See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/338625442

Applications of Artificial Intelligence in Oil and Gas Development

Article in Archives of Computational Methods in Engineering · January 2020

DOI: 10.1007/s11831-020-09402-8

CITATIONS READS

98 8,914

5 authors, including:

Hong Li Haiyang yu
CCDC Sulige Project Management Department, Erdos, Inner Mongolia 017000, China China University of Petroleum - Beijing
7 PUBLICATIONS 124 CITATIONS 79 PUBLICATIONS 1,425 CITATIONS

SEE PROFILE SEE PROFILE

Nai Cao He Tian

China University of Petroleum - Beijing 11 PUBLICATIONS 151 CITATIONS
16 PUBLICATIONS 192 CITATIONS
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Retention of Iron-Oxide Nanoparticles in Sandstone Rocks with High Salinity View project

Identification and Characterization of Waterflood-Induced Fractures in Tight Reservoirs View project

All content following this page was uploaded by Hong Li on 10 June 2021.

The user has requested enhancement of the downloaded file.

Archives of Computational Methods in Engineering (2021) 28:937–949
https://doi.org/10.1007/s11831-020-09402-8

ORIGINAL PAPER

Applications of Artificial Intelligence in Oil and Gas Development

Hong Li1 · Haiyang Yu1 · Nai Cao1 · He Tian1 · Shiqing Cheng1

Received: 10 July 2019 / Accepted: 2 January 2020 / Published online: 16 January 2020
© CIMNE, Barcelona, Spain 2020

Abstract
Artificial intelligence has been back on the stage of research works in all the walks in recently years, the sharply increase of
AI-based work have shown its potential to be a future direction for almost all disciplines. In oil and gas industry, AI technol-
ogy is also doubtlessly a new shining star that draws attention from researchers devoted themselves into it. In order to dig
up more about the applications of artificial intelligence in oilfield development for a hint of the future trend of this exciting
technology in oil and gas industry, literature investigation of a large amount of AI-based work reported has been conducted
in this work. Based on the investigation, the application of AI in important issues in oilfield development including oilfield
production dynamic prediction, developing plan optimization, residual oil identification, fracture identification, and enhanced
oil recovery are specifically investigated and summarized, the backs and cons of existing AI algorithms has been compared.
Based on the analysis and discussion, current situation of the application of AI in oilfield development is concluded, and
suggestions and potential directions of future work AI application in oil and gas developing are provided.

Abbreviations NARX Nonlinear auto regressive model with

ANFIS Adaptive network-based fuzzy inference system eXogenous
ANN Artificial neural network PCA Principal component analysis
APSO Adaptive particle swarm optimization PNN Polynomial neural network
BDA Big data analytics PSO Particle swarm optimization
BP Back propagation QPSO Quantum particle swarm optimization
CNN Convolutional neural network RMSE Root mean squared error
DM Data mining SD Standard deviation
FCM Fuzzy clustering method SOM Self-organizing maps
GA Genetic algorithm SRM Surrogate regulation model
GNN Graph neural network SVM Support vector machine
HIS Hybrid intelligent system WOB Weight on bit
Iot Internet of things technology
IPSO Improved particle swami optimization
LSSVM Least squares support vector machine 1 Introduction
MAPE Mean absolute percent error
ML Machine learning Artificial intelligence has a long history began since the year
MLPNN Multi-layer perceptron neural network 1950, when the British Mathematician Alan Turing asked
MSE Mean squared error the famous question “can machines think?” [1]. Artificial
intelligence (AI) was formally proposed and defined as a
new research field at the 1956 Dartmouth academic confer-
ence. Then came the first spring of artificial intelligence,
* Haiyang Yu when AI was quickly applied in various fields [2, 3]. Arti-
[email protected] ficial intelligence laboratories began to be set up in many
Hong Li countries, and experts back then believed that machines
[email protected] would soon replace humans in various areas. However, when
1
State Key Laboratory of Petroleum Resources
it came to the 1970s, due to the limitations of artificial intel-
and Prospecting, China University of Petroleum, ligence algorithms at that time, the development of artificial
Beijing 102249, China

13
Vol.:(0123456789)
938 H. Li et al.

intelligence was limited due to the inability to implement the application of AI technology in the petroleum field, in
large-scale or complex work. A few years later, with the order to organize regular relevant discussions. Based on the
widespread use of the “expert system”, AI began to flour- search result from the Onepetro platform, the number of
ish again [4]. However, since the “expert system” required articles on AI has increased significantly since 2000, whose
strong knowledge processing ability and high maintenance main algorithms include the artificial neural network (ANN),
cost, there was no significant breakthrough in the develop- fuzzy logic, support vector machine (SVM), hybrid intelli-
ment of AI. The rapid development of computers in the late gent system (HIS), genetic algorithm (GA), particle swarm
1990 s seems to have ushered in a new spring for AI. After optimization (PSO), etc. This suggests an increasing interest
more than 60 years’ ups and downs, AI came back to the of the researchers in the application of artificial intelligence
focus with the victory of AlphaGo over Lee Se-dol [1]. Then in the oil industry, and among all the algorithms, the ANN
in the year 2017, AlphaGo Zero showed up with high-speed is the most studied one (Fig. 1).
self-training without any human input, which aroused high At present, the application of AI in the oil and gas indus-
attention from all walks of life, and brought new thoughts to try is rapidly developing, as the concept of AI gradually
the development of AI in various fields [5]. With the devel- penetrates various stages of the industry, intelligent drill-
opment of cloud computing, big data, artificial neural net- ing, intelligent production, intelligent pipeline, intelligent
work, deep learning and other new technologies, it could be refinery, etc., and it will become the future research direc-
said that AI has achieved a new leap and changed our daily tion. With artificial intelligence algorithms, developers have
life as well. Driverless cars, accurate face recognition and developed a series of practical application technologies in
other artificial intelligence applications are no longer just exploration and development. In the field of exploration,
figments in the sci-fi movies. AI has been applied in almost utilization of the ANN method has already achieved good
all the aspects of daily life, as the core of the traditional results in reducing exploration risks and improving the suc-
energy industry, the oil and gas industry also embraces AI cess rate of exploration wells [8]. In the field of drilling,
to bring new technological breakthroughs for oil and gas new equipment such as automatic drilling rig and intelli-
exploration, development and production [6–8]. gent drill pipe have improved the drilling level and reduced
Application of AI in the field of petroleum engineering the cost significantly [9]. In the oilfield development, based
was proposed on the forum by international association as on the historical data of oilfield production to optimize the
early as in the 1970s. In 2009, SPE established the branch development plan is the main application mode of AI tech-
of “artificial intelligence in prediction analysis” to promote nology. In addition, AI has provided more accurate method

Fig. 1  AI the statistic of 400

increasing number of articles on
AI algorithms. Source: https​:// ANN
www.onepe​tro.org/ PSO
Fuzzy
SVM
300 GA
Numbers

200

100

0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Year

13
Applications of Artificial Intelligence in Oil and Gas Development 939

for fracturing scheme design and the selection of operating efficient and sustainable development of oilfield by compre-
wells and target layers [10]. hensive perception of oilfield dynamic change, automatic
Though many scholars have their achievements shared, manipulation, prediction and optimization of oilfield [13,
due to the abundant methods and contents of AI, there has 14]. AI oilfield plays a leading and guiding role in the infor-
been little specific summary and conclusion on the AI appli- matization of oil fields at all levels and is the future trend of
cation in the field of oilfield development. Production per- oil enterprises [15], for it can greatly reduce the oil produc-
formance prediction, development optimization, residual oil tion cost, improve the average oil field recovery, improve
identification, enhanced oil recovery, and correct identifica- the management efficiency of enterprises, and indirectly
tion and prediction of artificial fractures in unconventional promotes the economic and social development at the same
reservoirs, which are particularly important, are the core time. Trending to mature stage, the main future development
tasks in oilfield development. To embrace AI technology in direction of the digital oil field system is to deeply dig the
these works, it is of great significance to analyze and sum- oilfield data and integrate digital platform based on existing
marize the existing achievements. digital oilfield; to establish AI oilfield with the ability of
In order to dig up more about the applications of artificial prediction, warning, efficient analysis and optimization with
intelligence in oil and gas field development, in this work, the help of emerging technologies, such as internet of things
a great deal of related investigations have been conducted. (IoT) technology, the cloud computing technology, and the
Of course, the content in one single paper is limited, so we big data technology, etc. Deep integration of informatiza-
focused on certain important issues in oil and gas field devel- tion and industrialization is also a necessity to reduce costs
opment to narrow the scope. The issues discussed in this and to improve quality and efficiency. A truly competitive
paper included the comparison of main existing algorithms, oilfield should be one that can fully perceive, automatically
the application of AI in oilfield production dynamic predic- control, predict trends and optimize decisions [16, 17].
tion, in development optimization, in residual oil identifica- Many oil companies have launched intelligent oilfield
tion, in fracture identification, and the application of AI in projects to improve the quality of decision-making and man-
enhanced oil recovery. agement. Exploration and development multi-dimensional
In each section, the AI algorithms mainly adopted dealing environment platform software launched by Schlumberger
with specific issue were summarized, the backs and cons of can realize automatic drilling design utilizing IoT and other
which were compared, and the most suitable algorithm was technologies [18]. Kuwait national petroleum corporation’s
suggested. Finally, based on the analysis and investigation, digital oilfield (KwIDF) has been upgraded to an advanced
recommendations and potential directions of future work in intelligent workflow integrated aboveground and under-
the application of artificial intelligence in oilfield develop- ground systems [19]. In addition, AI-based intelligent man-
ment was proposed. agement assistants, such as oilfield robot, virtual oilfield
assistant and intelligent oilfield APP, can not only replace
human beings to deal with high-risk work, but also reduce a
2 From Digital Oil Field to AI Oil Field great many repetitive work [20] (Fig. 2).

The transition from digital oilfield to AI oilfield is inevitable.

The oil oilfield development involves huge data volume and 3 AI in Oil and Gas Development
unpredictable emergencies, without resource integration and
automatic management, ideal results could not be achieved The mainly adopted application of AI in reservoir devel-
and unexpected situations may even occur, though a good opment and exploitation is the optimization of develop-
many material and financial resource would be consumed. ment based on historical oilfield production data. Entering
Digital oilfield is advanced overall management supported the big data era, it is necessary for oilfield development to
by information technology, which includes a series of pro- fully explore the huge potential value of big data and to
cesses from exploration to production. This management can reveal hidden, previously unknown and potentially valuable
obtain data timely, share completely and achieve delicacy information, which is also one of the hot issues in AI oil-
management through analysis of production optimization field development research [21]. At present, AI has been
[11, 12]. Digital oilfield enterprises are therefore more effi- widely used in many industries (such as communication,
cient, creative and competitive. AI oilfield is an advanced e-commerce, etc.), but it still has a long way to go in the
version of digital oilfield, it is an advanced automatic iden- oil industry. This is mainly due to the particularity of data
tification treatment system that covers all the aspects in and applications in oilfield development. For example,
the oilfield based on the advanced computer technology, large data amount, data with high dimension and coupling,
automation technology, sensor technology and the profes- multi-source data with complex format, unstructured data,
sional technology. The treatment system could achieve more strong heterogeneity of the research objects, etc. [22]. With

13
940 H. Li et al.

Fig. 2  KwlDF workflow tools

architecture [19]

the development and application of big data and continu- Table 1  Evolutionary trend of various artificial intelligence tech-
ous improvement of various related algorithms, AI plays niques [23]
an increasingly important role in the field of oilfield devel- AI technique Year of developer
opment. Combined with other relevant new technologies development
(cloud computing, Internet of things, virtual reality, etc.),
Artificial neural network 1943 McCulloch and Pitts
new technologies and systems involved with AI will be con-
Fuzzy logic 1965 Lofti A. Zadeh
stantly proposed, which will surely become an important
Genetic algorithm 1970 John Holland
way to reduce costs and improve efficiency. Thus, researches
Case based reasoning 1977 Schank and Abelson
on AI and its application in oilfield development have a no
Support vector machines 1995 Vapnik V
doubt promising future and prospect. In this section, main
Particle swarm algorithm 1995 Eberhart and Kennedy
existing algorithms, the application of AI in oilfield produc-
Hybrid intelligent systems – –
tion dynamic prediction, in development optimization, in
residual oil identification, the in fracture identification, and
the application of AI in enhanced oil recovery are discussed
in detail respectively. mostly algorithm optimization. Since the current research
problems tend to be more complicated, which could be
3.1 Backs and Cons hardly solved by solely using one single algorithm, com-
bination of multiple algorithms and invention of new and
The proposer and year of the mainly adopted artificial intel- more intelligent algorithms have become the new targets
ligence algorithms are summarized by Agwu in Table 1 [23]. for researchers.
Almost all algorithms were proposed in the 1990s, and the Based on the history of artificial intelligence, in order
earliest and most mature one is the artificial neural network to better select a suitable algorithm and have it applied
algorithm, which has been applied in all walks of life with properly, understanding the advantages and disadvantages
good results. The purposes of the following algorithms are of each algorithm is more important than just knowing the

13
Applications of Artificial Intelligence in Oil and Gas Development 941

history itself. For this purpose, we summarize the advan- with fast matching speed and good precision, and further
tages and disadvantages of the commonly used algorithms become a new history matching method.
in the oil field as shown in Table 2 to provide a clear view Al-Thuwaini et al. [25] proposed a method for calcu-
of their advantages and disadvantages. ANN is the most lating root mean square (RMS), and performed history
commonly used and simplest algorithm, but it has higher matching based on artificial intelligence (ANN) combined
requirements of input parameters. PSO is easy to be real- with self organizing maps (SOM) method, with examples
ized and could be free from the problem information, but of history matching of model pressure and water contents
the accuracy is relatively low. The utilization of fuzzy logics as illustration. The SOM method can use geological and
doesn’t require mathematic model but its accuracy is low at hydraulic (static and dynamic) information to gather grids
the same time. SVM is suitable for small sample learning with similar behaviors, which can significantly improve
and is sensitive to real data. The GA method has a great the matching quality and reduce the number of times
parallelism, it could quickly search and be easily combined needed to achieve target matching.
with other algorithms, but has a more complex program- Mohaghegh et al. [26]. adopted data mining technol-
ming process and longer training time. Although the above ogy to conduct a top-down reverse modeling of shale
algorithms have certain limitations, suitable selection and reservoirs, which is quite different from the traditional
application of algorithm is the key to solve problems. bottom-up modeling from geological model to dynamic
prediction. AI and data mining (neural network, genetic
3.2 AI in History Matching of Oilfield algorithm and fuzzy logic, etc.) were carried out on the
established data to establish the cohesive modeling of the
The most important task in oil gas development is to pre- whole model, based on which the continuous fuzzy recog-
dict the future development based on the existing data and nition algorithm was used to transform discrete data into
make a reasonable development plan, history matching reservoir viscosity model. The top-down reverse modeling
is needed to facilitate the subsequent numerical simula- was validated with actual cases.
tion before prediction. However, due to the complexity of Costa et al. [27] used artificial neural network (ANN)
actual oilfield development that multiple factors interfere combined with GA to conduct production history match-
simultaneously, which increases the difficulty of history ing, which showed that the neural network could effectively
matching with conventional methods. With the develop- capture the nonlinearity of problems with a validation by
ment of AI, it has been proved to be feasible to apply synthetic reservoirs with actual reservoir characteristics. The
neural network to history matching of oilfield development work showed that neural network, as an agent model, has
[24]. With the history matching and training of the existing a good application prospect in history matching of oilfield
data, it can effectively capture the nonlinearity of problems with less simulation times and better fitting effect.

Table 2  The comparison between the advantages and disadvantages of algorithms

Method Advantages Disadvantages

ANN High classification accuracy, strong parallel distributed pro- Many parameters are required, such as network topology, initial
cessing ability, strong distributed storage and learning ability, values of weights and thresholds, output difficult to be inter-
strong robustness and fault tolerance to noise nerves, full preted, long training time, etc.
approximation to complex nonlinear relations, associative
memory function, etc.
PSO Free from the problem information, solve problem with real Low accuracy, prone to divergence, has certain dependence on
numbers, strong universality, few parameter adjustment, parameters, imperfection of the theory
simple theory, easy to achieve, collaborative search, fast
convergence
Fuzzy logics Precise mathematical models are not required, strong robust- Low accuracy, lack of systematic design
ness, easy to achieve
SVM Suitable for small sample machine learning problems; can Sensitive to missing data, no general solution for nonlinear
improve generalization performance, solve high dimensional problems
problems, solve nonlinear problems, and avoid Neural net-
work structure selection and local minimum point problem
GA Fast random search capability unconfined in problem domain, Complex programming, problem decoding is required after the
potential parallelism and good robustness, simple process, determination of optimal solution, accurate solution requires
use probability mechanism to conduct iteration with certain more training time, etc.
randomness, extensible and easy to be combined with other
algorithms

13
942 H. Li et al.

Shahkarmi et al. [28] developed the surrogate regulation be divided into methods need production data (oilfield
model (SRM) as the driver of history matching process with numerical simulation method, characteristic curve method,
the pattern recognition capability of artificial intelligence production decline method, material balance method, etc.)
and data mining (AI and DM). With history matching of 24 and the ones need not (analogy method, empirical for-
production wells in heterogeneous field had been operat- mula method and chart method, etc.) [29]. Although the
ing for 30 years, it showed that the high speed and accu- above methods have been applied in oilfield production
racy make SRM a fast and effective tool to assist history for years, they still have obvious limitations due to vari-
matching. ous and complex factors that affect the production index
The research on history matching has been relatively dynamic prediction. Prediction method based on AI for
mature but still needs to be further studied. It can be seen oilfield production performance has become an important
from the above studies that artificial neural network (ANN) research orientation because of its rapid development in
is commonly utilized to conduct training and history match- recent years. At present, a common way is to combine
ing oilfield development, and combine genetic algorithm neural network with fuzzy theory or intelligent algorithm
(GA) is often combined to optimize and improve the match- and thereby reach a great fitting accuracy, with static and
ing speed and accuracy. As we know, each step of oilfield dynamic production data involved. Production indexes are
development involves large number of parameters, which mainly used to evaluate the current development status of
is typical “big data”. It is urgent to develop artificial intel- oil fields (such as injection volume, cumulative production
ligence algorithms that can solve practical engineering prob- volume, etc.) and predict the dynamic change trend (such
lems to improve efficiency and reduce costs and risks. The as oil recovery rate, water cut rise rate, etc.) [30]. Eco-
problems or parameters targeted by various researchers are nomic indexes are mainly used to evaluate the input cost
uneven, but since current research has shown that the speed and economic benefits of oilfields. Artificial intelligence
and accuracy of history matching can be improved by arti- algorithms such as neural network has already had certain
ficial intelligence algorithm, meaning that AI application is achievements in the applications in oilfield production
feasible, what awaited to be done is in-depth and integration prediction, however, more practical and refined artificial
of research. Based on the summarization and discussion, for intelligence algorithm aiming at the particularity of data
history matching, we concluded and came up with the fol- gathered in the dynamic production of oil field, which is
lowing problems and potential suggestions: research worthy, is awaited to be further explored to real-
ize efficient oilfield development.
• The history matching process relies too much on artifi- To better understand the AI application in oilfield pro-
cial intelligence algorithm and lacks analysis of physical duction performance prediction at present, research results
significance of the results; in this aspect are summarized in Table 3 [31–39].
• The production law of oilfield development should be As shown in Table 3, many artificial intelligence algo-
summarized after history matching to make prediction rithms have been applied in oilfield development dynam-
and optimization of the developing plan; ics, the output parameters involved mainly focused on fluid
• The traditional history matching ignores the combina- production (oil production or total production) and water
tion with geological data, such as seismic data, logging content, while a little attention was paid on oil production
data, etc., so the adjustable parameters and their variation speed. The input parameters varied widely, from one sin-
range should be determined on the basis of fully under- gle parameter to large number of static and dynamic data.
standing the geological background. Based on this summarized situation, we came up with the
• The combination of AI and reservoir engineers can effi- following potential suggestions:
ciently optimize the best parameter combination from
multiplicity of solutions. • BP artificial neural network is the most widely used and
• The existing dynamic and static data should be deeply the most mature AI algorithm in production dynamic
mined, the artificial intelligence history matching system prediction at present, it could be optimized by combi-
should be formed by combining various algorithms, and nation with SVM and GA algorithm;
the relevant norms and standards should be established • The input objects and output parameters varied widely
by integrating resources of various platforms. in reported works, the correlation among parameters,
such as parameters that have important influence on the
3.3 AI in Production Dynamic Prediction of Oilfield output results and the correlation among them, should
be clarified;
Production index prediction plays an important role in • Prediction results should be summarized, reasonably
research on reservoir engineering and oilfield develop- analyzed and interpreted;
ment. Oil field production dynamic analysis methods can

13
Applications of Artificial Intelligence in Oil and Gas Development 943

Table 3  The application of AI in the dynamic prediction in oilfield development

References Method Input parameters Output parameters Errors

[31] ANN Gas prices; GDP growth rate; annual depletion; wells Gas production 0.0034 (MSE)
drilled; footage drilled and other properties.
[32] BP Temperature; pressure; superficial gas velocity; superficial Liquid holdup 8.544 (SD)
liquid velocity
[33] GNN + IPSO liquid producing capacity Water content 1.37% (MAPE)
[34] BP Remaining geological reserves; total number of production Monthly oil production 2% (MAPE)
wells, monthly injection–production ratio; kernel func- and liquid producing
tion; number of open injection wells, newly opened capacity
production wells and old wells with efficient treatment
[35] PCA + APSO + LSSVM Remaining geological reserves; injection–production ratio; Oil production 0.1485 (RMSE)
water content; number of open wells, open injection
wells, newly opened production wells and old wells with
efficient treatment
[36] ANN Porosity; velocity; horizontal permeability et al. Oil production cumulative –
[37] BP Deep; GR log; neutron log; density log; sonic log; deep Porosity; permeability –
resistivity log; dizgenesis
[38] MLPNN Distributed temperature sensing; distributed acoustic sens- Gas production –
ing; daily flowing time
[39] ANN + ANFIS Gamma ray; density; neutron; three different resistivity’s; Water sateration 0.07 (MSE)
caliper; porosity

• Production prediction is a dynamic process, and a Due to the mutual influence of technical, economic and
dynamic AI algorithm should be established to realize social indicators in the adjustment and optimization of
monitoring and prediction. developing plan, it is difficult to evaluate the performance
• AI has the potential to develop a multi-well dynamic of the plan with one single index. Therefore, it is necessary
prediction system considering inter-well and inter-layer to establish an efficient and effective method for develop-
interference on the basis of improving the prediction ing plan valuation and optimization. Traditional methods
accuracy of a single well. are relatively one-sided, so it is urgent to apply new intel-
• Existing big data should be fully exploited, the availabil- ligent optimization algorithm in solving the problems in
ity of those data should be dug up more to support more the oilfield development process.
accurate prediction. To better understand the AI application in oilfield devel-
oping plan optimization, research results in this aspect are
3.4 AI in Development Plan Optimization summarized in Table 4 [41–46].
Developing plan is crucial for oilfield development, due
In the 1960s, linear programming method has already to the complexity of the situation in the developing pro-
applied optimization in aiming at improving production effi- cess, its optimization should be done based on actual situa-
ciency of homogeneous reservoirs, some exploratory works tion. At present, AI application in developing plan optimi-
had been reported afterwards, but no attention was paid on zation are mainly to improve production or its rate with the
optimization of oilfield development back then. After the consideration of economic factors. As summarized above,
year 1985, the developing situation of oilfields changed, the input parameters in different work varied greatly, and
plan optimization was badly needed [40]. With long-term the adopted algorithm are also flexible, based on which,
water injection development, the oilfield gradually entered we put forward the following potential suggestions:
the period of high water cut, and great dissimilitude showed
in the heterogeneity between layers. These could be solved • The most commonly used algorithm for oilfield devel-
by recombination developing layers, the methods of which oping plan optimization is the ANN + GA mode, which
included fuzzy cluster analysis, grey correlation analysis, can be further improved in algorithm;
neural network algorithm, etc. Moreover, due to the pro- • The oilfield developing plan optimization should
gress of computer technology, evaluation and optimization include many aspects and multi-dimensions rather than
of developing plan began to draw more attention from major to evaluate solely from the economic aspect or produc-
companies in various countries since the oilfield developing tion result;
plan directly affected the life span and benefit of oilfields.

13
944 H. Li et al.

Table 4  The application of AI in the oilfield developing plan optimization

References Method Input parameters Output parameters

[41] BP Net present value; profit investment ratio; investment payback period; Expected value
total investment; average cost; cumulative oil production; oil and
gas ratio
[42] PNN +GA Realizable net present value; porosity; permeability; temperature; Gas production
initial pressure
[43] BP Final recovery; maximum recovery rate; net present value; total profit Expected value
and internal profitability; annual decline rate; total investment;
dynamic payback period and average unit cost
[44] NARX Energy production; energy use; final consumption expenditure; gross Predicting oil price movements
domestic product; gdp growth; gold price; oil rent
[45] QPSO 12 measures and unit price Annual production; cumulative production ratio
[46] FCM Fuzzy clustering parameters mainly include porosity, permeability Oil production; recovery and water content
variation coefficient, contained area, residual reserve abundance
and water flooding control extent

• Combining geological background, historical matching, oil saturation; digging potential residual oil, et al [47]. The
dynamic monitoring and economic benefit to achieve factors influencing the distribution of residual oil could
dynamic development plan optimization. be divided into two types, namely, the geological factors
• Instead of divergent research, a complete AI-based stand- (reservoir heterogeneity, structure, fault, etc.) and develop-
ard evaluation process and system should be established. ment factors (injection–production relationship, well pat-
tern distribution, production dynamic, etc.).The synthetic
3.5 AI in Identification in Oilfield Development effects of these factors leads to the diversity of residual
oil distribution. The displacement mechanism for different
3.5.1 The Application of AI in the Oilfield Development residual oil types are different, which set barriers to exca-
Identification vate residual oil and improve oil displacement efficiency.
Traditional modeling methods include deterministic mod-
AI identification technology has already been relatively eling and stochastic modeling, but they cannot predict res-
mature in face recognition, fingerprint recognition and ervoir parameters in time dimension [48]. Therefore, it is
other aspects, it also can play a significant role in the oil of great significance to integrate AI into the identification
industry. In oilfield developing process, in addition to of residual oil for both its great potential in the identifica-
development dynamic prediction and plan optimization, tion of residual oil morphology and model establishment.
residual oil identification could have the reservoirs to Some related works are shown in Table 5 [49–51].
be re-understood and provide necessary premise for the As shown in Table 5, applications of Ai in research on
adjustment of developing plan. The research methods of residual oil identification are not as so many as those in pro-
residual oil distribution mainly include geological method, duction dynamic prediction and developing plan optimiza-
reservoir engineering, well testing, numerical simulation tions. However, we still believe that it is a research worthy
method and laboratory experiment technique, etc. The orientation, the correct identification of residual oil is of
research works on residual oil mainly focuses on the dis- great significance to oil and gas development. The potential
tribution description of residual oil; monitoring remaining suggestions we concluded are as follow.

Table 5  The application of AI in the oilfield development identification

References Method Input parameters Output parameters

[49] BP Sand body type, injection–production relationship, connection status between The water flooded degree
the estimated well and the injection–water well, distance from the injection–
water well, and water injection status
[50] BP Net present value, profit investment ratio, investment payback period, total Expectations
investment, average cost, cumulative oil production, oil and gas ratio
[51] BP Concavity, roundness, aspect ratio, rectangularity, eccentricity, and radius Island shape, network shape, strip shape,
ratio column shape, plug shape, membrane
shape

13
Applications of Artificial Intelligence in Oil and Gas Development 945

• More research on AI should be conducted, the classifica- Research on fractures is one of the hot spots in petroleum
tion and displacement mechanism of residual oil should field, which mainly focuses on the morphological character-
be clarified to guide the development of residual oil. istics, property characteristics and classification character-
• Learning from more mature and advanced technical istics of fractures. The identification process is complicated
methods applied in other fields (such as route optimiza- because the reservoir contains natural fractures, artificial
tion in intelligent maps, rapid identification, etc.) could fractures and induced fractures, and the parameters involved
expand the thinking of o oilfield developing identifica- mainly rely on logging data. Application of AI in identifi-
tion. cation of fractures is a new and reliable research method.
Based on above discussion, we came up with the following
3.5.2 The Application of AI in Fracture Detection suggestions:

Great influences on developing effect of reservoirs, espe- • The seismic data has the advantage of continuous
cially the unconventional ones, could be brought by natural observation in space, so the inter well fractures may be
fractures and artificial fractures formed by fracturing, so the detected more accuracy by combining seismic data with
fracture identification work is great importance. Fractures logging data.
have always been a research topic in oilfield development. • Optimize and enrich the AI algorithm to improve the
Almost all the reservoirs have natural fractures, and in most identification accuracy to achieve dynamic identification;
unconventional reservoirs, such as shale gas and tight oil, • The integrated study of formation, opening and closing
many artificial fractures are formed by fracturing, making of fractures should be conducted with the application of
fractures in the entire reservoir a major factor affecting pro- AI.
ductivity [52]. The study on fractures could be generally • Reasonable analysis and interpretation of the identifica-
divided into qualitative and quantitative aspects–first quali- tion results should be made, and based on which reverse
tatively identify fractures and re-recognize the reservoirs guidance of optimization should be done.
and then quantitatively calculate the distribution of fractures
(azimuth, length, openness, fracture porosity, etc.). Although 3.5.3 The Application of AI in the Oilfield Diagnosis
many researches have been conducted on fracture identifica-
tion, due to the strong heterogeneity of reservoirs and com- In addition, safety ranks the priority in oilfield production.
plicated seepage characteristics caused by fractures, precise Measures and other operations are carried out on con-
identification of fractures remains to be a key problem await struction sites, which often rely on large number of data
to be solved. Some of the related works are summarized in monitoring systems. But due to the complexity and great
Table 6 [53–60]. uncertainty of underground conditions, it is difficult for

Table 6  The application of AI in fracture detection

References Method Input parameters Output parameters

[53] BP Orientation, aperture, surface roughness, alteration The conductive state of individual fracture
degree, mineral fillings, and other properties
[54] BP Neutron porosity, volume density, acoustic time dif- Fracture density in well log interpretation and core
ference, depth and shallow lateral resistivity, natural description
gamma spectrum
[55] BP Reflection intensity, Root Mean Squared amplitude Fracture density
and arc length properties, coherence, curvature, ant
body and instantaneous frequency
[56] ML +DBA +SVM Surface treating pressure, clean and slurry pump- Automatically classify hydraulic fractures
rates, surface and downhole amounts of 100 mesh
and 30/50 mesh sand proppants
[57] PSO + LSSVM Mud body integral number, resistivity differential No filling, argillaceous filling, siliceous filling and the
ratio, induced porosity, density ratio, acoustic time crystallization of filling
difference ratio and compensated neutron ratio
[58] CNN Borehole diameter, acoustic wave, neutron, density, Fracture reservoir types
gamma ray, shallow laterolog, deep laterolog
[59] ANN Bottom pressure; the regularization parameter Closure pressure
[60] ANN weight on bit; rotation per minutes; rate of penetra- Fracture pressure
tion; mud weight; pore pressure

13
946 H. Li et al.

traditional technology to spot the abnormal conditions in better nonlinear fitting ability and higher prediction accuracy
time. However, the AI algorithm can carry out deep learn- for the prediction of steam flooding development effect.
ing based on the background of big data and make timely Shi et al. [67] predicted fracturing effect (cumulative oil
judgments according to the actual situation, so as to improve production) based on grey correlation analysis and BP neu-
the diagnostic accuracy of staff and save the cost of time and ral network. Based on BP neural network, the selection and
economic. The identification involved are mainly the identi- optimization of water-controlled fracturing in bottom-water
fication of abnormal well conditions, indicator diagram and reservoir was conducted by Yang et al. [68], which had a
HSE warning in the production process. Some of the related more accurate optimization result and overcame the disad-
work are summarized in Table 7 [61–64]. vantages of previous nonlinear factors.
Timely identification and diagnosis of the problems The application of AI in EOR is mainly simulation pre-
encountered during the developing process is always the diction, which can greatly improve the efficiency and accu-
goal. With AI technology, fast identification could be real- racy. But it requires test data for verification. In addition, the
ized, some achievements have already been applied in fields, relationship between input and output parameters needs to
such as early warning with pressure detection and indicator be further studied. Though AI algorithms in EOR research
diagram. The follows are some suggestions based on above is still on its way to be mature, its application potential is
summarization and discussion. tremendous, thus it is a research orientation worth researcher
to pay close attention to.
• Improve the recognition accuracy and diagnosis accu-
racy of the algorithm, and make accurate interpretation
of results and right response to it; 4 Conclusion
• A thorough intelligent early warning response system
should be formed, which can response and solve prob- Based on the investigation of the application of AI in oil-
lems quickly. field development, it could be concluded that the intelligent
oilfield is on its way towards integration of business appli-
cation, coordination of decision and deployment, real-time
3.6 AI in EOR production management, visualization of comprehensive
research and sharing of information resources. AI oilfield
Since the 1950s, enhanced oil recovery (EOR) started its will eventually become an intelligent ecosystem integrated
important role in oilfield development. Up to the present the exploration, development, gathering, refining and man-
stage, the main EOR technologies applied in the field agement, etc. Based on the ecosystem, collaboration of all
include chemical flooding, thermal recovery, gas flooding, levels, regions and disciplines could be realized to extend the
etc. Exploring the application of AI in EOR may bring us life cycle of oilfield, improve the decision-making efficiency
new ideas and technological breakthroughs. and quality, reduce cost and increase economic benefit, and
Zhou et al. [65] used correlation analysis to select the finally fulfill the transition from digital oilfield to AI oilfield.
parameters could exert great impact on the EOR perfor- The following conclusions are drawn based on our investiga-
mance of polymer flooding, and applied polynomial regres- tion and discussion.
sion analysis and BP neural network to predict the nonlinear
and uncertain multivariable system. • Based SPE literature quantity survey, the number of AI-
Ni et al. [66] proposed a refined method to predict the related articles are increasing annually, research on AI
effect of steam flooding by a fusion of artificial bee swarm has become a research hotspot in recent years;
algorithm and radial basis function (RBF) neural network • ANN is the most commonly used and simplest algorithm,
algorithm, the work showed that the proposed method has a but it has higher requirements of input parameters. PSO

Table 7  The application of AI in the oilfield diagnosis

References Method Input parameters Output parameters

[61] ANN Active energy; kinematic viscosity; density; minimum pour point; molecular weight Wax appearance temperature
[62] BP Wellhead injection pressure Diagnostic output
[63] CNN Well fluid production, water content, oil pressure, casing pressure, stroke times, well- Probability of occurrence of dif-
head temperature, power consumption ferent fault types
[64] CNN Acquisition time, well ID, rod displacement, rod load, current, power, pressure, well- Diagnosis the working conditions
head temperature, pump depth, pump diameter, stroke length, frequency of stroke,
etc.

13
Applications of Artificial Intelligence in Oil and Gas Development 947

is easy to be realized and could be free from the problem 6. Liu W, Yan N (2018) Application and influence of artificial intelli-
information, but the accuracy is relatively low. Fuzzy gence in petroleum engineering area. Pet Technol Forum 4:32–40
7. Li DW, Shi GR (2018) Optimization of common data mining algo-
logics doesn’t require precise mathematic model but it rithms for petroleum exploration and development. Acta Pet Sin
has a low accuracy. SVM is suitable for small sample 39(2):240–246
training and is sensitive to real data. GA could search 8. Giuliani M, Cadei L, Montini M et al (2018) Hybrid artificial
quickly and be combined with other algorithms easily, intelligence techniques for automatic simulation models matching
with field data. In: Abu Dhabi international petroleum exhibition
but it requires long training time and complex program- & conference, 12–15 November. Society of Petroleum Engineers,
ming. Although all the algorithms have certain limita- Abu Dhabi, pp 1–11
tions, suitable selection and application of algorithm is 9. Hojageldiyev D (2018) Artificial intelligence in HSE. In: Abu
the key to solve problems. Dhabi international petroleum exhibition & conference, 12–15
November. Society of Petroleum Engineers, Abu Dhabi, pp 1–9
• BP neural network is the most widely used and most 10. Liu BJ (2015) Construction conception of intelligent oilfield. J
mature AI algorithm applied in oilfield. But to achieve Shengli Oilfield Party Sch 06:99–101
better performance in oil developing plan optimization, 11. Liu XD, Yang TT, Yan HY (2013) Forecast of oilfield indexes
GA or other algorithms should be combined; ANN is based on elman neural network and genetic algorithm. Comput
Mod 2:150–152
suitable for identification and diagnosis, combination of 12. Lien MO, Jansen JD, Brouwer DR (2006) Multiscale smart well
other algorithms could help to improve the result accu- management. In: Intelligent energy conference and exhibition,
racy and algorithm speed. 11–13 April. Society of Petroleum Engineers, Amsterdam, pp
• It is common for AI algorithms to rely too much on 1–11
13. Liu N (2014) Digital oilfield construction, smooth evolution to
data and ignore the physical relations between various intelligent oilfield. Silicon Val 4:191
parameters. In addition, dynamic AI research are far from 14. Yousef AA, Liu JS, Blanchard GW et al (2012) Smart waterflood-
enough. Thus, instead of blindly trust the results provided ing: industry. In: SPE annual technical conference and exhibition,
by AI, analysis and interpretation should be made to a 8–10 October. Society of Petroleum Engineers, San Antonio, pp
1–25
closed loop and solve practical problems more accurately. 15. Wang HF, Wang SJ, Zhu SB (2018) Conception and exploration
• The collection and processing of data is the key to realize of the smart oil and gas field construction in “internet +” era. Oil
the application of AI in oil and gas fields. Gas Field Surf Eng 8:1–8
• More efforts are still needed to realize AI oilfield, the 16. Anderson S, Barvik S, Rabitoy C (2019) Innovative digital inspec-
tion methods. In: Offshore technology conference, 6–9 May, Hou-
most important is to have the capacity to utilize and share ston, TX, pp 1–9
big data of oilfield and integrate intelligent systems at 17. Denney D (2006) To support digital oil fields. SPE 58(10):71–72
various stages. 18. Nadhan D, Mayani MG, Rommetveit R (2018) Drilling with digi-
tal twins. In: IADC/SPE Asia pacific drilling technology confer-
ence and exhibition, 27–29 August. Society of Petroleum Engi-
neers, Bangkok, pp 1–18
Acknowledgements The authors are grateful for financial support from 19. Rodriguez JA, Nasr H, Scott M et al (2013) New generation of
the National Natural Science Foundation of China (51874317) and the petroleum workflow automation: philosophy and practice. In: SPE
National Science and Technology Major Projects of China (Grant Nos. digital energy conference, 5–7 March. Society of Petroleum Engi-
2016ZX05037003). neers, The Woodlands, pp 1–13
20. Tan AA, Potts JK (1995) Digital log management system. SPE
Compliance with Ethical Standards 7(04):88–90
21. William WW, Balch RS, Stubbs BA (2002) How artificial intel-
Conflict of interest All authors declare no conflict of interest. ligence methods can forecast oil production. In: SPE/DOE
improved oil recovery symposium, 13–17 April. Society of Petro-
leum Engineers, Tulsa, pp 1–16
22. Gatta SR (1999) Decision tree analysis and risk modeling to
appraise investments on major oil field projects.In: SPE middle
References east oil show and conference. Information Engineering Research
Institute, USA, pp 187–202
23. Agwu OE, Akpabio JU, Alabi SB et al (2018) Artificial intelli-
1. Wang XL (2017) Application of artificial intelligence in oil and gence techniques and their applications in drilling fluid engineer-
gas industry. Mod Inf Technol 3(1):117–119 ing: a review. J Pet Sci Eng 167:300–315
2. Gilman H, Nordtvedt J E (2014) Intelligent energy: the past, the 24. Tang HK, Li XY (2015) BP neural network is applied to the pro-
present, and the future. In: SPE intelligent energy conference & duction dynamic analysis of an oil field. Oil Gas Field Surf Eng
exhibition, 1–3 April. Society of Petroleum Engineers, Utrecht, 4:20–21
pp 185–190 25. Al-Thuwaini J, Zangl G, Phelps RE (2006) Innovative approach
3. Yu WJ (2016) On the development trend and necessity of intel- to assist history matching using artificial intelligence. In: Intel-
ligent oilfield. Sci Wealth 10:500 ligent energy conference and exhibition, 11–13 April. Society of
4. Shi YJ (2016) Analysis of the research status of intelligent oilfield Petroleum Engineers, Amsterdam, pp 1–7
in China. Straits Technol Ind 12:81–83 26. Mohaghegh SD, Grujic OS, Zargari S, Dahaghi AK (2011) Mode-
5. Ma SX (2018) Technology focus: formation evaluation. SPE ling, history matching, forecasting and analysis of shale reservoirs
70(08):50. https​://doi.org/10.2118/0818-0050-JPT performance using artificial intelligence. In: SPE digital energy

13
948 H. Li et al.

conference and exhibition, 19–21 April. Society of Petroleum 44. Godarzi AA, Amiri RM, Talaei A et al (2014) Predicting oil price
Engineers, The Woodlands, pp 1–14 movements: a dynamic artificial neural network approach. Energy
27. Costa LAN, Maschio C, Schiozer DJ (2014) Application of Policy 68(5):371–382
artificial neural networks in a history matching process. SPE J 45. Sun H, Li PC (2016) Measures optimization for oil and water well
123:30–45 based on quantum particle swarm optimization. Comput Technol
28. Shahkarami A, Mohaghegh SD, Gholami V, Haghighat SA (2014) Dev 26(9):78–82
Artificial intelligence (AI) assisted history matching. In: SPE 46. Feng GQ, Pan LY, Kong B et al (2018) Hierarchical optimization
western North American and Rocky Mountain joint meeting, research based on fuzzy clustering analysis. Eval Dev Oil Gas
17–18 April. Society of Petroleum Engineers, Denver, pp 1–26 Reserv 3:30–39
29. Denney D (2011) Modeling history matching, forecasting, and 47. Roueché JN, Karacan CO (2018) Zone identification and oil satu-
analysis of shale-reservoir performance with artificial intelligence. ration prediction in a waterflooded field: residual oil zone, East
SPE 63(09):60–63 Seminole field, Texas, USA, Permian Basin. In: SPE improved
30. Zhu QJ, Gu YB, Lu SL et al (2002) Application of artificial neu- oil recovery conference, 14–18 April. Society of Petroleum Engi-
ral network model in oil resource prediction. East China Geol neers, Tulsa, pp 1–14
23(4):281–287 48. Denney D (2000) Artificial neural networks identify restimulation
31. Al-Fattah SM, Startzman RA (2001) Predicting natural gas pro- candidates. SPE 52(02):44–45
duction using artificial neural network. In: SPE hydrocarbon eco- 49. Du QL, Zhu LH (2009) Remaining oil distribution and its descrip-
nomics and evaluation symposium, 2–3 April. Society of Petro- tion technology during ultra-high water cut stage in La-Sa-Xing
leum Engineers, Dallas, pp 1–10 oilfi1eld. Daqing Pet Geol Dev 5:99–105
32. Osman ESA (2001) Artificial neural networks models for identify- 50. Liu J, Gao HY, Wang TT et al (2013) Parameters prediction of
ing flow regimes and predicting liquid holdup in horizontal mul- remaining oil of aII oil region in Daqing based on BP neural net-
tiphase flow. In: SPE middle east oil show, 17–20 March. Society works. Inner Mong Petrochem Ind 23:47–50
of Petroleum Engineers, Manama, pp 1–8 51. Mao GQ, Teng QZ, Wu Y et al (2014) Shape recognition of
33. Yan HY, Fu JY, Dong JH et al (2014) Forecast model of oilfield remained oil based on Bp neural network. J Terahertz Sci Electron
indexes based on IPSO GNN. Sci Technol Eng 14(15):197–202 Inf Technol 12(6):858–864
34. Xu SH, Bi CC, Zhang Y et al (2015) Oilfield development indi- 52. Denney D (2006) Treating uncertainties in reservoir-performance
cators prediction based on radial basis process neural network. prediction with neural networks. Soc Pet Eng 58(06):69–71
Comput Technol Autom 3:52–54 53. Thomas AL, La Pointe PR (1995) Conductive fracture identifi-
35. Feng GY, Han JX (2015) The oilfield production prediction model cation using neural networks. In: The 35th U.S. symposium on
based on principal component analysis and least squares support rock mechanics (USRMS), 5–7 June. American Rock Mechanics
vector machine. Comput Knowl Technol 11(31):144–147 Association, Reno pp 1–6
36. Gaurav A (2017) Horizontal shale well EUR determination 54. Zhang JC, Xing YZ, Zheng LH (2005) Using artificial intelligent
integrating geology, machine learning, pattern recognition and technique to identify fractures. Well Log Technol 29(1):52–54
multivariate statistics focused on the Permian basin. In: SPE liq- 55. Huang FX, Xia ZY, Gui HB et al (2016) The application of BP
uids-rich basins conference—North America, 13–14 September. neural network to DMT hill metamorphic fracture prediction.
Society of Petroleum Engineers, Midland, pp 1–19 Chin J Eng Geophys 13(4):483–490
37. Salem KG, Abdulaziz AAM, Abdel Sattar A, Dahab ASD (2018) 56. Anderson RN, Xie B, Wu L, Kressner AA, Frantz JH, Ockree
Prediction of hydraulic properties in carbonate reservoirs using MA, McLane MA (2016) Using machine learning to identify the
artificial neural network. In: Abu Dhabi international petroleum highest wet gas producing mix of hydraulic fracture classes and
exhibition & conference, 12–15 November. Society of Petroleum, technology improvements in the Marcellus shale. In: Unconven-
Abu Dhabi, pp 1–18 tional resources technology conference, pp 1–13
38. Ghahfarokhi PK, Carr T, Bhattacharya S, Elliott J, Shahkarami 57. Xie W (2017) Pso-LSSVM based identifying method for the
A, Martin K (2018) A fiber-optic assisted multilayer perceptron fracture-vug fillings. Daqing Pet Geol Dev 2:135–142
reservoir production modeling: a machine learning approach in 58. Wu ZY, Mo XW, Liu JH et al (2018) Convolutional neural net-
prediction of gas production from the Marcellus shale. In: SPE/ work algorithm for classification evaluation of fractured reser-
AAPG/SEG unconventional resources technology conference, voirs. Pet Explor 57(04):135–143
23–25 July, Houston, Texas, USA pp 1–10 59. Nande S (2018) Application of machine learning for closure pres-
39. Khan MR, Tariq Z, Abdulraheem A (2018) Machine learning sure determination. In: SPE annual technical conference and exhi-
derived correlation to determine water saturation in complex bition, 24–26 September. Society of Petroleum Engineers, Dallas,
lithologies. In: SPE Kingdom of Saudi Arabia annual technical pp 1–10
symposium and exhibition, 23–26 April. Society of Petroleum 60. Ahmed SA, Elkatatny S, Ali AZ, Abdulraheem A, Mahmoud M
Engineers, Dammam, pp 1–10 (2019) Artificial neural network ANN approach to predict fracture
40. Zhang ZX, Kang XF, Zhang FF (2012) Application of grey theory pressure. In: SPE middle east oil and gas show and conference,
for oil and gas reservoir evaluation program optimization. Adv 18–21 March. Society of Petroleum Engineers, Manama, pp 1–9
Mater Res 616–618:1008–1012 61. Adeyemi BJ, Sulaimon AA (2012) Predicting wax formation using
41. Kang XJ, Yuan AW, Gao J et al (2006) Neural network method artificial neural network. In: Nigeria annual international confer-
for comprehensive evaluation of oilfield development planning ence and exhibition, 6–8 August. Society of Petroleum Engineers,
scheme. Spec Oil Gas Res 13(2):48–50 Lagos pp 1–8
42. Park H, Lim JS, Kang JM, Roh J, Min B (2006) A hybrid arti- 62. Li TN (2015) Study on diagnosis model of oilfield abnormal well
ficial intelligence method for the optimization of integrated gas based on feedback dynamic neural network. Comput Technol
production system. In: SPE Asia pacific oil & gas conference and Autom 2:114–116
exhibition, 11–13 September. Society of Petroleum Engineers, 63. Liu BJ (2018) Research on diagnostic technique of indicator dia-
Adelaide, pp 1–9 gram based on CNN convolution neural network. J Xi’an Pet Univ
43. Xiao DR, Pan H (2010) Optimization of designing of the oil field 5:70–75
exploiting based on fuzzy mathematics and BP neural network. 64. Wang X, He Y, Li F, Dou X, Wang Z, Xu H, Fu L (2019) A work-
Microcomput Inf 26(6):209–211 ing condition diagnosis model of sucker rod pumping wells based

13
Applications of Artificial Intelligence in Oil and Gas Development 949

on big data deep learning. In: International petroleum technology 68. Yang ZH, Li ZP (2017) A new method for choice of water-control
conference 26–28 March, Beijing, China, pp 1–10 fracturing segments in horizontal wells based on the BP neural
65. Zhou CC, Li J, Zhang XG et al (2008) Predication for EOR by network system. Geol Explor 53(4):0818–0824
polymer flooding based on artificial neural network-comparison
between ANN and quadratic. Polynomial Stepwise Regres Method Publisher’s Note Springer Nature remains neutral with regard to
27(3):113–116 jurisdictional claims in published maps and institutional affiliations.
66. Ni HM, Liu YJ, Fan YC et al (2014) Optimization of steam flood-
ing injection and production based on improved particle swarm
optimization. Acta Pet Sin 35(1):114–117
67. Shi SZ, Yu HY, Sun ZL et al (2014) Forecast of fracturing effect
based on gray correlation analysis and BP neural network. J
Changjiang Univ (Self Publ Ed) 31:154–156

13

View publication stats