SPE 165297

Technical and Financial Evaluation of a Process of Cyclic Steam Injection

Using Horizontal Wells
C. A. Higalgo, Universidad Industrial de Santander; S. F. Muñoz, Universidad Industrial de Santander; L. R.
Oliveros, Universidad Industrial De Santander; C. E. Naranjo, ECOPETROL S.A

Copyright 2013, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Enhanced Oil Recovery Conference held in Kuala Lumpur, Malaysia, 2-4 July 2013.

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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily 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 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 SPE copyright.

Abstract
In order to determine the best operating scheme, and taking into account the increase in demand for oil and the need of
increasing the recovery of existing fields, enhanced recovery techniques are an alternative to the oil industry.

In Colombia, the stratified deposits of heavy oil have become the focus of oil exploration with EOR methods, including the
one which is conceptually evaluated in this paper. However, the production of these fields has been restricted to strata having
considerable thickness, forgetting about certain amounts of hydrocarbons from thin strata, representing significant potential for
increasing production and reserves.

In search of the best alternative for extracting heavy oil reserves are contained in thin strata reservoir, this paper shows the
technical and financial evaluation of applying cyclic steam injection using horizontal wells in these fields, using the injectivity
and the benefit of the location Horizontal wells, and viscosity reductions product with steam heating. To carry out this process,
It was evaluated the effect of each operational and reservoir parameters. Subsequently, due to the uncertainty of the
distribution of steam in the reservoir and the importance of heterogeneities in this effect, the numerical simulation was
performed with homogeneous and heterogeneous models, comparing productive behavior of each stage, to determine the best
model to use in optimizing operational parameters.

Once the best scenario was determined it was performed a sensitivity and process optimization, seeking operating conditions
that showed the best results with cyclic steam injection in horizontal wells, for further financial analysis that would provide a
comprehensive feasibility study .

Introduction
Colombia has an estimated OOIP 42,000 MMBE (million of boe), where about 12,000 MMBO (million barrels) of heavy oil
to correspond OOIP, located primarily in the eastern Basin of Llanos and Middle Magdalena Valley .The deposits have
common country mostly high heterogeneity stratification and, because of this, production is difficult, making very low
recovery factor and losing a large quantity of hydrocarbon recoverable. Therefore, it is important to use new IOR and EOR
techniques to improve the development of hydrocarbon production, and optimize the work done up to date.

IOR and EOR techniques are of great importance to increase reserves and generate extraction alternatives, where the use of
horizontal wells is presented as an IOR technology leaders in the industry, given its versatility and benefits to oil
development.Thus, the use of horizontal wells is a great alternative, as it is not a new technique, and experts say it greatly
increases the oil production rate and recovery factor. Furthermore, its versatility to reach new reserves, where there are
problems such as stratification, clays, fractures and dips, in order heterogeneities, avoiding the use of multiple vertical wells to
acquire a similar response of the reservoir, and reducing operating costs and investment, generates well accepted in the
industry and a large financial benefit in most projects.

Today has been quite helpful to use horizontal wells in thermal recovery in heavy oil reservoirs, due to the existence of a larger
contact area with the reservoir, allowing for better heat distribution and transport of fluids. Cyclic steam stimulation using
2 SPE 165297

horizontal wells depends on reservoir characteristics and the technical design and financial viability, and thus in this project is
to evaluate the technical and financial feasibility of applying the technique in fields with different conditions of heterogeneity
and for thin sands, taking into account the cycles of injection and the amount of steam, to achieve the best results and thus
show the benefits it can bring recovery technique to local and international oil industry.

Cyclic Steam Stimulation using Horizontal wells

One technique that has been developed in order to increase the temperature of the deposit, reduce the damage of the wells, the
oil viscosity decrease and increase the recovery factor is the cyclic steam injection, this technique has shown its applicability
having into account the additional recovery of hydrocarbon surface necessary facilities and the costs of its implementation,
which is a success as enhanced oil recovery method, figure 1.

With the advent of horizontal well technology, the production of heavy oil reservoirs, extra-heavy and bitumen has been
considerably good. Cyclic steam injection using horizontal wells has been a success for the industry. However, numerous
economic and technical factors may limit its commercial use, figure2.

One economic factor that may limit the process is the high cost of drilling a horizontal well compared to the cost of a vertical
one, which generally costs a third of a horizontal well, besides having high viscosities of heavy oil reservoirs and extra heavy,
where the technique is applied generally, represent a challenge for any oil company, as well as control of the steam distribution
in the reservoir as an uneven distribution of steam through the producing interval may cause the failure of the implementation .

Therefore, it is important to assess the feasibility of implementing this technique, taking into account the benefits and losses
generated by its implementation, taking as a starting point already made applications, which can be shown screening in Table
1.

Reservoir description
The simulation conceptual model was built considering average properties of Teca field, which is a Colombian Middle
Magdalena field located 160 kilometers northwest of Bogota and with an area of 2788 acres. The stratigraphy shows a
basement of igneous and metamorphic rocks of Cretaceous age, covered by a sedimentary sequence of Tertiary age not
consolidate, composed of fine grains, conglomerates, claystones, river sand, interspersed with shales.The sedimentary section
has a thickness ranging between 2000 and 3000 feet.

The site has two producing sands, the sand that is located at a depth between 1400 and 1800 feet, with a gross thickness of
approximately 390 feet, and has an average porosity of 29% and a maximum permeability of 1080 mD, and B sand which lies
in the range of 1920 to 2050 feet, with a thickness of approximately 130 feet, have a porosity of 28% and a maximum
permeability of 780 mD.

CONCEPTUAL MODEL SIMULATION

Conceptual models allow us to evaluate reservoir performance when subjected to different patterns of exploitation, such as
EOR techniques, regardless of the production history of the field and therefore the historical setting of certain variables. For
construction of the model are necessary as certain properties of the grid dimensions, petrophysical properties, PVT and
mechanical condition of the wells, so that these properties are considered as input and are based, for the purposes of this
model, on average data Teak field, others in the simulator data by default and the remaining production of the author.
Simulation grid dimensions.

The first step in the construction of the simulation model is to establish the simulation grid, which will implement the process
with their respective dimensions, and enter data as initial reservoir depth, thickness, reservoir pressure, among others, and data
of the rock as compressibility and thermal properties. Model building was done with a Cartesian meshing, which allows us to
model fluid flow satisfactorily and time savings in computation. The cells in the grid are: 40i 40j 8k, with dimensions of 50
feet and j, k and 30 feet to the first layer of clay (overburden), 2 feet for each of the six cells that represent the thickness and
net producer 30 feet to the lower clay layer (underburden), generating a normal area of 91.8 acres to implement the 4 wells, the
suggestion of the thesis director and other thesis in enhanced Oil recovery.The depth of the top layer is 1630 feet producer,
taking into account the average depth of the zone A teak field. The producer thickness is 12 feet, to a depth at the bottom of the
production area of 1642 feet. Figure 3 shows the basic simulation model was built in CMG Builder.

Table 2 shows the initial conditions were taken and modified from Teca field to carry out the process, modeling the behavior
typical of this deposit on their terms and / or features.

Rock properties
The rock properties are very important when choosing an operating scheme, due to certain conditions and rock properties that
favor one or another application of enhanced recovery, as shown in Table 3.
SPE 165297 3

Fluid properties it is important to outline that for generating the PVT model it was important to have a characterization of
existing models in the simulator, because many of them are not properly modeled PVT behavior of a heavy crude.After a
literature review on the matter, it was found that the equation of Beal and chew is what best fits the teak fluid model, since
applied for viscosities> 2000 cP (heavy oil) API in a range of 10 <API <20. From equations loaded into the simulator, the Beal
and Chew was the only one fit for heavy oil, so, despite being quite sensitive to the change of API, was chosen as the best at
what we were looking for initial conditions of the fluid.With the model Chew achieved Beal and adjust the viscosity of the
fluid with the field representative values for varying teak to 10.8 ° API as the best fit.

Other properties obtained with the synthetic PVT relationship as the gas - oil formation volume factor of the oil and oil density
were calculated from correlations used.

Another very important property is the reservoir fluid viscosity, as already mentioned heavy oil is used, by which is meant a
value greater than 2000 cP at reservoir conditions, being in the case of this simulation of 4050 cP @ 105 º F. The oil viscosity
is a property that is severely affected by temperature changes as the reservoir and the gas viscosity, which is generally
combined with the oil and reduce the overall viscosity of the fluid, the water has no significant influence on the viscosity oil.

Rock-fluid interaction.Other important properties are the result of fluid-rock interaction, as it determines the movement of
fluids in the reservoir; these data are the basis of numerical models for predicting performance of any EOR process.

Wells In this section it is portant to consider operational parameters as well spacing, their location, length horizontal wells,
flowing bottomwell pressure and other influential parameters. For the project were built 4 horizontal wells located in an area
of 91.8 acres, with spacing of 500 feet between wells and a horizontal length of 1575 feet. Moreover, the wells were placed on
half the thickness net producer seeking to create a steam chamber in the stratum ellipsoidal type producer, as seen in Figure 4
below. The spacing between wells was selected considering heated radio calculations for each well, where the behavior is
assumed for the steam distribution ellipsoid shaped because horizontal wells are being worked when thin.

In total 8 wells were implemented in the simulator, 4 producers and injectors 4, because the simulator does not recognize the
fact that produce and inject wells simultaneously. For this purpose, the 4 wells were assigned producers operating conditions,
where necessary sensitivity analysis to flowing bottomwell pressure (BHP), because the default data (28 psi) showed abnormal
behavior production, and low BHP carries a rapid depletion of the reservoir, and thus a high pressure drop.

Showed sensitivity to low BHP, in this case of 100 psi, you get the highest recovery factor and fluids are produced faster, so
that the recovery is accelerating and profits of the project are evident in less time. However, the sensitivity was shown also to a
lesser BHP, the reservoir quickly loses energy, whereby the length of the reservoir is much lower than with BHP's higher.

When performing sensitivity analysis BHP concluded that BHP was better than 200 psi, in consensus with the project
manager, because the recovery factor is high and the site is produced in less time. Further, although the reservoir pressure has
a substantial decrease, this trend of pressure drop is normal and therefore can carry out the field development.The operational
data of the wells shown in Table 5.

Similarly, the sensitivity in the BHP, reduction was performed at the initial injection pressure of 900 psi to 1400 psi, because
the initial pressure was very high for a normal process was injecting possibly inadequately above fracture pressure. Moreover,
it was found by some simulation runs for injection pressures of 1400 psi, 1200 psi, 900 psi and 800 psi the recovery factor is
the same, and therefore this parameter seems not to have much influence on the process development.

By looking at the parameters of the wells, it is important that the injection rate data at 800 bbl / day was selected considering
the deliverability of a generator, which in reality is 40 MMBTU / hour, so that a very high rate exceed the generation capacity
and therefore would require a second generator, which is economically unfeasible.

PRELIMINARY ASSESSMENT
For the simulation base case of the process of cyclic steam injection using horizontal wells, we obtained an additional recovery
factor of 1.4% compared to cold production, where it is important to highlight the contribution in producing wells horizontal,
due to contact with the reservoir and the high injectivity thereof. It is also important to note that for the cold production or field
sector had an Np of 457,799 bbls oil of about 1'983 to 312.397 bbls OOIP for a recovery factor of approximately 22.2%, while
for the hot production cyclic steam injection was obtained Np of 490,155 bbls oil, which shows an incremental production of
29,074 bbls injection in three cycles, for a recovery factor of approximately 23.6%.

As mentioned above, is carried out a cold production of 8 years trying to reproduce the normal behavior teak field, obtaining
itself a nearby reservoir pressure at 570 psi at the beginning of stimulation with steam.
4 SPE 165297

Injection scheme
In analyzing the development of the process and behaviors that generate production results, it was decided to implement an
alternate steam injection where two schemes were tested mainly the first injecting a well at a time and the second two wells
injecting at once.

The decision was made to stimulate two wells simultaneously, to get the benefit of the injectivity of horizontal wells and the
delivery capacity of the generator, to avoid some preferential flow and reduce operational costs wells, looking for an increase
in production of hydrocarbons and improved operational performance.The order to be used for stimulation of the 4 horizontal
wells is shown in Figure 5, and is done so in order to better exploit the benefits of heating the reservoir and the heating effect
of a well in another .

When deploying the steam injection is alternately possible to increase the recovery factor with respect to the cold production in
1.42% (Figure 6), noting that the change from the model which are injected simultaneously in all wells is very minimal, as are
fixed all operating conditions. Also be seen (Figure 7) that the rate of oil is maintained with peak production or field sector
differently, due to the difference in the times of injection - production, from the model with simultaneous injection into all
wells. When performing the steam injection is alternately pressure also undergoes significant change because cycles show a
higher pressure maintaining the sector, although the third cycle the lost energy reservoir when the cycle ends injection (Figure
8).

Figure 8 shows a drop in pressure below the normal fall in the third steam cycle, regarding the cold production, since the effect
of the steam is rapidly depleting the reservoir and the cycle is not generating stimulation Good maintenance and recovery of
hydrocarbons. Moreover, alternately inject can observe an increase in hydrocarbon production and thus an increased
production of water, as shown in Figures 9 and 10, which increases treatment costs and surface facilities in general.

The temperature behavior of the radius can see the influence of heating and a well in another, for this model was observed
almost no influence of the wells from each other, as seen in Figure 11 and 12. However, little influence between wells is that it
is difficult to appropriately represent the steam distribution behavior, because they require very robust refinements with the
cell capacity and the color scale permitted by the license is not possible.

In the saturation distribution can be observed that when the well is pumping, decreasing oil saturations near the well because
they are saturated with water, and when there is increased producing because fluids are routed drainage area the well (Figure
13). It is also noted that the effect of steam dramatically not exhaust the reservoir, since there are high oil saturations
postdeployment steam.

HETEROGENEOUS MODELS, CONTAINMENT AND COMPARISON WITH VERTICAL WELLS

Permeability is one of the most important parameters in the development of a steam injection process, taking into account that
this is attributed to the reservoir allowing fluid flow both injection and production. Therefore, the evaluation of the effect of
permeability is of great importance to choose the best scheme exploiting the deposit agreement.

To evaluate the effect of the permeability in the development process took into account the variation of the permeability in the
shaft direction and in the direction K, as objective functions for the recovery factor, the temperature distribution and
saturations.

CONFINED MODEL
Confinement is a problem inherent in the construction of numerical simulation models, because when you create a grid
simulation to represent an area of the site, is not taken into account the effect that causes the remaining stretch of the same
sector.

Among the effects that are not adequately represented, you can highlight the pressure drop due to production and supply of
fluids in those areas surrounding the sector. To solve this problem, we built a ring of 84 cells with high porosity properties that
simulate a large amount of fluid in the surrounding areas, and thus will cause a pressure maintenance.

To select the ring is known as confinement ring not certain tests were done where cells were included above, increasing the
area of the reservoir to 101.24 acres. In these cells were worked porosities of 0.8 and 0.6 in fraction, presenting the problem
that the OOIP was increased due to the large amount of fluids. Therefore, it exhibited a recovery factor decreased as a result of
which the drainage area of the wells did not vary.

In order to solve the drawback of increasing the model OOIP divided into two sectors that have not confinement ring and the
initial sector simulated field, as seen in Figure 17. Furthermore, there were two cases of sensitivity to the confinement with
porosities in fraction 0.8 and 0.6, for which test results cast model more convenient holding reservoir permeability constant
SPE 165297 5

(800 mD).

Observing the results of pressure drop and recovery factor of oil production rate, it can be seen that the confinement
accomplishes the goal of maintaining the reservoir pressure and increase oil production, is equally important to note that the
best model is the porosity fraction equal to 0.8, for maintaining reservoir pressure and steam cycles.

For the analysis of confinement is of great importance to create some sectors within the same sector that defines a portion of
reservoir, because the recovery factor would be representative to make a comparison with a model without a ring, as it
increases the original fluid volume in the simulator considers site, and therefore reduces the recovery factor of the field
represented.

MODEL WITH VERTICAL WELLS

Considering the results obtained with the base model and adjustments to the confinement, is appreciation for the third cycle
steam stimulation horizontal wells had problems, because the recovery factor was low and the pressure drop accelerated.

In order to evaluate the contribution of horizontal wells compared to vertical, the simulation was carried out of the area or
sector used for the four horizontal wells with 9 vertical wells.For simulation with vertical wells was considered as a design
criterion that each well drains an area of about 10 acres, being in the drainage area of 23 acres each horizontal well.Therefore,
the sector having an area of 91.8 acres 9 vertical wells were implemented, as this number of wells drain an area of 90 acres,
which means almost the entire length of the model (Figure 17).

For the injection of steam into the model is performed with vertical wells alternately injected in a star, so as to uniformly heat
the reservoir, so as to ensure that the next well to be stimulated this as far as possible above the well stimulated.In the proposal
steam injection was started by stimulating the well 3, then at 8.1, 6, 7, 2, 9, 4, and finally the shaft 5, as shown in Figure 18.

By modeling the field sector with vertical wells, it is appreciated that the vertical wells injectivity problems arise in this type
thin thicknesses (12 feet) and therefore several tests were performed to define the most appropriate injection rate (figure 19 ).

As seen in Figure 19, the injectivity problem occurs in the first injection cycle, therefore injectivity analysis focused mainly on
what the reservoir taking in this cycle, in order to have the same design for subsequent cycles and maintain stable all
parameters. Looking at the problem of injectivity was very important to know how was injecting each well, as we would
expect the biggest problems would arise in the first stimulated wells hence analysis was performed of the injectivity of each
well in their first cycle bbLeq with 250 / day and 150 bbleq / day of steam (Figure 20).

Looking at the results of Figure 20, shows that at an injection pressure of 1300 psi and a flow rate of injection of 250 barrels
equivalent, the reservoir does not have sufficient injectivity in any of the wells, to take the injected steam correctly, so it is
necessary to reduce the injection rate in order to achieve that all the steam injection well into the formation.

To achieve the injectivity into all wells of the model with vertical wells, after an analysis of the required injection rate, it was
found that the deposit allowed 150 bbl / day of steam injected per well (Figure 21), by what the designs suitable for injection
with vertical and horizontal wells with the operational specifications for the process are shown in Table 6.

The operational parameters used for the model with vertical wells and horizontal wells model were selected considering that
when compared they should inject the same energy to the reservoir, which explains why the injection time and injection rate
are different in wells vertical and horizontal. In figure 22 you can see the behavior of the recovery factor for the model with
vertical and horizontal wells.

Clearly, in Figure 22, it is evident that horizontal wells produce a more effective hydrocarbon recovery than vertical. However,
it also shows that the vertical wells have a better productive response to stimulation with steam, while with vertical because the
increased recovery factor of approximately 1.3% over the cold production in the horizontal magnification is approximately 0.5
% (Figure 23).

In conclusion, horizontal wells have technically improved production response for thin gauge fields. Undoubtedly these wells
to better drain the reservoir, and thus allow better hydrocarbon production. In addition, vertical wells have a great disadvantage
to the horizontal, because the injectivity and production have a low value because of limited contact with the reservoir.

OPTIMIZATION OF INFLUENCING PARAMETERS IN THE CYCLING STEAM STIMULATION USING

HORIZONTAL WELLS
The optimization can be defined as the analysis of influential parameters performed in one process, in order to determine the
best combinations that generate the reservoir response, given a objective functions. To perform an optimization is advisable to
6 SPE 165297

first conduct a sensitivity analysis to determine the influence of each variable in the process response.

Sensitivity to horizontal well length

Before performing sensitivity analysis and optimization is very important to verify how significant is the effect of the
horizontal length of the well, such verification was performed manually sensitivity where we used three different well lengths,
base length lengths of 1000 feet and 600 feet in the horizontal, as shown in Figure 24 below.

In Figure 24 it is observed the effect of horizontal wells in the response field production, being the case that best presents the
hydrocarbon recovery from the well with horizontal length of 1575 feet, confirming what is stated in the literature review on
the effect long lengths positive horizontal wellbore.

SENSITIVITY ANALYSIS CMOST AIDED AND OPERATIONAL PARAMETERS OF DEPOSIT

Sensitivity analysis to quantify and visualize the influence of certain variables in a process over an objective function. These
functions are generally objective recovery factor, cumulative oil production, cumulative water production, injected steam-oil
ratio produced, among others.

To perform this sensitivity analysis tool CMOST should be aware that it can be done in different ways, the first is to vary one
parameter while the others remain constant, this is called mono-sensitivity parameter, the second is evaluate the effect of some
parameters on one or more response variables, which according to the literature is that which is more efficient for the analysis,
since the first sensitivity does not investigate all possible combinations of the parameters that are part of the study, generating
some useful results for engineering purposes.

Analysis of sensitivity to operating parameters

Given the range of variation of each parameter, we proceeded to define the sampling method of "Latin Hypercube" generating
120 combinations to sensitize the parameters, and thus covering a large number of possible combinations that show a positive
and negative effect on the development simulation model.Figure 25.

Subsequently defined objective function, which in the case of this simulation were the recovery factor, production of oil and
accumulated water and the SOR, commonly used as response variables to analyze such processes (Figure 25). In Figure 26 we
can see the methodology implemented for performing a sensitivity analysis, which takes into account the general parts of the
development of an optimization.

Results of the sensitivity analysis to operational parameters for the analysis of the sensitivity results in the diagram is generally
used tornado, these are horizontal bar charts, which shows a sequence of estimates of the effect of all parameters on each
objective function in order upward when their values are changed between a maximum and a minimum.

As mentioned above, for sensitivity analysis parameters were chosen as shown in Table 7, in order to see the influence of the
cyclic process steam injection using horizontal wells. Before going into detail is important to note that results vary according
to selected ranges, and properties of both rock and reservoir fluid. That is, the answer as to the influence of each variable
depends on the difference between the values used and the response of the reservoir to the process.

As mentioned above, construction of sensitivity model CMOST, 120 runs were performed where sensitivity was observed for
the recovery factor, steam-oil ratio and cumulative water production, certain combinations show better results than the base
case because certain parameters which influence the process are highlighted with favorable values. Figures 27, 28 and 29 show
the percentages of influence of each sensitized parameters are objective functions; recovery factor, steam-oil ratio and
cumulative water production.

By observing the influence of the parameters on the recovery factor can be seen that the best case yields a recovery factor of
27,334% and 13,441% worse at the end of stimulation, including the recovery of oil in primary production, also it is evident
that the operational parameter is more influential flowing bottom hole pressure, as this directly affects parameter reservoir
depletion and hence the production of fluids, there is also the second most influential parameter is the injection time, and that
the longer the greater to the stimulated area and hence the production of fluids.

As the least influential parameter, the soak time does not significantly affect the recovery factor, but it is noteworthy that the
positive influence indicates a good pressure maintenance when this time is greater (Figure 27).

In analyzing the steam-oil ratio is evidence that steam-oil relations maximum and minimum are low, and are in the range for
viable projects of this type, where the most influential parameters are the injection time, number of cycles and rate injection
increases because the injected steam into the reservoir and hence the condensed water produced, and the effect of the steam
heating the reservoir shrinking production in response to pass cycles (Figure 28).
SPE 165297 7

With the cumulated production of water is observed with the greatest influence is the injection time, because the volume of
steam injected is increased and hence the condensed water produced is higher. However, it is also important to note that the
number of cycles, the injection rate and the flowing bottom hole pressure have a significant influence, where the first two
involve injection of steam that condenses over into the reservoir, and the third affects the amount of water produced due to the
pressure differential of the producing well (Figure 29).

Analysis of sensitivity to reservoir parameters

The reservoir parameters are important as they determine and limit the response of the reservoir both in cold production as for
implementations of any enhanced oil recovery method. We performed sensitivity analysis to determine the influence of certain
parameters of reservoir production response, but only seeks to determine the influence of the parameters on the response of the
reservoir, since for the present work conditions were maintained typical of a heavy oil field in Colombia.

To perform the analysis tool was used CMOST, CMG simulator, using the same methodology used for the sensitivity analysis
of operational parameters.

Sensitized parameters are:

 Sand Thickness
 Thickness clay
 Initial reservoir temperature
 Initial reservoir pressure
 Horizontal permeability
 Vertical permeability

When different combinations of reservoir properties found two extreme scenarios, where the pessimist shows a recovery factor
of 13,878% and 55,103% optimistic (Figure 99), the results make clear that the most influential parameters are temperature
and reservoir pressure, which results in the energy quantity of the deposit and therefore the production thereof answer, so at
higher temperatures and pressures, there will most favorable one production site.

It is also important parameters such as the thickness and horizontal permeability, as mentioned in the literature review, because
if you have the answer thinner gauges in the recovery factor with the technique will improve, but this does not mean a greater
amount of hydrocarbons produced.

The horizontal permeability is also very important since it is necessary a good distribution of the steam horizontally, taking
into account that the steam quickly reaches the vertical limits when injected into thin sands.

OPERATIONAL PARAMETERS OPTIMIZATION

For the numerical simulation of deposits, when required an optimization process, we have developed several programs that are
designed to predict reservoir behavior through various development schemes. CMOST, is an example of this type of program
that performs optimization tasks, applying statistics to perform procedures more efficiently and thus make better decisions.

Methodology used in the optimization.The general methodology used to perform optimization CMOST is presented below in
Figure 30, the methodology consists of six main stages, and was used for the selection of process development stage of cyclic
steam injection using horizontal wells in the conceptual model, represented by numerical simulation.

Determination of the base case and selection of the most influential parameters. most influential parameters in steam injection
process were established taking into account the sensitivity to operational parameters, production of the author and suggestions
of the thesis director.The parameters to be optimized are shown in Table 10 with their respective range of variation.

For steam injection rate took into account the specifications of the steam generator capacity in terms of delivery, the injection
pressure was determined considering that no problems will arise in the reservoir injectivity in the number of cycles used up to
4 cycles because it had strong pressure drops after this last cycle, and the soaking time and production economy and oil
deliverability of a cycle.

Objective Functions
To set the selling price of crude oil, was necessary to take into account the price of opportunity, the lifting cost, royalties and
taxes generated by oil production (Table 8). The table below shows the values required by the simulator to calculate the net
present value (NPV) (Table 9).
8 SPE 165297

Analysis of results
The optimization results are presented in tables CMOST removal parameters, a graph that shows the spread of all runs made
with respect to net present value and histograms for each parameter. The algorithm eliminates DECE parameters not
performing well and shown in red, indicating that generate a significant maximizing the NPV and therefore are not optimal for
deployment (Table 11).

The NPV scatterplot (Figure 30), presented in the ordinate axis the objective function, and the abscissa the number of runs
made by the simulator, which are represented as a series of dots.The black dot indicates the NPV for the base run, the red dots
represent the best combinations of operating parameters and blue color combinations that are not so convenient.As shown in
Figure 30, there runs the VPN registered over the base case, showing a process optimization.Importantly VPN often show a
positive and much higher than the base case, so that at this point of the implementation is attractive project steam injection in
horizontal wells

On the other hand, histograms describe the frequency with which the simulator used parameter value of each operation during
the simulation runs.As seen in Figure 31-a, 3 cycles were the most used by the simulator, thus suggesting that this variable is
the one with better effect on cyclic process steam injection in horizontal wells, the same effect can be for each parameter
determined from the analysis raised.

Analyzing the results of the optimization runs, it was decided that it is necessary to implement 3 to 4 cycles of steam injection,
as best benefits are generated by the incremental production and maintaining the reservoir with a normal pressure drop.

Regarding the injection pressure and rate, the best cases was for an injection pressure of 900 psi and a rate of 800 bbl / day per
well, because these parameters match perfectly, and higher rates if injecting the results will generate future economic benefits,
but instead will add costs, which can lead to failure of the investment project.

The injection times, and soaking production yield expected results considering the literature review, we can see that the best
injection time is 5 days, because it is dealing with a thin reservoir that does not require a great amount of energy to produce a
good response, the most frequent soaking time was 3 days, because it is sufficient for the transmission of energy to the
reservoir without affecting project economics, and the most appropriate production time is 300 days because it takes a good
use of the steam cycle.

FINANCIAL ANALYSIS
To carry out an enhanced recovery project is necessary to evaluate its feasibility by different criteria. For practical purposes in
a project of this type (HCSS) use two points of view, a technical and financial evaluation. The technical assessment takes into
account the study of the site and the equipment necessary for the implementation of the process, and the financial evaluation
determines whether the project's benefits worth the investment required to achieve them.

The oil industry has acquired the ability to drill horizontal wells at low cost and with less uncertainty in the positioning of
these, reducing operational risk and negative cash flows. The use of horizontal wells as a new production technique requires a
financial assessment, the main objective incremental production and increased reserves.

To make the financial evaluation of an investment project should be defined evaluation criteria, which are chosen taking into
account the impact generated in the evaluation of the feasibility of this, which for purposes of this research were worked
mainly three criteria are, the net present value, internal rate of return and payback time.

Considering the benefits and importance showing in the project financial evaluation, we decided to perform the analysis on the
STEAM software tool, created by the enhanced recovery group (GRM-UIS).
INITIAL INVESTMENT
The initial investment includes all costs that are generated in the initial stage of the project, it includes all costs for
construction and project development. For this project it was considered in the initial investment, the cost of the drilling of
four horizontal wells, and injection production lines and civil works. The project took the steam generator already existed,
because in the Teca field, where average data were taken for the project had already been implemented steam. It was also
thought that all computers were available and installed treatment. For the initial investment, as mentioned, we take into
account all expenses at the beginning of the project, as shown in Table 12.

Incomes
Income refers to the profit gained in the productive time of a project, generally associated with the production, so this project
is important to consider the incremental oil with steam injection with respect to the horizontal well production in cold, because
technically the largest contribution to production is by the use of horizontal wells.
SPE 165297 9

Consider whether injecting steam or not depends on the financial viability, for this case of evaluation, a standard price of oil by
56 U.S. $ / bbl.Therefore, in Table 13 shows the optimum case production cycles 3 and 4 with respect to the cold production
and revenue generated by each case.

In cyclic steam injection, there are three major groups representing expenditures, these are: operation, maintenance and taxes.
Within the operating costs include the generation water treatment and produced fluids, power consumption and natural gas.
The money for maintenance, designed to take optimum conditions the equipment and tools used in the operation. As for taxes,
are considered royalties and income tax, which is earmarked for the nation which in turn distributes them according to the
regulations. Table 14 shows the indicators used to calculate expenses on the project.

CASH FLOW
For financial analysis, is necessary cash flow, which accounted for the initial investment, income and expenses of the project,
giving an estimate of the cost of this. For a steam injection project should be counted as income in oil production, and as the
initial investment expenses, costs of treating water generation, produced fluids and other operating costs, maintenance costs
and taxes. As mentioned above, financial analysis was performed for the cold production, injection cycles 3 and 4 injection
cycles, with horizontal wells, so that in Figure 33 shows the cash flow for the three cases.

FINANCIAL EVALUATION RESULTS

Already known initial investment that takes into account, for this type of EOR process, the costs associated with drilling the
well and all the necessary elements for steam generation, infrastructure and civil works, and also known cash flows the project
can generate, such as the income generated from the sale of oil produced, and expenses, that deal broadly with treatment costs,
fuel and energy consumption, operation personnel and taxes have been can perform financial calculations to determine the
feasibility of implementation.

Among the methods for financial analysis of investment can be highlighted as the most used in the oil industry the following:
Net present value (NPV), which is a procedure to calculate the present value of a number of cash flows future, taking into
account an interest rate equal to or around the time considered, the rates of return (internal rate of return (IRR) and average
rate of return (TPR)), for determining the percentage of profit or gain received from the investment of capital, either before or
after taxes, and that work by trial and error considering present value equations, and the recovery time of the investment or
PAYBACK TIME (simple payback time of repayment adjusted), which is the time at which recovers the money initially
invested in the project, ie for positive net cash flows equal to the initial investment.These methods are used together, in order
to obtain a right decision in the project, and evaluate the alternative investment successful.

When performing financial analysis took into account the price of standard crude and treatment costs, used throughout the
project (56 USD / bbl). For financial analysis financial criteria were used as the NPV, IRR, TPR Simple PAYBACK and
adjusted PAYBACK; looking to have as much information to determine whether the project is financially feasible.

Table 15 shows the results of the financial analysis for each criterion for all three scenarios, which shows that the faster
scenario is the investment pays four injection cycles, since all three scenarios have the same time simulation (4015 days),
however the difference is not very far to cold production, which can generate risk and uncertainty in the implementation of
steam.

Looking at the financial criteria of Table 15 and the graphs generated in the STEAM software, a great similarity is evident in
the results of each of the criteria, so for the VPN and set PAYBACK only considered graphs show the production Cold and 4
injection cycles (figure 37 and 38).

In Figure 37, we see that the project is very good, because it pays high dividends quickly and generates also enables high
interest rates (i = 39%) proving beneficial for any investor.Figure 38 shows a similar behavior of VPN and set about
PAYBACK cold production, confirming its proposals IRR analysis, TPR and on benefit Simple PAYBACK vapor.

CONCLUSIONS
Analyzing graphics of recovery factor and reservoir pressure drop, it was observed that the third cycle onwards the process of
cyclic steam injection with horizontal wells becomes uneconomical, since there is not a good incremental production and
reservoir is depressurized drastically, below the normal behavior in cold production.

Noting the results of heterogeneous models posed, it is appreciated that variations of permeability parallel to the direction of
the well show a better response to steam cycles compared to homogeneous model and thus the wells must be located in the
direction of the greatest variation permeabilities.
10 SPE 165297

By sensitivity analysis and optimization, it was observed that although the high injectivity allow horizontal wells, cyclic steam
injection into reservoirs thin reservoirs must be done with low rates and for short periods, since the area to stimulate presents
losses large heat into upper and lower strata.

On the other hand, the financial analysis of the project allowed assessment cyclic steam injection with horizontal wells as
profitable. However, the benefit of steam in the hydrocarbon production, it does not generate significantly higher cash flows,
for cold production, so that it can be risky to implement the process.Con el fin de determinar el mejor esquema de explotación,
teniendo en cuenta el aumento de la demanda de hidrocarburos y la necesidad de aumentar el recobro de los campos existentes,
las técnicas de recobro mejorado son una alternativa para la industria petrolera.

En Colombia, los yacimientos estratificados de crudo pesado se han convertido en el foco de explotación petrolera con
métodos EOR, incluido el evaluado conceptualmente en este trabajo. Sin embargo, la producción de estos campos se ha
limitado a estratos que presentan espesores considerables, olvidando ciertas cantidades de hidrocarburos de yacimientos
delgados, que representan un potencial importante para el aumento de producción y reservas.

En búsqueda de la mejor alternativa de extracción de estas reservas de crudo pesado contenidas en espesores delgados, el
presente trabajo muestra la evaluación técnica y financiera de aplicar inyección cíclica de vapor usando pozos horizontales en
estos yacimientos, aprovechando la inyectividad y el beneficio de la ubicación de los pozos horizontales, y la reducción de
viscosidad producto del calentamiento con vapor. Para llevar a cabo este proceso, se evaluó el efecto de cada parámetro
operacional y de yacimiento. Posteriormente, debido a la incertidumbre de la distribución de vapor en el yacimiento y la
importancia de las heterogeneidades en este efecto, se realizó la simulación numérica del proceso con un modelo homogéneo y
modelos heterogéneos, comparando los comportamientos productivos de cada escenario, para determinar el mejor modelo a
usar en la optimización de parámetros operacionales.

Una vez determinado el mejor escenario de evaluación, se realizó la sensibilidad y optimización del proceso, buscando las
condiciones de explotación que mostraran los mejores resultados con la inyección cíclica de vapor en pozos horizontales, para
un posterior análisis financiero que brindara un completo estudio de factibilidad.

REFERENCES
Yessica Mateus, “Evaluación técnica y económica del uso de vapor para acelerar el factor de recobro en yacimientos estratificados de crudo
pesado”.

Colmenares K. Martínez E, “Análisis de riesgo e incertidumbre aplicados a la evaluación económica de proyectos de inyección de vapor”
Proyecto de grado, Universidad Industrial de Santander, 2008.

Gonzales, K., Bashbush, J.L., Rincon, A.: “Simulation Study of Steamflooding with Horizontal Producers Using PEBI Grids,” paper SPE
121488. 2009Energy Information Administration (2001) "Venezuela Offers Full Market Value to Encourage Foreign Investment in Oil.

Saavedra, Néstor Fernando, ICP. Retos de la Industria de Hidrocarburos en Colombia. En: Conferencia SPE-UIS Student Chapter (2011:
Bucaramanga).

Mendoza Humberto, Horizontal well steam stimulation: A Pilot Test in Western Venezuela, PDVSA-Maraven, Caracas, Venezuela. SPE
129-1998.

E. Escobar, P. Valko, W.J. Lee,M.G. Rodríguez. Optimization methodology for cyclic steam injection with horizontal wells.PDVSA
INTEVEP &texas A&M. SPE 65525.

C.F. GATES and B.G. HOLMES. Mobil Oil Corporation. “Thermal well completions and operations”.SPE 07v03p0419.
C.T. Doan & S.M. Farouq Ali. “Horizontal wells-are they worth the world?”, University of Alberta (Edmonton). SPE 30154.

Oscar J. Vargas Ordoñez. “revisión de los parámetros del empaquetamiento con grava y planteamiento para mejorar su eficiencia en los
pozos del campo Jazmin”. Universidad Industrial de Santander, UIS. Bucaramanga 2006.

Mannucci, J. E.: “Recobro Adicional de Petróleo por Métodos no Convencionales”. 1990. Capítulo 3.p. 14.

Computer modeling group, CMG, Computer Assisted History Matching, Optimization and Uncertainty Assessment Tool (CMOST).

Trevor Bennison. “Prediction of heavy oil viscosity”. AEA technology plc, presented at IBC heavy oil development conference, London 2- 4
December 2008.

G.C. Thakur. “Horizontal Well Technology – a key to improving reserves” Chevron Petroleum Techonology Company. PETSOC -99-10-05-
P.
SPE 165297 11

Table 1. Screening
SCREENING DE LA TÉCNICA HCSS
Porosity (%) > 20 Table 5. Operational parameters cycles
verticals horizontals
Kv/Kh (fracción) > 0.4
Injection pressure (psi) 1300 1300
So (%) > 55 Injection rate (bbl/dia) 150 1012
K (darcys) >1 Injection (dias) 15 5
Depth(ft) < 3000 Soaking (días) 4 4
Thickness (ft) 10 -100 Production (días) 200 200

Viscosity (Cp) > 300

Table 2. Initial condition

Tabla 6. Sensitive parameters
Property Valor
Temperature (ºF) 105 Sensitive parameters Mín. Base value Máx.
So (%) 80 Pwf (psi) 100 200 500
Injection pressure (psi) 700 900 1300
Sw (%) 20
# cycles 1 3 4
Presion @ 1600 ft (psi) 890
Injection (days) 5 15 20
Depth (pies) 1630
Net pay (pies) Soaking (days) 1 4 6
12
Production (days) 100 200 400
Steam quality (fraction) 0.5 0.65 0.8
Table 3. Rock properties Temperature- steam (ºF) 480 530 550
Propiedad Valor
Rate injection (bbl/day) 600 800 1200
Horizontal permeability (mD) 800
Vertical permeability (mD) 800 Skin factor 0 0 3
Porosity (%) 29 Horizontal length (ft) 600 1575 1575
Termal conductivity (Btu/(pies*day*F)) 24
-1 -4
Rock compressibility (psi ) 5.0 E
Sand volumetric heat capacity (Btu/(pies3*F)) 30
Shale volumetric heat capacity (Btu/(pies3*F)) 40
Table 7. Ramge to variation of parameters
# cycles 1 2 3 4

Table 3. Fluid Properties Injection pressure (psi) 700 900 1100 1300
Propiedad Valor
Viscosity @ 105 ºF 4050 Injection rate (bbl/día) 600 800 1000 1200
API 10.8
Molecular weigth (lb/lb mol) 600 Injection (días) 5 10 15 20
3
Oil density (lb/pies ) 60.988
Soaking (días) 1 2 3 4
Water
Standar conditions
Production (días) 100 200 300 400
12 SPE 165297

Table 124. Initial investment

Descripción Costos (US$)
Table 8. Selling Price determination
Wells 5.665.200
Valou (USD) (%)
Steam generation 0
Price 56
Distribution lines 21.500
Royalties 3,36 6%
Wáter treatment equipment injection 0
Lifting cost 4,77
Wáter treatment equipment production 0
Bruto 47,87
Civil works 70.000
Duty 16,75 35%
Total 5.756.700
NETO 32

Table 13. Income production

Table 9. VPN in CMOST Time Primary Income 3 Income 4 income
Costs Valor 42 (años) production (US$) cycles (US$) cycles (US$)
Wáter treatment (us/bw) 0,5 1 128.896 7.218.204 128.896 7.218.204 128.896 7.218.204
Steam generation ( us$/bew) 6 2 62.664 3.509.240 62.664 3.509.237 62.664 3.509.240
Annual discount rate (%) 10 3 48.787 2.732.093 48.787 2.732.093 48.787 2.732.093
4 41.876 2.345.062 41.876 2.345.062 41.876 2.345.062
Table 10. Optimal results matrix. 5 37.124 2.078.995 37.124 2.078.995 37.124 2.078.995
# cycles 1 2 3 4 6 33.549 1.878.748 33.549 1.878.748 33.549 1.878.748
Injection pressure (psi) 700 900 1100 1300 7 30.894 1.730.085 30.894 1.730.085 30.894 1.730.085
8 28.618 1.602.659 28.618 1.602.659 28.618 1.602.659
Injection rate (bbl/día) 600 800 1000 1200 9 26.881 1.505.378 35.939 2.012.595 35.939 2.012.595
Injection (días) 5 10 15 20 10 25.232 1.412.999 28.457 1.593.597 28.457 1.593.597
Soaking (días) 1 2 3 4 11 23.605 1.321.913 22.528 1.261.605 26.418 1.479.445
Production (días) 100 200 300 400 12 22.138 1.239.735 22.831 1.278.543 21.782 1.219.827

Table 14. Financial indicators

Table 11. Costs of drilling horizontal wells Descripción Costo
Descripción cantidad US$/cantidad Total (US$) Cost of gas 4 US$/MMBTU
Location 1 40.000 40.000
Wáter treatment equipment injection 6 US$/bbl
Ring rental (days) 20 30.000 500.000
Fuels and lubricants 20 480 9.600 Wáter treatment equipment production 0,5 US$/bbl
Mud drilling and additives 1 200.000 200.000 Duty 34%
Logs 1 16.000 16.000 Royalties 6%
Cementation 1 76.000 76.000
casing 1 100.000 100.000
Table 15. Finantial criteria results
Bits 1 20.000 20.000
Primary 3 cycles 4 cycles
Cannonade, packaging 1 120.000 120.000
production (3534) days (3841) days
Injection string , ft 3200 45 144.000
Injection head 1 20.800 20.800 TIR (%) 39,694 39,573 39,603
Downhole equipment 1 21.500 21.500 TPR (%) 24,599 25,099 25,242
Accessories 1 12.400 12.400 PAYBACK simple (years) 4,065 3,984 3,962
Others 1 1.000 1.000
Lifting system 1 135.000 135.000
TOTAL 1.416.300
SPE 165297 13

Figure 1. Reservir with shale/sand Figure 6. Recovery factor vs time

Figure 2. HCSS

Figure 7. Oil rate production vs time

Figure 3. Base model

Figure 8. Pressure vs time

Figure 4. Horizontal wells ubication

Figure 9. Water cut vs time

Figure 5. Injection scheme with horizontals wells
14 SPE 165297

Figure 10. Water rate vs time Figure 14. Reservoir pressure drop for models with no
confinament rings

.
Figure15. Recovery factor for models with no confinement ring
Figure 11. Temperatura distribution on the areal view

.
Figure 16. Oil rate models for models with no confinement ring
Figure 12. Distribution of sight areal loading

Figure 13. Confinament ring model .

Figure 17. Base model with vertical wells
SPE 165297 15

Figure 18. Injection scheme with vertical wells Figure 22. Recovery factor vs tiempo, horizontal wells vs vertical
wells

Figure 3. Sensitivity to steam injection rate for the model with

vertical wells

Figure 23. . Incremental recovery factor vs tiempo, horizontal

wells vs vertical wells

Figure 20. Injectivity of the Wells to a rate 250 bbl/ day

.
Figure21. Injectivity of the Wells to a rate 150 bbl/dia Figure 24. Sensitivity to the horizontal length
16 SPE 165297

Figure 26. Methodology for sensitivity analysis of variable.

Figure25. Objetive functions for the process sensitivity

Figure 28. sensitivity analysis with respetct to SOR

Figure 27. Sensitivity analysis regarding recovery factor
SPE 165297 17

Figure30. Results of the sensitivity analysis of reservoir conditions

Figure 29. Sensitivity analysis result regarding cumulative water regarding FR
production

Figure 31. Methodology used for the optimization process

Base case 

influence 
Best scenario
parameters 

Optimization
Tarjet 
function
18 SPE 165297

Figure 32. Scatter plot of the objective function VPN

Figure 33. Frequency histogram parameters evaluated

Figure 34. Methodology for a financial evaluation

initial investment

income and expenses

project cash flow

financial evaluation results
SPE 165297 19

Figure 35. Cash flow for production with horizontal Wells

CASH FLOW 
8,000,000 $

6,000,000 $
producción en frio,
4,000,000 $ ingresos
producción en
2,000,000 $ frio,egresos
3 ciclos de vapor,
US$

ingresos
0,000 $ 3 ciclos de vapor,
0 1 2 3 4 5 6 7 8 9 10 11 12 egresos
‐2,000,000 $ 4 ciclos de vapor,
ingresos
‐4,000,000 $ 4 ciclos de vapor,
egresos
‐6,000,000 $

‐8,000,000 $
Year

Figure 36. PAYBACK AND VPN and adjusted for cold production with horizontal wells