Вы находитесь на странице: 1из 10

Safety, Reliability and Risk Analysis: Theory, Methods and Applications – Martorell et al.

(eds)
© 2009 Taylor & Francis Group, London, ISBN 978-0-415-48513-5

Availability and reliability assessment of industrial complex systems:


A practical view applied on a bioethanol plant simulation

V. González, C. Parra & J.F. Gómez


Industrial Management PhD Program at the School of Engineering, University of Seville, Spain

A. Crespo & P. Moreu de León


Industrial Management School of Engineering, University of Seville, Spain

ABSTRACT: This paper shows a practical view about the behaviour of an industrial assembly in order to assess
its availability and reliability. For that intention it will be used such a complex system like a Bioethanol Plant.
A computerized model will help to create a realistic scenario of the Bioethanol Plant Life Cycle, obtaining an
estimation of the most important performance measures through real data and statistic inference. By this way, it
will be possible to compare and discuss the profit of different plant configurations using the model and following
the initial technical specifications. Basically, the Bioethanol Plant will be divided for that purposes in functional
blocks, defining their tasks and features, as well as their dependencies according to the plant configuration.
Additionally, maintenance information and data bases will be required for the defined functional blocks. Once
these data have been compiled and using any commercial software, it will be possible to carry out a model of the
plant and to simulate scenarios and experiments for each considered configuration. Parameters about availability
and reliability will be obtained for the most important functions, following different plant configurations. From
their interpretation, it will be interesting to consider actions that improve the availability and reliability of the
system under different plant functional requirements. Among other important aspects, it will be researchable
as well a sensitive analysis, i.e., the exploring on how parameters modifications have influence on the result or
final goal.

1 INTRODUCTION assessment in a plant. Once simulated the operational


cycles of each component and through their combi-
Nowadays, investors, engineers etc. have to take nation and dependencies, we will be able to obtain
into consideration a lot of requirements and condi- the whole system operation cycle. The scope of this
tions in order to avoid risks and hazards on industrial research will be the availability assessment of alterna-
systems. Components or subsystems have potential tives configurations in a plant according, not only to
failure modes which have to be in mind from the sys- predetermined maintenance strategies, but also to new
tem initial state and according to its operation modes, maintenance policies.
environmental conditions, failure times etc. . .
The failure modelling could be very complicated
because of dependencies or inferences among com-
2 PROCEDURE
ponents and, at the same time, the great amount
of required data. The intention here is to generate
2.1 Definition of the system configuration
a random number of events under a computerized
model which simulates the plant life scenario. There Description: It is needed to define functional blocks
are simulation methods that allow us taking into and their dependencies according to the plant config-
consideration important aspects on the system oper- uration and each block tasks or purposes.
ation like redundancies, stand-by nodes, preventive Result: Functional block list/diagram: function,
maintenance, repairing priorities. . . inputs, outputs, etc. Functional chart of the sys-
The use of such methods is rising up when one wants tem containing the relations among blocks and their
to predict the general availability or the economical reliability features.

687
2.2 Data compiling Grain (Starch)
Milling Hydrolysis Saccharifi-
Description: Maintenance information and data base Water
cation

will be required for each considered block.


Result: Schedule with preventive tasks, times, Evaporation Centrifuge Distillation Fermentation

failure ratios etc.


Drying Dehydration
2.3 Model construction
Description: Using any commercial software, it will Syrup
DDG Ethanol CO2
be possible to make a model of the plant.
Result: Plant computerized model. Figure 1. General diagram of the Bioethanol obtaining
process.
2.4 Simulation
Description: Using the model and with the above men-
tioned data we will simulate scenarios and experiments
for each considered configuration.
Result: Scenarios list, real and unreal events repli-
cation.

2.5 Results and analysis


Description: It will be able to calculate and discuss the
simulation results.
Results: Parameters about availability and reliabil- Figure 2. Diagram of the grain reception, cleaning, milling
ity will be obtained for the most important functions, and storage system.
following different plant configurations. From their
interpretation, we will consider actions to improve
the availability and reliability of the system under
plant functional requirements. Among other important Conversion of starch into bioethanol involves sev-
aspects, we will have as well the sensitive analysis, i.e., eral more process steps, as. It starts by making a
the exploring on how parameters modifications have ‘‘beer’’ from the milled grain, then distilling off the
influence on the result or final goal. alcohol followed by recovery of the residual solids and
recycle of water.
It is necessary to be care with each step of the pro-
3 SHORT DESCRIPTION OF THE cess to assure an efficient conversion, particularly due
BIOETHANOL PROCESS AND ITS to the fact that it is a biological process where unsuit-
RELATED BLOCK DIAGRAMS able reactions can happen causing loss in yield, and
also because different grains have different process
Ethanol is commonly produced by the fermentation requirements. Some of the most important matters to
of sugars. The conversion of starch to sugar, called have into account in the sequence of process steps are
saccharification, significantly expands the choice of below mentioned.
feedstock to include cereals as corn, wheat, barley Together with this description are also included
etc. Other alternative route to bioethanol involves the Block Diagrams of the main systems in a Bioethanol
enzymatic conversion of cellulose and hemi-cellulose. plant. These diagrams are only an approaching of an
Completing the picture, ethanol is also commercially ideal plant. They try to show briefly the main equip-
synthesised from petrochemical sources but this does ments, devices, as well as material flows inside the
not fall into the renewable category, which is the frame whole plant.
of this paper.
Basically, the chemical reactions involved in
bioethanol production can be simplified as follows: 3.1 Slurry preparation
The feed grain is milled to a sufficient fineness in order
to allow water access to all the starch inside each grain.
The meal is then mixed with warm water till a specific
concentration without generating excessive viscosities
downstream.

688
Figure 3. Diagram of the mashing, cooking, liquefaction
and fermentation system. Figure 4. Diagram of the distillation and dehydration
system.

3.2 Hydrolysis
The slurry temperature is raised up in order to accel-
erate the hydrolysis of the grain’s starch into solution.
Again there is an optimum depending on the grain
type—if the slurry is too hot, the viscosity is exces-
sive and if too cool, the required residence time for
effective hydrolysis is too long.

Figure 5. Diagram of the centrifugation, evaporation and


3.3 Saccharification drying system.
With enzymes, the dissolved starch is converted to sug-
ars by saccharification, but at a reduced temperature
parallel absorber and the almost pure ethanol is sent
which again is selected to achieve a balance between a
to product storage. The cycle is then repeated.
satisfactory reaction rate, and avoiding the promotion
of unsuitable side reactions and a subsequent loss in
yield. 3.7 Centrifugation
The residual slurry left after distillation is called
3.4 Fermentation ‘‘Whole Stillage’’ and it contains all the insoluble and
soluble non-starch components from the feed grain,
The slurry is cooled to a fermentation temperature and as well as the yeast which has grown during fermen-
held in large batch fermentation tanks for a specific tation. The bulk of the solids, termed ‘‘Wet Distillers
time. Fresh yeast is prepared in parallel and added Grains’’ (WDG) are removed by centrifuge leaving a
at the beginning of each batch fermentation cycle. ‘‘Thin Stillage’’.
The fermenting slurry is agitated and also circulated
through external exchangers to remove the heat gen-
erated by the fermentation process. On completion of 3.8 Evaporation
fermentation the batch is transferred to the Beer Well To minimise water consumption, a large portion of the
and the particular fermentation tank and associated Thin Stillage is recycled to the front of the process.
equipment are cleaned by the plant’s CIP system. Basically, by effect of the evaporation a portion of
water is recycled, and the residual ‘‘Syrup’’ that con-
3.5 Distillation tains around 30–35% solids is either blended with the
WDG or sold separately to animal feed.
The ‘‘beer’’ contains about 8–12% ethanol. It is contin-
uously pumped to the distillation unit which produces
an overhead stream of about 90% ethanol and water. 3.9 Drying
Ethanol and water form a 95% azeotrope so it is not WDG is commonly dried as a mixture with the syrup to
possible to reach 100% by simple distillation. about 10% moisture by gas fired rotary dryer. This by-
product, called ‘‘Dried Distillers Grains and Solubles’’
(DDGS) is sold to animal feed.
3.6 Dehydration
The 90% overhead stream is passed through an
3.10 CO2 Recovery
absorber containing a molecular sieve which traps the
ethanol while letting the water pass through. Once the Fermentation simultaneously generates carbon diox-
bed in the absorber is full, the feed is switched to a ide which is collected and scrubbed to recover any

689
4.1.3 System availability and unavailability
For the system to be available, each subsystem should
be available. Thus:

Aseries = A1 · A2

Conversely, the unavailability is:


Figure 6. Diagram of the pelletizing system.
UAseries = 1 − Aseries = 1 − (1 − UA1 ) · (1 − UA2 )
ethanol. It can either be processed into a by-product = UA1 + UA2 − UA1 · UA2
or released to atmosphere.
4.1.4 System mean down time for repairable
subsystems
If two subsystems are both repairable, one with mean
4 SUMMARY OF FORMULAS down time MDT1 and the other MDT2 , then, the mean
AND CONCEPTS REGARDING down time for the two subsystems in series will be as
RELIABILITY AND AVAILABILITY follows. At any instance in time, the system is in one
of the 4 states:
The intention here is to summarise the formulation
regarding the reliability and availability concepts for • Both subsystems functional.
two subsystems (non-identical) in series or in parallel. • Only subsystem #1 is non-functional.
Additionally, a system with n identical subsystems in • Only subsystem #2 is non-functional.
parallel is also here taken into consideration, where • Both subsystems are non-functional.
the system will be declared as failed if m or more
The last 3 cases are responsible for the system
subsystems fail (the m_out_n case).
being non-functional. It is assumed that the 4th case
The system characteristics to be formulated here are
has negligible probability. Given the system is down,
such as ‘‘system failure rate’’, ‘‘system Mean Time
the probability that it is because the subsystem #1 is
Between Failure’’, ‘‘system availability and unavail-
non-functional is obviously:
ability’’, and ‘‘system mean down time’’. We under-
stand as ‘‘Failure (f)’’ the termination of the ability of λ1
a component or system to perform a required function. .
Therefore, the ‘‘Failure Rate (λ)’’ will be the arithmetic λ1 + λ2
average of failures of a component and/or system per
unit exposure time. The most common unit in relia- Since subsystem #1 needs MDT1 to repair, the
bility analyses is hours (h). However, some industries repair time associated with repairing subsystem #1 is
use failures per year then:
λ1
∗ MDT1
4.1 Two subsystems in series λ1 + λ2
4.1.1 System failure rate A similar expression is true for subsystem #2.
Just for one subsystem, the failure rate is λ1 . The Summing them up, one gets:
probability of failure in dt is λ1 dt. For two sub-
systems in series, the probability of failure in dt MTBF1 · MDT2 + MTBF2 · MDT1
is (λ1 dt + λ2 dt ). The system failure rate is thus MDTseries =
MTBF1 + MTBF2
(λ1 + λ2 ) : λseries = λ1 + λ2
The reliability function is:
4.2 Two subsystems in parallel
R(t) = exp[−(λ1 + λ2 )t]. Here the two subsystems are repairable. The mean
down times are MDT1 and MDT2 .
4.1.2 System MTBF
4.2.1 System failure rate
From the exponential form of the reliability function,
If the system just consists of subsystem #1, then the
it is obvious that:
system failure rate is λ1 . The probability of failure
in dt is λ1 dt. Adding subsystem #2 in parallel, the
MTBF1 · MTBF2 probability for system failure in dt is λ1 dt reduced by
MTBFseries = 1/(λ1 + λ2 ) =
MTBF1 + MTBF2 the probability that the subsystem #2 is in the failure

690
state. The probability to find the subsystem #2 in the MDT1 MDT2
UA1 · UA2 = ·
failure state is given by: MTBF1 MTBF2

MDT2 Consequently:
MTBF2 + MDT2
MDT1 · MDT2
MDTparallel =
Assuming that: MDT1 + MDT2

MDT2 << MTBF2 4.3 M out of N parallel subsystems

And using: If a system consists of n parallel, identical subsystems


and the system is down if there are m or more subsys-
1 tems down, then, the formulas for system failure rate,
MTBF2 = system MTBF, system availability, and system mean
λ2
down time will be the following ones.
Then, the reduced failure rate for subsystem #1 is
then given by: λ1 · λ2 · MDT2 4.3.1 System failure rate
Likewise, the reduced failure rate for subsystem #2 If the system just consists of subsystem #1, then the
is: λ1 · λ2 · MDT1 system failure rate is λ. The probability of failure in
Consequently: λparallel = λ1 · λ2 · (MDT1 + MDT2 ) dt is λdt. To have a system failure, we need to have
other (m-1) subsystems in the failure state. The chance
4.2.2 System MTBF that any one subsystems is in the failure state is given
Taking the approach that the inverse of the failure rate by MDT/(MTBF+MDT), or (MDT/MTBF), if we
is MTBF (true for exponential distribution), one gets: assume MDT<< MTBF. To find (m−1) subsystems
in the failure state, the probability is:
m−1
MTBF1 · MTBF2

MDT
MTBFparallel = 1/λparallel =
MDT1 + MDT2 MTBF
It is noted that if the two subsystems are not There are n−1 Cm−1 ways to group (m-1) subsystems
repairable, then the MTBF for the parallel case is the out of (n-1) subsystems. Also, it is possible to choose
sum of the individual MTBF’s. any subsystem to be the #1 subsystem in the analysis.
Putting all together, one has:
4.2.3 System availability and unavailability
For the system to be available, either subsystem should MDT m−1
be available. Thus: λm_out_of _n = λ · ( ) n−1 C m−1 · n
MTBF
n!
Aparallel = A1 + A2 − A1 · A2 = λm · MDT m−1 (2)
(n − m)!(m − 1)!
Conversely, the unavailability is:
This is the failure rate for exactly m subsystem
UAparallel = 1 − Aparallel = 1 − (A1 + A2 − A1 · A2 ) failures. The failure rate for more than m subsystem
failures is going to be smaller by a factor of (λ · MDT ).
= (1 − A1 ) · (1 − A2 ) = UA1 · UA2 (1) For a consistency check, we consider n = m = 2.
This is a system consisting of two parallel, identical
4.2.4 System mean down time for repairable subsystems. When m = 2 subsystems fail, the system
subsystems fails. Eq. (2) for this case is λsystem = λ2 · (2 · MDT )
From the definition of: which agrees with the above mentioned formula.
MDT MDT
Unavailability = ≈ 4.3.2 System MTBF
MTBF + MDT MTBF Taking the approach that the inverse of the failure rate
is MTBF (true for exponential distribution), one gets:
It is possible to get the MDT for the parallel case
by using Eq.(1) above. MTBFm_out_of _n = 1/λm_out_of _n
MDTparallel MDTparallel MTBF m
UAparallel = = MTBF1 ·MTBF2 = n!
MTBFparallel MDT +MDT
1 2 (n−m)!·(m−1)! · MDT m−1

691
4.3.3 System availability and unavailability 5.1 System failure rate
For the system to be available, at least (n-m+1)
• 3CO2 = λMilling + λFermentation
subsystems should be available. Thus:
• 3Ethanol = λMilling + λFermentation + λDistillation +
λDehydration
n
X n! • 3DDGS = λMilling + λFermentation + λDistillation +
Am_out_of _n = Ai (1 − A)n−i λCentrifugation + λDrying
i=n−m+1
(n − i)! · i!
• 3Syrup = λMilling + λFermentation + λDistillation +
λCentrifugation + λEvaporation
Using the following equality:

n
Here it has been included the Hydrolysis and Sac-
n
X n! charification in the same process area as Fermentation.
1 = [A + (1 − A)] = Ai (1 − A)n−i
i=0
(n − i)! · i!

It is possible to rewrite the availability as:


5.2 System MTBF system availability (A)
MTBFMilling ·MTBFFermentation
n−m
X n! • MTBFCO2 = MTBFMilling +MTBFFermentation
Am_out_of _n = 1 − Ai (1 − A)n−i MTBFCO2 ·MTBFDis+Deh
(n − i)! · i! • MTBFEthanol =
i=0 MTBFCO2 +MTBFDis+Deh ,
n!
≈1− (1 − A)m
m! · (n − m)! Where:
MTBFDistillation ·MTBFDehydration
MTBFDis+Deh = MTBFDistillation +MTBFDehydration
And the unavailability is given (again, for
MDT << MTBF), by:
MTBFCO2 ·MTBFDis+Cen+Dry
• MTBFDDGS = MTBFCO2 +MTBFDis+Cen+Dry ,
n!
UAm_out_of _n = UAm
m! · (n − m)!
Where:
MTBFDis+Cen ·MTBFDrying
4.3.4 System mean down time for repairable MTBFDis+Cen+Dry = MTBFDis+Cen +MTBFDrying , Where:
subsystems MTBFDistillation ·MTBFCentrifugation
MTBFDis+Cen = MTBFDistillation +MTBFCentrifugation
From the definition of:

MDTm_out_of _n MTBFCO2 ·MTBFDis+Cen+Eva


UAm_out_of _n = • MTBFSyrup =
MTBFm_out_of _n MTBFCO2 +MTBFDis+Cen+Eva ,

It is possible to get the MDT for the m_out_of_n Where:


MTBFDis+Cen ·MTBFEvaporation
case by using the above mentioned formulas for MTBFDis+Cen+Eva = MTBFDis+Cen +MTBFEvaporation , and as
UAm_out_of _n and MTBFm_out_of _n . above mentioned:
Consequently:

MDT MTBFDistillation · MTBFCentrifugation


MDTm_out_of _n = MTBFDis+Cen =
m MTBFDistillation + MTBFCentrifugation

5 FORMULATION APPLIED TO THE BLOCK


DIAGRAMS
5.3 System availability (A)
Taking into consideration the General diagram of the • ACO2 = AMilling · AFermentation
Bioethanol obtaining process included at the begin- • AEthanol = ACO2 · ADistillation · ADehydration
ning of Section 3, it is possible to summarize basically • ADDGS = ACO2 · ADistillation · ACentrifugation · ADrying
the Reliability characteristics of the Bioethanol plant • ASyrup = ACO2 · ADistillation · ACentrifugation · AEvaporation
in the following formulas for a configuration in series:

692
5.4 System Mean Down Time (MDT) which can be applied to analyze the reliability charac-
teristics of the whole complex system, in this case, a
• MDTCO2
Bioethanol Plant.
MTBFMilling ·MDTFermentation +MTBFFermentation ·MDTMilling
= MTBFMilling +MTBFFermentation

• MDTEthanol 6 CONCLUSION
MTBFCO2 ·MDTDis+Deh +MTBFDis+Deh ·MDTCO2
= , With this research we pretend to improve the esti-
MTBFCO2 +MTBFDis+Deh
mations, demonstrating as well how requirements
Where expressed in initial technical specifications can be
MDTDis+Deh incompatible or even impossible to accomplish for
determined plant configurations. That means, avail-
MTBFDistillation ·MDTDehydration +MTBFDehydration ·MDTDistillation
= MTBFDistillation +MTBFDehydration
ability expectations on proposed configurations of the
whole plant could be lower, having higher reliability
or mantenability on each functional block, following
• MDTDDGS the technical requirements in effect.
MTBFCO2 ·MDTDis+Cen+Dry +MTBFDis+Cen+Dry ·MDTCO2 Additionally, reasonable estimations will be pro-
= MTBFCO2 +MTBFDis+Cen+Dry ,
vided for the production availability, which can be
Where delivered to the final customer in a more realistic
engineering project. These estimations will be based
MDTDis+Cen+Dry on validated calculations of functional blocks consid-
MTBFDis+Cen ·MDTDrying +MTBFDrying ·MDTDis+Cen ered for the model simulation, showing moreover the
= MTBFDis+Cen +MTBFDrying ,
importance and opportunity of a sensitive analysis. It
And can also be decisive for the final selection the plant
technical configuration.
MDTDis+Cen At the same time, this study can also be used to
MTBFDistillation ·MDTCentrifugation +MTBFCentrifugation ·MDTDistillation adjust some initial requirements in the plant technical
= MTBFDistillation +MTBFCentrifugation specification. Once the data have been introduced in
the model, they can be adjusted according to the real
• MDTSyrup =
equipments included in the offer. By the way, it is pos-
MTBFCO2 ·MDTDis+Cen+Eva +MTBFDis+Cen+Eva ·MDTCO2 sible to study logistical aspects like spare parts amount
MTBFCO2 +MTBFDis+Cen+Eva ,
in stock.
Where Finally, not only the availability and reliability are
important, but also the costs estimation is a key factor.
MDTDis+Cen+Eva Therefore, an extension of this study could be to trans-
MTBF
Dis+Cen ·MDT
Evaporation +MTBF
Evaporation ·MDTDis+Cen fer the information provided in this research to a life
= MTBFDis+Cen +MTBFEvaporation , cycle cost model, with the intention to assess globally
and as above mentioned, the plant.
MDTDis+Cen

=
MTBFDistillation ·MDTCentrifugation +MTBFCentrifugation ·MDTDistillation ACKNOWLEDGEMENTS
MTBFDistillation +MTBFCentrifugation
The author would like to thank the reviewers of the
Once developed these formulas for the main parts of paper for their contribution to the quality of this work.
the process, it is possible to continue breaking down
these as deeper as wanted. It is clear that each area
(milling, fermentation, distillation. . . ) in the plant REFERENCES
has actually its own configuration for its different
devices and equipments, with its specific combina- Asociación Española de Mantenimiento. 2005. El Manten-
tion for these subsystems in series or in parallel (see imiento en España: Encuesta sobre su situación en las
for instance the different block diagrams also included empresas españolas.
in Section 3). Bangemann T., Rebeuf X., Reboul D., Schulze A., Szymanski
J., Thomesse J.P., Thron M., Zerhouni N.. 2006. Proteus-
Therefore, and as it has just mentioned, the relia- Creating distribuited maintenance systems through an
bility characteristics can be also broken down till such integration platform. Computers in Industry, Elselvier.
a detail level where is possible to apply real values Benoît Iung 2006. CRAN Laboratory Research Team
for such units. There are published many Data Bases PRODEMAS in Innovative Maintenance and Depend-
which include real values for process equipments, and ability. Nancy University—Nancy Research Centre

693
for Automatic Control (CRAN). CNRS UMR 7039 Hansen, M.T.; Noria, N.; Tierney, T. 1999. What’s your
(http://www.cran.uhp-nancy.fr). strategy for managing knowledge?. Harvard Business
Bourne, M. & Neely, A. 2003. Performance measurement Review.
system interventions: the impact of parent company ini- Henley, E.J. and Kumamoto, H. 1992 Probabilistic Risk
tiatives on success and failure. Journal of Operation and Assessment: Reliability Engineering, Design & Analysis.
Management. ISBN 0 87942 290 4. IEEE Press, US.
Campbell J.D. & Jardine A. 2001. Maintenance excellence. Høyland A & Rausand M, 1994. System Reliability Theory.
New York: Marcel Dekker. 2001. Models and Statistical methods. ISBN 0-471-59397-4.
Carter, Russell A. 2001. Shovel maintenance gains from Wiley-Interscience.
improved designs, tools and techniques. Elsevier Engi- Inc, Renewable Fuels Association. 2001. Ethanol Plant
neering Information. Development Handbook. BBI International. USA
Center for Chemical Process Safety (CCPS)Guidelines for Intelligent Maintenance Centre 2007. www.imscenter.net.
Process Equipment Reliability Data, with Data Tables. Iserman R. 1984. Process fault detection based on modelling
ISBN: 0-8169-0422-7. and estimation methods. Automatica.
Clark J. 1995. Managing Innovation and Change: People, ITSMF, IT Service Management Forum 2007. ITIL v3.
Technology and Strategy. Business & Economics. Information Technology Infrastructure Library. ITIL v2.
Crespo M.A., Moreu de L.P., Sanchez H.A. 2004. Ingeniería Information Technology Infrastructure.
de Mantenimiento. Técnicas y Métodos de Aplicación a Jardine A.K.S., Lin D., Banjevic D. 2006. A review on
la Fase Operativa de los Equipos. Aenor, España. machinery diagnostics and prognostics implementing
Crespo M.A. 2007. The Maintenance Management Frame- condition based maintenance. Mech, Syst. Signal Process.
work. Models and Methods for Complex Systems Main- Jharkharia S. & Shankarb R. 2005. Selection of logistics
tenance. Londres, Reino Unido. Springer. service provider: An analytic network process (ANP)
Dale & Plunkett 1991. Quality Costing. Chapman Hall. approach. International Journal of Management Sciente,
Dandois, P.A. & Ponte, J. 1999. La administración del Omega 35 (2007) 274–289.
conocimiento organizacional. El management en el siglo Kaplan, Robert S. & David P. Norton 1996. The Balanced
XXI. Scorecard: Translating Strategy Into Action . Boston,
Davenport T. 1993. Process innovation: Reengineering MA: Harvard Business School Press.
work through Information Technology. Harvard Business Kent Allen 1990. Encyclopedia of Computer Science and
School Press. Technology. CRC Press.
David John Smith. 2005. Reliability, Maintainability and Klein, M.M. 1994. The most fatal reengineering mistakes.
Risk: Practical Methods for Engineers. ISBN 0-750- Information strategy: The Executive’s J. lO(4) 21–28.
66694-3 Butterworth-Heinemann. Lee J. 1995. Machine performance monitoring and proac-
Deardeen, J. Lilien, G. and Yoon, E. 1999. Marketing tive maintenance in computer-integrated manufacturing:
and Production Capacity Strategy For Non-Differentiated review and perspective. International Journal of Computer
Products: Winning And Losing At The Capacity Cycle Integrating Manufacturing.
Game. International Journal of Research In Marketing. Lee J. 2004. Infotronics-based intelligent maintenance sys-
Deming W. Edwards 1989. Calidad, productividad y compet- tem and its impacts to close-loop product life cycle
itividad: la salida de la crisis. Madrid, Ediciones Díaz de systems. Proceedings of de IMS’2004 International Con-
Santos. ference on Itelligent Maintenance Systems, Arles, France.
Dixon J.R. 1966. Design engineering: inventiveness, anal- Levitt Joel. 2003. Complete Guide to Preventive and Predic-
ysis, and decision making. New York, McGraw-Hill, tive Maintenance. Industrial Press.
Inc. Davis, M. 1988. Applied Decision Support. Prentice Hall,
Dyer R.F. & Forman E.H. 1992. Group decision support Englewood Cliffs .
with the Analytic Hierarch Process. Decision Support Mitchell Ed., Robson Andrew, Prabhu Vas B. 2002. The
Systems. Impact of Maintenance Practices on Operational and
Earl M.J. 1994. The New and the Old of Business Pro- Business Performance. Managerial Auditing Journal.
cess Redesign. Journal of Strategic Information Systems, Mobley Keith 2002. An Introduction to Predictive Mainte-
vol. 3. nance. Elsevier.
EN 13306:2001. Maintenance Terminology. European Stan- Moubray J. 1997. Reliability-centered Maintenance. Indus-
dard. CEN (European Committee for Standardization), trial Press.
Brussels. Nakajima Seiichi 1992. Introducción al TPM (Manten-
European Foundation for Quality Management. 2006. EFQM imiento Productivo Total). Productivity Press.
Framework for Management of External Resources. By Neely, A.D., Gregory, M. and Platts, K. 1995. Performance
EIPM—EFQM. Measurement System Design—A Literature Review and
Gelders, L., & Pintelon, L. 1988. Reliability and mainte- Research Agenda. International Journal of Operations and
nance" in: Doff, R.C. and Nof, S.J. (ed.), International Production Management.
Encyclopedia of Robotics, Application and Automation, Nonaka & Takeuchi 1995. The Knowledge-Creating Com-
Wiley, New York. pany. USA: Oxford University Press.
Goldratt E. 1997. Cadena Crítica. Ediciones Diaz de Santos. Patton, J.D. 1980. Maintainability and Maintenance Manage-
Hammer & Champy 1993. Reengineering the Corporation. ment. Instrument Society of America, Research Triangle
Harper Business. Park, NC.
Hammer M. 1990. Reengineering Work: Don’t Automate Peters T. & Waterman H.R. Jr. 1982. ‘‘In Search of Excel-
Obliterate. Harvard Business Review. lence’’.

694
Pintelon L.M. & Gelders L.F. 1992. Maintenance manage- Turban E. 1988. Decision Support and Expert Systems:
ment decision making. European Journal of Operational Managerial Perspectives. New York: Macmillan.
Research. UNE 66174 2003. Guide for the assessment of quality man-
Porter, M. 1985. Competitive Advantage. Free Press. agement system according to UNE-EN ISO 9004:2000
Prusak Laurence 1996. The Knowledge Advantage. Strategy standard. Tools and plans for improvement. UNE.
& Leadership. UNE 66175 2003. Systems of Indicators. UNE.
Ren Yua, Benoit Iung, Herv!e Panetto 2003. A multi-agents UNE-EN ISO 9001:2000. Quality management
based E-maintenance system with case-based reasoning systems—Requirements. International Organization for
decision support. Engineering Applications of Artificial Standardization.
Intelligence 16 321–333. Wireman, T. 1991. Total Productive Maintenance. Industrial
Saaty, T.L. 1977. A Scaling Method for Priorities in Hier- Press.
archical Structures. Journal of Mathematical Psychology, Yan S.K. 2003. A condition-based failure prediction and
15: 234–281, 1977. processing-scheme for preventive maintenance. IEEE
Saaty, T.L. 1980. The Analytic Hierarchy Process. McGraw Transaction on Reliability.
Hill. Zhu G., Gelders L. and Pintelon L. 2002. Object/objective-
Saaty, T.L. 1990. How to make a decision: The ana- oriented maintenance management. Journal of quality in
lytic hierarchy process. European Journal of Operational maintenance engineering.
Research.
Shu-Hsien Liao 2005. Expert system methodologies and
applications—-a decade review from 1995 to 2004.
Elselvier. Expert Systems with Applications 28 93–103.

695

Вам также может понравиться