This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.

Digital Object Identifier 10.1109/ACCESS.2017.Doi Number

Applying Reliability Centered Maintenance

(RCM) to Sampling Subsystem in Continuous
Emission Monitoring System
Yuan-Jian Yang1, Xin-Yin Zhang1, Zhi-jie Zhao1, Gui-hua Wang1, Yan-Jiang He2, Yi-Liang
Wu2 and Jian Li2
1
School of Safety Engineering, Chongqing University of Science and Technology, Chongqing, P.R. China
2
Chongqing Natural Gas Purification Plant PetroChina Southwest Oil & Gas field Company, Chongqing, P.R. China

Corresponding author: Yuan-Jian Yang (e-mail: yuanjyang@hotmail.com).

This research was partially sponsored by Natural Science Foundation of Chongqing, China (Grant No. cstc2019jcyj-msxmX0434), and the Foundation of
Chongqing University of Science and Technology (Grant No. ck2017zkyb009).

ABSTRACT Continuous emission monitoring system (CEMS) is widely used in the field of environmental
protection, but the maintenance routines of it consist mostly of pre-defined activities. Reliability-centered
maintenance (RCM) uses a systematic approach to develop maintenance strategies that take into account
factors such as system life, operating efficiency, and maintenance cost-effectiveness. This paper is
motivated to apply RCM to sampling subsystem in CEMS. Based on the failure mode and effect analysis
(FMEA), the RCM logic decision used to identify the reasonable maintenance mode is implemented.
Subsequently, an average availability model is proposed to determine the maintenance interval cycle.
Finally, the maintenance outline is formed, which can be then further used as valuable information to
improve equipment reliability and develop maintenance strategies.

INDEX TERMS continuous emission monitoring system, sampling system, reliability centered
maintenance, failure mode and effect analysis, maintenance strategy.

Reliability is one of the most common concepts in system

I. INTRODUCTION
With the severe affection of stationary pollution, the
emissions from factories should be continuously monitored.
While continuous emission monitoring system (CEMS), as
an optical, mechanical and electrical integration equipment
for monitoring emissions from stationary pollution, is
widely used in the field of environmental protection. With
the use of CEMS, the components and contents of the
smoke emissions can be obtained in real-time, thus
avoiding environment accidents [1,2]. However, the
problems such as frequent failure and high maintenance cost
on the system, especially sampling subsystem bring a lot of
trouble [3].
VOLUME XX, 2017
science, which is used to develop maintenance engineering
and general industrial quality of service [4,5]. Proper
maintenance to subsystems and components is of great
importance to the reliability of the system. Insufficient
maintenance may reduce the performance of the system to
some extent and increases the loss caused by failure, while
excessive maintenance may increase the cost of maintenance
[6]. Accordingly, carrying out a reliability-centered
maintenance strategy can balance system reliability and
maintenance cost, avoiding insufficient and excessive
maintenance [7].
Reliability-centered maintenance was early applied to the
aviation industry and can be traced back to the maintenance
planning of the American aviation industry since the 1960s

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access
[8,9]. The first application of RCM theory in the aviation
industry effectively improvs the reliability of equipment.
Since then, many fields, especially the civil aviation industry
in the United States, have used modern science and
technology to explore the basic laws of equipment
maintenance [10,11]. By the end of the 1960s, the ATA (Air
Transport Association of America) created the MSG
(Maintenance Steering Group), a task force to review the
application of existing methods and maintenance techniques
for aircraft maintenance [12]. In MSG-1 and MSG-2, the
maintenance methods such as regular maintenance,
condition-based maintenance, and condition monitoring were
introduced, and it was successfully applied to the formulation
of the Boeing 747 maintenance program [13].
In 1978, the report "Reliability-centered Maintenance"
written by Stanley-Nowlan and Howard-Heap from ATA
took RCM theory a step forward, and it also became the
basis for the later application of RCM [14]. Later in 1980,
the western civil aviation community took advantage of
RCM, leading to the appearance of MSG-3. Then the wide
range use of the MSG-3 results in the publication of RCM-II,
a proven and accepted methodology used in a variety of
industries [15].
Today, the RCM theory is the internationally used method
to determine equipment preventive maintenance demand and
optimize maintenance strategies [16,17]. Besides, RCM is a
tool commonly used to study system failure patterns so that
propose optimal maintenance strategies to ensure system
reliability, and finally ensure safety in the most economical
way [18-21].
RCM has been continuously developed and improved in
practice, stands out from many equipment maintenance
theories, and is widely used in all walks of life and has
achieved significant results. A lot of efforts have been
recently dedicated to the application of RCM, which in
different industrial fields. Guevara and Souza [22] presented
the reliability-based method aiming to improve the
availability of gas turbine installed in the maintenance of
combined-cycle thermoelectric power plant. Wang [23]
conducted research on the condition assessment and
condition prediction of high-speed train equipment, and
established a condition-based maintenance optimization
strategy model that takes into account non-new and
economic requirements. Jae-Haeng Heo et al [24] found the
optimal maintenance strategies for transmission systems
from a large class of possible maintenance scenarios by using
genetic algorithm. Katharina Fischer et al [25] identified the
most relevant functional failures, their causes, and suitable
preventive measures for wind turbines based on statistical
analysis and practical experience. Chen and Zhang [26]
described the implementation of reliability centered
maintenance in China’s nuclear energy field. Vishnu [27]
proposed an AHP based framework for reliability centered
maintenance strategy selection for process plants and it is
validated with the maintenance history data. Farzaneh et al
VOLUME XX, 2017
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
[28] proposed game theory to identifying the system
reliability performance, which is applied to the
reliability-centered maintenance in electric power systems.
Mohamed et al [29] presented a reliability-centered
maintenance approach that is combined with evolutionary
optimization to develop optimal maintenance plans for
hospital facilities.
For the CEMS-2000 commonly used in natural gas
purification, in the past, maintenance routines consisted
mostly of pre-defined activities carried out at regular
intervals, which is not scientific. For complex
electromechanical systems, maintenance should be carried
out in a systematic way, taking into account factors such as
system life, operating efficiency, and maintenance
cost-effectiveness. However, to the best of the authors’
knowledge, no reference can be found on the maintenance
strategy of CEMS. Therefore, this paper aims to obtain the
system maintenance strategy by applying the RCM method.
The remaining paper is organized as follows. In Section II,
the working principle and composition of CEMS are
discussed. Section III introduces each part of RCM approach
in detail. Finally, section IV presents RCM in sampling
subsystem, and the results are discussed.
II. WORKING PRINCIPLE OF CEMS
It takes CEMS-2000, which is widely used in natural gas
purification plant as an example. The CEMS-2000 is
composed by a sampling system, an on-line analyzer, a data
acquisition processing system and auxiliary facilities [30,31].
The sampling system consists of four parts, including
sampling extraction, sampling transport, sampling treatment
and exhaust stream disposal. The principle of the sampling
system is to filter the solid papers and impurities in the flue
smoke emissions, and further to send the filtered sample to
the on-line analyzer for component measurement [32].
The on-line analyzer contains OMA-2000 gas analyzer,
LDM-100 dust monitor, TPF-100 temperature pressure flow
integrated transmitter, flue gas humidity measurement
module and oxygen meter. The subsystem can not only
monitor the concentration of SO2, NOX and O2 in the flue,
but also can monitor flue gas flow rate, temperature, pressure
and humidity. It can also calculate the discharge rate, total
emission and the conversion of relative concentration.
The data acquisition processing system is composed of
four parts: the junction box, the IPC, the display and data
processing software. It is responsible for the collection and
processing of monitoring data. Monitoring data is transmitted
to the government's environmental protection department
and enterprise through the back-end system.
The auxiliary facilities include PLC control system, safety
environmental protection facilities and public utilities. The
power supply system provides a stable supply for the CEMS.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

The system composition of CEMS is shown in Figure 1, Figure 2.

and the reliability block diagram of the CEMS is shown in

Continuous Emission Monitoring System

Date acquisition Auxiliary facilities

Sampling system On-line analyzer
processing subsystem

Sampling Sampling OMA-2000 TPF-100 Flue gas humidity Industrial Public

Data PLC
extraction treatmentt analyzer Temperature measurement PC utilities
processing control Safety
pressure flow module.
software system environmental
Exhaust LDM-100 integrated
Sampling Oxygen protection
stream dust transmitter. Display
transport facilities
disposal monitor meter

FIGURE 1 The system composition of CEMS.

Sampling Sampling Exhaust stream LDM-100 OMA-2000

Sampling transport
extraction treatmentt disposal Dust monitor gas analyzer

TPF-100
Flue gas humidity
Data processing temperature pressure
Display IPC Oxygen meter measurement
software flow integrated
module.
transmitter.

Safety
PLC control system environmental Public utilities
protection facilities

FIGURE 2 The reliability block diagram of CEMS.

operation of the system should be clearly informed based on

III. RCM APPROACH the historical fault and maintenance data [35].
Information Collection and
Reliability centered maintenance is to ensure the security and System Selection

reliability of the equipment by using the minimum resource

consumption based on the equipment design features,
Functional and Failure
functional structure, failure modes and consequence analysis. Analysis

In order to optimize the maintenance of the equipment, the Important functional part Unimportant functional part

logic decision method is applied to determine the

maintenance work content, maintenance method, FMEA Analysis FMEA Analysis

maintenance cycle of each component, and reliability

centered maintenance strategy is formed finally. Failure Mode Failure Effects
The RCM analysis process consists of the following steps:
collecting research object data, dividing subsystems and
Logic Decision and
selecting key components, carrying out FMEA (Failure Maintenance Selection

Mode and Effects Analysis), applying RCM logic decision

diagrams for maintenance mode decision, and forming a Design Maintenance Maintenance Failure
Change Measures Cycle Mode
maintenance plan strategy [33,34]. The procedure of RCM
methodology is presented in Fig.3.
Preventative Maintenance Restorative Maintenance
Outline Outline

A. INFORMATION COLLECTION AND SYSTEM

SELECTION
Comprehensive Maintenance
In this part, the components and subsystems of the systems Outline

being researched should be determined. Besides, the FIGURE 3 RCM analysis procedure diagram.

VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access


n
B. FMEA t
In this part, the function, failure mode and failure cause of MTBF = i =1 (6)
each components should be detailed, then criticality of each n
components should be calculated and sorted. The frequency where t is the failure interval; n represents the total
(O), detection (D), and severity (S) of the FMEA mode are times of fault.
scored, and the scoring standard is 1-10 points. The higher
the score, the more serious the degree. The value of RPN can D. RCM LOGIC DECISION
be obtained from the formulation: RPN = O  D  S [36]. The logic decision analysis for important functional
equipment is the focus of RCM analysis. It is to select the
C SYSTEM RELIABILITY EVALUATION INDEX appropriate maintenance method for the failure mode in
The exponential distribution model is often applied to FMEA and determine the appropriate maintenance level.
electronic devices and components. Therefore, we assume Usually, it is achieved by the RCM logic decision diagram.
that the sampling system is subject to an exponential The analysis process of the RCM logic diagram is divided
distribution. The cumulative distribution function and into two parts, and the first part is to decide whether
probability density function of the exponential extension preventive is needed through the types of functional fault
distribution is given by: effects. For the functional fault effects, there are 4 categories,
F ( t ) = 1 − e- t
including Hidden(H), Safety(S), Operation(O) and
(1)
Economy(E). By analyzing the failure mode according to the
f (t ) =  e − t
(2) RCM fault effect classification logic, the fault effects can be
classified, and the fault impact classification decision process
The corresponding reliability function is defined as
is shown in Fig.5.
follows:

R(t ) = e−t (3) Failure mode

where  is the failure rate, and the failure rate function is

expressed as: Does the fault N
have a significant
 (t ) =  (4) effect on the
system?

For repairable systems, MTBF is often used to represent

Y
two adjacent fault interval working hours [20]. For the
systems subject to an exponential distribution, the MTBF can Evident fault
effect
be described as:
1
MTBF = (5)
 N
Does the failure
The research object is the performance change of a single affect system safety?

system running overtime. The failure process can be

N
Does the fault
represented by a random point process without considering affect the system's
Y operability?
the repair time. The Fig.4 depicts the relationship between
fault time points of T0 , T1 , T2 … Tn and Tn +1 . The Y

system fault interval is equal to the difference between Tn −1 Safety Operation Economy Hidden
and Tn called ti .
FIGURE 5. Fault impact classification logic diagram.
t1 t2 t3 ti
Also, the RCM logic decision contains the second step,
T0 T1 T2 T3 Tn-1 Tn Tn+1 T which identifies the specific maintenance mode, and to select
the appropriate preventive maintenance by answering the
FIGURE 4. Repairable system usage process.
questions in turn. The maintenance mode decision process is
The MTBF point estimation method is taken as follows: as shown in Fig.6.
From Fig.6, it can be seen that the logical sequence of
maintenance mode is selected according to the maintenance
VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

mode from easy to difficult and the maintenance cost from personnel [37]. Therefore, the maintenance cycle is
low to high, which reflects the economy of RCM introduced by the mathematical model. In this section,
maintenance theory. The basic idea of it is the different fault periodic maintenance modes are selected to calculate the
impact modes after FMEA. Therefore, the emphasis of component maintenance interval cycle. Under the premise of
developing maintenance measures for different types of fault considering the availability of equipment, the maintenance
impact is different. cost should be reduced as much as possible. And the optimal
maintenance calculation model of the equipment is selected
to formulate a reasonable maintenance strategy.
Step 0: The common periodic maintenance model is based on task
Is the care and Y Care and
reliability, based on maximum availability and minimum
maintenance maintenance
effective? maintenance costs [38]. For exponential distribution systems,
calculation models based on maximum availability and
N minimum maintenance costs are not applicable. So, in this
section, the average availability model is selected to calculate
Step 1:
the maintenance cycle.
Is the condition- Y Condition-based For repairable equipment which follows an exponential
based maintenance maintenance distribution, within the maintenance cycle T , it is known
effective? to us that the life distribution of the equipment is F (t ) , its
reliability is R(t ) = 1 − F (t ) , the failure rate is  (t )
N and the probability density function is f (t ) . Then the
average life expectancy E (t ) is:

E (T ) =  t  f (t )dt
Step 2:
Y (7)
Is the periodical Periodical
0
recovery recovery
effective? And it is known from the reliability theory that:

dF (t ) −dR(t )
N f (t ) = = = − R(t ) (8)
dt dt
Step 3:
Is the periodical Y Periodical Combine the Eq. (7) and (8) to get:
 
E (T ) =  t   R(t ) dt =  R(t )dt
replacement replacement
(9)
effective? 0 0

Then the equipment average availability is:

N
E (T ) (10)
A2 (T ) =
Step 4:
Y
T
Is the periodical Periodical Since life follows an exponential distribution, brings Eq.
inspection inspection (3), (4), and (5) into Eq. (9) to get:
effective?
E (T ) 1 − e − t
N A2 (T ) = = (11)
T T
Corrective − t
maintenance Take e for Taylor’s formula expansion, so that:

N (T )2 (T )3 (−T )n (12)

e− t = 1 − T + − + +
Redesign 2  n
Combine the Eq. (11) and Eq. (12) to get:

(T ) 2 (T )3 ( −T ) n

FIGURE 6. RCM maintenance mode decision logic diagram.

T − + − +
E. MAINTENANCE INTERVAL CYCLE A2 (T ) = 2  n
Due to the flexibility and failure frequency, the maintenance
T (13)

(−T ) n −1
=
interval cycle is not accurate enough to be determined only
by the experience and qualitative method of the management
n =1 n
VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

(−T ) n −1 where T is the maintenance interval cycle time.

When the value is already small enough to
n
be negligible, so Eq. (13) can be simplified to: IV. CASE STUDY
T (T )2
A2 (T ) = 1 − + (14) A. FMEA FOR SAMPLING SUBSYSTEM
2 
The FMEA is the starting point of RCM analysis [39]. It is
To solve T by Eq. (14), so that: clear that there were found the functions, failure modes,
failure cause and the severity, thus the measures can be made
1.5 − 6 A2 − 3.75 to prevent and control the failures. Table I gives the FMEA
T= (15)
 result of the sampling system.
TABLE I
FMEA OF THE SAMPLING SYSTEM

Number Components Function Failure mode Failure cause O S D RPN Compensation Measures
High particulate matter in
the flue gas or excessive
1 Blockage 8 3 2 48 Check and clean regularly
accumulation of dust
inside the probe
Damage to the filter
element caused by Replace or reverse blow
2 7 5 4 140
long-term alternating heat regularly
Particulate matter
and cold
Sampling more than 20μm in
"O" type ring filling or
probe flue gas and Corrosion of seal surface
3 3 5 3 45 silicone oil, regular
sampling extraction of probe
Breakage replacement
Choose high elastic sealing
Failure deformation of material, smear lubricant
4 3 5 4 60
sealed O-ring evenly over the grooves and
seals
Check and replace
5 Unqualified probe quality 3 4 4 48
regularly
Keep the filter core at
High particulate matter in
6 Blockage 5 5 5 75 constant temperature and
the flue gas
dry
Unreasonable temperature Adjust the heat tracing
7 5 2 5 50
Flue gas setting temperature regularly
Sampling condensation Poor heating effect of Check and replace
8 Sampling transport 5 2 4 40
tube heating cable regularly
Check the sealing condition
Untightened pipe
9 3 4 2 24 of filter element and joint
connections
Leakage regularly
Check, clean and replace
10 Aging pipeline 4 3 3 36
regularly
Check the solenoid valve
11 Back blowing on the Circuit breaker 2 3 5 30
Solenoid Non action of regularly
probe and filter at a
valve solenoid valve Check and replace
12 time Coil damage 2 3 3 18
regularly
After the scoring from the experts of natural gas The most harmful failure cause is the damage to the filter
purification plant, the value of O, D and S are obtained, element, it is followed by the high particulate matter in the
then the value of RPN can be sorted as R12< R9< R11< R10< flue gas, while the least harmful is non action of the
R8< R3< R1= R5< R7< R4< R6< R2. solenoid valve. The result of the analysis provided above is
in accordance with the actual.
VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

According to the FMEA results of the sampling system, Therefore, in the same way, the reliability index of other
the weak link of it can be found, and the causes and modes components can be obtained, as shown in Table Ⅲ.
of system failure are checked. Thus carrying out different TABLE Ⅲ
FAILURE RATE OF THE COMPONENTS OF SAMPLING SYSTEM
maintenance strategies for different components, which
provides the basis to the RCM. Part name Failure mode MTBF(h)  (h-1)
B. RELIABILITY EVALUATION RESULTS Blockage of probe
2628 3.8052E-04
The most serious failure mode probe filter blockage is filter
sampling probe
taken as an example. Based on the maintenance Sampling probe
8156 1.2261E-04
management records, the fault interval time can be obtained breakage
through fault statistics, screening and calculation. The fault Blockage of
18940 5.2798E-05
interval time data is shown in Table Ⅱ. sampling tube
TABLE Ⅱ condensation in
sampling pipe 29785 3.3574E-05
The FAULT INTERVAL TIME sampling tube
Years Fault interval time (h) sampling tube
36457 2.7230E-05
leakage
2015 2792, 1089, 2789
electromagnetic electromagnetic
2016 2156, 3386, 3670 36053 2.7737E-05
valve valve inaction
2017 1378, 2787, 3467
2018 2564, 2876, 1894, 2956 C. LOGIC DECISION ANALYSIS
2019 2994 Starting with the FMEA, after a series of questions from
RCM logic decision, the fault impact classification and
Then after calculation, the value of MTBF is obtained as
specific maintenance modes are obtained, as shown in
follow:
Table IV.
MTBF = 2628h
TABLE Ⅳ
THE RESULT OF LOGIC DECISION

Logic decision
Fault impact Maintenance
Number Part name Failure mode
classification method
0 1 2 3 4

Blockage of Periodical
1 S NO NO NO NO YES
Sampling probe filter Inspection
Probe Sampling probe Periodical
2 S NO NO NO NO YES
breakage Inspection
Blockage of Periodical
3 S NO NO NO NO YES
sampling tube Inspection
Sampling Condensation in Periodical
4 S NO NO NO YES
Pipe sampling tube Replacement
Sampling tube Periodical
5 S NO NO NO NO YES
leakage Inspection
Solenoid Non action of Periodical
6 H NO NO NO YES
Valve solenoid valve Replacement
Different maintenance modes of the same system may The proposed average availability model provides a
obtain different maintenance methods after analysis. reasonable methodology to calculate maintenance intervals
Because the results of the analysis are obtained from the under different availability. From the Eq. (15), with the
systematic and scientific RCM logic decision analysis of parameter λ already known, the relationship between
each failure mode, the timing, cause and impact of each availability and maintenance interval can easily be found.
failure mode are different, and the results of maintenance After the combination of fault data, the recommended
decision will be different [40]. maintenance cycle curve for each failure mode is obtained,
as shown in Fig.7.
D. DETERMINATION OF MAINTENANCE INTERVAL
CYCLE
VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

FIGURE 7. Recommended maintenance interval cycle with different availability.

outline of important functional units include failure modes,
E. RELIABILITY CENTERED MAINTENANCE OUTLINE maintenance modes, maintenance work content and
maintenance interval cycle. Setting the equipment available
FOR SAMPLING SUBSYSTEM
to be maintained above 95%, then the maintenance strategy
After completing the RCM analysis, summarize the above of each component of the sampling system is shown in
information to form a reliability-centered maintenance Table V.
outline. The main contents of the maintenance strategy
TABLE V
MAINTENANCE OUTLINE FOR THE SAMPLING SYSTEM
Failure Maintenance Maintenance Maintenance
Part name
Mode Method Content Cycle(h)

Blockage of probe Periodical Check and clean

272
filter Inspection regularly/Replacement
Sampling Probe
Sampling probe Periodical Regularly replace/Use other heating
845
breakage Inspection measures to preheat the backflush

Blockage of Keep the filter element at a constant

Periodical
sampling tube temperature and dryness/Clean it 1962
Inspection
regularly
Adjust the heating temperature
Sampling Pipe Condensation in Periodical
/Replace the fault heating cable, 3086
sampling tube Replacement
thermostat, etc.
Sampling tube Periodical Check the filter element and seal at the
3777
leakage Inspection joint regularly

Non action of Periodical Fasten the thread/ If it is damaged,

Solenoid Valve 3735
solenoid valve Replacement replace the coil in time.

FMEA outcome, the maintenance mode decision is carried

V. CONCLUSION
The reliability-centered maintenance (RCM) is applied to
sampling subsystem in continuous emission monitoring
system (CEMS). The main cause of failure and weak
components are identified through FMEA. Then based on the
VOLUME XX, 2017
out. To keep the high availability of the system, an average
availability model is introduced to obtain a reasonable
maintenance interval cycle. Finally, the maintenance outline
for sampling system is formed, which provides a theoretical
basis to the factories. The results of the maintenance methods

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access

of the components from sampling subsystems decided [10] J. Liu, “Reliability Centered Maintenance in the Study on the
according to the RCM theory are more targeted than the Turnout of High Speed Railway,” China Academy of Railway
traditional maintenance method in the past, in that there are
different maintenance modes for different components. In the Sciences, 2014.
past, for CEMS-2000, maintenance methods mainly [11] Q. Meng, “A Method Based-on RCM Used in the Spot Welding
considered manufacturers' guidance and experience factors Robot Station Maintenance,” Shanghai Jiao Tong University, 2011.
and did not start from the fault consequences of actual
[12] J. Endrenyi et al., "The present status of maintenance strategies and
equipment operation. Compared with it, the results of the
maintenance method decision obtained in this paper are more the impact of maintenance on reliability," in IEEE Transactions on
scientific and reliable. Power Systems, vol. 16, no. 4, pp. 638-646, Nov. 2001.
[13] Backlund, F. and Akersten, P., "RCM introduction: process and
REFERENCES
requirements management aspects", Journal of Quality in
[1] U.S.Environmental Protection Agency, “Plain English Guide to the
Maintenance Engineering, Vol. 9 No. 3, pp. 250-264, 2003.
Part 75 Rule,” Washington, DC: Clear Air Markets Division of
[14] F. S. Nowlan and H. F. Heap, Reliability centered maintenance. San
U.S.Environmental Protection Agency, 2005.
Francisco, America: United Airlines Publications, 1978．
[2] Q. Wang and K. Yang, “The monitoring technology and its
[15] J Moubray, Reliability-centered maintenance, Industrial Press Inc.,
application of Continuous Emission Monitoring System(CEMS),”
2001.
Chemical Industry Press, vol. 8, no. 12, pp. 45-49, 2015.
[16] X. M. Dong, “Theory and Practice of Modern Technical Equipment
[3] H. B. Yan, M. Feng, “Discussion on the Problems of Flue Gas
Maintenance: Application in Track Train System,” Southwest
Monitoring in Thermal Power Plants,” Shanxi Electric Power, vol.
Jiao-tong University Press, 2009.
10, no. 5, pp. 20-21, 2009
[17] A Chateauneuf, F Cocheteux, F Deffarges, and F Sourget,
[4] R. Djurić and V. Milisavljević, “Investigation of the relationship
“Reliability analysis of screwed connections in high-speed trains,
between reliability of track mechanism and mineral dust content in
considering fatigue, corrosion, and imperfect maintenance
rocks of lignite open pits,” Eksploatacja i
operations,” Proceedings of the Institution of Mechanical Engineers
Niezawodnosc-Maintenance and Reliability, vol. 18, no. 1, pp.
Part O Journal of Risk and Reliability September, vol. 225, no. 3,
142-150, 2016.
pp. 293-306, 2011.
[5] Y. J. Yang, W. Peng, S. P. Zhu, and H. Z. Huang, “A bayesian
[18] W V Jaarsveld, R Dekker, “Spare parts stock control for redundant
approach for sealing failure analysis considering the non-competing
systems using reliability centered maintenance data,” Reliability
relationship of multiple degradation processes,” Eksploatacja i
Engineering & System Safety, vol.96, no.11, pp. 1576-1586, 2011.
Niezawodnosc-Maintenance and Reliability, vol. 18, pp. 10-15,
[19] Y. X. Ji, G. H. Yang, L. X. Wang, and C. J. Li, “A reliability
2016.
centered maintenance approach for substation maintenance
[6] J. H. Seo and D. S. Bai, “An optimal maintenance policy for a
optimization,” 2015 International Conference on Energy and
system under periodic overhaul,” Elsevier Science Publishers B. V,
Mechanical Engineering, pp. 529-535, 2016.
vol. 39, pp. 373-380, 2004.
[20] M. A. E Suryono, and C. N. Rosyidi, “Reliability Centred
[7] X. Y. Li, H. Z. Huang, and Y. F. Li, “Reliability analysis of phased
Maintenance (RCM) Analysis of Laser Machine in Filling Lithos at
mission system with non-exponential and partially repairable
PT X,” IOP Conference Series: Materials Science and Engineering,
components,” 2018 Reliability Engineering & System Safety, vol.
vol. 319, 2018.
175, pp. 119-127, 2018.
[21] O. Samuel, F. Ishola and O. Ajayi, “A Reliability-Centered
[8] J. Endrenyi et al., "The present status of maintenance strategies and
Maintenance study for an Individual Section Forming Machine,”
the impact of maintenance on reliability," in IEEE Transactions on
Machine Design, vol.6, no. 4, pp. 50, 2018.
Power Systems, vol. 16, no. 4, pp. 638-646, Nov. 2001.
[22] F. J. Carazas Guevara and G. F. Martha de Souza, "RCM
[9] L. Bertling, R. Allan and R. Eriksson, "A reliability-centered asset
Application for Availability Improvement of Gas Turbines Used in
maintenance method for assessing the impact of maintenance in
Combined Cycle Power Stations," in IEEE Latin America
power distribution systems," in IEEE Transactions on Power
Transactions, vol. 6, no. 5, pp. 401-407, Sept. 2008.
Systems, vol. 20, no. 1, pp. 75-82, Feb. 2005.
VOLUME XX, 2017

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.2980630, IEEE Access
[23] L. Z. Wang, “Research on Reliability-centered maintenance
Decision and Support System for High-speed Train Equipments,”
Beijing Jiaotong University, 2011.
[24] J. Heo et al., "A Reliability-Centered Approach to an Optimal
Maintenance Strategy in Transmission Systems Using a Genetic
Algorithm," in IEEE Transactions on Power Delivery, vol. 26, no. 4,
pp. 2171-2179, Oct. 2011.
[25] K. Fischer, F. Besnard and L. Bertling, "Reliability-Centered
Maintenance for Wind Turbines Based on Statistical Analysis and
Practical Experience," in IEEE Transactions on Energy Conversion,
vol. 27, no. 1, pp. 184-195, March 2012.
[26] Y. Chen and T. Zhang, "Application & development of
Reliability-centered Maintenance (RCM) in China's nuclear energy
field," 2012 International Conference on Quality, Reliability, Risk,
Maintenance, and Safety Engineering, Chengdu, 2012, pp. 543-548.
[27] Vishnu C R, Regikumar V, “Reliability based maintenance strategy
selection in process plants: a case study,” Procedia technology, vol.
25, pp. 1080-1087, 2016.
[28] F. Pourahmadi, M. Fotuhi-Firuzabad and P. Dehghanian,
"Application of Game Theory in Reliability-Centered Maintenance
of Electric Power Systems," in IEEE Transactions on Industry
Applications, vol. 53, no. 2, pp. 936-946, March-April 2017.
[29] Salah M, Osman H, Hosny O. “Performance-based
reliability-centered maintenance planning for hospital facilities,”
Journal of Performance of Constructed Facilities, vol.53, no.1,
2018.
[30] W. Yang, “Study on Continuous Emission Monitoring System
(CEMS) applied to resources and environmental,” Dalian
University of Technology, 2013.
[31] B. Shi, Z. H. Ma, and J. M. Jiang, “Research on Continuous
Emission Monitoring System in Thermal Power Plant,” Applied
Mechanics & Materials vol. 427, pp. 1117-1120, 2013.
[32] C. M. Lü, “Research on the Application of Continuous Emission
Monitoring Systems based on DOAS,” Tianjin University, 2013.
[33] X.S Jia, “Reliability centered maintenance (RCM) development
trend,” Journal of the college of ordnance engineering, vol.6, no. 14,
pp. 3-6, 2002.
[34] Moradi S, Vahidinasab V, Kia M, et al. “A mathematical
framework for reliability‐centered maintenance in microgrids,”
International Transactions on Electrical Energy Systems, vol.29,
no.1, 2019.
VOLUME XX, 2017
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
[35] Gupta G, Mishra R P. “Identification of critical components using
ANP for implementation of reliability centered maintenance,”
Procedia CIRP, vol.69, pp. 905-909, 2018.
[36] J. M. Cai,. X. M. Li, and. G. H. Yang, “The risk priority number
methodology for distribution priority of emergency logistics after
earthquake disasters,” International Conference on Management
Science & Industrial Engineering IEEE, pp. 560-562, 2011.
[37] Rahmati S H A, Ahmadi A, Karimi B. “Multi-objective
evolutionary simulation based optimization mechanism for a novel
stochastic reliability centered maintenance problem,” Swarm and
Evolutionary Computation, vol.40, pp. 255-271, 2018.
[38] F. Fang, Z. Zhao, C. Huang, X. Zhang, H. Wang and Y. Yang,
"Application of Reliability-Centered Maintenance in Metro Door
System," in IEEE Access, vol. 7, pp. 186167-186174, 2019.
[39] C. Xie, M. F. Sun, and J. M. Du, “Research on the Reliability
Centered Maintenance Plan of a Launching System,” International
Journal of Plant Engineering and Management, vol.16, no.4, pp.
243-248, 2011.
[40] Tang Y, Liu Q, Jing J, et al. “A framework for identification of
maintenance significant items in reliability centered maintenance,”
Energy, vol.118, pp. 1295-1303, 2017.
Yuan-Jian Yang received his doctoral degrees in
Mechanical and Electronic engineering from University
of Electronic Science and Technology of China. He is
currently a master’s supervisor of the safety
engineering of Chongqing University of Science and
Technology. His research interests include reliability and safety of
complex system.
Xin-Yin Zhang is currently a master degree candidate
at the Chongqing University of Science and Technology.
Her research interests include reliability analysis and
evaluation.
Zhi-Jie Zhao is currently a master degree candidate at
the Chongqing University of Science and Technology.
His research interests include reliability analysis and
evaluation.