Expert Systems with Applications 36 (2009) 8676–8681
Contents lists available at ScienceDirect
Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa
Real-time turbine maintenance system Tung-Liang Chen * Department of Industrial and Information Management, National Cheng Kung University, Tainan 701, Taiwan, ROC
a r t i c l e
i n f o
Keywords: Radio frequency identification Total productive maintenance Turbine
a b s t r a c t Reliable power generation and low maintenance costs are the major goals of power plant administration. This goal, in fact, can be achieved by a proper turbine maintenance policy. This study presents a model for total productive maintenance to enhance the efficiency of power plant equipment. A probabilistic failure analysis model is used to determine the optimal turbine maintenance cycle. Additionally, the costs savings achieved by using radio frequency identification (RFID) technology is demonstrated in an operational maintenance model. 2008 Elsevier Ltd. All rights reserved.
1. Introduction Turbine startup is one of the critical problems in the operation of electrical power plants (Marco, Lopez, Flores, & Garcia, 2003). Therefore, the quality of turbine-generated power is related to maintenance system and management policies, which increase the core competitiveness of a power plant. A preventive maintenance monitoring system is important for maximizing equipment availability and reliability (Tam, Chen, & John, 2007). Although many studies have analyzed control charts for equipment maintenance, few have demonstrated how to extend the life of equipment (Badia, Berrade, & Campos, 2002). To minimize abnormal power system conditions, proper equipment maintenance is vital (Colban & Thole, 2007). In 1977, the first Taiwan nuclear power plant began to supplement the power provided by existing hydroelectric and fossilfueled and thermal power plants. Therefore, the operational importance of turbines was heightened and realized (Akturk & Gurel, 2007). However, the turbine is a complex multi-axle system consisting of a high-pressure generator, two low-pressure generators and eight exciter rotors. The steam turbine blade is extremely complex since it must be flexible enough to change shape during operation in response to cold temperatures and the dynamic coupling effect (Ricardo, 2007). Reducing environmental damage and increasing turbine efficiency are essential issues (Parka, Jungb, & Yum, 2000). Long term degradation of turbine efficiency may be difficult to detect. Hence, preventive maintenance is essential. An effective performance monitoring system must monitor more than just a single process in order to clarify the interaction among processes (Parka et al., 2000).
* Tel.: +886 75572286; fax: 886 75570070. E-mail address:
[email protected] 0957-4174/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2008.10.019
This research therefore proposes a preventive maintenance monitoring system and simultaneously considers the relationships between real-time system control and material testing. 2. Literature review Effective maintenance is essential for effective power plant operation and economic viability. 2.1. Turbine characteristics A steam turbine is a mechanical device that extracts thermal energy from pressurized steam and converts it into useful mechanical power (Ricardo, 2007). The steam turbine has almost completely replaced the reciprocating engine in power generation, primarily because of its greater thermal efficiency and higher power-to-weight ratio. Even, combined cycle gas turbine will design and implement to detect on-line and diagnose anomalies as soon as possible in the dynamic evolution of the behavior of a power plant (Arranz, Cruz, Sanz-Bobi, Ruiz, & Coutino, 2008). Also, because the turbine generates rotary motion rather than requiring a linkage mechanism to convert reciprocating to rotary motion, it is particularly suited for driving an electrical generator. Approximately 86% of all electricity in the world is generated by steam turbines. The steam turbine achieves thermodynamic efficiency by using multiple stages of steam power capture. A governor is essential for controlling turbines, which must be run up gradually to prevent damage. Some applications, such as electricity generation, require precise speed control. Most systems include a mechanism for overspeed trip when excessive acceleration of the turbine rotor closes the nozzle valves controlling the flow of steam to the turbine. If this fails, the turbine may continue accelerating until it is destroyed, often spectacularly (Akturk & Gurel, 2007).
T.-L. Chen / Expert Systems with Applications 36 (2009) 8676–8681
Turbines are expensive and require precision manufacturing methods as well as high quality materials. Electrical power stations use large steam turbines driving electric generators to produce most of the electricity throughout the world. These centralized stations include fossil fuel, nuclear, geothermal, solar thermal electric and biomass power plants. The turbines used for electric power generation are often directly coupled to their generators. Generators must rotate at constant synchronous speeds according to the frequency of the electric power system. The most common speeds are 3000 revolutions per minute for 50 Hz systems and 3600 revolutions per minute for 60 Hz systems. Most large nuclear sets rotate at half those speeds and use 4-pole rather than the more common 2-pole generators. As Fig. 1 shows, gas, steam and water turbines have a casing around the blades that contains and controls the working fluid. The turbine transfers system energy into machine power by rotary action. The action of this paper discusses real-time monitoring technologies and the data control model developed to address these utility needs. A device similar to a turbine but operating in reverse is a compressor or pump. The axial compressor in many gas turbine engines is a common example. Because of the importance of the turbine in multi-axle systems, equipment maintenance is vital. Turbines may be single-grade turbine and multi-cylinder grades turbine in several cylinders. Turbines may be single-axle and mounted on an axle at all levels or double-axle and mounted on a parallel-axle at all levels. Other variations are vapor-congealing turbines, heating turbines, pressing-type turbines and saturationsteam turbines (Parka et al., 2000). Innovative utility-scale blades have been proposed to reduce the costs of power generation. The added energy achieved by increasing the size of utility-grade wind turbines produces diminishing returns unless blade weight (which is nonlinear) is substantially reduced. The challenge is developing new methods of generating energy with lighter components, such as the swept star blade. Other weight-reducing concepts such as carbon spar caps, offaxis carbon fibers that facilitate bend-twist coupling and new ‘‘structural” airfoils have been incorporated in small-scale prototype blades and are currently in testing. 2.2. Real-time analysis Real-time monitoring systems are needed to ensure the correctness and timeliness of information. Information from RFID (radio frequency identification) tags can be routed to a database in realtime. Data written on the tag also provides a permanent record. Every tag has a basic identification code which can be read by RFID readers. Read/write tags have storage capability for libraries to
8677
write additional information for circulation purposes, such as adding a physical location that can be used in conjunction with automatic book sorters. For example, RFID tags can also be read while an item is in motion of power plant, using RFID readers to checkin returned items while on a conveyor belt reduces staff time (Ruff & Hession-Kunz, 2001; Strassner & Chang, 2003). The tags are easily programmed in-house by library workers and are designed to perform consistently for the life of the item they identify. Unlike bar codes, RFID tags do not require line-ofsight or physical contact in order to be read. Therefore, more than one tag can be read simultaneously, a characteristic referred to as ‘‘anti-collision”. Anti-collision enables workers to rapidly take inventory simply by walking down an aisle of books with a hand-held RFID reader. All vendors surveyed in these studies used this anti-collision feature. Beyond speeding up inventory, handheld readers can also help workers remove books and locate misplaced items. Also, using digital tags and readers in conjunction with archives, special collections or government documents can minimize the physical handling of rare or delicate items since they can be tracked via RFID while still housed in opaque storage boxes (Strassner & Chang, 2003). Without engaging RFID-related professional services, the successful performance of any RFID technology implementation beyond the most basic application is seriously jeopardized. Ideally, RFID technology would be plug and play. The reality is, even for an apparently uncomplicated implementation of this technology, literally hundreds of variables must be considered. Successfully managing the physics of RFID (for instance, the physical characteristics of radio waves, materials and the surface consistency of those materials), is essential for system success, not to mention for optimal system performance. The best-designed, most robust RFID devices cannot perform optimally within a particular environment unless they have been configured, tuned and installed specifically for that environment and its related variables. Chang and Lin (2005) examined how to supervise turbine blades operation auto mutually. A turbine blade is impacted to work by the stream. For inspection of turbine blades, planning the measuring stream flow paths is largely an attempt to generate an appropriate process in order to prevent an unexpected collision between the static blade and the movement blade. He also discussed how to adjust the angle between blades and shafts and how to position the blade and shaft to avoid abnormal conditions. Exceptional conditions occurring in nuclear power plants may be much more dangerous than usual. The radio frequency identification is therefore put on the blades of the high-speed turbine so it can instantaneously indicate the position of the turbine. The systematic frequency of RFID adjusts so the reader can receive the tag data (Chow, Choy, Lee, & Lau, 2006). If the reader receives no unusual data, the turbine is still operating. If it receives not tag
Fig. 1. Turbine.
8678
T.-L. Chen / Expert Systems with Applications 36 (2009) 8676–8681
Fig. 2. Tags inside turbine blade.
data, the status of the turbine is unusual and requires appropriate analysis, treatment or modification (Rusin, 2007). Fig. 2 shows RFID tags installed on the blades. Therefore, power plants are undertaking erection and commissioning, overhauling, renovation, refurbishment, servicing and preventive maintenance of steam turbines, turbo generators and power plants along with auxiliary systems such as turbo blowers and turbo feed pumps on a turnkey basis. 2.3. Preventive maintenance When warming up a steam turbine for use, the main stream stop valves (after the boiler) have a bypass line to allow superheated steam to slowly bypass the valve before heating the lines in the system along with the steam turbine. Also, a gear engaged when no steam is being applied to the turbine slowly rotates the turbine to maintain the even heating required to prevent uneven expansion and rotor bowing. After first rotating the turbine by the turning gear and allowing time for the rotor to assume a straight plane (no bowing), the turning gear is disengaged, and steam is admitted to the turbine. For most utility and industrial steam turbines, a starting and loading chart is included in the unit instruction manual. The starting and loading chart is used to guide turbine operators in loading their units not only to minimize rotor and shell thermal stresses, but also to minimize the chances of the rotor heating faster than the shell. When starting a shipboard steam turbine (marine unit), steam is normally admitted to the astern blades located in the LP (low-pressure) turbine, then to the ahead blades slowly rotating the turbine at 10–15 revolutions per minute (RPM) to slowly warm the turbine. Problems with turbines are rare, and maintenance requirements are limited. Any rotor imbalance can cause vibration which, in extreme cases, can cause a blade to detach and possibly punch through the casing. Also, it is essential that the turbine be turned with dry steam. Exposure of the blade to water from the steam (moisture carryover) can rapidly impinge and erode the blades, possibly leading to imbalance and catastrophic failure. Also, water entering the blades is likely to destroy the thrust bearing in the turbine shaft. To prevent this, controls and baffles in the boilers ensure high quality steam. Condensate drains are also installed in the steam piping leading to the turbine. Maintenance can be classified as preventive or collective maintenance. Preventive maintenance is essential for: (1) equipment known to be degraded or equipment which is easily degraded and (2) equipment which may suddenly
breakdown during operation. Correct maintenance and skills are needed to ensure the system is in normal condition (Chin, Duan, & Tang, 2006; Tadashi, Hiroyuki, & Shunji, 2001). If maintenance is performed on a fixed schedule (cyclic) maintenance, the cycle can be modified for the desired level of maintenance. The proposed maintenance model optimizes the balance of performance and cost in power plant operation. Fig. 3 is the system management model. The reader examines errors identified by the operation equipment, and the tags transfer the error information to the control module for analysis, which then transmits the results to the alarm module to record the data or alarm. 2.3.1. Unusual situation First, the following notation is used to describe the models in this paper. a variation factor i The Ith prevention maintenance T failures interval time of preventive maintenance under tags permit the greatest upper lower limits change range ZK Dt the time of preventive maintenance interval equipment effective after prevention maintenance; Ca The Ith prevention maintenance execution time Zi "(t) availability of the equipment permit the greatest error eK the equipment prevention maintenance labor cost of minCPP imum maintenance each time the equipment prevention maintenance material cost of CPE minimum maintenance each time CPP + CPE CPM If the operating conditions are normal, the production program is not adjusted. To determine the change factor (a) and required error reduction, let YK be the correct tag position, and let ZK be the reader receiving the information P+ZK (upper control limit); otherwise, 6ZK (lower control limit) indicates the reader cannot receive information and assumes the tag position has random error eK. That is, YK = ZK ± eK, when YK cannot be received indicates eK is beyond the control area. Carefully controlling equipment maintenance and real-time monitoring of the equipments is priority concern. In Eq. (2), Ca is a new effective change factor (a), and 0 < a 6 1. The value b represents the time-adjustment factor, as 0 < b, and i is the age of equipment
8679
T.-L. Chen / Expert Systems with Applications 36 (2009) 8676–8681
System management model Equipment
Monitor
Alarm
Equipment module
Control module
System set up module
Unusual judge module
Control room
Information transmit
Alarm module
Print module Information Flow Fig. 3. Prevent maintenance system model.
f ðDt; NÞ ¼ CðDt; NÞ C a ¼ ð a 8Þ
bi
ð1Þ
i ¼ 1; 2; . . . ; N 1
ð2Þ
2.3.2. Analysis The power plant turbine is unable to collect real-time information. Assume the reader has received unusual information following Weibull as failure probability density function (Saygin, 2007). Let CPM represent maintenance cost, CPP represent maintenances cost of labor and CPE represent maintenances cost of material
C PM ¼ C PP þ C PE
ð3Þ
If T, is failures interval time of preventive maintenance under, Dt is the time of preventive maintenance interval in the minimum expectations cost, CC is the maintenance cost for single circulation as in (4)
E½CðT; DtÞ ¼ C PP E½T=Dt þ C PE E½T=Dt þ C C E½1=T
Z Dt 1 hðtÞdt C mr N Dt 0 ( N1 Z ðkþ1ÞDt K X X þ ½hðði 1ÞðDt sÞ þ DtÞ
CðDt; NÞ ¼
kDt
i¼1
þ ðN 1ÞC PP
þðN 1ÞC PE þ C c
Dt ¼
91 b > > > > > > > > > =
8 > > > > > > > > > <
ðN 1ÞC PP þ ðN 1ÞC pE þ C C 22 3b 2 33b > > Pi Pi > > > > i g i1 g > PN 6 > > j¼1 j1 j¼1 j1 7 > > > 4 5 4 5 > > C ðb 1Þ 4 5 mr > > i¼1 h h > > ; :
ð8Þ Since steam turbine power is becoming as important as nuclear and thermal power, so parameter b can input a different variation number value. For the thermal power plant, the turbine pivot number is 3600 RPM whereas the nuclear turbine pivot numbers is 1800 RPM. Therefore, the variation number value measured by the reader can be adjusted. 3. Thermal power plant implementation
) )
hði ðDt sÞÞ þ hðt k sÞ dt
2.3.3. Improvement Appropriately planning the maintenance of unusual equipment is essential. Planning can minimize cost. Loss of efficiency must be carefully measured, but preventive maintenance is also important. Considering the shape parameter b, characteristic life h, (4) and (5) determine maintenance spacing interval Dt, as (8)
ð4Þ
If one operation parameter maintains spacing interval Dt, the equipment fault is within Weibull probability distribution. The partial differential method is adopted to calculate optimal solution Dt. The optimal solution Dt value determines the preventive maintenance cycle and preventive maintenance method, as shown in (5)
k¼1
Regardless of whether nuclear/thermal power plants of the steam turbine of the power plant maintained at interval Dt is from one year to 2 years, the average for 1.5 years is determined, namely Dt = 18 (months).
ð5Þ
This study examines continuously operating equipment. Based on the parameter, set and maintenance time, this case focuses on a total production and maintenance model to minimize expected cost.
Utilizing Eq. (5), one can calculate N (optimal number of prevent maintenance tasks) as (6)
3.1. Calculation
N ¼ min CðDt N ; NÞ;
This case examines how changing the variables Dt and N can minimize expected cost. The goal is to adjust total production
N ¼ 1; 2;
ð6Þ
Let (7) g acts as 0 6 gi 6 1 as improvement factor and gi be minimum maintenance at 0; gi = 1 is complete maintenance, a is the adjusted parameter for cost, b is the adjusted parameter for time, b > 0; as i larger, the equipment improvement factor gi is smaller as well
b1 C þ C PE gi ¼ a PP CC
i ¼ 1; 2; . . . ; N 1
ð7Þ
Table 1 Equipment parameter. Maintenance cost
Adjust parameter
CPP
CPE
Cmr
Cc
a
b
5000
5000
50,000
5,000,000
1
0.001
8680
T.-L. Chen / Expert Systems with Applications 36 (2009) 8676–8681
Assume the equipment failure rate satisfies the Weibull probability distribution. The shape parameter b for a nuclear power plant is equal to 2; in a fossil fuel power plant, b equals 50. The scale parameter h is 100. Parameters a and b are adjusted to optimize the maintenance cost of equipment. Tables 1 and 2 show the relevant parameters, used in Eq. (8). Firstly, Table 1 shows the prevention and maintenance workforce costs CPE, the prevention and maintenance equipment costs CPE, the minimum repair costs Cmr and single-cycle maintenance costs CC. Secondly, there exists adjustment parameters a and b, shape parameter b and scale parameter h. Based on these assumptions and the distributed Weibull failure rate, the improvement factors and expected cost per time are analyzed.
Table 2 The analysis of b parameter. No.
1
2
3
4
5
b h Dt N T C(Dt, N)
2 100 155 22 3413 3052
5 100 70 19 1343 4820
10 100 60 18 1096 5240
20 100 55 18 1000 5438
50 100 52 18 951 5543
maintenance in response to variances. Moreover, it is possible to determine optimal maintenance time, rationally increase maintenance frequency and effectively control maintenance cost. Table 3 The turbine blade gap of protective ring and rabbet. Blade number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 *
Inlet end
Outlet end
Difference grow up analyses
2002
2006
2002
2006
0.65 0.72 0.40 0.50 0.50 0.47 0.00 0.66 0.53 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.98 0.50 0.55 0.61 0.00 0.88 0.55 0.64 0.65 0.70 0.30 0.31 0.00 0.44 0.40 0.45 0.00 0.00 0.00 0.35 0.43 0.45 0.40 0.50 0.00 0.50 0.50 0.50 0.00 0.14 0.00 0.28 0.36 0.57 0.45 0.35 0.35 0.45
0.89 0.99 0.40 0.55 0.60 0.60 0.45 0.80 0.85 0.45 0.40 0.65 0.70 0.70 0.45 0.45 0.55 0.60 0.65 0.80 0.40 0.60 0.70 0.80 0.95 1.00 0.30 0.45 0.48 0.70 0.75 0.75 0.60 0.70 0.65 0.45 0.60 0.65 0.35 0.65 0.75 0.75 0.75 0.75 0.63 0.50 0.45 0.50 0.55 0.80 0.70 0.55 0.50 0.70
0.23 1.92 0.44 0.44 0.44 0.21 0.00 2.80 1.06 1.00 0.42 0.35 0.00 2.17 0.77 0.55 0.49 0.28 0.16 2.40 0.64 0.45 0.20 0.20 0.20 2.95 0.87 0.87 0.87 0.85 0.25 2.90 0.65 0.61 0.45 0.34 0.34 3.22 0.51 0.51 0.44 0.25 0.00 2.52 0.40 0.40 0.31 0.28 0.22 2.40 0.53 0.53 0.58 0.10
0.30 2.64 0.50 0.50 0.40 0.30 0.20 3.44 0.90 0.75 0.55 0.40 0.00 2.52 0.80 0.65 0.65 0.30 0.25 2.32 0.65 0.55 0.35 0.25 0.25 2.31 0.80 0.60 0.60 0.60 0.30 3.55 0.70 0.70 0.60 0.50 0.25 3.97 0.50 0.50 0.35 0.45 0.15 3.30 0.50 0.50 0.35 0.30 0.23 3.05 0.55 0.65 0.60 0.00
The distance between axial and radial turbine apart P 0.6 mm.
Inlet end
Outlet end
mm
%
mm
%
0.24 0.27 0.00 0.05 0.10 0.13 0.45 0.14 0.32 0.45 0.40 0.65* 0.70* 0.70* 0.45 0.05 0.43 0.10 0.10 0.19 0.40 0.28 0.15 0.16 0.30 0.30 0.00 0.14 0.48 0.26 0.35 0.30 0.60* 0.70* 0.65* 0.10 0.17 0.20 0.05 0.15 0.75* 0.25 0.25 0.25 0.63* 0.36 0.45 0.22 0.19 0.23 0.25 0.20 0.15 0.25
36.92 37.50 0.00 10.00 20.00 27.66 N/A 21.21 60.38 N/A N/A N/A N/A N/A N/A 10.00 43.88 20.00 18.18 31.15 N/A 31.82 27.27 25.00 46.15 42.86 0.00 45.16 N/A 59.09 87.50 66.67 N/A N/A N/A 28.57 39.53 44.44 12.50 30.00 N/A 50.00 50.00 50.00 N/A 257.14 N/A 78.57 52.78 40.35 55.56 57.14 42.86 55.56
0.07 0.72* 0.06 0.06 0.04 0.09 0.20 0.64* 0.16 0.25 0.13 0.05 0.00 0.35 0.03 0.10 0.16 0.02 0.09 0.08 0.01 0.10 0.15 0.05 0.05 0.64* 0.07 0.27 0.27 0.25 0.05 0.65* 0.05 0.09 0.15 0.16 0.09 0.75* 0.01 0.01 0.09 0.20 0.15 0.78* 0.10 0.10 0.04 0.02 0.01 0.65* 0.02 0.12 0.02 0.10
30.43 37.50 13.64 13.64 9.09 42.86 N/A 22.86 15.09 25.00 30.95 14.29 N/A 16.13 3.90 18.18 32.65 7.14 56.25 3.33 1.56 22.22 75.00 25.00 25.00 21.691 8.05 31.03 31.03 29.41 20.00 22.41 7.69 14.75 33.33 47.06 26.47 23.29 1.96 1.96 20.45 80.00 N/A 30.95 25.00 25.00 12.90 7.14 4.55 27.08 3.77 22.64 3.45 100.00
T.-L. Chen / Expert Systems with Applications 36 (2009) 8676–8681
As Table 2 shows, when shape parameter b is increasing, the optimal interval time T and prevention and maintenance point are decreasing. Because of the abatement of total time, the cost per time as well as prevention and maintenance point are also boosted. As a result, a smaller b significantly impacts Dt, N and C(Dt, N). Conversely, a larger b has less effect. The prevention and maintenance point N can be entered into Eq. (6) to find the maintenance interval time Dt. Entering N = 1 into Eq. (8) gets the following Eq. (9):
Cðh; 1Þ ¼
8 > > > > > < > > > > > :C mr ðb 1Þ
"" P1 i
CC
g j¼1 j1 h
#
"
i1
P1
g j¼1 j1
h
##b
91 b > > > > > = > > > > > ;
ð9Þ
Entering N = N + 1 into Eq. (8) gets Dt. After entering Dt and N into cost function C(Dt, N), if C(Dt, N) < C(Dt, N 1), then C(Dt, N) is the minimum cost. Entering N = N + 2 into Eq. (8) gets the CðDt; NÞ > CðDt; N 1Þ.The improvement factor g in Eq. (7) ranges from 0 to 1. As g approaches 1, the effect is enhanced. As parameter a and prevention maintenance cost, CPM increase, g also increases. The two parameters a and CPM have the same impact on g. Similarly, as b increases, g decreases. 3.2. Results and discussion This case examines how a mechanical equipment crack may change between axial and radial turbine components. Table 3 examines the crack value between the protective ring of the blade and the rabbet in 54 mechanical equipment blades in 2002. Because the power plant data was insufficient to enable immediate correction, the machine was shut down for maintenance in 2006. The difference between the axial and radial crack can be determined artificially. As Table 3 shows, the inlet end is numbered 12, 13, 14, 33, 34, 35, 41 and 45, and the outlet end is numbered 2, 8, 26, 32, 38, 44 and 50 P 0.6 mm. Operations began in 2002. In 2003, the gap between import and export changed from zero to fifteen blades. Stopping the machines for repair at that time would have substantially enhanced industrial safety at minimal cost. Turbine operations started in 2002. In 2004, if the gap between import and export changed from zero to fifteen blades, then the machines would have been stopped and repaired. The maintenance cost would have been 5240, and the following maintenance interval would have been 60. Industrial safety would have substantially increased. When operations started in 2002, the maintenance cost was 4820, and subsequent maintenance would have been 70 in 2005. The increased interval time accelerated the disabling of the machine. When operations started in 2002, the maintenance cost was 3052, and the subsequent maintenance interval would have been 155 in 2006. The increased interval time accelerated the disabling of the machine without risking industrial safety. In this case, if the fifteen blades were initially embedded with RFID and supervised on an annual bias, preventive action could have been taken. Doing so would not only have reduced injuries,
8681
but also the cost associated with the shutdown. Restated, the maintenance interval was too long to respond to the bias immediately. If the maintenance interval were shorter, the machines would have been stopped for repair; also, the time when the machine becomes disabled cannot be accurately estimated. 4. Conclusion A steam turbine operates at a continuous high-speed, and the high flow rate of steam per area generates substantial power output. Envisioning the development of the giant turbines is an important idea, and extending the life of the blades is the key to efficient use of the giant turbines. This study examined the feasibility of using a RFID technique to devise a total preventive maintenance system. The model demonstrates how to manage controllable factors so as to improve variants and enhance efficiency. Also, it avoids operation when the variants are high. As a result, the production prevention systems can extend the life of equipment by performing effective preventive maintenance at lower operating costs. This research demonstrates that the TPM model can improve power plant operations and efficiency. References Akturk, M. S., & Gurel, S. (2007). Machining conditions-based preventive maintenance. International Journal of Production Research, 45(8), 1725–1743. Arranz, A., Cruz, A., Sanz-Bobi, M. A., Ruiz, P., & Coutino, J. (2008). DADICC: Intelligent system for anomaly detection in a combined cycle gas turbine plant. European Journal of Operational Research, 34, 2267–2277. Badia, F. G., Berrade, M. D., & Campos, C. A. (2002). Optimal inspection and preventive maintenance of units with revealed and unrevealed failures. Reliability Engineering and System Safety, 78, 157–163. Chang, H. C., & Lin, A. C. (2005). Automatic inspection of turbine blades using a 3axis CMM together with a 2-axis dividing head. International Journal of Advanced Manufacturing Technology, 26, 789–796. Chin, K. S., Duan, G., & Tang, X. (2006). A computer-integrated framework for global quality chain management. International Journal of Advanced Manufacturing Technology, 27, 547–560. Chow, H. K. H., Choy, K. L., Lee, W. B., & Lau, K. C. (2006). Design of a RFID case-based resource management system for warehouse operations. Expert Systems with Applications, 30, 561–576. Colban, W., & Thole, K. A. (2007). Experimental and computational comparisons of fan-shaped film cooling on a turbine vane surface. ASME, 129(1), 23–31. Marco, A., Lopez, A., Flores, C. H., & Garcia, E. G. (2003). An intelligent tutoring system for turbine startup training of electrical power plant operators. Expert Systems with Applications, 24, 95–101. Parka, D. H., Jungb, G. M., & Yum, J. K. (2000). Cost minimization for periodic maintenance policy of a system subject to slow degradation. Reliability Engineering and System Safety, 68, 105–112. Ricardo, F. M. B. (2007). Experimental evaluation of active flow control mixed-flow turbine for automotive turbocharger application. ASME, 129(1), 44–52. Ruff, T. M., & Hession-Kunz, D. (2001). Application of radio frequency identification systems to collision avoidance in metal nonmetal mines. IEEE Transactions on Industry Applications, 37(1), 112–117. Rusin, A. J. (2007). Technical risk involved in long-term operation of steam turbines. Reliability Engineering and System Safety, 92, 1242–1249. Saygin, C. (2007). Adaptive inventory management using RFID data. International Journal of Advanced Manufacturing Technology, 32, 1045–1051. Strassner, B., & Chang, K. (2003). Passive 5.8-GHz radio-frequency identification tag for monitoring oil drill pipe. IEEE Transactions on Microwave Theory and Techniques, 51(2), 356–364. Tadashi, D., Hiroyuki, O., & Shunji, O. (2001). Optimal control of preventive maintenance schedule and safety stocks in an unreliable manufacturing environment. International Journal of Production Economics, 74, 147–155. Tam, A. S. B., Chen, W. M., & John, W. H. P. (2007). Maintenance scheduling to support the operation of manufacturing and production assets. International Journal of Advanced Manufacturing Technology, 34, 399–407.
