sensors

Article
An Energy Saving System for a Beam Pumping Unit
Hongqiang Lv 1 , Jun Liu 2, *, Jiuqiang Han 1 and An Jiang 1
1 School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an 710049, China;
hongqianglv@mail.xjtu.edu.cn (H.L.); jqhan@mail.xjtu.edu.cn (J.H.); bisgoon@163.com (A.J.)
2 School of Electrical Engineering, Xi’an Jiaotong University, Xi’an 710049, China
* Correspondence: eeliujun@gmail.com; Tel.: +86-29-8266-8665 (ext. 165)

Academic Editors: Mihai Lazarescu and Luciano Lavagno

Received: 26 March 2016; Accepted: 9 May 2016; Published: 13 May 2016

Abstract: Beam pumping units are widely used in the oil production industry, but the energy
efficiency of this artificial lift machinery is generally low, especially for the low-production well and
high-production well in the later stage. There are a number of ways for energy savings in pumping
units, with the periodic adjustment of stroke speed and rectification of balance deviation being
two important methods. In the paper, an energy saving system for a beam pumping unit (ESS-BPU)
based on the Internet of Things (IoT) was proposed. A total of four types of sensors, including load
sensor, angle sensor, voltage sensor, and current sensor, were used to detect the operating conditions
of the pumping unit. Data from these sensors was fed into a controller installed in an oilfield to adjust
the stroke speed automatically and estimate the degree of balance in real-time. Additionally, remote
supervision could be fulfilled using a browser on a computer or smartphone. Furthermore, the data
from a practical application was recorded and analyzed, and it can be seen that ESS-BPU is helpful in
reducing energy loss caused by unnecessarily high stroke speed and a poor degree of balance.

Keywords: energy saving; beam pumping unit; stroke speed adjustment; balance degree estimation;
surface dynamometer card; pump dynamometer card; Internet of Things

1. Introduction
Beam pumping units have been widely used in the oil production industry, which serves about
two-thirds of artificially-lifted oil wells [1,2]. When the natural drive pressure in a well is too low to
encourage crude oil to the surface, artificial pumping units can be employed to lift the fluid from the
reservoir. Even for the flowing wells, which contain enough energy for crude oil to rise to the surface
without any stimulation in the beginning, the natural drive pressure will reduce over time, and an
artificial lift process is also needed in the later stage of these oil wells to increase their production life.
Therefore, the artificial pumping unit is generally necessary for most oil wells [3]. There are nearly
920,000 oil wells all over the world, and about 87% of them are operated with the help of artificial
pumping units [4]. Different types of artificial pumps can be found in oil production fields, such as
sucker rod pumps, hydraulic pumps, plunger piston pumps, electrical submersible pumps, and so
on. Among them, sucker rod pumping units are well-known and generally understood, which can
be further divided into two categories including beam pumping units and non-beam pumping units.
The beam pumping unit accounts for approximately 71% of the artificial lift machinery market in the
oil production field [4,5].
The energy saving of the beam pumping unit plays an important role in solving the problems of
low profit, energy crisis, and environmental pollution during oilfield development. Beam pumping
units have a huge installed capacity around the world, and it is responsible for one-third [6] of the
total energy consumption in oilfields. Unfortunately, the energy efficiency of this kind of artificial
lift machinery is obviously low, although there is a certain degree of difference according to various

Sensors 2016, 16, 685; doi:10.3390/s16050685 www.mdpi.com/journal/sensors

Sensors 2016, 16, 685 2 of 13

well conditions. For example, in China, there are at least 100,000 beam pumping units in use, the total
installed capacity can reach up to 3500 MW, and the power consumption is approximately more than
10 billion kWh per year [7]. However, the energy efficiency of this artificial lift machinery ranges only
from 12% to 23% [8]. It is considered that the energy saving of beam pumping units has a lot of room
for improvement and a good application prospect.
There are three common ways in which the energy saving of beam pumping units can be carried
out for efficient oil production. Firstly, improvement of the mechanical structure. The mechanical
structure of the beam pumping unit is adjusted for energy saving, such as a double-horsehead pumping
unit, curved beam pumping unit, front-mounted pumping unit, and other heterotypic pumping
units [9,10]. Secondly, replacement of traditional motors. The traditional three-phase asynchronous
motor on the beam pumping unit is replaced by a specially-designed motor to improve the energy
efficiency, such as high-slip motor [11] and double-fed induction motor [12]. Finally, installation of
additional control devices. A novel electronic controller is introduced to adjust the operating state of
the beam pumping unit according to the conditions of the oil well for energy saving, such as a motor
control device with a power-off control strategy [13] and an energy saving system based on a wavelet
neural network [14]. It can be seen that only an additional controller is needed to be installed without
any change of the beam pumping unit itself in the third way.
The automatic adjustment of the stroke speed of the beam pumping unit is one of the most
effective means of energy saving in the oil production field. For a high-production oil well, it is usually
concerned with increasing the stroke speed of the pumping unit as much as possible to improve the
efficiency of oil production. However, for low-production wells and high-production wells in the later
stage, it is unacceptable that the economic value of the oil that is pumped from the reservoir to the
surface in each stroke is not greater than the cost of the consumed electricity. Some strategies have
been adopted to solve this problem. A novel power-off control strategy was proposed to serve for the
energy saving of the pumping unit [13] and a dual pulse width-modulation frequency converter was
introduced to keep the sucker rod running in an up-slowly-and-down-fast mode [7].
The balance deviation rectification of the beam pumping unit is also important for energy saving.
Once a pumping unit is well balanced, it is not only beneficial to safety, but can also convert the excess
energy of the motor into the gravitational potential energy of the counterweight on the beam during
the downstroke, and could help the motor to lift a heavier load with the conversion of this gravitational
potential energy during the following upstroke. On the contrary, if the counterweight of the beam
pumping unit is not properly matched, the motor would have to consume more power and convert
a part of it into the vibration and noise of the pumping unit. Recently, some new devices and ideas have
been proposed to solve this problem, such as an energy feedback device [15] and a redesigned beam
pumping unit [16], but these options have a much higher cost compared with the balance deviation
rectification of the original pumping unit. In practice, there are several alternative methods that can be
used to detect the degree of balance, such as ratios of time, peak current, total power consumption,
and peak torque between the downstroke and upstroke of the pumping unit. These methods have
high practical value, but the manpower cost is large without the help of an automatic balance degree
estimation system.
In this paper, an online energy saving system for beam pumping units (ESS-BPU) was proposed.
The stroke speed could be adjusted automatically according to the conditions of oil wells, the degree
of balance could be estimated in real-time, and the operating states can be monitored online using
a browser on a computer or smartphone so as to make a timely response to an abnormal situation
when necessary. The structure of the rest of the paper is organized as follows: Section 2 introduces the
principle of this energy saving system, including the overall architecture, design ideas, and necessity of
the Internet of Things (IoT); in Section 3, the methods that were used to determine the optimal stroke
speed, estimate the degree of balance, and evaluate the performance of pumping unit were explained.
Section 4 describes the hardware and software components of the energy saving system. In Section 5,
Sensors 2016, 16, 685 3 of 13
Sensors 2016, 16, 685 3 of 13

discussed, including the installation and result analysis. Finally, the conclusions are given in
Section
the 6.
application of this system in oilfield is discussed, including the installation and result analysis.
Finally, the conclusions are given in Section 6.
2. Principle of the Energy Saving System
2. Principle of the Energy Saving System
The overall architecture diagram of an ESS-BPU is shown in Figure 1. Efforts have been made
fromThe
the overall architecture
following diagram
three aspects of an ESS-BPU
to design the onlineissystem
shown for
in Figure 1. Efforts
the energy have
saving of been
beammade from
pumping
the following
unit. three aspects to design the online system for the energy saving of beam pumping unit.
‚ Automatic
Automatic adjustment
adjustmentofofthethestroke speed.
stroke A stroke
speed. speedspeed
A stroke determination unit and
determination unita motor
and a control
motor
unit are combined to carry out the automatic stroke speed adjustment. In
control unit are combined to carry out the automatic stroke speed adjustment. In the speedthe speed determination
unit, the load and
determination displacement
unit, the load andofdisplacement
the polished rod of the
of the beam rod
polished pumping
of theunit
beam arepumping
measuredunit by
aare
load sensor and an angle sensor, respectively. Then, these two types of data
measured by a load sensor and an angle sensor, respectively. Then, these two types of data are fed into a wave
equation to aget
are fed into a pump
wave dynamometer
equation to get a pumpcard,dynamometer
and the optimal card,stroke speed
and the is determined
optimal stroke speed so as
is
to maximizeso
determined theasarea of the pump
to maximize dynamometer
the area of the pump card. In the motorcard.
dynamometer control unit,
In the a frequency
motor control
converter is used to
unit, a frequency adjust theismotor
converter used toflexibly
adjusttothemake
motorsureflexibly
that thetopumping
make sure unitthat
is running at the
the pumping
optimal stroke speed.
unit is running at the optimal stroke speed.
‚ Estimation of the degree of balance.
balance. The input voltage and current of the three-phase motor of
the beam pumping unit is measured by a set of voltage and current sensors. Then, the real-time
active power of this motor is calculated, and the ratio of the total power consumption consumption between
the downstroke and upstroke is regarded as the indicator to assess the balance degree of the
pumping
pumping unit.
unit.
‚ Remote supervision ofofthe
Remote supervision the operating
operating states.
states. TheThe remote
remote supervision
supervision platform
platform consists
consists of twoofparts,
two
parts, including
including a database
a database and web and webunit,
server server
as unit, as awell
well as remote as amonitoring
remote monitoring
unit. The unit. The
database
database
and and web
web server server
unit unit
receive receive
the the status
real-time real-time
datastatus
sent data
from sent from oilfields
different differentby oilfields
means by of
means ofPacket
General General Packet
Radio Radio(GPRS),
Service Servicestores
(GPRS), stores
them them
in the in the database
database and deliversand web
delivers
pages web
to
pages toterminals.
remote remote terminals.
The remoteThemonitoring
remote monitoring
unit allows unit allows
clients toclients
monitor to the
monitor the operating
operating states of
states of pumping
pumping units using units using aon
a browser browser on a or
a computer computer
smartphone or smartphone
and make aand make
timely a timely
response to
response to
abnormal abnormal situations.
situations.

Figure 1. Overall architecture diagram of the energy saving system.

ESS-BPU could
ESS-BPU could be considered as
be considered as aa practical
practical application
application of
of IoT in the
IoT in the energy
energy saving
saving of
of pumping
pumping
units. The energy saving device in the oilfield and remote supervision platform online
units. The energy saving device in the oilfield and remote supervision platform online are integrated are
integrated together to serve the energy saving of the
together to serve the energy saving of the pumping unit. pumping unit.
The energy
The energy saving
saving device
device aims
aims to
to reduce
reduce the
the unnecessary
unnecessary energy
energy consumption
consumption by means of
by means of
automatic stroke speed adjustment and balance degree estimation. A total of four types
automatic stroke speed adjustment and balance degree estimation. A total of four types of sensors of sensors
are employed to detect the operating conditions of beam pumping unit, including the load and
Sensors 2016, 16, 685 4 of 13

are employed to detect the operating conditions of beam pumping unit, including the load and
displacement of a polished rod, as well as the input voltage and current of motor. The controller
receives data from these sensors, calculates the optimal stroke speed, estimates the degree of balance,
and adjusts the motor to make sure that the pumping unit is running at the optimal stroke speed.
However, the energy saving efficiency of this device will be greatly reduced while the pumping
unit works under abnormal conditions, such as sand production from the oil well, leakage of
tubing and paraffin deposits on pipe walls, and the existing device lacks the ability to detect these
complex anomalies. Therefore, it is necessary to collect and analyze the operating data manually
and periodically.
The remote supervision platform is designed for online monitoring of operating states and timely
responses to abnormal situations. Firstly, the pumping unit is often located in an inaccessible area and
has a relatively dispersed geographical distribution, this centralized monitoring platform can greatly
reduce the frequency of manual inspections over a long distance. More importantly, the information
on this platform is conductive to the judgment of abnormal situations and helpful to recovery of
the energy-saving efficiency of the device. The real-time, historical, and statistical data hosted in
the database server can be accessed through a terminal browser in some kind of meaningful form,
which allows clients to make a comprehensive analysis remotely. Thus, it is convenient to identify, or
even predict, which kind of abnormal condition the pumping unit is or will be under so that a timely
response can be made and the energy-saving efficiency of this device could be recovered as soon
as possible.

3. Methods

3.1. Method of Stroke Speed Determination

The optimal stroke speed is determined with the help of a pump dynamometer card of the beam
pumping unit. There are two kinds of common dynamometer cards that can be used for computer
diagnosis of the pumping unit, including a surface dynamometer card and a pump dynamometer
card [17]. A surface dynamometer card is a curve of load vs. displacement at the suspension point of
the polished rod throughout a complete stroke cycle, but it is often difficult to understand how the
downhole pump is operating from this kind of dynamometer card [18,19]. A pump dynamometer
card is a curve of load vs. displacement at the downhole pump throughout a complete stroke cycle.
It could be employed to diagnose the downhole operating conditions of the pumping unit, including
pump performance, tubing problems, leaking pump valves, fluid pound, and so on [20]. The area of
the pump dynamometer card is one of the most important methods to determine the applied work of
a downhole pump and the fluid yield of the pumping unit [20,21]; that is to say, the optimal stroke
speed could be determined so as to maximize the area of the pump dynamometer card.
Several methods have been proposed for the calculation of the pump dynamometer card,
one of the most common and effective methods is to solve the boundary value problem of a wave
equation given by Gibbs, based on the data corresponding to the surface dynamometer card [22,23].
The technique considers the polished rod as a transmission line, the downhole pump is a transmitter,
the suspension point of the polished rod at the surface is a receiver, the load and displacement at the
transmitter can be derived from the corresponding signals at the receiver.
The final solution of this wave equation is as follows:

n
$
σ0 υ0 ř
& Upx, tq “ 2EA x ` 2 ` pOn pxqcosnωt ` Pn pxqsinnωtq
’
’
„ n “1  (1)
n
’
% Fpx, tq “ EA
’ σ0
`
ř 1 1
pOn pxqcosnωt ` Pn pxqsinnωtq
2EA
n “1

where Fpx, tq and Upx, tq are the dynamic load and displacement at the depth x down along the
polished rod for time t separately, σ0 and υ0 are Fourier coefficients associated with the load and
Sensors 2016, 16, 685 5 of 13

displacement curves measured at the suspension point of the polished rod respectively, E and A are
the elastic modulus and cross-sectional area of the polished rod, n is the number of truncated Fourier
series, ω “ 2πT in which T is the period of a complete stroke cycle, and the definitions of On pxq, Pn pxq,
On1 pxq, and Pn1 pxq can be found in Gibbs’s original paper [22,23].
In this work, the load vs. time curve Fpx, t |x “ xs q and displacement vs. time curve Upx, t |x “ xs q
at the suspension point of the polished rod were measured by a load sensor and an angle sensor,
respectively, and the surface dynamometer card was acquired first. Then, the dynamic model given
ˇ
by Gibbs was employed to transform Fpx, t |x “ xs q and Upx, t |x “ xs q at the surface to Fpx, t ˇx “ x p q
ˇ
and Upx, t ˇx “ x p q at the downhole pump separately, and the pump dynamometer card was obtained.
Finally, the optimal stroke speed was determined to maximize the area of this pump dynamometer
card. Here xs denotes the position of the suspension point at the surface, and x p indicates the position
at the downhole pump down along the polished rod. In practice, the stroke speed was increased by
a specific incremental step, and the speed which can achieve the peak area was ultimately selected as
the optimal stroke speed.

3.2. Method of Balance Estimation

There are several laws that can be used to estimate the balance degree of beam pumping unit, such
as ratios of time, peak current, total power consumption, and peak torque between the downstroke
and upstroke. Among these methods, the time and current laws may be easier to implement, but their
accuracies are relatively low. The torque law could achieve a higher accuracy, but the measurement of
the torque of a motor is difficult. Additionally, the power law may be more practical compared with
the other laws.
In the paper, this power law was chosen to evaluate the balance degree of the beam pumping unit.
The input voltage Uptq and current Iptq of the motor were measured by a set of voltage and current
sensors, then the active power in a stroke cycle was obtained, and the ratio of total power consumption
between the downstroke and upstroke was calculated to assess the balance degree of the pumping
unit. The ratio λ is defined as: ş
pUptq ˆ Iptq ˆ cosϕqdt
tP T
λ “ şd (2)
pUptq ˆ Iptq ˆ cosϕqdt
tPTu

where Td and Tu are the half periods of the downstroke and upstroke in a complete stroke cycle,
respectively, and T “ Td ` Tu . In practice, the degree of balance of the beam pumping unit is
considered acceptable when λ P r0.8, 1.0s.

3.3. Method of Performance Evaluation

Several metrics could be used to evaluate the performance of an energy saving system. In the
paper, the active power saving rate ηa , reactive power saving rate ηr , and total energy saving rate ηt
were adopted to quantify the energy saving efficiency of the proposed system. The definitions of these
three metrics are as follows: $
’ ηa “ W1W ´W2
ˆ 100
’
’ 1
&
Q1 ´ Q2
ηr “ Q ˆ 100 (3)
’ 1
% ηt “ W1 ´W2 `Kq pQ1 ´Q2 q ˆ 100
’
’
W `Kq Q
1 1

where W1 and W2 are the active power consumption for lifting a ton of liquid up one hundred meters
before and after the use of the energy saving system, respectively, Q1 and Q2 are the reactive power
consumption for lifting a ton of liquid up one hundred meters before and after the use of this system
separately, and Kq is the reactive power economic equivalent.
where W1 and W2 are the active power consumption for lifting a ton of liquid up one hundred
meters before and after the use of the energy saving system, respectively, Q1 and Q2 are the
reactive power consumption for lifting a ton of liquid up one hundred meters before and after the
Sensors this 16,
use of 2016, separately, and K q is the reactive power economic equivalent.
685
system 6 of 13

4.
4. Composition
Compositionof
ofthe
theEnergy
EnergySaving
SavingSystem
System
ESS-BPU couldbe
ESS-BPU could beregarded
regardedasas a practical
a practical application
application of IoT
of IoT technology
technology in energy
in energy savingsaving of
of beam
beam pumping unit. The hardware of ESS-BPU mainly consists of a series of sensors
pumping unit. The hardware of ESS-BPU mainly consists of a series of sensors and a controller in and a
controller
oilfield as in oilfield
well as wellnumber
as a certain as a certain number ofwith
of computers computers
differentwith different
standards standards
online. online. The
The software of
software of ESS-BPU can be divided into two parts, including the software embedded
ESS-BPU can be divided into two parts, including the software embedded in the controller and in the
the
controller
supervision and the supervision
platform online. platform online.

4.1. Hardware

4.1.1. Main Hardware Components

The main hardware components of ESS-BPU are presented in Figure 2. It can be seen that the
energy saving device in the oilfield is primarily
primarily comprised
comprised of of four
four types
types of
of sensors
sensors and
and aa controller.
controller.
These sensors
sensorsinclude
includeload
loadsensor, angle
sensor, anglesensor, voltage
sensor, sensor,
voltage and and
sensor, current sensor.
current The controller
sensor. is the
The controller
center
is the for calculation
center and control
for calculation andofcontrol
this energy saving
of this system,
energy savingwhich mainly
system, contains
which a stroke
mainly speed
contains a
determination unit, motor control unit, balance degree estimation unit, human machine
stroke speed determination unit, motor control unit, balance degree estimation unit, human interaction
unit, andinteraction
machine GPRS module.unit, and GPRS module.

Figure 2. Main hardware components of the energy saving system.

Figure 2. Main hardware components of the energy saving system.

4.1.2. Sensors
4.1.2. Sensors
The sensors are employed to detect the operating conditions of the pumping unit
The sensors are employed to detect the operating conditions of the pumping unit automatically
automatically and continuously in the oilfield, thus the appropriate ones with certain mechanical
and continuously in the oilfield, thus the appropriate ones with certain mechanical and electrical
and electrical properties should be chosen to satisfy the needs of actual measurement. Table 1 gives
properties should be chosen to satisfy the needs of actual measurement. Table 1 gives some key
some key specifications of the four types sensors involved in this energy system.
specifications of the four types sensors involved in this energy system.
Table 1. Specifications of the four types of sensors.
Table 1. Specifications of the four types of sensors.
Sensor
Parameter Sensor
Parameter Load Sensor Angle Sensor Voltage Sensor Current Sensor
Load Sensor Angle Sensor Voltage Sensor Current Sensor
Measurement range 0~120 KN 0~360° 0~500 V 0~100 A
Measurement
Degree range
of precision 0~120
0.95%KN 0~360˝
0.05° 0~500
1.0%V 0~100
1.0%A
Degree of precision 0.95% 0.05˝ 1.0% 1.0%
Temperature
Temperaturerange
range −30~70℃
´30~70°C −40~85℃
´40~85°C −40~70℃
´40~70°C −40~70℃
´40~70°C

It is known that noise usually has a negative impact on the follow-up processing and analysis
of signals, especially for the weak analog signal [24]. In this study, the output analog signals of these
sensors are often mixed with a variety of noise caused by different kinds of external interference, thus,
measures were taken in two ways to reduce this noise before the analog signals arrive at the controller.
For each sensor terminal, a closed metal shell was used to shield noises in the process of analog signal
Sensors 2016, 16, 685 7 of 13

It is known that noise usually has a negative impact on the follow-up processing and analysis
of signals, especially for the weak analog signal [24]. In this study, the output analog signals of
Sensors 2016, 16, 685 7 of 13
these sensors are often mixed with a variety of noise caused by different kinds of external
interference, thus, measures were taken in two ways to reduce this noise before the analog signals
arrive at the controller.
amplification. For each
For the analog signalsensor terminal,
transmission viaathe
closed metal
wired way,shell was used
the cables with to shield
metal noises
shields were in
the process
chosen of analog
to connect signal
sensor amplification.
terminals and theFor the analog signal transmission via the wired way, the
controller.
cables with metal shields were chosen to connect sensor terminals and the controller.
4.1.3. Controller
4.1.3.The
Controller
controller in the oilfield is the core hardware component of the ESS-BPU, which can fulfill the
tasksThe
of sensor datain
controller receiving, stroke
the oilfield speed
is the coreadjustment, balance degree
hardware component of theestimation, humancan
ESS-BPU, which machine
fulfill
interaction,
the tasks ofand data data
sensor exchange with the
receiving, sever.
stroke Its functional
speed adjustment,block diagram
balance and physical
degree appearance
estimation, human
can be found
machine in Figures
interaction, and3 data
and 4exchange
respectively.
with The following
the sever. are the main
Its functional functions
block diagram and
and features of
physical
this controller:
appearance can be found in Figure 3 and 4, respectively. The following are the main functions and
features of this controller:
‚ Sensor data receiving. Sensor data is the basis of subsequent processing and analysis. A filter
 Sensor was
circuit data designed
receiving.toSensor
reduce data
theisnoises
the basis
from of environment,
subsequent processing
especiallyand fromanalysis. A filter
the frequency
circuit was designed to reduce the noises from environment, especially from
converter in this controller, itself. Then the filtered signals were fed into the analog-to-digital the frequency
converter in
converters forthis controller, itself. Then the filtered signals were fed into the analog-to-digital
digitization.
‚ converters for digitization.
Data calculation and motor control. The stroke speed determination unit was dedicated to
 Data calculation
calculation and motor
of the optimal control.
stroke speedThe stroke speed
by maximizing thedetermination
area of the pump unitdynamometer
was dedicated to
card.
calculation of the optimal stroke speed by maximizing the area of the pump
The motor control unit was used to adjust the motor speed flexibly by means of a frequency dynamometer card.
The motortocontrol
converter make unit
sure was
that used to adjustunit
the pumping the motor
running speed
at theflexibly
optimalby stroke
means speed,
of a frequency
and the
converter to make sure that the pumping unit running at the optimal
balance degree estimation unit was devoted to the estimation of balance degree of pumping stroke speed, and the
balance degree estimation unit was devoted to the estimation of balance
unit. In the controller, a dual-CPU embedded subsystem based on digital signal processor (DSP) degree of pumping
unit.advanced
and In the controller, a dual-CPU
RISC machine (ARM) embedded
was designed subsystem based the
to implement on tasks
digitalofsignal processor
calculation and
(DSP) and advanced
control, respectively. RISC machine (ARM) was designed to implement the tasks of calculation
‚
and control, respectively.
Human machine interaction. A LCD, a key board and several buttons on the front panel were
 Human machine interaction. A LCD, a key board and several buttons on the front panel were
used for the interaction between human and machine. The LCD was employed for the display of
used for the interaction between human and machine. The LCD was employed for the display
menu, real-time status data, curves of dynamometer cards, and so on. The keyboard was mainly
of menu, real-time status data, curves of dynamometer cards, and so on. The keyboard was
responsible for authentication and parameter setting, and the buttons were devoted to manual
mainly responsible for authentication and parameter setting, and the buttons were devoted to
operation in an emergency, such as system reset, emergency stop, and power switch.
manual operation in an emergency, such as system reset, emergency stop, and power switch.
‚ Data exchange with the server. The network connectivity is essential for a practical application
 Data exchange with the server. The network connectivity is essential for a practical application
of IoT technology in energy savings of a beam pumping unit. Considering the poor network
of IoT technology in energy savings of a beam pumping unit. Considering the poor network
conditions in an oilfield and the cost of long-range wireless communication, a GPRS module was
conditions in an oilfield and the cost of long-range wireless communication, a GPRS module
chosen to implement the real-time data exchange with the server online, instead of a 3G module
was chosen to implement the real-time data exchange with the server online, instead of a 3G
or even the Internet.
module or even the Internet.

Figure 3. Functional
Figure block
3. Functional diagram
block of the
diagram controller.
of the controller.
Sensors 2016, 16, 685 8 of 13
Sensors 2016, 16, 685 8 of 13

(a) (b)
Figure
Figure 4.
4. Physical
Physical appearance
appearance ofof the
the controller
controller for
for stroke
stroke speed
speed adjustment
adjustment and
and balance
balance estimation.
estimation.
(a) Front panel of the controller; and (b) inner construction of the controller.
(a) Front panel of the controller; and (b) inner construction of the controller.

4.2. Software
4.2. Software
4.2.1.
4.2.1. Embedded
Embedded Software
Software
The
The software
software embedded
embedded in in the
the controller
controller can
can bebe divided
divided into
into two
two parts
parts due
due to
to its
its dual-CPU
dual-CPU
architecture based on DSP and ARM. The software embedded in
architecture based on DSP and ARM. The software embedded in DSP is mainly used for theDSP is mainly used for the
digitization
digitization of analog
of analog signals as wellsignals as well as the
as the calculations calculations
of area of area of pump
of pump dynamometer carddynamometer card
and ratio of total and
power
ratio of total power consumption between the downstroke and upstroke.
consumption between the downstroke and upstroke. The software embedded in ARM is primarily The software embedded
in ARM is
devoted toprimarily
the overalldevoted
system to thescheduling,
task overall system
suchtask scheduling, such
as communication as the
with communication
DSP, human with the
machine
DSP, humandata
interaction, machine interaction,
exchange with the data exchange
server underwith the server
the help under
of a GPRS the help
module, of motor
and a GPRScontrol
module,
by
and motor control by means of a frequency converter. That is to say, the software embedded
means of a frequency converter. That is to say, the software embedded in the DSP and ARM is generally in the
DSP and ARM
responsible for is
thegenerally
calculationresponsible for tasks
and control the calculation
separately.and control tasks separately.

4.2.2. Remote
Remote Supervision
Supervision Platform
The remote
remote supervision
supervisionplatform
platformconsists
consistsofof
twotwo parts,
parts, including
including a database
a database and and
web web server
server unit,
unit,
as wellasaswell as a remote
a remote monitoring
monitoring unit. The unit. The database
database and web and web
server unitserver
on theunit on the
Internet Internet is
is responsible
responsible
for receiving, forstorage,
receiving,
and storage,
deliveryand delivery
of the of the
real-time real-time
status data of status
pumping dataunits
of pumping
sent fromunits sent
different
from different
oilfields by meansoilfields by The
of GPRS. means of GPRS.
remote The remote
monitoring monitoring
unit allows users tounit allows
access users
the data for to
anyaccess the
specified
data for any specified pumping unit in the platform and make
pumping unit in the platform and make a timely response to abnormal situations. a timely response to abnormal
situations.
A browser/server architecture is used in the remote supervision platform, so that the access to
A browser/server
this platform architecture
and information queryiscould
used inbe the remote
quite supervision
convenient. Theseplatform,
operations so can
thatbe
the accessout
carried to
this platform and information query could be quite convenient. These operations
using a browser on a computer or smartphone by any client with appropriate permission. The clients can be carried out
using a browser
can easily log intoonthea computer
platform viaor the
smartphone
Internet tobyapprove
any clientandwith appropriate
register a certainpermission.
beam pumping The
clients can then
unit. Only easily
canlog into the
various platformofvia
information thethe Internet
pumping to be
unit approve
queriedand andregister
read, justa like
certain
any beam
other
pumping unit unit.inOnly then can There
the platform. various areinformation
four kinds of ofinformation
the pumping unit befrom
available queried and read, just
the browser.
like any other pumping unit in the platform. There are four kinds of information available from the
‚ Static information. The information that will not change over a long period of time is declared
browser.
on the first child page. For example, the location, type, and various mechanical parameters of
 Static information. The information that will not change over a long period of time is declared
beam pumping unit, as well as the latest maintenance date, name of the person in charge, contact
on the first child page. For example, the location, type, and various mechanical parameters of
number, and so on.
beam pumping unit, as well as the latest maintenance date, name of the person in charge,
‚ Real-time status. A series of measured values in the latest action cycle can be retrieved from the
contact number, and so on.
database server in real-time, including the load and displacement of the polished rod, as well as
 Real-time status. A series of measured values in the latest action cycle can be retrieved from
the database server in real-time, including the load and displacement of the polished rod, as
Sensors 2016, 16, 685 9 of 13
Sensors 2016, 16, 685 9 of 13

wellinput
the as the inputand
voltage voltage
currentandofcurrent of motor.
motor. The Thedata
real-time real-time data is
is analyzed analyzed
and processedand processed
before being
before being
displayed displayed in
on webpages onanwebpages
acceptablein form,
an acceptable form, on
such as curves such as curves
a graph on a in
or figures graph or
a table.
figures in a table. The latest surface dynamometer card, stroke speed, ratio
The latest surface dynamometer card, stroke speed, ratio λ, peak/valley load, peak/valley current  , peak/valley
load,other
and peak/valley current
meaningful itemsand other
can meaningful
be found on theitems
secondcan be found
child page. on the second child page.
‚ Historical data. All kinds of status data within the past five days is
Historical data. All kinds of status data within the past five days temporarily retained
is temporarily retained inin the
the
database server, so that the historical status of pumping unit in a valid specified period can be
database server, so that the historical status of pumping unit in a valid specified period can be
queried and
queried and displayed
displayed on on the
the historical
historical data
data query
query page.
page.
‚ Statistical results.
Statistical results. The
Theoptimal
optimaloperation
operationstate
state since
since thethe beginning
beginning andand statistical
statistical results
results per
per day
daypermanently
are are permanentlystoredstored in the database
in the database server.
server. These These
data can bedata
usedcanto be
makeused to make a
a comparative
analysis and generate statistical reports of any number of days, and the reports couldthe
comparative analysis and generate statistical reports of any number of days, and reports
be printed
could be printed
out, if necessary. out, if necessary.

5. Results
Resultsand
andDiscussion
Discussion

5.1. Installation in Oilfields

ESS-BPUs have
ESS-BPUs have been installed
installed and
and applied
applied in a number of oilfields located in the northwest of
China, such as the Huachi oilfield in Gansu province, the Tuha oilfield in Xinjiang province, and the
Wuqi oilfield
oilfield in
in Shaanxi
Shaanxiprovince.
province.AAbeam
beampumping
pumpingunit unitand
andthetheenergy-saving
energy-savingdevicedeviceinstalled onon
installed it
areare
it shown
shown in Figure 5. Some
in Figure important
5. Some detailsdetails
important are magnified, including
are magnified, the actual
including theassembled positions
actual assembled
for the four
positions fortypes of sensors
the four types ofand the controller.
sensors It can be seen
and the controller. It canthat
bethe load
seen sensor
that was sensor
the load embeddedwas
in the carrier
embedded in bar
theofcarrier
the pumping unit,
bar of the the angle
pumping sensor
unit, the was
anglefixed at one
sensor wasend of the
fixed at central
one endrotating
of the
shaft of rotating
central the walking beam,
shaft thewalking
of the voltage and
beam,current sensorsand
the voltage werecurrent
mounted at thewere
sensors output power lines
mounted of
at the
the frequency converter, and the controller was employed to replace the original
output power lines of the frequency converter, and the controller was employed to replace the control cabinet for
the pumping
original controlunit. The sensors
cabinet for themeasured
pumpingthe operating
unit. state ofmeasured
The sensors the pumping the unit automatically
operating state of and
the
converted unit
pumping it intoautomatically
analog signals in real-time.
and convertedThe controller
it into analogreceived
signalsthein configuration
real-time. Theinformation
controller
entered bythe
received theconfiguration
installer, and implemented the tasksby
information entered of the
analog signal digitization,
installer, and implemented calculation, control,
the tasks of
and communication
analog with the
signal digitization, server online.
calculation, control, and communication with the server online.

Figure 5.
Figure The energy
5. The energy saving
saving device
device installed
installed on
on aa beam
beam pumping
pumping unit,
unit, with
with the
the actual
actual assembled
assembled
positions for
positions for the
the four
four types
types of
of sensors
sensors and
and the
the controller.
controller.
Sensors 2016, 16, 685 10 of 13
Sensors
Sensors
Sensors 2016,
Sensors2016,
2016, 16,
2016,16,
16, 685
16,685
685
685 10
10 of
10of
10 13
of13
of 13
13
Sensors
Sensors
Sensors
Sensors 2016,
2016, 16,
16, 685
685 10
10 of
of 13
13
Sensors 2016,
Sensors 2016, 16,
2016,
2016, 16, 685
16,
16, 685
685
685 10
10 of
10
10 of 13
of
of 13
13
13
5.2.
5.2. Validation
5.2. Validation
5.2. Validation
5.2. Validation
Validation of
of theof the
the Utility
the Utility
ofUtility
of the Utility
Utility of
of ESS-BPU
of ESS-BPU
of ESS-BPU
of ESS-BPU
ESS-BPU
5.2.
5.2. Validation
5.2. Validation of
Validation of the
of the Utility
the Utility of
Utility of ESS-BPU
of ESS-BPU
ESS-BPU
5.2.
5.2. Validation
Validation of
of the
the Utility
Utility of
of ESS-BPU
ESS-BPU
The
The
The actual
The actual
The actual actual
actual data
datadata
data
data for
for
for the
for the
for the No.
the No.
the No.
M45
No.
No. M45
M45
M45
M45 beam
beam
beam pumping
pumping
beampumping
beam pumping unit
pumping unit
unit
unit in the
unit in
in
in the
the
the Tuha
theTuha
Tuha
Tuha oilfield
Tuhaoilfield
oilfield
oilfield
oilfield was
was
was
was
was recorded
recorded and
and
recorded
recorded
recorded andand
and
The
The
The actual
actual
actual data
data
data for
for
for the
the
the No.
No.
No. M45
M45
M45 beam
beam
beam pumping
pumping
pumping unit
unit
unit in
in
in the
the
the Tuha
Tuha
Tuha oilfield
oilfield
oilfield was
was
was recorded
recorded
recorded and
and
and
The
The
analyzed
analyzed
analyzed
analyzed actual
actual
in
in
in
in data
data
order
order
order
order to
to
to
to for
for the
the
validate
validate
validate
validateNo.
No.
the
theM45
M45
the
the beam
beam
utility
utility
utility
utility of
of
of
of pumping
pumping
this
this
this
this unit
unit
system,
system,
system,
system, in
in the
the
including
including
including
including Tuha
Tuha
the
the
the
the oilfield
oilfield
stroke
stroke
stroke
stroke was
was
speed
speed
speed
speedrecorded
recorded and
and
adjustment,
adjustment,
adjustment,
adjustment,
analyzed in order
analyzed
analyzed
analyzed in
in
in to validate
order to the
validateutility
the of this
utility system,
of this including
system, the
including stroke
the speed
stroke adjustment,
speed balance
adjustment,
analyzed
analyzed
balance
balance
balance
balance in order
in
degree
degree
degree
degree
order
order
order to
to validate
to
to validate
validate
validate
estimation,
estimation,
estimation,
estimation,
the
the
the
remote
remote
remote
remote
utility
utility of
utility
utility of this
of
thesupervision,
of
supervision,
supervision,
supervision,
this
this
this system,
system,
system,
andsystem,
and
and
and
including
including
including
including
efficiency
efficiency
efficiency
efficiency
the
the stroke
the
the
evaluation.
evaluation.
evaluation.
evaluation. stroke speed
stroke
stroke speed adjustment,
speed
speed adjustment,
adjustment,
adjustment,
degreebalance
estimation,
balance
balance degreeremote
degree
degree supervision,
estimation,
estimation,
estimation, remote
remote and efficiency
supervision,
remote supervision,
supervision, and
and evaluation.
efficiency
and efficiency
efficiency evaluation.
evaluation.
evaluation.
balance degree
balance degree estimation,
estimation, remote
remote supervision,
supervision, and
and efficiency
efficiency evaluation.
evaluation.
Table
Table
Table 2.
Table 2.
2. Results
2. Results
Results of
Results of
of the
of the
the automatic
the automatic
automatic stroke
automatic stroke
stroke speed
stroke speed adjustment
speed adjustment
speed using
adjustment using
adjustment this
using this
using energy
this energy
this saving
energy saving
energy system
saving system
saving on
system on
system the
on the
on the
the
Table
Table 2.
2. Results
Results
Results of
of
of the the
the automatic
automatic stroke
stroke speed
speed adjustment
adjustment using
using this
this energy
energy saving
saving system
system on
on the
the
Table No.
2.
Table
Table
Table
Table
No.
No.
2.
2. Results
2.
2.
M45
No.M45
M45
M45
Results
Results
Results
beam
beam ofautomatic
of
of
of the
the
pumping
beampumping
beam pumping
pumping
stroke
the automatic
the automatic
automatic
automatic
unit.
unit.
unit.
unit.
speed
stroke
stroke
stroke
stroke adjustment
speed
speed
speed
speed adjustment using
adjustment
adjustment
adjustment using
using thisenergy
using this
using this
this
this energysaving
energy
energy
energy saving
saving
saving
saving system
system
system
system
system on
on on
the the
on the
on the
the
No.
No.
No. M45
No.
No. M45
M45
beam
M45
M45 beam
beam pumping
pumping
pumping
beam
beam pumping unit.
unit.
unit.
pumping unit.
unit.
No. M45
No. M45 beam
beam pumping
pumping unit.
unit.
Area
Area
Areaof
Area of Pump
ofPump
of Pump
Pump
Stroke
Stroke
Stroke
Stroke Pump
Pump Dynamometer
PumpDynamometer
Pump Dynamometer
Dynamometer Area
Area
Area of
of
Area of Pump
Pump
of Pump
Pump
Stroke
Stroke Surface
Surface Dynamometer
Dynamometer Card
Card Pump
Pump Dynamometer
Dynamometer Area
Area of
of Pump
Pump
Dynamometer
Area of Pump
Dynamometer
StrokeStroke
Stroke
Stroke
Stroke
Speed
Speed
Speed
Surface
Surface
Surface
Surface
Surface
Surface
Dynamometer
SurfaceDynamometer
Dynamometer
Dynamometer Card
DynamometerCard
Card
Card
Card Pump
Pump
Pump
Pump
Pump Dynamometer
Dynamometer
Dynamometer
Dynamometer
Dynamometer
Card
Card Card Dynamometer
Dynamometer
Dynamometer
Dynamometer
Speed
Speed
Speed
Speed Surface Dynamometer
Surface Dynamometer Card
Dynamometer
Dynamometer Card
Card
Card Card
Card
Card
Card Dynamometer
Dynamometer
Dynamometer
Dynamometer
Dynamometer
Card
Card Card
Speed
Speed
Speed
Speed Card
Card
Card
Card Card
Card
Card
Card
Card
Card
Card
Card

4 min ´1 −1 50,717.81
50,717.81
50,717.81
4444minmin
min
−1
−1
min−1−1 50,717.81
50,717.81
444444 min
min
min
min
−1
−1
−1 50,717.81
50,717.81
50,717.81
50,717.81
min
min −1
−1 50,717.81
50,717.81

51,623.65
51,623.65
51,623.65
51,623.65
5555min
5 min min
´1
min
−1
−1
−1
min−1−1 51,623.65
51,623.65
51,623.65
51,623.65
51,623.65
555555 min
min
min −1
min−1
−1 51,623.65
51,623.65
min
min −1
−1

6666minmin
min
´1
−1
−1
−1
min−1−1 52,245.02
52,245.02
52,245.02
52,245.02
6 min
666666 min
min
min −1
min−1
−1 52,245.02
52,245.02
52,245.02
52,245.02
52,245.02
min
min −1
−1 52,245.02
52,245.02

7777minmin
min
−1
−1
−1
min−1−1 43,410.45
43,410.45
43,410.45
43,410.45
777777 min
7 min min
min
´1 −1
min−1
−1 43,410.45
43,410.45
43,410.45
43,410.45
43,410.45
min
min −1
−1 43,410.45
43,410.45

8888minmin
min
−1
−1
−1
min−1−1 35,882.41
35,882.41
35,882.41
35,882.41
888888 min
min
min −1
min−1
´1 −1 35,882.41
35,882.41
35,882.41
35,882.41
8 min min
min −1
−1 35,882.41
35,882.41
35,882.41

For
For
For the
For the
the automatic
the automatic
automatic adjustment
automatic adjustment
adjustment of
adjustment of
of stroke
of stroke
stroke speed,
stroke speed, an
speed, an
speed, incremental
an incremental
an stroke
incremental stroke
incremental frequency
stroke frequency
stroke strategy
frequency strategy
frequency was
strategy was
strategy was
was
For
For
For the
the
the automatic
automatic
automatic adjustment
adjustment
adjustment of
of
of stroke
stroke
stroke speed,
speed,
speed, an
an
an incremental
incremental
incremental stroke
stroke
stroke frequency
frequency
frequency strategy
strategy
strategy was
was
was
For
For
employed
employed
employed
employed the
theto
to
to
to automatic
automatic
find
find
find
find the
the
the
the adjustment
adjustment
optimal
optimal
optimal
optimal stroke
stroke
stroke
stroke of
of stroke
stroke
speed
speed
speed
speed which
whichspeed,
speed,
which
which can
can
can
can an
an incremental
incremental
maximize
maximize
maximize
maximize the
the
the
the stroke
stroke
area
area
area
area of
of
of
of frequency
frequency
the
the
the
the pump
pump
pump
pump strategy
strategy
dynamometer
dynamometer
dynamometerwas
was
dynamometer
employed
For theFirst
employed
employed to
to
to find
automatic
find
find the
the
the optimal
adjustment
optimal
optimal stroke
stroke
stroke speed
of stroke
speed
speed which
speed,
which
which can
can
can maximize
anmaximize
incremental
maximize the
the
the area of
strokethe pump dynamometer
frequency strategy was
employed
employed
card.
card. Firstto
card. First
card. First to
of
of find
offind
of all,
all, the
all, the
all, the optimal
theoptimal
the energy
energy stroke
energy stroke
energy saving
saving speed
savingspeed
saving device
device
device
devicewhich
which can
can
installed
installed
installed
installed maximize
maximize
on
on
on the
on the
the the area
the
beam
the beam
beam
beam
area
area
area of
of the
of
of
pumping
pumping
pumping
pumping
the
the
the pump
pump
pump
pump
unit
unit
unit
unit
dynamometer
dynamometer
dynamometer
dynamometer
was
was
was
was configured
configured
configured
configured
card.
card.
card.
employed First
First
First of
of
of all,
all,
all, the
the
the energy
energy
energy saving
saving
saving device
device
device installed
installed
installed on
on
on the
the
the beam
beam
beam pumping
pumping
pumping unit
unit
unit was
was
was configured
configured
configured
card.to
card. findof
First
First oftheall,optimal
all, the stroke
the energy
energy speed
saving
saving which
device
device can maximize
installed
installed on the
on the beamthe pumping
beam area of the
pumping pump
unit
unit was dynamometer
was configured
configured
card. First of all, the energy saving device installed on the beam pumping unit was configured
according to the conditions of oil well and pumping unit. The device was allowed to measure and
Sensors 2016, 16, 685 11 of 13

control once every 20 min, each containing 100 sampling points. The minimum allowable stroke
speed was assigned to 3 min´1 , the maximum allowable stroke speed was set to 10 min´1 , and the
incremental step was chosen as 1 min´1 . Then the surface dynamometer card was acquired based
on the load and displacement data measured at a specific stroke speed, the corresponding pump
dynamometer card was obtained by solving the Gibbs’s wave equation, and the optimal stroke speed
was finally determined to be 6 min´1 so as to the area of the pump dynamometer card could reach
the peak value 52,245.02. Table 2 illustrates some of the important points in the process of finding the
optimal stroke speed. Since the apply work of downhole pump and the fluid yield of pumping unit
could be estimated by the area of the pump dynamometer card, it can be inferred that the device has
an ability to avoid energy loss caused by unnecessary high stroke speed while ensuring the liquid
yield of the pumping unit. The ability has also been verified in other field tests, however, it was found
that this method, which determines the optimal stroke speed so as to maximize the area of the pump
dynamometer card, still has some limitations. It cannot be well applied to some special issues, such as
sand production from oil wells, leakage of tubing, and paraffin deposit on pipe walls.
For the estimation of the degree of balance, two different states of the beam pumping unit before
and after the balance adjustment were compared. The ratio λ given by the energy saving device and
the motor power parameters measured by a power testboard under the two states were shown in
Table 3. It can be seen that the maximum power, minimum power and average power of the motor
when λ “ 0.92 are all smaller than those when λ “ 0.65. That is to say, the motor consumes relatively
less power while the value of this ratio is within the balance range, and vice versa. The comparison
result could be indirect evidence of the effectiveness of the ratio in the evaluation of the balance degree
of pumping unit.

Table 3. Balance ratio and the corresponding motor power parameters under two different balance
states of the No. M45 beam pumping unit.

Ratio λ Maximum Power Minimum Power Average Active Power Average Reactive Power
0.65 11.46 KW ´5.88 KW 8.24 KW 5.73 KW
0.92 7.90 KW ´3.14 KW 6.83 KW 4.75 KW

The static information, real-time status, historical data and statistical results of this beam pumping
unit can be retrieved from the remote supervision platform. The static information, such as the location,
type, and various mechanical parameters, was declared on the first child page. The real-time status,
including the surface dynamometer card, stroke speed, ratio λ, peak/valley load, peak/valley current,
and other meaningful items for the latest action cycle, were represented on the second child page
(Figure 6a). The historical data query page was very similar to this real-time status page except for the
extra selection of a past time period. For the statistical results, there are a number of options to choose
according to the prevailing concerns, herein the motor power in the past five days were selected to
be shown in Figure 6b. It is hoped that the information in this remote supervision platform could be
conducive to comparative analysis and timely response.
The overall energy saving efficiency of the proposed system varies with individual well condition,
type of pumping unit, and different time periods. In the paper, this efficiency of the system on
the No. M45 beam pumping unit was evaluated based on the actual statistical data. The power
consumption and liquid yield data over a two-month test period was recorded and compared.
The statistical results show that, due to the use of this system, the active and reactive power saving
rate of this pumping unit can reach up to 15.45% and 13.83%, respectively, and the average total energy
saving rate is 15.32%. It is noteworthy that the power consumption of this system itself, especially the
frequency converter, has been taken into account in this efficiency evaluation.
Sensors 2016, 16, 685 12 of 13

Figure 6. Partial information of the No. M45 beam pumping unit in the remote supervision platform.
(a) Real-time status; and (b) statistics of motor power.

6. Conclusions
In this paper, an energy saving system for a beam pumping unit (ESS-BPU) in oil production field
was proposed. ESS-BPU has three main functions, including automatic adjustment of the stroke speed,
estimation of balance degree and remote supervision online. The real-time status data from a total of
four types of sensors is fed into a controller installed in oilfield to implement the tasks of stroke speed
adjustment and balance degree estimation, these information is transmitted to a remote supervision
platform by means of General Packet Radio Service (GPRS), and clients are allowed to monitor the
operating states of pumping unit using a browser on a computer or smartphone so as to make a timely
response to abnormal situations when necessary. One of the advantages of this proposed system is
the combination of intelligent instrument and Internet of Things (IoT) in the energy saving of beam
pumping unit, we can stay in the office, monitor the concerned conditions of pumping unit revealed
by this energy saving device from the sensor data, and respond to it only when necessary. In addition,
it has been found that several methods have been put forward for automatic regulation of balance and
estimation of liquid yield of a beam pumping unit. It is hoped that some of them which can fulfill the
actual needs very well could be introduced, so that the energy budget of each pumping unit could be
allocated reasonably to optimize the global energy consumption in oilfields.

Acknowledgments: This work was supported by grants from China Postdoctoral Science Foundation
(No. 2015M580851) and Fundamental Research Funds for the Central Universities (No. XJ20150420).
Author Contributions: The work presented in this paper is a collaborative development by all of the authors.
In particular, J.H. is the main researcher of this project and proposed the system; H.L. carried out a literature survey,
developed the hardware, and wrote the paper; J.L. developed the algorithm and the software; A.J. contributed to
data collection and modification of the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Ma, Z.; Peng, S.L.; Qu, Z.Z.; Li, J. The detailed calculation model of the friction between sucker rod and the
liquid in the sucker rod pump lifting system of heavy oil. Appl. Mech. Mater. 2014, 694, 346–349. [CrossRef]
2. Liu, F.; Patel, A. Well failure detection for rod pump artificial lift system through pattern recognition.
In Proceedings of the International Petroleum Technology Conference, Beijing, China, 26–28 March 2013.
3. Goswami, S.; Chouhan, T.S. Artificial lift to boost oil production. Int. J. Eng. Trends Technol. 2015, 26, 1–5.
[CrossRef]
4. Xing, M.; Dong, S. A new simulation model for a beam-pumping system applied in energy saving and
resource-consumption reduction. SPE Prod. Oper. 2015, 30, 130–140. [CrossRef]
5. Brown, K.E. Overview of artificial lift systems. J. Pet. Technol. 1982, 34. [CrossRef]

1
Sensors 2016, 16, 685 13 of 13

6. Peng, Y.; Yan, W.; Wu, H. The study on energy saving and output enhancement of pumping well based on the
theory of qfd/triz. In Proceedings of the International Conference on Energy and Environment Technology
(ICEET’09), Guilin, China, 16–18 October 2009; pp. 449–453.
7. Lu, J.; He, J.; Mao, C.; Wu, W.; Wang, D.; Lee, W.-J. Design and implementation of a dual pwm frequency
converter used in beam pumping unit for energy saving. IEEE Trans. Ind. Appl. 2014, 50, 2948–2956.
[CrossRef]
8. Feng, C. Synthetic description of status and energy-saving method on oil-abstracting engines in oilfield.
Inner Mongolia Petrochem. Ind. 2008, 24, 017.
9. Li, D.; Fan, J.; Wang, L.; Kou, C. Curved beam pumping unit. Oil Forum 2013, 4, 63–65.
10. Zou, L.-Q.; Zhou, L. Dynamic performance analysis of double-horse heads pumping unit. Oil Field Equip.
2012, 8, 005.
11. Xiang, F. High-slip motors for sucker rod pumping units. Chin. Pet. Mach. 1986, 4, 004.
12. Lu, Q.; Fan, C.; Ye, Y.; Zhu, Z. A high torque density permanent magnet motor for oil pumping unit.
J. Appl. Phys. 2008, 103, 07F108. [CrossRef]
13. Luo, Y.; Luo, Y.; Wang, Y. A multifunction energy saving device with a novel power-off control
strategy for beam pumping motors. In Proceedings of the 2009 Record of Conference Papers-Industry
Applications Society 56th Annual Petroleum and Chemical Industry Conference, Anaheim, CA, USA,
14–16 September 2009; pp. 1–7.
14. Tian, J.; Gao, M.; Zhou, S.; Zhang, F. Energy-saving control system of beam-pumping unit based on wavelet
neural network. In Proceedings of the Fourth International Conference on Natural Computation, Jinan,
China, 18–20 October 2008; pp. 509–513.
15. He, J.P.; Mao, C.X.; Lu, J.M.; Yang, J.W. Design and implementation of an energy feedback digital device
used in elevator. IEEE Trans. Ind. Electron. 2011, 58, 4636–4642. [CrossRef]
16. Jiang, S.; Yang, X.; Yan, X.; Li, B. Simulation analysis on a sort of beam-pumping unit with new balance
methods. Acta Petrolei Sinica 2005, 26, 117–120.
17. Foley, W.; Svinos, J. Expert adviser program for rod pumping (includes associated paper 19367). J. Pet. Technol.
1989, 41, 394–400. [CrossRef]
18. Schirmer, P.; Gay, J.; Toutain, P. Use of advanced pattern-recognition and knowledge-based system in
analyzing dynamometer cards. SPE Comput. Appl. 1996, 3, 21–24. [CrossRef]
19. Everitt, T.A.; Jennings, J. An improved finite-difference calculation of downhole dynamometer cards for
sucker-rod pumps. SPE Prod. Eng. 1992, 7, 121–127. [CrossRef]
20. Tripp, H. A review: Analyzing beam-pumped wells. J. Pet.Technol. 1989, 41, 457–458. [CrossRef]
21. Miska, S.; Sharaki, A.; Rajtar, J. A simple model for computer-aided optimization and design of sucker-rod
pumping systems. J. Pet. Sci. Eng. 1997, 17, 303–312. [CrossRef]
22. Gibbs, S.; Neely, A. Computer diagnosis of down-hole conditions in sucker rod pumping wells. J. Pet. Technol.
1966, 18, 91–98. [CrossRef]
23. Gibbs, S. Predicting the behavior of sucker-rod pumping systems. J. Pet. Technol. 1963, 15, 769–778. [CrossRef]
24. Zhang, W.D.; Guan, L.G.; Zhang, G.J.; Xue, C.Y.; Zhang, K.R.; Wang, J.P. Research of doa estimation based on
single mems vector hydrophone. Sensors 2009, 9, 6823–6834. [CrossRef] [PubMed]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).