International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

Profit-Function of Two Similar Warm Standby System subject to ’s failure for failing to implement sufficient safety checks and failed to follow good industry practice with different repair facilities

Ashok Kumar Saini BLJS COLLEGE, TOSHAM (BHIWANI) HARYANA, INDIA

White City, for failing to implement sufficient safety Abstract : "The Metronet PPP contracts to upgrade the Tube left the DfT without effective means of protecting the checks despite being ordered to do so by TfL. taxpayer. Metronet’s failure led to a direct loss to the In April 2005 the Commissioner of Transport for taxpayer of between £170 million and £410 million. The London, , pressed for an urgent review of the DfT’s work with the , TfL and London PPP, describing its performance as "bordering on Underground on a long term solution will need to improve governance and risk management in the new arrangements disaster". A week later the chief executive of Metronet they are intending to put in place to protect the was sacked, after complaints that it had made £50m taxpayer."The Comptroller and Auditor General, 5 June profit despite being behind on all its major works. By 2009 April 2005, it had started work on only 13 station refurbishments (instead of 32 as scheduled), and was In this paper we have taken failure of Metronet’s failure for failing to implement sufficient safety checks and failed more than a year behind on the refurbishment of 78 to follow good industry practice. When the main unit fails District line trains. It was also behind on its track then warm standby system becomes operative. Failure of replacement programme, having completed 28 km Metronet’s caused by failed to follow good industry instead of the anticipated 48 km. practice cannot occur simultaneously in both the units and after failure the unit undergoes Type-I or Type-II or Type- In March 2005 the House of Commons Transport Select III repair facility immediately. Applying the regenerative Committee noted that "Availability is the most point technique with renewal process theory the various important factor for Tube travellers. All the infracos reliability parameters MTSF, Availability, Busy period, needed to do to meet their availability benchmarks was Benefit-Function analysis have been evaluated. to perform only a little worse than in the past. On most Keywords: Warm Standby, failure of Metronet’s failure lines, they did not even manage that." for failing to implement sufficient safety checks and failed In November 2006, Metronet were heavily criticised by to follow good industry practice, first come first serve, the PPP arbiter, Chris Bolt, over their performance from MTSF, Availability, Busy period, Benefit -Function. 2003 to 2006. His analysis included criticism that INTRODUCTION Metronet had not performed in an economic or efficient manner, and had failed to follow good industry practice. Metronet Rail was one of two infrastructure companies (the other being Ltd) in a public-private In this paper we have taken failure of Metronet’s failure partnership with . for failing to implement sufficient safety checks and failed to follow good industry practice with different Metronet was responsible for the maintenance, renewal, repair facilities. When the main operative unit fails then and upgrade of the infrastructure on nine London warm standby system becomes operative. Failure of Underground lines from 2003 to 2008. This included Metronet’s caused by failed to follow good industry track, trains, signals, civil work and stations. From 18 practice can’t occur simultaneously in both the units and July 2007 to 26 May 2008, the company was in after failure the unit undergoes repair facility of Type- II administration and on 27 May 2008, the company by ordinary repairman or Type III, Type IV by responsibilities were transferred back into public multispecialty repairman immediately when failure of ownership under the authority of . Metronet’s failure for failing to implement sufficient In June 2009 the National Audit Office estimated that safety checks and failed to follow good industry the failure of the Metronet PPP contract cost the practice. The repair is done on the basis of first fail first taxpayer up to £410m adding that "most of the blame for repaired. Metronet's collapse lay with the consortium itself."

Metronet was declared at fault by an accident investigators' report into a May 2004 derailment at ______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 41 International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

Assumptions Metronet’s failure for failing to implement sufficient safety checks 1, 2 3 are constant failure rates when failure due warm standby, failure of Metronet’s failure for failing to 2(FGIPFfgipf, urII , Onsscf) The operative unit failed to implement sufficient safety checks and failed to follow follow good industry practice and undergoes repair of good industry practice respectively. The CDF of repair type II and the standby unit becomes operative with no time distribution of Type I, Type II and multispecialty failure of Metronet’s failure for failing to implement repairmen Type-III, IV are G1(t), G2(t) and G3(t) G4(t). sufficient safety checks

1. The failure of Metronet’s failure for failing to 3(FGIPFfgipf, urIII , Onsscf) The first unit failed to follow implement sufficient safety checks and failed to good industry practice and under Type-III multispecialty follow good industry practice is non- repairman and the other unit is operative with no failure instantaneous and it cannot come simultaneously of Metronet’s failure for failing to implement sufficient in both the units. safety checks

2. The repair starts immediately after failure of 4(SSCF sscf,uR1 , SSCF sscf,wrI) The unit failed due to Metronet’s failure for failing to implement SSCF resulting from failure of Metronet’s failure for sufficient safety checks and failed to follow good failing to implement sufficient safety checks under industry practice and works on the principle of repair of Type- I continued from state 1and the other first fail first repaired basis. The repair facility unit failed due to SSCF resulting from failure of does no damage to the units and after repair units Metronet’s failure for failing to implement sufficient are as good as new. safety checks is waiting for repair of Type-I.

3. The switches are perfect and instantaneous. 5(SSCFsscf,uR1 , FGIPFfgipf,wrII) The unit failed due to SSCF resulting from failure of Metronet’s failure for 4. All random variables are mutually independent. failing to implement sufficient safety checks is under 5. When both the units fail, we give priority to repair of Type- I continued from state 1and the other operative unit for repair. unit failed to follow good industry practice is waiting for repair of Type- II. 6. Repairs are perfect and failure of a unit is detected immediately and perfectly. 6(FGIPFfgipf, uRII , SSCFsscf ,wrI) The operative unit failed to follow good industry practice is under repair 7. The system is down when both the units are non- continues from state 2 of Type –II and the other unit operative. failed due to SSCF resulting from failure of Metronet’s Symbols for states of the System failure for failing to implement sufficient safety checks is waiting under repair of Type-I. Superscripts O, WS, SSCF, FGIPF, 7(FGIPFfgipf,uRII , SSCFsscf,wrII) The one unit failure due Operative, Warm Standby, failure of Metronet’s failure to failed to follow good industry practice is continued to for failing to implement sufficient safety checks and be under repair of Type II and the other unit failed due failed to follow good industry practice respectively to SSCF resulting from failure of Metronet’s failure for Subscripts nsscf, sscf, fgipf, ur, wr, uR failing to implement sufficient safety checks is waiting for repair of Type-II. No failure of Metronet’s failure for failing to implement sufficient safety checks, failure of Metronet’s failure for 8(SSCFsscf,urIII , FGIPFfgipf, wrII) The one unit failure of failing to implement sufficient safety checks, and failed Metronet’s failure for failing to implement sufficient to follow good industry practice, under repair, waiting safety checks is under multispecialty repair of Type-III for repair, under repair continued from previous state and the other unit failed to follow good industry practice respectively is waiting for repair of Type-II. Up states– 0, 1, 2, 3, 10 ; Down states – 4, 5, 6, 7,8,9,11 9(SSCFsscf,urIII, FGIPFfgipf, wrI) The one unit failure of Metronet’s failure for failing to implement sufficient regeneration point – 0,1,2, 3, 8, 9,10 safety checks is under multispecialty repair of Type-III States of the System and the other unit failed to follow good industry practice waiting for repair of Type-I 0(Onsscf, WSnsscf) One unit is operative and the other unit is warm standby and there is no failure of Metronet’s 10(Onsscf , FGIPFfgipf, urIV ) failure for failing to implement sufficient safety checks The one unit is operative with no failure of Metronet’s of both the units. failure for failing to implement sufficient safety checks and warm standby unit failed to follow good industry 1(SSCFsscf, urI , Onsscf) The operating unit failure of Metronet’s failure for failing to implement sufficient practice and undergoes repair of type IV. safety checks is under repair immediately of Type- I and standby unit starts operating with no failure of

______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 42 International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

11(Onsscf , FGIPFfgipf, uRIV )

The one unit is operative with no failure of Metronet’s MTSF = E[T] = (s) failure for failing to implement sufficient safety checks and warm standby unit failed to follow good industry s=0 practice and repair of type IV continues from state 10. ’ ’ = (D1 (0) - N1 (0)) / D1 (0) Transition Probabilities = ( + ( p01 + p0,10 p10,1) +( p02 + p0,10 p10,2)( Simple probabilistic considerations yield the following + µ )+ µ p / (1 - (p + p p ) p - (p + expressions: 3 10 0,10 01 0,10 10,1 10 02 p0,10 p10,2) p23 ) - p0,10 p10,0 p01 = 1 / 1 + 2 +3, where p02 = 2 / 1 + 2 +3 , p0,10 = 3 / 1 + 2 +3 휇0 = 휇01+ 휇02 +µ0,10 , * * (4) (5) p = pG (  )+q G (  ) 휇1 = 휇10 + 휇11 + 휇12 , 10 1 1 2 2 , (7) (6) * (4) 휇2 = 휇23+휇28 + 휇29 , p14 = p- pG1 ( 1) = p11 , * (5) µ10= µ10,0 + µ10,1+ µ10,2 p15 = q- q G1 ( 2) = p12 , * * p23 = pG2 ( 1)+q G2 ( 2) , Availability analysis p = p- pG *(  ) = p (6) , 26 2 1 29 Let M (t) be the probability of the system having started p = q- qG *(  ) = p (7), i 27 2 2 28 from state i is up at time t without making any other p30 = p82 = p91 = 1 * * regenerative state. By probabilistic arguments, we have p0,10 = pG4 ( 1)+q G4 ( 2) * (11) − t − t − t p10,1 = p- pG4 ( 1) = p10,1 M0(t) = 푒 1 푒 2 푒 3 * (11) p10,2 = q- q G4 ( 2) = p10,2 (1) -  t , M1(t) =p G1(t) e 1 We can easily verify that -  t M2(t) =q G2(t) e 2 , p01 + p02 + p03 = 1, -  t (4) (5) M3(t) = G3(t), M 10(t) = G4(t) e 3 p10 + p14 (=p11 ) + p15 (=p12 ) = 1, (6) (7) p23 + p26 (=p29 ) + p27 (=p28 ) = 1 p30 = p82 = p91 = 1 The point wise availability Ai(t) have the following (11) (12) p10,0 + p10,1 (=p10,1) + p10,2 (=p10,2 ) = 1 (2) recursive relations

And mean sojourn time is A0(t) = M0(t) + q01(t)[c]A1(t) + q02(t)[c]A2(t) + q0,10(t)[c]A10(t) µ = E(T) = 0 (5) A1(t) = M1(t) + q10(t)[c]A0(t) + q12 (t)[c]A2(t)+ (4) Mean Time to System Failure q11 (t)[c]A1(t) , (7) Ø0(t) = Q01(t)[s] Ø1(t) + Q02(t)[s] A2(t) = M2(t) + q23(t)[c]A3(t) + q28 (t)[c] A8(t) + (6) Ø2(t)+ Q0,10(t)[s] Ø10(t) q29 (t)] [c]A9(t) A3(t) = M3(t) + q30(t)[c]A0(t) Ø1(t) = Q10 (t)[s] Ø0(t) + Q14(t) + Q15(t) A8(t) = q82(t)[c]A2(t) Ø2(t) = Q23 (t)[s] Ø3(t) + Q26(t) + Q27(t) Ø3(t) = Q30(t)[s] Ø0(t) A9(t) = q91(t)[c]A1(t) Ø10(t) = Q10,0(t)[s] Ø10(t) + Q10,2(t)[s] (11) A10(t) = M 10(t) + q 10,0(t)[c]A 0(t) + q10,1 (t)[c]A1(t)+ q Ø1(t)+ Q10,2(t)[s] Ø2(t) (3-6) (11) 10,2 (t)[c]A2(t) (8-14) We can regard the failed state as absorbing Taking Laplace Transform of eq. (8-14) and solving for Taking Laplace-Stiljes transform of eq. (3-6) and solving for

* = N2(s) / D2(s) (15) ø0 (s) = N1(s) / D1(s) (7) where where * * * * * (4) (7 N1(s) = {Q01 + Q0,10 Q10,1 } [ Q14 (s) + Q15 (s) ] + N2(s) ={ 0,10 10+ 0 } [{1 – 11 }{1- 28 82 * * * * * (5) (6) {Q02 + Q0,10 Q10,2 } [ Q26 (s) + Q27 (s) ] }- 12 29 * * * * * * D1(s) = 1 - {Q01 + Q0,10 Q10,1 } Q10 - {Q02 + Q0,10 (11) * * * * * 91 ] + { 01+ 0,10 10,1 }[ 1 Q10,2 } Q23 Q30 - Q0,10 Q10,0 (7) (5) Making use of relations (1) & (2) it can be shown that {1 – 28 82} + 12 23 3+ 2]+{ 02 + 0,10 * (11) (4) (6) ø0 (0) =1 , which implies that ø0 (t) is a proper 10,2 } [{ 23 3}{1 – 11 }+ 29 91 1] distribution. (4) (7 (5) (6) D2(s) = {1 - 11 }{1- 28 82 }- 12 29 91 -

{ 01+ 0,10 ______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 43 International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

(11) (7) (5) R (t) = S (t) + q (t)[c]R (t) 10,1 }[ 10 {1 – 28 82} + 12 23 30 ] – { 8 8 82 2

02 + 0,10 R9(t) = S9(t) + q91(t)[c]R1(t) (11) (11) (4) (6) R10(t) = S10(t) + q 10,0(t)[c]R0(t) + q10,1 (t)[c]R1(t)+ q 10,2 }{[ 23 30 {1 – 11 }+ 29 91 10] (11) 10,2 (t)[c]R2(t) (19-25) (Omitting the arguments s for brevity) where

The steady state availability -  t S1(t) =p G1(t) e 1 ,

A0 = -  t S 2(t) =q G2(t) e 2

= = S3(t) = S8(t)= S9(t) = G3(t)

S10(t) = G4(t) (26) Using L’ Hospitals rule, we get Taking Laplace Transform of eq. (19-25) and solving for

A0 = = (16)

Where = N3(s) / D2(s) (27)

(4) (7) where N2(0) ={p0,10 10 (0)+ 0 (0) } [{1 – p11 }{1- p28 }- (5) (6) (11) (7) (11) p12 p29 ] + { p01+ p0,10 p10,1 }[ 1(0){1 – p28 } N 3(s) ={ 01 + 0,10 10,1 }[ 푆 1(1 – +p (5) p (0)+ (0)]+{ p +p p (11)} [{p 12 23 3 2 02 0,10 10,2 23 (7) (5) (4) (6) 28 82} + 12 [ 푆2 + 23 푆3+ 3(0)+ 2(0) }{1 –p11 }+ p29 1(0)] (7) (6) ’ (7) (5) (6) 28 푆 8+ 29 푆 9)]]+ { 02 + 0,10 D2 (0) =µ0[p10 (1- p28 }+ p12 p23 ]+ µ1[p29 + p01 p23 - (7) (11) (4) p0,10 {p10,0{1- p28 }+p23 p10,2 p23}]+ µ2[(1-p11 ) - (11) (7) (6) (11) (5) 10,2 } [ { 푆2+ 23푆3 + 28 푆8 + 푆9 29 )(1- p01 p10 -p0,10 (p10 - p10 p10,2 + p12 p10,0 )] } + µ3 (4) (6) (5) (11) (4) 11 )+ 푆 1 29 91] + 0,10 [p23[p12 {p01 + p0,10 p10,1 }+(1 – p11 }{ p02 + p0,10 (11) (7) p10,2 }]+ µ8 [p28 (1- p0,10 p10,0 - p10{ p01+ p0,10 (7) (4) (6) (11) (6) (5) 푆 10 [{1- 28 82 }{1- 11 }- 29 p10,1 })] + µ9 [p29 { p12 (1- p0,10 p10,0 +( p02 + p0,10 (11) (6) (5) p10,2 })] + µ10 [p29 { p12 (1- p0,10 p10,0 +( p02 + p0,10 (5) (11) 91 12 ] p10,2 })] and D (s) is already defined. and 2 (Omitting the arguments s for brevity) µ3 = µ30 , µ9 = µ91 , µ8 = µ81

The expected up time of the system in (0,t] is In the long run, R0 = (28)

(t) = Where

(11) N 3(0) ={p01 +p0,10 p10,1 }[ 푆 1(1 – So that (17) (7) (5) (7) (6) p28 } +p12 [ 푆 2 +p23 푆 3+p28 푆 8+p29 푆 9)]]+ {p02 (11) (7) (6) The expected down time of the system in (0,t] is +p0,10 p10,2 } [ { 푆 2+ p 23푆 3 +p 28 푆 8 + 푆 9 p29 )(1- (4) (6) (7) (4) (6) p11 )+ 푆 1p29 ] + p0,10 푆 10 [{1-p28 }{1- p11 }- p 29 p (5) (t) = t- (t) 12 ] ’ So that (18) and D 2 (0) is already defined. The expected busy period of the server when there is The expected busy period of the server when there is failure due to failure of Metronet’s failure for failing to failure due to failure of Metronet’s failure for failing implement sufficient safety checks and failure due to to implement sufficient safety checks and failure due failed to follow good industry practice in (0,t] is to failed to follow good industry practice in (0,t]-R0

R0(t) = q01(t)[c]R1(t) + q02(t)[c]R 2(t) + q0,10(t)[c]R10(t) (t) = So that (5) R1(t) = S1(t) + q10(t)[c]R0 (t) + q12 (t)[c] R2 (t) + (4) The expected number of visits by the repairman q11 (t)[c]R1(t) Type-I or Type-II for repairing the identical units in (7) R2(t) = S2(t) + q23(t)[c]R3(t) + q28 (t) (0,t]-H0 (6) R8(t) +q29 (t)][c]R9(t) H0(t) = Q01(t)[s][1+ H1(t)] +

R3(t) = S3(t) + q30(t)[c]R0(t) Q02(t)[s][1+H2(t)]+Q0,10(t)[s] H10(t)]

______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 44 International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

(5) (11) H1(t) = Q10(t)[s]H0(t)] + Q12 (t)[s] where N5(0) = {p 01+ p 0,10 p10,1 } (4) H2(t) + Q11 (t)] [s]H1(t) , (5) (11 (4) (7) p 12 + { p 02+ p 0,10 p10,2 } {1 – p 11 }] H2(t) = Q23(t)[s]H3(t) + Q28 (t) [s] (6) H8(t) +Q29 (t)] [c]H9(t) The expected number of visits by the multispecialty H3(t) = Q30(t)[s]H0(t) repairman Type-III for repairing the identical units H8(t) = Q82(t)[s]H2(t) in (0,t]-Y0 H9(t) = Q91(t)[s]H1(t) (11) Y0(t)=Q01(t)[s]Y1(t)+Q02(t)[s] Y2(t) +Q0,10(t)[s] H10(t) = Q10,0(t)[s]H10(t)] + Q10,1 (t)[s]H1(t)]+ (11) [1+Y10(t)] Q10,2 (t)[s] H2(t)] (29-35) (5) Taking Laplace Transform of eq. (29-35) and solving for Y1(t) = Q10(t)[s]Y0(t) + Q12 (t)[s] (4) Y2(t) + Q11 (t)] [s]Y1(t) , (7) Y 2(t) = Q23(t)[s]Y3(t) + Q28 (t) [s] = N4(s) / D3(s) (36) (6) * * (4)* (7)* * Y8(t) +Q29 (t)] [c]Y9(t) N4(s) = { Q01 + Q02 }[ { 1 – Q11 }{1-Q28 Q82 } – (5)* (6)* * Q12 Q29 Q91 ] Y3(t) = Q30(t)[s][1+Y0(t) ]

And Y8(t) = Q82(t)[s]Y2(t) (4)* (7)* * (5)* (6)* D3(s) = {1 – Q11 } { 1- Q28 Q82 } – Q12 Q29 Y9(t) = Q91(t)[s]Y1(t) * * * * * (11)* * Q91 ](1- Q0,10 Q10,0 )-{ Q01 + Q0,10 Q10,1 }[ Q10 { 1 – (11) (12) (7)* * (5)* * * * * Y10(t)=Q10,0(t)[s]Y0(t)+ Q10,1 (t)[s] Y1(t) + Q10,2 (t)[s] Q28 Q82 }+ Q12 Q23 Q30 ] – {Q02 + Q0,10 (11)* * * (4)* (6)* * * Y2(t) (47-53) Q10,2 }[ Q23 Q30 {1 – Q11 }+ Q29 Q91 Q10 ] Taking Laplace Transform of eq. (47-53) and solving (Omitting the arguments s for brevity) * forY0 (s),we get In the long run, * Y0 (s) = N6(s) / D3(s) (54) ’ H0 = N4(0) / D3 (0) (37) * (4)* (5)* * N6(s) = Q0,10 [{1 – Q11 }(1- Q28 Q82 } - (5)* (6)* * * * * where Q12 Q29 Q91 {1- Q0,10 Q,10,0 }+{Q02 + * (11)* * * (4)* * (6)* (4) (7) (5) (6) Q0,10 Q10,2 }[ [ Q23 Q30 {1 – Q11 }+ Q10 Q29 N4(0) ={1–p0,10}[ {1 – p 11 } { 1- p 28 } – p 12 p 29 ] * Q91 ] The expected number of visits by the multispecialty (Omitting the arguments s for brevity) repairman Type-III for repairing the identical units in (0,t]-W0 In the long run, ’ W0(t)=Q01(t)[s]W1(t)+Q02(t)[s] W 2(t) + Q10,0(t)[s] W10(t) W 0 = N6(0) / D3 (0) (55) (5) (4) (7) (5) (6) W 1(t) = Q10(t)[s]W 0(t)] + Q12 (t)[s] where N6(0) = p 0,10[{1-p 11 }{1- p28 }- p12 p29 ] (4) (5) (11 (4) W 2(t) + Q11 (t)] [s]W1(t) , p 12 + { p 02+ p 0,10 p10,2 } {1 – p 11 }] (7) W 2(t) = Q23(t)[s]W 3(t) + Q28 (t) [s] Benefit- Function Analysis (6) W 8(t) +Q29 (t)] [c]W9(t) The Benefit-Function analysis of the system considering mean up-time, expected busy period of the system under W (t) = Q (t)[s][1+W (t) ] 3 30 0 failure of Metronet’s failure for failing to implement W 8(t) = Q82(t)[s][1+W2(t) ] sufficient safety checks and failed to follow good industry practice, expected number of visits by the W (t) = Q (t)[s][1+W (t) ] 9 91 1 repairman for unit failure. The expected total Benefit- (11) W10(t)=Q10,0(t)[s]W0(t)+ Q10,1 (t)[s] W1(t) + Function incurred in (0,t] is (12) Q10,2 (t)[s] W2(t) (38-44) C(t) = Expected total revenue in (0,t] Taking Laplace Transform of eq. (33-39) and solving for - expected busy period of the server when there is failure due to failure of Metronet’s failure for failing to implement sufficient safety checks and failed to follow = N5(s) / D3(s) (45) good industry practice in (0,t] * * (11)* (5)* * * N5(s) = {Q01 + Q0,10 Q0,10 }[Q12 [ Q23 Q30 + (5)* * (6)* * * * (11)* - expected number of visits by the repairman Type- I Q28 Q82 + Q29 Q91 ] + {Q02 + Q0,10 Q10,2 }[ [ * * (5)* * (6)* * (4)* or Type- II for repairing of identical the units in (0,t] Q23 Q30 + Q28 Q82 + Q29 Q91 {1 – Q11 }] - expected number of visits by the multispecialty (Omitting the arguments s for brevity) repairman Type- III for repairing of identical the units In the long run, in (0,t] ’ W 0 = N5(0) / D3 (0) (46) ______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 45 International Journal of Recent Advances in Engineering & Technology (IJRAET) ______

- expected number of visits by the multispecialty to follow good industry practice, the MTSF, steady state repairman Type- IV for repairing of identical the units availability decreases and the Profit-function decreased in (0,t] as the failure increases.

C = = REFERENCES [1] Dhillon, B.S. and Natesen, J, Stochastic Analysis = K1A0 - K 2R0 - K 3H0 - K 4W0 –K5Y0 of outdoor Power Systems in fluctuating where environment, Microelectron. Reliab. ,1983; 23, 867-881. K1 - revenue per unit up-time, [2] Kan, Cheng, Reliability analysis of a system in a K2 - cost per unit time for which the system is busy under repairing, randomly changing environment, Acta Math. Appl. Sin. 1985, 2, pp.219-228. K3 - cost per visit by the repairman type- I or type- II for units repair, [3] Cao, Jinhua, Stochastic Behaviour of a Man Machine System operating under changing K4 - cost per visit by the multispecialty repairman environment subject to a Markov Process with Type- III for units repair two states, Microelectron. Reliab. , 1989; 28, pp. 373-378. K5 - cost per visit by the multispecialty repairman Type- IV for units repair [4] Barlow, R.E. and Proschan, F., Mathematical CONCLUSION theory of Reliability, 1965; John Wiley, New York. After studying the system, we have analyzed graphically [5] Gnedanke, B.V., Belyayar, Yu.K. and Soloyer , that when the failure rate due to Metronet’s failure for A.D., Mathematical Methods of Reliability failing to implement sufficient safety checks and failed Theory, 1969; Academic Press, New York.

Fig. The State Transition Diagram Up-State Down-State

regeneration point 

______ISSN (Online): 2347 - 2812, Volume-3, Issue -7, 2015 46