Journal Homepage: http://ijcpe.uobaghdad.edu.

iq
Iraqi Journal of Chemical and Petroleum
Engineering
Vol. 25 No. 2 (June 2024) 161 – 174
EISSN: 2618-0707, PISSN: 1997-4884

Gas Lift Optimization for Zubair Oil Field Using Genetic

Algorithm-Based Numerical Simulation: Feasibility Study

Saad Hazem O. Al-Mansory a, *, Omar Al-Fatlawi b, c, Ali Kadkhodaie d

a Basra Oil Training Institute, Production Department, Basra, Iraq
b Petroleum Engineering Department, College of Engineering, University of Baghdad, Baghdad, Iraq
c WA School of Mines: Minerals, Energy and Chemical Engineering, Curtin University, WA, Australia
d Earth Sciences Department, Faculty of Natural Sciences, University of Tabriz, Tabriz, Iran

Abstract

The gas-lift method is crucial for maintaining oil production, particularly from an established field when the natural energy of the
reservoirs is depleted. To maximize oil production, a major field's gas injection rate must be distributed as efficiently as possible
across its gas-lift network system. Common gas-lift optimization techniques may lose their effectiveness and become unable to
replicate the gas-lift optimum in a large network system due to problems with multi-objective, multi-constrained & restricted gas
injection rate distribution. The main objective of the research is to determine the possibility of using the genetic algorithm (GA)
technique to achieve the optimum distribution for the continuous gas-lift injection flows in the structure of the network of Zubair, oil
field with 10 gas-lift injected wells. This will be done through numerical simulation and modeling studies. The overall enhancement
of the filed production rate is found to have increased from 15767 STB/day to 19847 STB/day. The well's reservoir pressure and
water cut sensitivity studies are carried out to study the possible impacts of these elements upon the well and its efficiency through
the course of the field. Our understanding of the potential benefits of utilizing gas lift techniques in a field from a technical and
economical point of view is deepened by the use of examples from economic analysis. Furthermore, even though the idea of
employing GA in this manner is not new, this work discusses GA-based optimization methodologies for increasing the oil production
rate by using gas lifting in a Zubair oilfield. In order to assign gas injection rates to specific wells in a network throughout the field
using limited gas injection rates, the model for optimization will be laid out step-by-step making it simple to understand and employ
as a guide, especially for the front-line production technicians involved in the development and design of gas-lift systems.

Keywords: Gas Lift Optimization; Gas Lift; Numerical Simulation; Genetic Algorithms; Zubair Oil Field.

Received on 06/10/2023, Received in Revised Form on 03/02/2024, Accepted on 04/02/2024, Published on 30/06/2024

https://doi.org/10.31699/IJCPE.2024.2.15

1- Introduction
are among the limitations that come with gas lift
The oil wells start to produce as a result of the operation. These restrictions must be considered during
reservoir's natural depletion mechanism. Artificial lift the optimization process. It is a difficult task to take these
methods are implemented when the natural energy constraints into account in order to determine the best lift
decreases [1]. When the reservoir pressure is insufficient gas injection rate allocation for each well in a network
to supply a well with a sustainable oil production rate, the [9]. Increasing the rate of gas injection in a well results in
artificial lift technique is frequently utilized to improve oil an increase in oil output; however, if the rate is increased
production rates. As energy demand has increased and too much, production declines because of the extra
mature fields' pressures have decreased, artificial lift intrusive frictional constraints: this phenomenon causes
techniques have become more and more important [2-4]. the oil rate versus gas injection rate curve, also referred
The gas lift technique is the only type of artificial lift that (GLPC), to take on a dome-like shape [10].
involves injecting gas primarily through valves into the Gas allocation problems typically involve fixed
tubing through an annulus in order to reduce the fluid's parameters due to earlier selection and installation
density and, as a result, reduce the necessary bottom decisions. Consequently, the rate of injected gas becomes
pressure, which increases the rate at which oil is produced the only variable element in these calculations [11]. A
[5-7]. facility's changed parameters, such as injecting depth,
The gas-lift method is crucial for maintaining oil output tube diameter, and compressor, may be optimized when
when natural reservoir energy runs out, especially from methods for optimization are used in the design phase
older fields. Many huge fields use gas lift techniques to [12].
increase economic output [8]. Al-Fatlawi et al. (2015) set out to explore the potential
Gas injection rate, injection pressure, availability for lift of using gas lift as a solution for increasing oil production
gas, compressors capabilities and water handling facilities
*Corresponding Author: Email: [email protected]
© 2024 The Author(s). Published by College of Engineering, University of Baghdad.
This is an Open Access article licensed under a Creative Commons Attribution 4.0 International License. This permits users to
copy, redistribute, remix, transmit and adapt the work provided the original work and source is appropriately cited.
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

in Iraq's giant oil field. To achieve this, they devised a restriction in this technique. The limitations of the
developed model that considered complexity due to facilities are not integrated [18].
shortage data of each of PVT, vertical pressure drops, The equal-slope approach was created by Kanu, Mach
wellhead temperature, and productivity index. Then, they and Brown, 1981 and is focused on optimizing profit in
ran two scenarios using this model: one that aimed to continual gas lift systems by weighing liquid output in
maximize oil production with a set bottom hole pressure terms of profit and the injection of gas (compression) in
constraint and another that maintained the same constraint terms of cost. The next step is determining the appropriate
while keeping the gas-oil ratio in check [2]. allocation rate that maximizes oil production while
Odah et al. (2023) focused on maximizing oil minimizing the price of gas utilization. According to the
production in an Iraqi oil field using gas lift optimization. equal slope technique's guiding principle, profit
The study identified six key variables that impact oil maximization will happen when the income from
production: wellhead pressure, reservoir pressure, water additional oil is equal to or greater than the expense from
cut, tubing diameters, casing pressure, and depth of further gas injection. This method employs a manual
injection. The primary variable of interest was the amount process that involves drawing several tangents with
of injected gas. Using PROSPER software, the study identical slopes on every one of the efficiency curves that
analyzed a network of 16 wells in the Buzrgan oil field were previously built at various gas injection rate values
and sought to identify the optimal gas injection rate for [18].
each well. The objective was to determine the highest Rashid et al. (2012) described a few optimization
possible oil production rate for the field through gas lift techniques that based their conclusions on the Gas Lift
optimization [13]. Performance Curve (GLPC). These approaches attempted
The subject of gas lift optimization has been studied to utilize the GLPC in a way that simultaneously
using a variety of optimization techniques. According to maximizes oil production and minimizes injection costs.
Peixoto et al. (2015), these algorithms can broadly be Numerous optimization strategies have incorporated
categorized into two groups: numerical approaches and financial factors like compression and net profit after
meta-heuristic methods. The conventional numerical deducting injection costs [16]. The linear programming
methods, such the equal slope method, are usually thought method was used by Fang and Lo, 1996 for allocating
of as being based on repeated plots or computations with injection of gas with a variety of rate of flow restrictions
absolute results. The difficulty with these methods is that [19]. In the case of when at least two wells share a flow
they become more complex as the number of parameters line, the effects of back-pressure are frequently
increases [14]. disregarded so that a semi-steady state concept can be
The AI heuristic methods are alternative methods for taken into account for the solution. For the assessment of
gas lift optimization that are based on random selection a person's well-being, such an approach might be
such that their various iterations arrive at different reasonable. When there are flow interactions among wells
conclusions. Their advantage is that they can solve in the system, coupled well solutions with a surface
complicated issues much more quickly and effectively networks approach appear to be more effective because
than numerical methods, particularly in current problems they take into account the impact of backpressure in the
where there are many input parameters [15]. system when distributing gas lift rate [20].
To allocate the lift injection of gas rate across all wells Khamehchi and Mahdiani (2017) and Bergeron et al.
as effectively as feasible, different gas-lift optimization (1999) have utilized the Newton Reduction Method in
methods have been developed within certain facility their studies which proved to be faster and more efficient
restrictions [16]. The multivariate optimization strategies compared to the previous methods. The method faced the
for the optimization of continuous gas lift systems problem of fast convergence due to the interdependent
comprising numerous wells in a network, however, are wells effect but the results were precise enough [15, 20].
typically not modeled by standard gas lift optimization By combining full-network solutions with sequential
methods. The backpressure impact brought on by wells quadratic programming (SQP), Dutta-Roy and
sharing the network infrastructure presents another Kattapuram, 1997 created a nonlinear model [21]. Wang
challenge. and Litvak, 2004 explored using an iterative strategy to
For instance, a single well analysis method might use tackle the gas lift distribution problem until the network's
the nodal analysis idea to generate a single well using overall minimum lift efficiency was reached [22].
accurate temperature and pressure surveys together with Nader et. al. (2008) has developed an innovative
an appropriate multiphase flow correlation. This approach production optimization model that could revolutionize
entails separating every well from the others under the the way we handle offshore gas fields. Unlike previous
presumption of steady wellhead or gathering-system approaches, this new model takes into account the entire
pressures. This assumption might be true in some system, from wells to compressors, providing a
circumstances, including when wells are choked at the comprehensive and holistic view of the complex network.
wellhead to maintain stable wellhead pressure, several The result is an optimization process that not only
wells are gathered at the separator to operate under the maximizes production and reduces costs but also makes
same pressure system of control, or the pipeline network the operation more efficient. The model has already been
is predominated by the decrease in pressure across the tested in real-world scenarios, demonstrating its ability to
well's tubing [17]. The cap on gas injection rate is the key increase revenue while cutting expenses [23].

162
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

Epelle and Gerogiorgis (2019) built a Real-Time the continuous gas-lift injection rate allocation in a
Production Optimization (RTPO) using a robust method network system of a Middle Eastern oil field with 43 gas-
based on the Basic-Open-Source Mixed Integer lift injected wells. This research conducts reservoir
(BONMIN) framework applied to mixed-integer pressure and water cut sensitivity studies to assess the
nonlinear programming (MINLP) on a numerical model impact of these parameters on well performance and
of the output system while taking into account a number production life cycle. Furthermore, sample economics
of operational limitations [24, 25]. analyses are performed to examine the potential technical
Researchers have explored various artificial lift and economic benefits of implementing gas lift
optimization methods, including Particle Swarm techniques in the field [9].
Optimization, Ant Colony Optimization, and Genetic Al-Janabi et al. (2021) this work focuses on applying
Algorithm (GA) to address the challenge of gas lift the Genetic Algorithm (GA) as a gas lift optimization
optimization [26]. algorithm to the difficult problem of maximizing the oil
Hamedi et al. (2011) proposed a novel gas lift production of a Middle Eastern oil field with 20
optimization method that demonstrated promising results production wells and a finite amount of gas to be injected.
in maximizing oil output while efficiently allocating This is done through numerical modeling and simulation
compressed gas. The method incorporates a particle studies to optimally allocate the gas lift injection rate [32].
swarm optimization algorithm that allocates the ideal gas Ahmed et al. (2023) examined the effect of gas lift design
injection rate for each well within a group, along with a and optimization on production results in the Mishrif
new gas lift performance curve-fit method that formation of the Halfaya oil field. A detailed production
significantly reduces computation time and gas volume. network nodal analysis model was constructed using
These improvements are particularly significant in gas lift PIPESIM Optimizer and field data from seven wells, with
operations, where compressed gas is a scarce and the model being calibrated to accurately reflect the field
expensive resource [27]. conditions. The Mishrif Formation in the Halfaya oil field
Ghaedi et al. (2014) proposed a Continuous Ant Colony is one example of a complicated reservoir system. This
Optimization (CACO) algorithm to allocate an optimal research highlights the importance of strategically
amount of gas to each well within a group for three designed and optimized gas lift systems for improving oil
different oil fields, each with a varying number of wells. recovery and operating efficiency. The achievement of
A comparison between the total oil production rates these ideal parameters was made possible by the
resulting from gas allocation to the wells using the CACO application of the PIPESIM Optimizer, which is based on
algorithm and previous studies utilizing other methods the concepts of genetic algorithms [33].
reveals that the CACO approach achieves superior gas The earlier described optimization techniques might
allocation to the wells [28]. only be effective under very specific conditions. Due to
Numerous academics have examined the Genetic the lack of consideration for back pressure and the
approach (GA) [9], the approach we choose to apply in challenges associated with integrating facility and
this study, and its application to the problem of gas lift production restrictions, these conventional techniques are
optimization. Martinez et al., 1994 were one of the first not well suited for handling wells that are part of the same
investigators who utilized the GA algorithm to address the network system. Using these methods to describe and
issue of gas lift optimization, and the outcomes of the simulate a big network system accurately is quite difficult
research showed that the mathematical method that was because it takes a lot of calculation time and frequently
utilized proved to be effective [29]. yields false results. Another aspect of poor solutions is
Ray and Sarker (2007) proposed modifications to the locating a local answer to the issue. Consequently, using
Genetic Algorithm and the formulation of gas lift dynamic global algorithms may be required to address
optimization problems as multi-objective models that these problems, as suggested by Buitrago et al. [34], they
offer improved performance in solving these complex offered derivative-free algorithms that implemented a
optimization problems. By treating reservoir optimization heuristic method for distributing gas lift while resolving
as a multi-period problem, this research also provides a the crucial problems of non-instantaneous flows (NIF)
methodology for generating flexible production plans that and unsmooth curvature for a significant number of wells.
better reflect the dynamics of oil extraction over time. In order to properly integrate all of these restrictions of
Furthermore, the population shrinking technique gas lift difficulties, the Genetic Algorithm, or GA,
incorporated in the GA enables the maintenance of optimization is the main emphasis of this study on
diversity in both objective and variable space, which improving the gas lift.
results in more robust and effective solutions to gas lift In our study, the genetic algorithm (GA) as an
optimization problems [30]. optimization algorithm a network simulator is utilized
Khamehchi et al. (2009) used the genetic algorithm to using the PIPESIM software in order to construct the
build and improve the efficient aspects while taking into model with the available 10 production wells in the field
account the (injection pressure and water cut) as a and we created a new model for matching PVT
constraint in order to achieve the highest production rate information, matching calculations for vertical pressure
[31]. drops, making sensitivity analyses (water cut sensitivity
AlJuboori et al. (2020) investigated the feasibility of and reservoir pressure sensitivity) of productivity index
using the Genetic Algorithm (GA) technique to optimize and wellhead temperature variations, achieving the best

163
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

gas lift design, and determining the best values for (children), while those with the smallest fitness value
injected gas rate and production of oil. have the lowest probability [40].
The most important step in the GA is crossover, which
2- Genetic Algorithm involves mating every pair of individuals who were
chosen as parents in the selection step by switching their
In addition to using natural selection to solve genes at one or two points of crossover in order to
optimization problems, the genetic algorithm is based on recombine the chromosomes and create the new offspring
the principles of genetic recombination processes, which (generation) [41]. Three different types of crossovers
include the main purposes of selection, crossovers, and exist.
mutation. Holland, 1992 was the first person to introduce  Single Point Crossover
the genetic algorithm, or GA, concept to solve both  Double Point Crossover
constrained and unconstrained optimization problems  Uniform Crossover
through the method of natural selection centered upon The crossover between two good individuals will not
Darwin's theory of the fittest function and the always produce a better individual but as the parents are
fundamental ideas of evolutionary science [35]. good (selected based on high fitness rank), the child will
Due to the fundamental distinction between the genetic have a high probability of being good, and in the case of
algorithm, or GA, and other conventional methodologies, offspring not being good enough they will be distant in
it was regarded as one of the most recent methods of the next selection.
optimization. The majority of GA solutions are global  The GA operator known as mutation maintains
optimums because GA employs a population of chosen diversity. It is a low-probability step that results in
points to avoid the optimal local solution [36]. Discrete the flipping of a few random genes on the newly
and integer problems are extremely suited to be addressed formed spring chromosomes. The mutation
by GA since the genetic algorithm depicts the chosen probability can range from zero (no mutation) to one
variables by utilizing a sequence of binary integers (0,1) (mutation after each iteration). The mutation process
for imitating the chromosomes in genetics [37, 38]. plays an important part in keeping a population's
The imitation of biology in the application of the GA variation, and although it prolongs the convergence
led to the importing of several terms and defections from process, mutation prevents premature convergence.
biology to the applications like: (population, generation, In the event that the initial population missed certain
gene, chromosome, mutation, and crossover) and to be areas of the solution that haven't been attained by
able to fully understand the process a well-established both the selection and crossover steps, mutation
knowledge of the terminology is required [39]. provides the probability. As a result, utilizing
 The Gene, the smallest building block in a string mutation allows GA to reach a better conclusion [40].
(individual) of GA, is made up of the binary values  When the population as a whole does not create new
(0,1) that stand in for every parameter (variables) in children (offspring) who are distinct from previous
the assembled chromosomes. generations (the population has converged), we have
 Chromosomes are a long string of numbers that are reached the phase known as termination. This occurs
produced when numerous genes are combined to when the genetic code solves the problem by offering
form an individual; each individual represents a a set of answers [32].
potential solution to the current issue because it is Gen and Cheng (1999) provided an example of how
denoted by a value that is numerical. And as conventional procedures, which employ a single-point
observed by Ray and Sarker [30]. answer and a deterministic sequencing of a derivative-
 The population is represented by the collection of based objective function, mostly provide local optimum
solutions (chromosomes), and the population size solutions. The population-based GA solution will avoid
identifies the quantity that distributes the gas lift rate sliding into the local optimum by maintaining multipoint
[30]. to multipoint approaches for each generation. In this
 The Fitness Function may be described as a score of strategy, the likelihood of the best, most fit solution
how physically fit an individual was and their outlasting the least fit one is higher. Therefore, the GA
capacity to compete with one another. This can be will typically produce a global optimum outcome [36].
accomplished by assigning each candidate a fitness According to Sarker et al., 2003, GA provides the
score, and candidates will be chosen based on that following advantages over traditional optimization
number. The task that the GA seeks to optimize is techniques for modeling gas lift method optimization
known as the fitness function, and in the context of problems [42]:
gas lift, this function refers to the oil flow rate that • Unlike other optimization techniques, GA does not
the method of optimization seeks to maximize [40]. require assuming the concavity, convexity, or
 The selection is the procedure of identifying the continuity of the algorithm.
fittest chromosome (individuals) and evaluating them • GA can incorporate some established techniques,
in terms of fitness function, allowing those with the such as the hybrid algorithm, to increase its
greatest fitness ranked to have the greatest possibility effectiveness in resolving the constraints present in
of being the parents who produce new offspring gas lift situations.

164
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

• GA may provide multiple-point solutions, such as the Initial population is the first step in the process and
first, second, third, etc., but conventional approaches consists of a group of individuals (i.e., solution) inside the
can only provide the most suitable single point search space. An individual is defined by a variety of
answer. This feature is particularly important for combinations of the "genes," which are put together to
decision-makers in the oil field sector because some form the chromosome (or solution) and are referred to as
of them might need to consider other potential design variables (or parameters). A determination of each
options. GA can provide three of the best options for solution's objective function, a measure of fitness, is
deciding the gas lift technique's optimal solution; made. A mating pair, which is utilized to reproduce future
each of these can be tailored to the choice maker's populations, is created by pairing each solution with
goals. another. Through the crossover of the values of the
• GA has the ability to integrate the function of parents, the mating pairs create new solutions called
penalties while answering all optimization problems offspring that have the value of the design variables.
according to constraints by recommending a change Additionally, some of the values in the new solution
to the structure to provide a workable solution that undergo random mutation. The vast range of factors can
satisfies the multicasting objective functions; be evaluated with the help of mutations, which can delay
• Compared to a traditional technique, GA typically the onset of convergence. Although its optimization
offers a quicker answer to an optimization challenge. allows for one solution to define the level of control, the
Even occasionally, a brief approximation solution GA technique's procedure is random. The steps for
requires a small generation [43]. applying the GA approach to model the gas lift problem
• A crucial consideration in selecting this method for are outlined below:
resolving the gas lift optimization problem is the 1. Creation of the first population: As seen in Fig. 1,
ability of GA to resolve multi-objective or lifting gas injection rates are regarded as
multivariate optimization problems because it is chromosomes, and 4 chromosomes (A, B, C, and D)
advantageous in that it offers a variety of options in a are produced with the gas injecting rate specified [9].
single simulation run while also taking into account 2. When evaluating fitness function, oil flow rate—
gas-lift constraints like the water supply and gas which depends on lift gas injection rate—is taken
handling system, flowing bottom well pressure, and into account.
the gas injection rates [32, 44-46]. 3. The optimization process is completed if the results
• The starting point solution is created randomly in this satisfy the optimization requirements.
method without the inclusion of any technique, which 4. In the absence of this, the solution is changed by
reduces the time required to get the starting point choosing two chromosomes.
solutions and the amount of work required to provide 5. performing crossover for each pair of chosen
the initial point data to start optimization. chromosomes with a probability lower than the
• In The objective function may need to be adjusted in crossover threshold.
the oil sector due to the ongoing changes in oil wells 6. Chromosome mutation, if necessary to create a new
and field data, particularly production data for each generation.
well like GOR, water cut, and oil flow rate. In this 7. Calculating the standard deviation of the values for
situation, GA can easily handle such an adjustment fitness of new generations.
and offer the best option. 8. If there is no restriction on the total number of new
Five fundamental phases make up GA techniques: generations, the problem is solved iteratively until the
starting population, fitness function, selection, crossover, best answer is obtained.
and mutation.

Fig. 1. The Workflow of GA to Obtain the Optimum Results [9]

165
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

3- Numerical Modeling and Simulation formation volume factor Additionally, since pressure and
temperature have a significant impact on the
The implementation of the GA method in a real case compressibility of gases, it is difficult to find a correlation
can effectively assess its power in modeling the problems that can be used across the entire pressure and
of gas lift well considering all numerical simulation temperature spectrum [47]. The wellhead pressure, oil
constraints being discussed above to find the feasibility of production rate, and a few test points were compared to
the use of such technique on a daily basis in the oil fields. the production data that was provided. The secret to
This work vastly involves modeling and numerical performing a precise nodal analysis is finding the
simulation for the Zubair oil field using the GA-based matching multiphase flow correlation [48]. Therefore, the
optimization tool. The model considered 10 wells of this bottom hole flow pressure, Pwf, is computed using
field, each of which is being installed with a gas-lift multiphase correlations utilizing the production data that
injection system in a network. GA uses the survival of is now available and the PVT parameters. The PIPESIM
best solution principles depending on generating the simulator then used the generated Pwf data to establish the
initial solutions, finding the results, selecting the best best history-matching tubing flow correlations. It is
solutions of multi-objective and multi-constrained interesting to note that Zubair Oil Field's matching flow
functions, and improving the solutions to obtain the correlation was found to be different. For vertical flow,
optimum result. Hagedom & Brown and Beggs & Brill updated
correlations that are closest to the available test points
4- Field Background
(measured data) are the best. For horizontal flow, Beggs
The Zubair oilfield is 20 kilometers southwest of the Al- & Brill modified correlations are the best.
Basra city in southern Iraq. It was discovered in 1949, and
6- Building Well Model
development began in 1951. Anticline (60 × 15 km) is the
field's shape, and it has three domes (after combining In order to construct the well model, numerous well
Hammar and Shuaiba into one dome): Hammar- Shuaiba, characteristics must be defined, including fluid properties,
Rafydia, and Safwan. The producing layers are situated at perforation, casing and tube information, PVT data, and
around a depth of (2250, 3100, 3180, 3360) m with initial perforation test point information. It has been discovered
reservoir pressure around 5400 psi. This field consists of that these data vary from well to well. Model schematic
three reservoirs: third pay, fourth pay, and upper shale. after construction using the PIPESIM simulator is shown
The third pay reservoir in the Zubair formation will be in Fig. 2. The location, type, casing information,
taken into consideration for this study; it is composed of perforation features, tubing size, nodal point, line of flow
sandstone and is located at a depth of 3180 meters. This diameter and temperature, and wellhead pressure are all
field has a number of gas lift wells to keep the reservoir entered before the well model is built. By adding this
pressure. Deep wells approximately 3200 meters deep data, a well model is created, which is then history-
with a depletion drive mechanism cause the reservoir matched in order to get ready to simulate the field's gas
pressure to drop very quickly, which lowers oil lift performance.
production rates and increases water cuts in some wells.
As a result, attention is drawn to the gas lift method with
the goal of maximizing oil production due to the decrease
in average pressure in the reservoir and well performance.
5- Well Flow Model
Building a numerical simulation model is usually
incredibly difficult because of the lack of necessary data,
especially when the model is meant to represent the entire
area. It might be difficult to accurately determine the
flowing bottomhole pressure (Pwf), perform pressure-
volume-temperature (PVT) investigations of reservoir
fluids, discover matching PVT correlations, and
determine the outflow performance (VLP) correlation for
the wells. According to the PVT measurement
experiments of wells in this field, the field PVT data
derived from lab data are taken into consideration as a
median for the entire field, especially as the disparity
between them appears to be minor based on the lab data.
Second, the data was further calibrated using the fluids in
the reservoir of Zubair Oil Field after matching the PVT
data with appropriate correlations. The precise
measurement of the gas compressibility factor is a
prerequisite for the exact computation of all gas Fig. 2. A Well Schematic (Well-ZA) After Being Built by
properties, including viscosity, gas density, and gas Simulation

166
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

7- Gas Lift Design Data function or error envelope that enables the modeling of an
extra equal space valve at the conclusion of the indicated
The associated gases, which are produced in the same operational valve to accommodate any design fault [32,
field as the oil, is thought to be the source of the gas 49]. This plan helps to resolve any ambiguity in the gas
injection. According to reports, this field produces gas at lift design by accounting for any error in the flow's
a rate of 33 MMscf per day on average. Initially multiphase correlation in both the vertical and horizontal
segregated through a separator, the associated gases are directions, as well as any uncertainty in a well's
then processed, gathered, and recycled once more by productivity index, as shown in Fig. 3.
being injected back into the oil wells. According to Fig. 3 shows the proposed design for a well with a gas
completion programs, 3.5′′ tubing is used to complete lift and activated bracketing feature.
every well. According to Al-Fatlawi et al., 2015 the Laing, 1991 provides support for this claim by
functioning gas lift valve is positioned 100 feet above the implementing this characteristic in two oil fields, which
packer. According to the field data, the temperature at the increased oil production and enhanced well performance.
surface is about 115°F. To ensure a single-phase inflow Additionally, a flexible design that can account for
into the well, a minimum Pwf of 50 psi over the point of unexpected or low well productivity index is already
bubble pressure (Pb) is assumed [2]. Additionally, the taken into consideration [50].
PIPESIM simulator contains a gas lift valve bracketed

Fig. 3. Depicts the Proposed Gas Lift Design for the Production Well, with the Bracketing Function Enabled

8- Building Field Model network model, and they are all taken into account for gas
lift optimization utilizing the GA approach in order to
The surface network simulation is constructed using the spread the gas injection across the network's wells as
PIPESIM simulator when the well modeling is complete, efficiently as possible.
as illustrated in Fig. 4. Ten producing wells are part of the

Fig. 4. The Network Wells Are Clearly Visible

167
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

9- Sensitivity Analyses The results will show the importance of applying the
optimization technique to the gas lift process and using a
Sensitivity assessments were conducted to examine the multi-constrains algorithm is highly important to be able
potential effects of changing reservoir pressure and water to handle all the different constraints especially when
cut upon the functionality of natural flow and gas lift working with a large number of wells and for the case of
wells, particularly over the long term when reservoir the limited amount of injection gas in order to achieve the
pressure is expected to decline and water cut to rise. best production rate with the available gas in hand.

9.1. Water cut sensitivity

In order to look into the impact of water cuts on both

the well and field performance, water cut sensitivity
experiments were carried out for both natural flow and
gas lift wells. In this analysis, several water cuts that
range from 10% - 50% were taken into account. In order
to contrast the good behavior in every instance without
favoring any particular procedure, the sensitivity analysis
was carried out for each well after and before the use of
gas lift at the exact same wellhead pressure.

9.2. Reservoir pressure sensitivity

Fig. 5. Comparison between the Production of Natural
The goal of the pressure in the reservoir sensitivity Flow Wells and the Gas Lifted Wells
research was to ascertain how reservoir pressure depletion
might affect the good performance profile; identify its
potential influence in the field production life cycle for
both natural flowing and gas lift wells; and investigate
whether the gas lift wells can have any positive influence
in relation to the improvement of good performance in the
case of reservoir pressure depletion. The average reservoir
pressure considered for the sensitivity studies were 3500
psi, 3250 psi, 3000 psi, 2750 psi, 2500 psi, 2250 psi, 2000
psi, 1750 psi, 1500 psi, 1250 psi, and 1000 psi for a given
minimum wellhead pressure of 250 psi. The wellhead
pressure of 250 psi is assumed to be the minimum
pressure required to support the surface system
requirement (e.g., separator system).

10- Results Analysis and Discussions Fig. 6. Comparing the Results of Gas Lift Before and
After the Optimization
10.1. Simulation results
Before the optimization work, we noticed that Well-ZF
Both naturally flowing wells and continuously lifting required the pumping of 5 MMscf/day gas into it, and its
gas wells were simulated. GA method is used to optimize production was 2500 STB/day. However, after the
gas lift performance at a 33 MMscf/day maximum gas lift optimization technique, we discovered that it no longer
injection rate. The simulation results for naturally flowing requires the pumping of gas, and despite this, its oil
wells and gas lift wells compared in Fig. 5 demonstrate production increased to 3300 STB/day as a result of the
that there is a significant increase in oil production rate neighboring Nearby wells, because when a gas lift was
(% of the increase in oil production) from each well for made, it eliminated the negative effect of the Nearby
this field as compared to natural flowing wells. The wells on it in the network system.
overall increase of filed production rate is found to be
increased from 15767 STB/day to 19847 STB/day. 10.2. Water cut sensitivity results
In order to compare the effect of the optimization
technique, Fig. 6 will provide a comparison between the Fig. 7 shows the findings of a study on water cut (WC)
two cases before and after the optimization to show the sensitivity for 10 oil wells that were both naturally
difference in the distribution of the gas injection rate for produced wells and gas-lifted wells. Fig. 9 confirms what
every well and the effect that increased the oil production was predicted, showing that for natural flow and gas lift
from (17242 STB/day) in the case of before optimization wells, the rate of oil production falls as the water cut
to (19847 STB/day) after the use of the optimization (WC) increases. Fig. 7 also shows that the increased WC
algorithm. necessitates a greater gas lift injected rate. However, it

168
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

appears that the use of gas lifts has greatly increased the flow and gas lift, which can identify the improvement of
pace of overall oil production. It is discovered that the use sensitivity caused by applying the various types of
of gas lift results in a very significant overall output artificial lift, is shown in Table 1, respectively. From
boost, particularly at a greater water cut. Table 1, it is evident that most wells in a natural flow
condition will typically shut down at (3187 psi) on
average. On the other side, we can observe that the gas lift
approaches were successful in lowering the operating
pressure in the reservoir and those wells will continue to
operate. For the gas lift method, the average pressure at
which the wells are going to shut down is (586 psi). Gas
lift may be viewed as an important justification for using
methods of lifting to maintain pressure because wells
remain able to flow out under lower pressures if the
pressure in the reservoir is reduced.

Table 1. Reservoir Pressure at which the Well Shuts

Down for Each Case
Well Name Natural Flow (psi) Gas Lift (psi)
Well-ZA 2950 572
Fig. 7. Results of Water Cut Sensitivity Well-ZB 2396 572
Well-ZC 3259 571
10.3. Reservoir pressure sensitivity results Well-ZD 3082 571
Well-ZE 2600 573
Well-ZF 5100 607
The goal of the reservoir pressure sensitivity study was Well-ZG 2990 571
to ascertain the minimum average reservoir necessary for Well-ZH 3137 607
a well to stay productive as well as to evaluate the impact Well-ZI 3700 607
of lowering reservoir pressure on the efficiency of 10 oil Well-ZJ 2655 607
wells. This sensitivity begins at the present reservoir
pressure of approximately 3500 psi and continues until it To be able to visualize the difference in the results of
reaches the pressure value where both oil-producing wells the pressure sensitivity a graphical illustration of the
and gas lift wells are unable to flow. In the beginning, effect of the gas lift method on maintaining the
nodal analysis was done at various reservoir pressures to production of the wells even in the cases of reservoir
figure out the average minimum reservoir pressure needed pressure reduction can be seen in Fig. 8 which represents
to maintain production at the specified wellhead pressure the outflow performance of the wells against different
of 250 psi for all cases. values of reservoir pressure which can then help to
The mean reservoir the pressure at which a well will identify the well operation point.
shut down for both of the cases of naturally occurring

Fig. 8. The Effect of Changing the PR for a Well Installed with Gas Lift

The results indicate that wells equipped with a gas lift utility in sustaining high rates of production for many
continue to produce at relatively high rates despite low years.
reservoir pressures, confirming the gas lift economic

169
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

11- Economic Analysis 12- Conclusions

To determine whether applying the gas lift optimization The main goal of this paper was to determine whether
method in the real world would be profitable, economic using a GA-based optimization strategy in numerical
analysis was used. From Nakashima and Camponogara, simulation for daily optimization of gas lifting wells in a
2006, the equation employed in this analysis is large field with a complex network structure was feasible.
represented by Eq. 1 [51]. Oil and gas prices of $55 per Using numerical simulation and the GA approach, one of
STB and $5,500 per MMscf, respectively, were used as the massive Middle Eastern oilfields’ continuous lifts of
the basis for the analysis. According to Huh et al., 2010, gas injection rate is optimized over 10 wells. This paper
the costs associated with disposing of water, running a gas includes a thorough discussion of the GA principles, the
lift, and other operations are each estimated to be $1 per mathematical model, as well as the process for carrying
barrel of water, $3,500 per million standard cubic feet, out simulation investigations.
and $8 each STB produced well [52]. Through the presentation of optimization and sensitivity
𝑁𝑒𝑡 𝑃𝑟𝑜𝑓𝑖𝑡 = ∑𝑛𝑖[𝑞𝑜𝑖 (𝑝𝑜 − 𝑐𝑜𝑝 ) + (𝑝𝑔 × 𝑞𝑔 ) − ( 𝑞𝑤𝑖 × 𝑐𝑤𝑖 ) −
analyses, the study also provided insight into future
prediction and expanded understanding from both a
(𝑄𝑔𝑖,𝑖𝑛𝑗 × 𝑐𝑔,𝑖𝑛𝑗 )] (1)
technical and economic standpoint. An example economic
Where: 𝑛: The Total Number of Wells. 𝑞𝑜𝑖: Rates of Oil analysis was also carried out to shed light on the benefits
Production (STB/day). 𝑞𝑔: Rates of Gas Production of employing gas lift techniques in depleting drive
(MMscf/day). 𝑞𝑤𝑖: Water Production Rate (STB/day). 𝑄𝑔𝑖, reservoirs, especially in the event of an increasing water
𝑖𝑛𝑗: Lifting Gas Injection Rates (MMscf/day). 𝑝𝑜: The cost
cut and severely low reservoir pressure.
of oil ($/STB). 𝑐𝑜𝑝: Operational Costs ($/STB) for each The study's findings lead to the following conclusions:
STB of oil. 𝑝𝑔: Gas Price ($/MMscf). 𝑐𝑤: Cost of disposal • The optimal way to distribute gas injection rates for
of water ($/STB). 𝑐𝑔, 𝑖𝑛𝑗: Cost of lift gas injection maximizing oil production rates appears to be
($/MMscf) predicted using genetic algorithms, or GAs, for
Applying the above prices in Eq. 1 would result in a net modeling many of the concurrently running wells
profit of (1269475 $/day) for the gas lift case compared to into a system of networks.
a value of (934576 $/day) for the case of natural flow • For wells with relatively greater water cuts, gas lift
which is considered a great profit and a great value procedures prove to be more advantageous.
provided by the increase of production from the gas lift • In the event of pressure depletion in the reservoir, gas
case, it’s worth mentioning that in the case of natural flow lift can be helpful in extending the field's life cycle
the equation was modified to remove the term of the gas and enhancing well production performance by
injection rate and it is price. enabling a well to continue operating at even low
seeing the outcomes of the OPEX estimates for the two flows the wellbore bottom hole the pressure (P wf) that
instances of natural flowing and gas lift to make it easier it is designated minimum wellhead pressure.
to compare the various outcomes, Fig. 9 is intended to • The sample economic analysis shows that, compared
show the oil production rates for every well in each to the use of natural wells, the gas lift approach can
instance as well as the net profit of every well in each significantly boost each well's daily crude oil
case. production rate and, consequently, its net profit.
Fig. 9 compares the daily net revenue without or with
gas lift as well as the total increase in the daily crude oil Acknowledgment
production rates for 10 wells. According to the computed
results, the gas lift method can significantly raise each The authors are grateful to Schlumberger Company for
well's daily oil production rates and, consequently, its net approving their use of the PIPESIM academic simulator.
profit when compared with naturally producing wells.
Nomenclature

AI Artificial Intelligence
GA Genetic algorithm
GLPC Gas lift performance curve
IPR Inflow performance relationship
OPEX Operational Expenditure
PI Productivity index
Pwf Bottom hole flowing pressure
Pwh Well Head Pressure
qgi,inj Gas lift injection rate, Mscf/day
qo Total oil production, STB/day
qoi Individual well oil production rate, STB/day
SQP Sequential quadratic programming approach
VLP Vertical lift performance
Fig. 9. Natural Flowing and Gas Lift are Compared with
Flow Rate and Net Profit

170
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

References [10] H. Hussein, A. Al-Durra, and I. Boiko, "Design of

gain scheduling control strategy for artificial gas lift
[1] M. S. Al-Jawad and D. J. Ottba, "Well performance in oil production through modified relay feedback
analysis based on flow calculations and IPR," test," Journal of the Franklin Institute, Article vol.
Journal of Engineering, vol. 12, no. 03, pp. 822-841, 352, no. 11, pp. 5122-5144, 2015,
2006, https://doi.org/10.31026/j.eng.2006.03.28 https://doi.org/10.1016/j.jfranklin.2015.08.007
[2] O. F. Al-Fatlawi, M. Al-Jawad, K. A. Alwan ,A. A. [11] M. R. Mahdiani, E. Khamehchi, and A. A. Suratgar,
Essa, D. Sadeq, and A. J. Mousa, "Feasibility of gas "Using modern heuristic algorithms for optimal
lift to increase oil production in an Iraqi giant oil control of a gas lifted field," Journal of Petroleum
field," in Society of Petroleum Engineers - SPE North Science and Engineering, Article vol. 183, 2019, Art
Africa Technical Conference and Exhibition 2015, no. 106348,
NATC 2015, 2015, pp. 1401-1412, https://doi.org/10.1016/j.petrol.2019.106348
https://doi.org/10.2118/175862-MS [12] J. N. M. de Souza, J. L. de Medeiros, A. L. H. Costa,
[3] M. M. Hossain and M. D. Bin Mohd Ismail, and G. C. Nunes, "Modeling, simulation and
"Potential application of Downhole Gas Compressor optimization of continuous gas lift systems for
to improve productivity for gas reservoir," in Society deepwater offshore petroleum production," Journal
of Petroleum Engineers - International Petroleum of Petroleum Science and Engineering, Article vol.
Technology Conference 2013, IPTC 2013 : 72, no. 3-4, pp. 277-289, 2010,
Challenging Technology and Economic Limits to https://doi.org/10.1016/j.petrol.2010.03.028
Meet the Global Energy Demand, 2013, vol. 5, pp. [13] H .A. Odah, M. J. Hamed, M. S. Reshk, and D. J.
4201-4213. https://doi.org/10.2523/IPTC-16982-MS Sadeq, "Optimization of gas lift production of oil
[4] M. D. M. Ismail and M. Hossain, "The application of wells," in AIP Conference Proceedings, Baghdad,
downhole gas compression to improve productivity 2023, vol. 2651, https://doi.org/10.1063/5.0111682
for depleted natural gas reservoirs," The APPEA [14] A. J. Peixoto, D. Pereira-Dias, A. F. S. Xaud, and A.
Journal, vol. 53, no. 1, pp. 369-374, 2013, R. Secchi" ,Modelling and extremum seeking control
https://doi.org/10.1071/AJ12032 of gas lifted oil wells," in IFAC-PapersOnLine, 2015,
[5] S. R. Ahmed and D. J. Sadeq, "Maximizing vol. 28, 6 ed., pp. 21-26,
Production Profits: Optimizing Gas Lift Design in the https://doi.org/10.1016/j.ifacol.2015.08.004
Halfaya Oil Field," Iraqi Journal of Chemical and [15] E. Khamehchi and M. R. Mahdiani, "Gas Allocation
Petroleum Engineering, vol. 25, no. 1, pp. 169-180, Optimization Methods in Artificial Gas Lift ",ed:
2024, https://doi.org/10.31699/IJCPE.2024.1.16 Springer Publishing Company, Incorporated, 2017.
[6] M. Ghaedi, C. Ghotbi, and B. Aminshahidy, https://doi.org/10.1007/978-3-319-51451-2
"Optimization of gas allocation to a group of wells in [16] K. Rashid, W. Bailey, and B. Couët, "A survey of
gas lift in one of the Iranian oil fields using an methods for gas-lift optimization," Modelling and
efficient hybrid genetic algorithm (HGA)," Simulation in Engineering, Article vol. 2012, 2012,
Petroleum Science and Technology, Article vol. 31, Art no. 516807, https://doi.org/10.1155/2012/516807
no. 9, pp. 949-959, 2013, [17] Q. Lu and G. C. Fleming, "Gas lift optimization
https://doi.org/10.1080/10916466.2010.535081 using proxy functions in reservoir simulation," SPE
[7] M. Ghaedi, B. Aminshahidy, and C. Ghotbi, Reservoir Evaluation and Engineering, Article vol.
"Improving gas allocation optimization to a group of 15, no. 1, pp. 109-119, 2012,
wells in gas lift using an efficient hybrid genetic https://doi.org/10.2118/140935-PA
algorithm (HGA)," Energy Sources, Part A: [18] E. P. Kanu, J. Mach, and K. E. Brown, "ECONOMIC
Recovery, Utilization and Environmental Effects, APPROACH TO OIL PRODUCTION AND GAS
Article vol. 36, no. 21, pp. 2361-2375, 2014, ALLOCATION IN CONTINUOUS GAS LIFT,"
https://doi.org/10.1080/15567036.2011.569835 JPT, Journal of Petroleum Technology, Article vol.
[8] A. S. Langvik and L. Dzubur, "Optimization of oil 33, no. 10, pp. 1887-1892, 1981,
production-applied to the marlim field," Master, https://doi.org/10.2118/9084-PA
Industrial Economics and Technology Management, [19] W. Y. Fang and K. K. Lo, "A Generalized Well-
Norwegian University of Science and Technology, Management Scheme for Reservoir Simulation," SPE
2012. Reservoir Engineering (Society of Petroleum
[9] M. AlJuboori, M. Hossain, O. Al-Fatlawi, A. Kabir, Engineers), Article vol. 11, no. 2, pp. 116-120, 1996,
and A. Radhi, "Numerical Simulation of Gas Lift https://doi.org/10.2118/29124-PA
Optimization Using Genetic Algorithm for a Middle [20] T. Bergeron, A. Cooksey, and J. S. Reppel, "New
East Oil Field: Feasibility Study," in International automated continuous gas-lift control system
Petroleum Technology Conference, Dhahran, improves operational efficiency," in Society of
Kingdom of Saudi Arabia, 2020, vol. Day 3 Wed, Petroleum Engineers - SPE Mid-Continent
January 15, 2020, D031S095R001. Operations Symposium 1999, MCOS 1999,
https://doi.org/10.2523/IPTC-20254-MS Oklahoma City, 1999, https://doi.org/10.2118/52123-
ms

171
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

[21] K. Dutta-Roy and J. Kattapuram, "A New Approach [32] M. A. Al-Janabi, O. F. Al-Fatlawi, D. J. Sadiq, H. A.
to Gas-Lift Allocation Optimization ",in SPE Mahmood, and M. A. Al-Juboori, "Numerical
Western Regional Meeting, California, 1997, vol. All Simulation of Gas Lift Optimization Using Artificial
Days, SPE-38333-MS. Intelligence for a Middle Eastern Oil Field,"
https://doi.org/10.2118/38333-MS presented at the Abu Dhabi International Petroleum
[22] P. Wang and M. Litvak, "Gas lift optimization for Exhibition & Conference, 2021.
long-term reservoir simulations," in Proceedings - https://doi.org/10.2118/207341-MS
SPE Annual Technical Conference and Exhibition, [33] S. R. Ahmed, D. J. Sadeq, and H. B. Mahmud,
2004, pp. 2913-2920, https://doi.org/10.2118/90506- "Optimization of Gas Lifting Design in Mishrif
ms Formation of Halfaya Oil Field," Iraqi Journal of
[23] M. S. Nadar, T. S. Schneider, K. L. Jackson, C. J. N. Chemical and Petroleum Engineering, vol. 24, no. 3,
McKie, and J. Hamid, "Implementation of a total- pp. 45-54, 2023,
system production-optimization model in a complex https://doi.org/10.31699/IJCPE.2023.3.5
gas lifted offshore operation," SPE Production and [34] S. Buitrago, E. Rodríguez, and D. Espin, "Global
Operations, Article vol. 23, no. 1, pp. 5-13, 2008, Optimization Techniques in Gas Allocation for
https://doi.org/10.2118/103670-pa Continuous Flow Gas Lift Systems," presented at the
[24] E. I. Epelle and D. I. Gerogiorgis, "Mixed-Integer SPE Gas Technology Symposium, 1996.
Nonlinear Programming (MINLP) for production https://doi.org/10.2118/35616-MS
optimisation of naturally flowing and artificial lift [35] J. H .Holland, Adaptation in natural and artificial
wells with routing constraints," Chemical systems: an introductory analysis with applications
Engineering Research and Design, vol. 152, pp. 134- to biology, control, and artificial intelligence. MIT
148, 2019/12/01/ 2019, press, 1992.
https://doi.org/10.1016/j.cherd.2019.09.042 https://doi.org/10.7551/mitpress/1090.001.0001
[25] M. A. AL-Hejjaj, D. J .Sadeq, and O. Al-Fatlawi, "A [36] M. Gen and R. Cheng, Genetic algorithms and
Review of the Electrical Submersible Pump engineering optimization. John Wiley & Sons, 1999.
Development Chronology," Iraqi Journal of https://doi.org/10.1002/9780470172261
Chemical and Petroleum Engineering, vol. 24, no. 2, [37] W. Barnett, C. Chiarella, S. Keen, R. Marks, and H.
2023, https://doi.org/10.31699/IJCPE.2023.2.14 Schnabl, "Commerce, complexity, and evolution,"
[26] M. A. M. Al-Janabi and O. Al-Fatlawi, "Gas lift Cambridge University Press, 2000 .
optimization: A review," in AIP Conference [38] M. M. Fischer and Y. Leung, GeoComputational
Proceedings, 2022, vol. 2443, modelling: techniques and applications. Springer
https://doi.org/10.1063/5.0091901 Science & Business Media, 2013.
[27] H. Hamedi, F. Rashidi, and E. Khamehchi, "A novel [39] R. L. Haupt and S. E. Haupt, Practical genetic
approach to the gas-lift allocation optimization algorithms. John Wiley & Sons, 2004.
problem," Petroleum Science and Technology, https://doi.org/10.1002/0471671746
Article vol. 29, no. 4, pp. 418-427, 2011, [40] O. F. Al-Fatlawi, "Numerical simulation for the
https://doi.org/10.1080/10916460903394110 reserve estimation and production optimization from
[28] M. Ghaedi, C. Ghotbi, and B. Aminshahidy, "The tight gas reservoirs," PhD, WA School of Mines:
Optimization of Gas Allocation to a Group of Wells Minerals, Energy and Chemical Engineering, Curtin
in a Gas Lift Using an Efficient Ant Colony University, 2018.
Algorithm (ACO)," Energy Sources, Part A: [41] M. Gen and L. Lin, "Genetic Algorithms and Their
Recovery, Utilization, and Environmental Effects, Applications," in Springer Handbook of Engineering
vol. 36, no. 11, pp. 1234-1248, 2014/06/03 2014. Statistics, H. Pham Ed. London: Springer London,
https://doi.org/10.1080/15567036.2010.536829 2023, pp. 635-674. https://doi.org/10.1007/978-1-
[29] E. R. Martinez, W. J. Moreno, J. A. Moreno, and R. 4471-7503-2_33
Maggiolo, "Application of Genetic Algorithm on the [42] R. Sarker, J. Kamruzzaman, and C. Newton,
Distribution of Gas-Lift Injection," in SPE Latin "Evolutionary optimization (EvOpt): a brief review
America/Caribbean Petroleum Engineering and analysis," International Journal of
Conference, 1994, vol. All Days, SPE-26993-MS. Computational Intelligence and Applications, vol. 3,
https://doi.org/10.2118/26993-MS no. 04, pp. 311-330, 2003,
[30] T. Ray and R. Sarker, "Genetic algorithm for solving https://doi.org/10.1142/S1469026803001051
a gas lift optimization problem," Journal of [43] M. Sakawa and K. Kato, "Genetic algorithms with
Petroleum Science and Engineering, vol. 59, no. 1, double strings for 0–1 programming problems,"
pp. 84-96, 2007/10/01/ 2007, European Journal of Operational Research, vol. 144,
https://doi.org/10.1016/j.petrol.2007.03.004 no. 3, pp. 581-597, 2003,
[31] E. Khamehchi, F. Rashidi, B. Karimi, P. Pourafshary, https://doi.org/10.1016/S0377-2217(02)00149-2
and M. Amiry, "Continuous gas lift optimization
using genetic algorithm," Australian Journal of Basic
and Applied Sciences, vol. 3, no. 4, pp. 3919-3929,
2009.

172
 S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174

[44] R. Sarker, K.-H. Liang, and C. Newton, "A new [49] J. M. Mach, E. A. Proano, H. Mukherjee, and K. E.
multiobjective evolutionary algorithm," European Brown, "A New Concept in Continuous-Flow Gas-
Journal of Operational Research, vol. 140, no. 1, pp. Lift Design," Society of Petroleum Engineers
12-23, 2002/07/01/ 2002. Journal, vol. 23, no. 06, pp. 885-891, 1983,
https://doi.org/10.1016/S0377-2217(01)00190-4 https://doi.org/10.2118/8026-PA
[45] E. Zitzler, K. Deb, and L. Thiele, "Comparison of [50] C. M. Laing, "Gas-Lift Design and Performance
Multiobjective Evolutionary Algorithms: Empirical Analysis in the North West Hutton Field," Journal of
Results," Evolutionary Computation, vol. 8, no. 2, pp. Petroleum Technology, vol. 43, no. 01, pp. 96-102,
173-195, 2000, 1991, https://doi.org/10.2118/19280-PA
https://doi.org/10.1162/106365600568202 [51] P. Nakashima and E. Camponogara, "Optimization of
[46] M. S. AL-Jawad and O. F. Hassan, "Prediction of lift-gas allocation using dynamic programming,"
optimum separation conditions for sequential field IEEE Transactions on Systems, Man, and
separation system," Journal of Engineering, vol. 11, Cybernetics - Part A: Systems and Humans, vol. 36,
no. 03, pp. 541-552, 2005, no. 2, pp. 407-414, 2006,
https://doi.org/10.31026/j.eng.2005.03.09 https://doi.org/10.1109/TSMCA.2005.855754
[47] O. Al-Fatlawi, M. M. Hossain ,and J. Osborne, [52] S. Huh, C. Park, J. M. Kang, and S. Kim, "The
"Determination of best possible correlation for gas Economic Optimization of a Continuous Gas Lift
compressibility factor to accurately predict the initial System Considering Lift and Cycle Efficiency in a
gas reserves in gas-hydrocarbon reservoirs," Mature Oil Field," Energy Sources, Part A:
International Journal of Hydrogen Energy, vol. 42, Recovery, Utilization, and Environmental Effects,
no. 40, pp. 25492-25508, 2017/10/0, 2017, vol. 32, no. 17, pp. 1614-1624, 2010/07/09 2010,
https://doi.org/10.1016/j.ijhydene.2017.08.030 https://doi.org/10.1080/15567030902842228
[48] U. A. H. Qasim and D. J. Sadeq, "Nodal Analysis of
Naturally Flowing Wells in Faihaa Oil Field,
Yamama Formation," The Iraqi Geological Journal,
vol. 57, no. 1B, pp. 122-139, 2024,
https://doi.org/10.46717/igj.57.1B.10ms-2024-2-19

173
 ‫‪S. H. O. Al-Mansory et al./ Iraqi Journal of Chemical and Petroleum Engineering 25, 2 (2024) 161 - 174‬‬

‫عملية الرفع بالغاز لتحسين حقل الزبير النفطي باستخدام المحاكاة العددية القائمة على‬
‫الخوارزمية الجينية‪ :‬دراسة الجدوى‬

‫‪4‬‬
‫سعد حازم عمران المنصوري ‪ ،* ،1‬عمر فالح الفتالوي ‪ ،3 ،2‬علي كخديدي‬

‫‪ 1‬معهد البصرة للتدريب النفطي‪ ،‬قسم اإلنتاج‪ ،‬البصرة‪ ،‬العراق‬

‫‪ 2‬قسم هندسة النفط‪ ،‬كلية الهندسة‪ ،‬جامعة بغداد‪ ،‬بغداد‪ ،‬العراق‬
‫‪ 3‬مدرسة غرب أستراليا للمناجم‪ :‬هندسة المعادن والطاقة والكيمياء‪ ،‬جامعة كيرتن‪ ،‬غرب أستراليا‪ ،‬أستراليا‬
‫‪ 4‬قسم علوم األرض‪ ،‬كلية العلوم الطبيعية‪ ،‬جامعة تبريز‪ ،‬تبريز‪ ،‬إيران‬

‫الخالصة‬

‫تعتبر طريقة رفع الغاز ضرورية للحفاظ على إنتاج النفط‪ ،‬خاصة من الحقول الحالية عندما يتم استنفاد‬
‫الطاقة الطبيعية من المكامن‪ .‬لزيادة إنتاج النفط‪ ،‬يجب توزيع معدل حقن الغاز في الحقل الرئيسي بأكبر قدر‬
‫ممكن من الكفاءة عبر نظام شبكي لعملية رفع الغاز‪ .‬قد تفقد تقنيات تحسين رفع الغاز الشائعة فعاليتها وتصبح‬
‫غير قادرة على تكرار رفع الغاز األمثل في نظام شبكة كبير بسبب مشاكل في توزيع معدل حقن الغاز متعدد‬
‫األهداف والمقيد؟ الهدف الرئيسي من البحث هو تحديد إمكانية استخدام تقنية الخوارزمية الجينية (‪ )GA‬لتحقيق‬
‫التوزيع األمثل لتدفقات حقن الرفع الغازي المستمر في هيكل شبكة حقل الزبير النفطي الذي يضم ‪ 10‬آبار حقن‬
‫الرفع الغازي‪ .‬وسيتم ذلك من خالل دراسات المحاكاة والنمذجة العددية‪ .‬تم العثور على الزيادة اإلجمالية في‬
‫معدل اإلنتاج المقدم من ‪ 15767‬ستب ‪ /‬يوم إلى ‪ 19847‬ستب ‪ /‬يوم‪ .‬يتم إجراء دراسات ضغط خزان البئر‬
‫وحساسية قطع المياه لدراسة التأثيرات المحتملة لهذه العناصر على البئر وكفاءته من خالل مجرى الحقل‪ .‬يتم‬
‫تعميق فهمنا للفوائد المحتملة الستخدام تقنيات رفع الغاز في الحقل من وجهة نظر فنية واقتصادية من خالل‬
‫استخدام أمثلة من التحليل االقتصادي‪ .‬عالوة على ذلك‪ ،‬على الرغم من أن فكرة استخدام ‪ GA‬بهذه الطريقة‬
‫ليست جديدة‪ ،‬فإن هذا العمل يناقش منهجيات التحسين المستندة إلى ‪ GA‬لزيادة معدل إنتاج النفط باستخدام رفع‬
‫الغاز في حقل نفط الزبير‪ .‬من أجل تعيين معدالت حقن الغاز آلبار محددة في شبكة في جميع أنحاء الحقل‬
‫باستخدام معدالت حق ن غاز محدودة‪ ،‬سيتم وضع نموذج التحسين خطوة بخطوة مما يسهل فهمه واستخدامه‬
‫كدليل‪ ،‬خاصة لفنيي اإلنتاج في الخطوط األمامية المشاركين في تطوير وتصميم أنظمة رفع الغاز‪.‬‬

‫الكلمات الدالة‪ :‬تحسين رفع الغاز‪ ،‬رفع الغاز‪ ،‬محاكاة رقمية‪ ،‬الخوارزميات الجينية‪ ،‬حقل الزبير النفطي‪.‬‬

‫‪174‬‬