Sie sind auf Seite 1von 10

2321-7871

Weekly Science Research Journal

Vol-1, Issue-14, 24th October 2013

Primary Article

Down Time Analysis In Process Industry-A Case


Study
Dnyaneshwar R.Thawkar

ABSTRACT
The growths of present day industries are forced towards the use of more
complex system. The production system consists thousands of parts and
components and the failure of one or more component may lead to affect the
entire production system. So with increase in automation and usage of
complex systems, evaluation of reliability has recently been recognized for
effective maintenance. Reliability has been evaluated mostly for electronic
components since and presently we are going to calculate Reliability for
components in process industry. For this purpose, the break down time is
calculated from sugar Industry. The data to be analyzed and reliability is
calculated for each component by weibull distribution in probability is
analyzed and criticality of each component is found by using FMEA (Failure
Mode Effect Analysis) software to suggest some of the preventive maintenance
schedules.
Keywords:Reliability, Weibull, Availability, MTBF, FMEA.
1.0 INTRODUCTION
The study was carried in sugar mills limited is situated NAGAPP
ATTINAM district in Tamil Nadu. The crushing capacity of the plant is 3500
TCD (Tones of cane Crushed per day). The various processing sections of the
factory are namely a) Loading section b) Mill section c) boiling section. The
preliminary discussion with the official's concerned reveal that the plant is
often giving rise to problems due to breakdown of various components; and
hence, it is decided to carry out failure analysis in this industry, and to suggest
measures to improve the availability of the components The objective of the
study is to estimate the availability, reliability of the components system and
also to perform the FMEA analysis to identify the critical components to
prepare PM (Preventive Maintenance) schedule by officials.
1.1 METHODOLOGY
The methodology followed to achieve is as follows:
1.The historical failure data, the down time and the availability of the
equipment is collected.
2.From the past failure data, the down time and the availability of the
equipment is calculated.
3.Weibull statistical distribution is used to model the failure history.
4.Using the parameters of weibull distribution the estimation of reliability has
been carried out.
5.Using FMEA software package, the criticality index of all components is
calculated.
6.The maintenance schedules are prepared in such a way that the system will
operate at or below minimum failure rate.
1.2 MEAN TIME BETWEEN FAILURES (MTBF)
The mean time between failures refers to average time of breakdown
until the device beyond Repair. The mean time between failures is one of the
Page No-1

Dnyaneshwar R.Thawkar
From
Asst. professor
M.Tech.(Industrial Engineering)
Umrer college of Engineering, Umrer
The Article is published on October
2013 issue & available at
www.weeklyscience.org
DOI: 10.9780/ 2321-7871/1142013/42

2321-7871

th

Vol-1, Issue-14, 24 October 2013

Down Time Analysis In Process Industry-A Case Study


useful terms in maintenance and reliability analysis.
Operating Time
MTBF = _______________
Number of failures
1.3 DATA COLLECTION
Failure data of the plant are collected refers to average time of breakdown until the
device beyond repair. The mean time between failures is one of the useful terms in
maintenance and reliability analysis. MTBF = Operating Time Number of failures.
1.4 AVAILABILITY
It is doable to outline 3 sorts of availableness reckoning on the time parts we have a
tendency to soak up to consideration. These are,Inherent availableness is that the chance
that a system or instrumentation shall operate satisfactorily once used below expressed
conditions in a perfect support settingwithout consideration for any schedule or preventive
maintenance at any given time .deliver the goods availableness within the definition of
inherent availableness we have a tendency to thought-about MTBF that doesn't take
in to account the period caused by maintenance. If this can be conjointly taken into
consideration, we get the achieved availableness, that is outlined to the chance that a
system or instrumentation shall operate satisfactorily once used below expressed
conditions in a perfect support environment at any given time. If any real time operation, we
have a tendency to can't scale back body period and provide downtime to zero. a definite
quantity of delay can continually be caused by time parts like these, and if they're taken in
to account it. Operational Availability of the system be the probability that a system or
equipment shall operate satisfactorily and in an actual supply environment at any given
time.
MTBF
A0 = _____________
MTBF + MDT
Where MTBF Mean Time Between Failures, MDT Mean Down Time
1.5 ESTIMATION OF OPERATIONAL AVAILABILITY
Downtime of Components
Downtime is the non-productive time of the machine. Downtime of the components
are calculated and tabulated with the help of the data collected. Table 4.2 shows the
downtime of the various components from March'04 to May'04. For example, the downtime
of the CANRCARRIER during March'04 is observed to be 6.45 hrs: min, which is the sum of
the breakdown hours on various occasions during the month of March'04. Similarly for all
other components for various periods, the want of cane time are calculated and tabulated.
Table 4.3 presents the data with respect to the number of failures for each component
month wise.
1.6 RELIABILITY ESTIMATION
Definition of Reliability
Reliability, which is a measure of quality, is an essential element at each stage of the
equipment manufacturing procedure through design and production to final delivery to the
user. Reliability, It is simplest form, means the probability that a failure may not occur in a
given time interval. A more rigorous definition of reliability is a follows Reliability of a unit
(or product ) is the probability that the unit performs its intended function adequately for a
given period of time under the stated operating conditions or environment. Reliability
characteristics, such as probability of survival, mean time to failure, availability, mean
down time and frequency of failures are some of the measures of system effectiveness.
Apart from the above factors, reliability does change due to other factors like quality,
workmanship, manufacturing process, material, storage, handling, engineering changes,
and deviations in production, inspection and test.

Page No-2

Vol-1, Issue-14, 24th October 2013

2321-7871
Down Time Analysis In Process Industry-A Case Study

Application of Weibull Distribution


In 1951, Weibull suggested a simple empirical expression, which represents a great varies
of actual data. The weibull cumulative distribution function is given by
-(t / )
F(t)
= 1-e t > , > 0
= 0, otherwise
Where ,
Scale parameter
Shape factor
Location factor
There various functions are given as
R (t) = Reliabilities = exp[-(t- ) / ) ]
(t) = Failure rate = ] / ((t- ) / ) -1
The constants appearing in these expressions represent.
MTBF of components
Mean Time Between Failures is referred to as the average time to satisfactory operation of
the system. This term is useful to carryout the maintenance and reliability analysis.
Operating time
MTBF
= ________________
No. of failures
Total Available Time Non-operating time
= _______________________________________
No. of failures
For example:
Component Name
Total available time

= Cane Carrier
= (31+30+31) x 24 hrs
= 2208 hrs.
Non-operating time
= Total breakdown time of cane carrier
= 13.00 hrs:min[FromTable 4.2]
Number of Failures= 10[From Table 4.3] 2208- 13.00
MTBF =

2208 13.00
_____________
10
= 219.50

Mean Downtime
The statistical mean of downtime d1, d2,d3, .. including supply time and
administrative downtime is called mean downtime. This mean downtime is concentrated on
breakdown time maintenance time and non-availability of components Table 4.5 shows the mean
downtime of various components.
For Example:
Month
Component
Downtime due to Breakdown
General down time
Total down time

= March'04
= Cane Carrier
= 13.00 hrs: min [Table 4.2]
= 156.40 hrs: min [Table 4.4]
= 13.00 + 156.40
= 163.25 hrs: min

Mean down time = 163. 25+ 71.00 + 179.50


_____________________
3
= 138.32 hrs.

Page No-3

Vol-1, Issue-14, 24th October 2013

2321-7871
Down Time Analysis In Process Industry-A Case Study

Availability of components
It is possible to define three types of availability of an equipment hours in
actual environments operational availability is defined as to be the probability that a
system of equipment shall operate satisfactorily when used under stated condition in
an actual environment at any given time.
A0

Mean time between failure


= ____________________________
Mean time between failure + Mean Down Time

For Example:
Component Name
MTBF
MDT

= Cane Carrier
= 219.50 hrs
= 138.32 hrs

Mean down time

219.50
= ________________
219.50 + 138.32
= 61.34%

Thus Availability of all components is calculated and tabulated in the TABLE


4.6 Locating constant defining the starting point or origin of the distribution can
be thought as a gurantee period in which no failure can occur or it can be thought of
as the minimum life.
Scaling constant, stretching the distribution along the time axis.
Shape parameter, which decides the shape pattern of failure.
When dealing with failure rates, the weibull shape parameter is special
importance as it describes the mode of failure.
For example, if
< 1, it indicates that the failure rate is a decreasing function of time and is
characterizes as an early failure phase.
=1, it means the failure rate is constant overtime, as is the case for exponential
distribution.
> 1, it means that the failure rate is increasing with time and can be characterized
as the out phase.
1.7 FAILURE TIME DISTRIBUTION: CALCULATION PROCEDURES
The following procedure is adopted to calculate the required parameters.
i) Class interval
Tmax Tmin
I
= ____________________
1 + 3...31og N
Where,
T max max. Time between failures
T min min Time between failures
N Total number of failures in the test time
ii) Percentage of failure
Number of failures in the time, interval
% failed =
___________________________________
Total Number of failures
iii) Cumulative percentage of failure
n
F(t)= fi
Page No-4

Vol-1, Issue-14, 24th October 2013

2321-7871
Down Time Analysis In Process Industry-A Case Study
I
=1
n Corresponding time interval sequence number.
iv) Reliability
R (t) Reliabilities = exp [-(t-) / )]
REALIABILITY OF COMPONENTS
COMPONENT NAME: CANE CARRIER
From Weibull Graph
= 1410
= 1.8
/ = 0.888
Therefore
= 1252.08 hrs
T=+
= 1252.08+ 0
= 1252.08 hrs.
R (t) = exp [- (1252.08-0) /- 1410) 1.8]* 100
= 44.6%

Thus Reliability is calculated for each component and it is tabulated in the TABLE.5.1.
From calculation of Reliability for all components it is found that New boiler has low Reliability
1.8 FMEA ANALYSIS AND PROCESS
The following logical steps should be followed when an FMEA,
1.Identify the product or system components.
2.List all possible failure modes of each component.
3.Set down the effects that each mode of failure would have on the over all function of products
system.
4.List all the possible causes of each failure mode.
5.Asses numerically the failure modes on a scale from 1 to 10.
6.Experience and Reliability data from company is given a input to determine the values, on a
scale 1-10 for severity (S), Occurrence (O) 7ecion (D).
SEVERITY (S)
Severity is the assessment of the seriousness of the effect of potential failure of system,
subsystem or component severity is applicable only to effect of failure mode severity is rated by
ranking by which 1 is for no effect and 10 for the most severe (serious) effect.
OCCURRENCE (O)
Occurrence is the probability that one of the specific cause/mechanism of failure will
occur.The likelihood of occurrence is assessed as 1 for least chance of occurrence and 10 for
highest Chance of occurrence.
DETECTION (D)
It is the relative measure of difficulty of detecting the failure before the product or service
is used by customer. If the design control certainly detect cause/mechanism of failure then it is
ranked or it is difficult to detect then it is ranked. Thus when these inputs are given, the results
generated from the software are tabulated in the table 6.1. The criticality index for each
component is generated and ranked according to the critical failure modes. This indicates the
relative priority of each failure mode and to concentrate on preventive activity of critical
components.
1.9 CONCLUSION
Availability of data regarding the equipment Reliability is of considerable benefit to
industry in many situations. Knowledge of life expectancy and wear out characteristics of the
system components leads to the development of sound Maintenance and appropriate provision
for spare parts and stand by equipment. It is possible to correlate equipment Reliability
withmaintenance requirements. Machines made by man, at any time succumb to fail. To attain
maximum productivity it is necessary that man minimize failure. This can be achieved by proper
Page No-5

2321-7871

th

Vol-1, Issue-14, 24 October 2013

Down Time Analysis In Process Industry-A Case Study


maintenance and timely replacements of some parts of machines or at times the whole
equipment. The timely replacement improves the machine Reliability as well as Availability
and made it useful for a achieving high productivity. An attempt had made in this thesis to
study the failure pattern and down time of the components. The critical components have
also been identified by carrying out FMEA analysis. It would be economical if the failures of
these critical components are minimize by proper preventive maintenance measures.
2.0 REFERENCES
1.Govil. A.K. (1983), Reliability Engineering, Tata, Mcgraw Hill Company Ltd., New Delhi.
2.Lewis W.W.(1987), Introduction to Reliability Engineering, John Willey & sons, Newyork.
3.Malik M.A.K(1979), A note on the physical Meaning of the Weibull Distrubition, IEEE
Trans. On Reliability, Vo 1. 24, No.1, Page:95.
4.Sinha S.K.(1980), Life Testing And Reliability Estimation, Willey Eastern Limited, India.
5.Srinath. L., Reliability Engineering . Balagurusamu E., Reliability Engineering.
TABLE 4.7 : VALUE OF TMAX AND TMIN
Sr. No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

NAME OF THE COMPONENT


CANE CARRIER
MILLS
RAKE/INTER CARRIER
BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRUP STORAGE TANK
OTHER UNITS IN BOILING HOUSE

TMAX (HRS)
601.45
44.20
395.20
568.45
1478.00
1313.25
767.00
785.75
881.45
1541.40
819.00
776.15
1132.60
943.45
522.20

T MIN
(HRS)
39.15
1.50
0.30
32.00
9.00
14.30
1.25
267.15
153.00
14.35
1.05
1.15
0.45
9.30
27.20

Table 4.1: Failure Data (fro example) Data Wise March 04


DATE
02.03.04
03.03.04
04.03.04
05.03.04

FROM
7.25
8.40
--8.30
9.25
9.50

TO
7.45
9.15
--8.55
9.40
16.50

TOTAL HOURS
0.20
0.35
--0.25
0.15
7.00

Page No-6

REASON
To reduce Juice Level in the Evaporator
To reduce Juice in the evaporator
Holiday
Holiday
Boiler water Scarcity
Kicker Jamming
11nd mill top roller shell circumference
crack.

Vol-1, Issue-14, 24th October 2013

2321-7871
Down Time Analysis In Process Industry-A Case Study
TABLE 4.2 : DOWN TIME OF COMPONENTS
S.NO.

NAME OF THE COMPONENT

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

CANE CARRIER
MILLS
RAKE / INTER CARRIER
BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRP STORAGE TANK
OTHER UNITS IN BOILIBG
HOUSE

MARCH
(HRS)
6.45
10.25
3.45
0.45
63.25
0.30
2.00
-8.45
2.55
9.00
1.45
5.00
2.15
1.30

APRIL
(HRS)
5.10
6.25
4.35
4.15
-2.05
2.05
2.50
0.25
-2.45
0.15
-1.30
2.05

MAY
(HRS)
1.05
1.35
2.30
0.40
0.30
3.15
1.10
0.55
1.10
0.55
1.55
1.55
6.25
2.55
3.20

TOTAL
(HRS)
13.00
18.25
10.50
5.40
63.55
5.50
5.15
3.45
10.20
3.50
13.40
3.55
11.25
6.40
6.55

TABLE 4.3: NUMBER OF FAILURES OF COMPONENTS

SR.
NO.

NAME OF THE
COMPONENT

MARCH
(HRS)

APRIL
(HRS)

MAY
(HRS)

1
2

CANE CARRIER
MILLS

2
6

5
6

3
2

TOAL No.
OF
FAILURES
10
15

RAKE / INTER CARRIER

15

4
5
6
7
8
9
10
11
12
13
14

BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRP STORAGE TANK
OTHER UNITS IN BOILIBG
HOUSE

2
12
1
3
-2
4
5
4
7
3

5
-1
3
2
1
-4
1
-1

2
1
4
2
1
2
2
3
5
4
4

9
13
6
8
3
5
6
12
10
11
8

15

TABLE 4.4: GENERAL DOWN TIME


MONTH
MARCH 04
APRIL 04
MAY 04

GENERAL
CLEANING
(HRS)
19.35
15.45
58.45

WANT OF
CANE (HRS)

HOLIDAY
(HRS)

TOTAL
HOURS

65.05
2.05
58.45

72.00
48.00
120.00

156.40
65.50
178.45

Page No-7

2321-7871

th

Vol-1, Issue-14, 24 October 2013

Down Time Analysis In Process Industry-A Case Study


TABLE 4.5 : MEAN DOWN TIME OF COMPONENT

S.
NO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

NAME OF THE
COMPONENT
CANE CARRIER
MILLS
RAKE / INTER CARRIER
BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN
BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRP STORAGE TANK
OTHER UNITS IN
BOILIBG HOUSE

TOTAL DOWN TIME (HRS : MIN)


MARCH
APRIL
MAY
(156.40)
(65.50)
178.45)
163.25
71.00
179.50
167.05
72.15
180.20
160.25
70.25
181.15
157.25
70.05
179.25
220.05
65.50
179.15
157.10
67.55
182.00
158.40
67.55
179.55
156.40
68.40
179.40
165.25
66.15
179.55
159.35
65.50
179.40
165.40
158.25
161.40
158.55
158.10

68.35
66.05
65.50
6720
67.55

180.40
180.40
185.10
181.40
182.05

MEAN DOWN
TIME
(HRS:MIN)
138.32
140.20
137.22
135.52
155.30
135.55
135.17
135.13
137.38
135.15
138.05
135.30
137.33
136.12
136.30

TABLE 4.6 : AVAILABILITY OF COMPONENTS

S.
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

NAME OF THE COMPONENT


CANE CARRIER
MILLS
RAKE / INTER CARRIER
BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRP STORAGE TANK
OTHER UNITS IN BOILIBG
HOUSE

Page No-8

MTBF
(HRS)
6.45
10.25
3.45
0.45
63.25
0.30
2.00
-8.45
2.55
9.00
1.45
5.00
2.15
1.30

MDT
(HRS)
138.32
140.20
137.22
135.52
155.30
135.55
135.17
135.13
137.38
135.15
138.05
135.30
137.37
136.12
136.30

AVAILABILITY
(%)
61.34
51.08
51.64
61.40
51.57
73.03
67.08
84.47
76.19
73.11
57.04
61.97
59.30
66.91
69.76

Vol-1, Issue-14, 24th October 2013

2321-7871
Down Time Analysis In Process Industry-A Case Study

TABLE 5.1 : RELIABILITY OF COMPONENTS


S. NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

NAME OF THE
COMPONENT
CANE CARRIER
MILLS
RAKE / INTER CARRIER
BAGASSE CARRIER
NEW BOILER
ELECTRICALS
OLD BOILER
LOW PRESSURE BOILER
POWER TURBINE
OTHER UNITS IN BOILER
RAW JUICE PUMP
EVAPORATOR
PAN
SYRP STORAGE TANK
OTHER UNITS IN BOILIBG
HOUSE

Page No-9

1410
1300
1000
1350
410
1500
1200
1600
1500
610
1100
1300
1048
1200
1400

1.8
1.4
1.97
2.5
0.60
2.27
1.6
2.59
2.61
1.00
2.00
1.2
1.8
1.8
2.25

AVAILABILITY
(%)
44.6
41.07
45.4
49.16
24.68
48.24
43.03
50.32
55.12
36.79
45.69
38.59
44.52
44.45
47.69

2321-7871

th

Vol-1, Issue-14, 24 October 2013

Down Time Analysis In Process Industry-A Case Study


TABLE 6.1 : FMEA ANALYSIS

Boiler

Traveling

New Boiler

Pressure
drop

Low pressure
Boiler

Pressure
drop

Evaporator

SO2 Gas
line

PAN

PAN
waiting

Bagasse carrier

Drive chain

Mills

Conveyor
belt cut

Other unit of
boiler
Rake / Inter
Carrier

Pipe line
Rake
elevator

Raw Juice pump

Pipe line

Syrup storage tank

Gas line

Other units in
Boiling house

Nozzle

Power turbines

Turbine

Electrical

Alternator

Cane carrier

Cane
carrier
plate

Combustion of
coal is affected
Low
Production of
steam
Low
Production of
steam
Leakage gas
Production of
sugar is
affected
Discharging of
Bagasse is
affected
Discharging of
Bagasse is
affected
Leakage of
juice
Drive is
affected
Leakage of
juice
Leakage of
juice
Leakage gas
Power
generation is
reduced
Affect the
bagasse
movement
Drive is
affected

Low quality of coal

180

Water Scarcity

100

Water Scarcity

96

Corrosion effect

56

Limited level of
syrup in tank

40

Jamming of bagasse

24

Worn out of belt

18

Corrosion effect

14

Improper tightening

12

Corrosion effect

12

10

Corrosion effect

Corrosion effect

12

Pressure reduction

13

Motor tripped

14

Corrosion on the
plate

15

Page No-10

11

Das könnte Ihnen auch gefallen