+ All Categories
Home > Documents > Prez Availability 5

Prez Availability 5

Date post: 07-Apr-2018
Category:
Upload: zeki-guer
View: 219 times
Download: 0 times
Share this document with a friend

of 27

Transcript
  • 8/6/2019 Prez Availability 5

    1/27

    RELIABILITY AND DIAGNOSTIC

    Faculty of Electronic Engineering and

    Technologies

    Anna Andonova

    E-mail: [email protected]

  • 8/6/2019 Prez Availability 5

    2/27

    Definition for repairable systems that

    are required to operate continuously i.e. round the clock

    are at any random point in time either operating or down because of failure

    are being worked upon so as to restore their operation in minimum time

    Possible states of the system:

    operating

    in repair

    Definition - probability that a system is operating satisfactorily at any random

    point in time t, when subject to a sequence of up and down cycles which

    constitute an alternating renewal process

    Combination of

    reliability parameters

    maintainability parameters

    AVAILABILITY THEORY

  • 8/6/2019 Prez Availability 5

    3/27

    AVAILABILITY THEORY

    a single equipment which is to be operated continuously

    a record is kept on when the equipment is operating or down overa period of time

    describe its availability as a random variable defined by

    a distribution function H(A)

    expected value availability is simply the average value

    of the function over all possible values of the variable

    a systems steady state availability ensemble of equipments

    at any particular time the number of equipments that are in state 0

    (available) to be NP0 NP0/N = P0N = total number of equipments and P0 = total equipment (N) in state 0

  • 8/6/2019 Prez Availability 5

    4/27

    System availability can be defined in the following ways:

    Instantaneous Availability: A(t)

    Mission Availability: A m(t2 - t1) average availability, AAV

    Steady State of Availability: AS

    Basic Concepts

  • 8/6/2019 Prez Availability 5

    5/27

    System availability

    Basic Concepts

    Fig. 6. The relationship between

    instantaneous, mission, and

    steady state availabilities as

    a function of operating time

  • 8/6/2019 Prez Availability 5

    6/27

    System availability

    Achieved Availability: AA

    includes all

    repair time (corrective and preventive maintenance time)

    administrative time

    logistic time

    Intrinsic Availability: Ai function of the basic equipment/system design

    does not include

    administrative time

    logistic time

    Basic Concepts

  • 8/6/2019 Prez Availability 5

    7/27

    System availability

    Achieved Availability: AA

    includes all

    repair time (corrective and preventive maintenance time)

    administrative time

    logistic time

    Intrinsic Availability: Ai function of the basic equipment/system design

    does not include

    administrative time

    logistic time

  • 8/6/2019 Prez Availability 5

    8/27

    The basic concepts of the Markov process

    state of the system (e.g., operating, nonoperating)

    state transition (from operating to nonoperating due to failure, or from

    nonoperating to operating due to repair)

    Markov process is defined by a set of probabilities pij which define the probability of transition from any

    state i to any state j

    The most important features of any Markov model

    transition probability pij depends only on states i and j and is completely

    independent of all past states except the last one, state I

    pij does not change with time

    Availability Modeling (Markov Process Approach)

  • 8/6/2019 Prez Availability 5

    9/27

    Availability Modeling (Markov Process Approach)

    Fig. 6. The relationship betweeninstantaneous, mission, and

    steady state availabilities as

    a function of operating time

  • 8/6/2019 Prez Availability 5

    10/27

    Assumptions in system availability modeling utilizing the Markov process

    The conditional probability of a failure occurring in time (t, t + dt) is l dt

    The conditional probability of a repair occurring in time (t, t + dt) is m dt

    The probability of two or more failures or repairs occurring simultaneously

    is zeroThe most important features of any Markov model

    Each failure or repair occurrence is independent of all other occurrences

    (failure rate) and (repair rate) are constant (e.g., exponentiallydistributed)

    Availability Modeling (Markov Process Approach)

  • 8/6/2019 Prez Availability 5

    11/27

    Single Unit Availability Analysis (Markov Process Approach)

    Markov graph for single unit

    S0 = State 0 = the unit is operating and available for use

    S1 = State 1 = the unit has failed and is being repaired

    = failure rate

    = repair rate

    transition matrix

    the conditional probability of failure in (t, t + dt) is dt

    the conditional probability of completing a repair in (t, t + dt) is dt

  • 8/6/2019 Prez Availability 5

    12/27

    Single Unit Availability Analysis (Markov Process Approach)

    For example the probability that the unit was in state 0 (operating) at time t and remained

    in state 0 at time t + dt is the probability that it did not fail in time dt, or (1 - ) dt

    the probability that the unit transitioned from state 0 (operating) to state 1

    (failed) in time t + dt is the probability of systems failure in time dt, or dt

    the probability that it was in state 1 (failed) at time t and transitioned to state 0

    (operating) in time dt is the probability that it was repaired in dt, or dt

    the probability that it was in state 1 (failed) at time t and remained in state 1 at

    time t + dt is the probability that it was not repaired in dt, or (1 - ) dt

    The single units availability

  • 8/6/2019 Prez Availability 5

    13/27

    Single Unit Availability Analysis (Markov Process Approach)

    The differential equations describing the stochastic behavior of this system canbe formed by

    the probability that the system is in state 0 at time t + dt is derived from the

    probability that it was in state 0 at time t and did not fail in (t, t + dt)

    the probability that it was in state 1 (failed) at time t and transitioned to state 0

    (operating) in time dt is the probability that it was repaired in dt, or dt

    the probability of being in state 1 at time t + dt is derived from the probability

    that the system was in state 0 at time t and failed in (t, t + dt); or it was in

    state 1 at time t, and the repair was not completed in (t, t + dt)

    the coefficients of these equations represent the columns of the transition matrix

  • 8/6/2019 Prez Availability 5

    14/27

    Single Unit Availability Analysis (Markov Process Approach)

    find the differential equations by defining the limit of the ratio

    differential - difference equations

    which yields

    If the system was initially in operation,the initial conditions are P0(0) = 1, P1(0) = 0,

    and the solutions are

    (23)

    1. Transforming Equation (23) into LaPlace transforms under the initial conditions

    2. Simplifying (24)

  • 8/6/2019 Prez Availability 5

    15/27

    Single Unit Availability Analysis (Markov Process Approach)

    3.S

    olving these simultaneously for P0(s) yields

    4. Therefore

    5. Taking the inverse Laplace transform

    (25)

  • 8/6/2019 Prez Availability 5

    16/27

    Single Unit Availability Analysis (Markov Process Approach)

    If the system was initially failed,the initial conditions are P0(0) = 0, P1(0) = 1,

    and the solutions are

    As t becomes very large, Eqs. (25) and (26) become equivalent.This indicates that after the system has been operating for some

    time its behavior becomes independent of its starting state.

    (26)

    The transient term becomes negligible when

    For a mission of (t1 - t2 ) duration, the mission availability is

  • 8/6/2019 Prez Availability 5

    17/27

    Single Unit Availability Analysis (Markov Process Approach)

    The steady state availability, As , is

    Therefore Eq. (26) becomes

    as

    The steady state availability becomes

    Usually is much larger in value than , and As may be written as

    The transient part decays relatively fast and becomesnegligible before

    If is substantially greater than , then the transient part

    becomes negligible before

  • 8/6/2019 Prez Availability 5

    18/27

    Single Unit Availability Analysis (Markov Process Approach)

    Fig.8. Single unit availability with repair

  • 8/6/2019 Prez Availability 5

    19/27

  • 8/6/2019 Prez Availability 5

    20/27

    Single Unit Availability Analysis (Markov Process Approach)

    The same technique described for a single unit canbe applied to different equipment/system reliability

    configurations, e.g., combinations of series and

    parallel units.

    As the systems become more complex, the

    mathematical manipulations can be quite laborious.

    The important trick is to set up the Markov graph

    and the transition matrix properly; the rest is just

    mechanical.For example

    For the most general case of n equipments and r repairmen where r = n,

    the steady state availability, As , is

  • 8/6/2019 Prez Availability 5

    21/27

    Single Unit Availability Analysis (Markov Process Approach)

    The same technique described for a single unit canbe applied to different equipment/system reliability

    configurations, e.g., combinations of series and

    parallel units.

    As the systems become more complex, the

    mathematical manipulations can be quite laborious.

    The important trick is to set up the Markov graph

    and the transition matrix properly; the rest is just

    mechanical.For example

    For the most general case of n equipments and r repairmen where r = n,

    the steady state availability, As , is

  • 8/6/2019 Prez Availability 5

    22/27

    The same technique described for a single unit canbe applied to different equipment/system reliability

    configurations, e.g., combinations of series and

    parallel units.

    As the systems become more complex, the

    mathematical manipulations can be quite laborious.

    The important trick is to set up the Markov graph

    and the transition matrix properly; the rest is just

    mechanical.For example

    For the most general case of n equipments and r repairmen where r = n,

    the steady state availability, As , is

  • 8/6/2019 Prez Availability 5

    23/27

  • 8/6/2019 Prez Availability 5

    24/27

    Reliability vs. Maintainabilitytradeoff example

    Figure 9 represents a system consisting of five major subsystems in aseries arrangement. The MTBF andMTTR of this system are

    Fig.9. Block

    diagram of a

    series system

  • 8/6/2019 Prez Availability 5

    25/27

    Reliability vs. Maintainability

    the maintenance time ratio equals

    which is the sum of the maintenance ratios of the five serial subsystems

    then(33)

    highest maintenance time ratios

    the first recourse could be the adding of a parallel redundant subsystem

    to Subsystem 3

    two cases may have to be considered:

    1) the case where no repair of a failed redundant unit is possible until both fail

    and the system stops operating

    2) repair is possible while the system is operating

    for a single repair crew

  • 8/6/2019 Prez Availability 5

    26/27

    Reliability vs. Maintainability

    The value of A3 = 0.946 of the redundant configuration corresponds to a

    maintenance time ratio of

    The whole system maintenance time ratio now becomes

    System availability A as compared with 0.752 without redundancy is

    Redundancy will not increase availability and may even reduce it,

    even though it increases system reliability

    To achieve gains in availability, repair of failed redundant units must be possible

    while the system is operating. This is called availability with repair.

  • 8/6/2019 Prez Availability 5

    27/27

    Reliability vs. Maintainability

    A method of straightforward trade-off between reliability and maintainability

    based on a requirement that the inherent availability of the

    system must be at least A = 0.99,

    the MTBF must not fall below 200 hr,

    and the MTTR must not exceed 4 hr

    Fig.9. Reliability-maintainability

    trade-offs


Recommended