected
to have a mixture of oil-injected fluid with a higher API
gravity than the initially produced. However the result was
totally opposite and the produced API gravity was lower than
the initial value (Fig. 3). To simulate that result, it was
assumed that any of the fluid injected was produced back and
the low API was the result of water content as
hidden emulsions.
5
23
API Gravity
R-B1
R-B2
21
19
17
15
13
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Volume Index
Figure 3 API Behaviour on Wells Treated without Sepapar-P
To reproduce and explain the produced fluid API gravity
on well R-B1 during the first test, it was assumed that the
analysed sample contains 20% water as hidden emulsion,
while for well R-B2 it was not possible to make the same
assumption because it would be necessary to have a 50%
water emulsion which is too high in comparison with the
average Miocene water cut production. However, the initial
API gravity (16.5º) and the commingled completion La Rosa –
Lower Lagunillas Superior (LLs), suggest that the well is
producing simultaneously from both reservoirs. Then, the
explanation for such reduction on API gravity is that the
solvent injection changed for a short period the commingled
production ratio. This particular behaviour from R-B2 was
again observed in well that R-A1 also produces from LLs but
treated with the wetting agent.
After producing back the volume injected (Production
index = 0,7-0,8), well R-B1 stabilized its API gravity on their
original values and no longer effects were observed on API
gravity values. However, on well R-B2, the solvent injection
still affected the commingled production ratio making the API
gravity decrease and increase with 1º API variance until
finally stabilized on its original value (16.5º), Figure 3.
The flow conditions into the well did not improved by the
solvent injection; as a consequence the two wells were always
producing pumped-off and the pump efficiency did not
change. Similarly to the API gravity, well R-B1 had a
viscosity increment during the first production test
(consequent of an API gravity reduction) and further on
returned to the original value, same as the API behaviour
(Fig. 4). Regarding of well R-B2, it was more evident the
changes on the commingled production ratio induced by the
solvent injection (Fig.4). These changes were consistent with
the changes on API gravity. Thus, viscosity increases when
API gravity decreases (as expected). Finally, after producing 4
times the total volume injected, the viscosity returned to its
original value.
 6 SPE 82252

700 In order to calculate the ratio injected/reservoir fluid in the

R-B1
produced oil, it was assumed that the wetting agent dispersed
Viscosity, mPa-s

R-B2
600 all hidden emulsions, and then oil samples were totally
500
dehydrated (0% water content) before API analyses. Thus, the
produced API gravity by well was match by combining
400 reservoir oil and injected fluid. During the first production
test, it was noticed that wells R-A2 and R-A3 show a good
300
mixture (22% and 28% by GC analyses) on the reservoir-
200
injection fluids and by that causing an increase of the API
gravity, a decrease of viscosity and an increment in the oil rate
100 during the first production period.
After the first production test, on wells R-A2 and R-A3,
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 the API gravity started to decrease and as expected, stabilized
Volume Index for volume index values greater than 1. For volume index > 1,
API gravity was equal to the original value before treatment.
Figure 4 Viscosity Behaviour on Wells Treated without Sepapar-P This behaviour is shown on Figure 5. For well R-A1, the
behaviour was different, API gravity was constantly
Since there were no improvements on the viscosity neither decreasing (and viscosity increasing) until volume index = 1.5,
in the pump efficiency, no production increments were from there, API gravity increase and viscosity decrease until
achieved. The changes on the commingled production ratio stabilizing on their original values. Well, R-A1 showed this
(due to the solvent injection), induced a viscosity reduction. particular behaviour because it is the only well treated with the
As a consequence, the production out of the two wells wetting agent and completed commingled with the heavy oil
decreases to approx. 13 bls/d. formation LLs. Our interpretation is, that the solvent injection
changed the contribution ration between the two producing
Wells Treated with Wetting Agent. formations in favour of LLs. Then, the presence of more heavy
The addition of water and the wetting agent Sepapar-P, create oil in the produced fluid mixture, results into an API gravity
a more viscous treatment mixture. As a result, more time was reduction and a viscosity increment until volume index > 1.5,
needed during the preparation. Wells treated with Sepapar-P when the API gravity and viscosity of well R-A1, started to
(R-A1, R-A2 and R-A3) also showed higher injection pressure return to their original values.
(150 to 300 psi).
For the three wells treated (R-A1, R-A2 and R-A3), the 24
R-A1
API Gravity

fluid level behaviour was as described: right after the R-A2

injection, the fluid level was at the wellhead. After that, during 22
R-A3

the shut-in period (soaking time), the level was lowering

constantly until the top of the liner depth. From there, fluid
20
levels information was not available anymore since the sonic
wave could not go deeper. For this reason, during the
production time, all acoustic shots were interpreted as the top 18

of the liner. Thus, real production fluid level was not

available. However, data from pump efficiency was good to 16
identify those cases were the fluid level was at the pump.
The average API gravities of the main components used 14
for the fluid injection preparation are as follows: Aromatic 0 1 2 3 4 5 6 7 8 9 10
solvent 33°, gas-oil 37° and water 10°. With the known Volume Index

composition and volumes used for the mixture preparation, the

average API gravity of the injection fluid was calculated and it Figure 5 API Behaviour on Wells Treated with Sepapar-P
was 30-32ºAPI in all cases. This gravity was significantly
greater than the oil gravity of the oil produced by each well On wells R-A2 and R-A3, viscosity (Figure 6) was
before the injection (15-21ºAPI) (R-A1: 15,5°, R-A2: 21°, R- affected differently than API gravity. While API gravity
A3: 15,7°). rapidly returned to its original value, Viscosity initially
One or two days after the initial production, the first fluid decreases and immediately after the first test, increases
sample was collected. Since by that time any well had not constantly until finally stabilize after producing 3 to 7 times
produced more than 20% of the total injected fluid, it was the total volume injected.
expected that the produced oil API would be close to the
injected fluid values. However, it was at least 1 or 2 ºAPI
higher. This result proved that reservoir fluids were mixed
with the injected fluid and produced back but not as in the
expected ratios.
 SPE 82252 7

600 30
Viscosity, mPa-s

Fluid Comp. After Treatment, %Vol

R-A1
R-A1 R-A3
R-A2
R-A3 25 Disolin 88 Disolin 88
Diesel Diesel
400
20

15
200

10

0 5
0 2 4 6 8 10
Volume Index
0
0 2 4 6 8 10
Figure 6 Viscosity Behaviour on Wells Treated with Sepapar-P
Volume Index

Hidden emulsions were the cause of this viscosity
behaviour. Thus, while the high API gravity components (gas-
oil and Disolin 88) did not have any effect on the produced oil
API gravity nor on the viscosity, the small amount of
demulsifier and wetting agent were enough to disperse the
agglomerated sand phase (Fig. 1) present in water, thus
produce the demulsification. The amount (and the effect) of
the demulsifier and Sepapar-P in the reservoir was decreasing
with production. Then, hidden emulsions start growing again,
producing this increment on viscosity (Figure 6).
On well R-A1, the viscosity behaviour was driven by the
Production Rate, BOPD
Figure 7 Dielsel and Disolin content in the produced fluid by GC
The effect of the Sepapar-P caused the elimination of the
hidden emulsions, thus, a fluid viscosity reduction. By this
reduction, the flow into the well is easier producing a fluid
level and a pump efficiency increment and finally an
increment on the production rate. This situation was observed
in all three wells (Fig. 8).
120

R-A1
solvents (gas-oil and Disolin 88) inducing a change on the R-A2
contribution ratio from the producing commingled zones, R-A3

increasing the heavy oil fraction from LLs formation. As a 80

consequence, the viscosity increased. After volume index >
1.5, solvents were produced back and viscosity returned to its
original value.
For well R-A1 the GC-results from back production are 40
shown in Figure 7. In all analysed samples the total amount of
treatment fluid was not higher than 5 %Vol. After a volume
index of 1,1 GC analysis detected only traces of the treatment
fluid. From the wells treated with the wetting agent, R-A1 was 0

the only one completed commingled with LLs and top of 0 1 2 3 4 5 6 7 8 9

Volume Index
10

Eocene (Table 1). Then, the little detection of the injected

fluid in the produced oil, could be explained if one of this
formations (LLs or top of Eocene) behave as a thief zone.
The fluid from well R-A3 exhibits at start-up a
concentration of approx. 26 %Vol (Fig. 7). During the trail the
analysed treatment fluid is not constantly decreasing and
different humps in concentration were detected. This is the
result from two zones producing commingled (LLi & La
Rosa). Up to a volume index of approx. 5, treatment fluids
were still detected in the samples. This behaviour was found
analogous in the producing rate, where the highest rate
correlates with the highest fraction of treatment fluid (1st test)
and later this portion decreased to a minimum value
(~4%Vol.). For volumes index greater than 5, the treatment
fraction stabilised and some traces of fluid treatment were still
found. This traces were enough to maintain a little production
rate increment. It was also observed that after treatment, when
increasing the volume index the GC chromatograms were
influenced by heavy crude oil components (LLi is completed
in this well).
Figure 8 Production Behaviour on Wells Treated with Sepapar-P
In well R-A3 the solvents injected (Disolin 88 and gas-oil)
diluted the formation oil. As a consequence, the API gravity
increased from 16 to 20ºAPI and the oil viscosity reduced
from 220 to 60 cP. This effect, was the reason for R-A3
having the highest fluid production increment (from 30 to 115
Bls/d). This was the only well that still had a production
increment (after production index > 8), producing at a rate of
45 Bls/d (15 Bls/d increment) (Fig. 8).
The use of the wetting agent Sepapar-P was a successful
method to avoid hidden emulsions formation. Thus, two of the
wells treated: R-A2 and R-A3 respond with a fluid production
rate increment. This production increment was observed until
the wells produce over 3 to 5 times the total volume injected.
Well R-A1 did not receive enough treatment due to the
existence of a thief zone (commingled completion). Then,
8 SPE 82252
production rate did not increment as expected and rapidly
return as before treatment.
Conclusions
The presence of hidden water-in-oil emulsions in the Miocene
oil reservoirs were detected during normal oil production. The
high aromatic content in the Cabimas oil and the small
difference in water / oil density among other effects are the
cause for the stabilized emulsion.
Three wells were selected (R-A1, R-A2 and R-A3) to treat
for hidden emulsions by using a wetting agent into an organic
solvent system. As a result, the use of the wetting agent was a
successful method to destroy existing hidden emulsions in two
of the wells treated: R-A2 and R-A3 where fluid production
rate increased up to three times the rate before treatment. The
exception was well R-A1 due to the existence of a thief zone
that avoids receiving enough fluid.
Another factor that was believed to be reducing the well
production capacity was the presence of organic solids
plugging the Miocene slotted liner-gravel packed completions.
To work out this problem, another two wells (R-B1 and R-B2)
were treated using the same organic solvent system but
without the wetting agent. Those two wells did not show any
change in production rate.
Treatment preparation and pump operation (between 250
and 400 bls/well) for all five wells were completed in a single
week. All these field processes were simple and carried out
with no operational/safety problems.
To ensure success by avoiding cross flow between treated
and untreated zones during future treatments, selected wells
must be completed exclusively on “La Rosa” and/or “Icotea”
formations (avoiding LLs formation) where treatment with the
wetting agent was successful.
Nomenclature
GC = Gas Chromatograph
TDS = Total Disolved Solids
∆ρ = Diferential Density
BS&W = Water and Solid Content
TD = Total Depth
ppm = Parts per Million
References
1. Espinoza, R. and Kleinitz, W.: “Cabimas Field Miocene
Reservoirs Chemical Treatments Final Report,” Preussag
Energie (Feb.2002) 2.
2. Espinoza, R. et al.: “Cabimas Field Reservoirs Integrated
Study and Plan of Development,” Preussag Energie
(Mar. 2001) 24.
3. Cuiec, L.E.: “Evalusation of Reservoir Wettability and its
Effect on Oil Recovery” in Interfacial Phenomena in
Petroleum Recovery, edited by N.R. Morrow, Surfactant
Science Series, Vol. 36, M. Dekker, Inc 1991
4. Jones, T.J.; Neustrader, E.L.; Whittingham, K.P.: “Water
in Crude Oil Emulsion Stability and Emulsion
Destabilization by Chemical Demulsifiers”, J. Can. Pet.
Technol. 1978, p. 100 – 107
5. Pal, R.; Yan, Y.; Masliyah, J.: “Rheology of Emulsions”
in Emulsions - Fundamentals and Applications in the
Petroleum Industry, edited by Schramm L.L; Advances in
Chemistry Series 231; ACS 1992; p. 131 – 170
6. Graham, D.E.; Stockwell, A.; Thompson, D.G.:
“Chemical Demulsification of Produced Crude Oil
Emulsions“ in Chemicals in the Oil Industry, edited by
P.H. Ogden, Royal Society of Chemistry 1983
7. Liebold, G.; Rehberg, W.: “Die Verwendung von
Netzmitteln bei der Entparaffinierung von Erdölförder-
sonden”; Erdoel-Erdgas-Zeitschrift, 1969, p. 190 – 195
8. Abbot, W. et al.: Petroleum Engineering Handbook, SPE,
TX (2000) 56-1.
9. Satter, A., Baldwin, J. and Jespersen, R.: Computer –
Assisted Reservoir Management, Penwell, CO (2000) 41.