Escolar Documentos
Profissional Documentos
Cultura Documentos
FAX: 1-281-852-3749
http://www.barringer1.com
Crow/AMSAA Plots
Most data fit a power law distribution which
gives a straight line on log-log paper
We had Duane plots, then Duane/AMSAA
plots, and now Crow/AMSAA plots
The methodology handles mixed failure
modes and forecast future failures
Results are simple and easy to use log-log
plots of failures vs timeeasy to explain,
easy to make, and easy to forecast failures
Barringer & Associates, Inc. 2003
Figure 1
Y=100X^(-0.3219)
Y=105.3268X^0.7739
Units Produced
Cum Aveg Cost (hrs)
Units Produced
T. P. Wrights
Idea For
Learning
Curves
Which Later
Drives
James Duanes
Idea For
Reliability
Growth Plots
Y=105.3268X^(-0.2260)
Units Produced
Barringer & Associates, Inc. 2003
Table 1
Maintenance Interventions
Month
1995
1996
1997
January
35
12
8
February
32
13
3
March
28
12
15
April
30
11
5
May
41
11
10
June
30
11
9
July
16
15
8
August
9
7
18
September
21
8
7
October
14
8
9
November
12
10
7
December
11
10
8
Total =
288
130
96
TPM Began August 1995
Actual
Failure
Data
From A
Chemical
Plant In
Brazil
Figure 2
1000
Crow/rgr
Repair Reduction
After TPM
100
Slight Improvement
Before TPM
Updated
Yr 2003
M01 D07
HPB
10
1
10
100
Big Improvement
After TPM
Before TPM
N = 34.65*t0.947
After TPM
N = 77.49*t0.529
Forecast at t = 40 months
N = 34.65*400.947= 1140
Forecast at t = 40 months
N = 77.49*400.529= 545
=1140 - 545 = 595 failures
avoided
7
1995
35
32
28
30
41
30
16
18
21
14
12
11
288
1996
1997
'98 Fcst
12
8
8
13
3
7
12
15
7
11
5
7
11
10
7
11
9
7
15
8
7
9
7
7
8
7
7
8
9
7
10
7
7
10
8
7
130
96
85
TPM Began August 1995
Barringer & Associates, Inc. 2003
'99 Fcst
7
7
6
6
6
6
6
6
6
6
6
6
74
8
Failures Represented By
Monthly Maintenance Costs
Table 3
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Figure 3
$10,000,000
$1,000,000
100
1000
10
Figure 4
Maintenance Cost (Total Cumulative US$)
$ 20,000,000
Crow/rgr
$ 18,000,000
11 Month
Cum Savings
$2,222,000
Dec 2002
C$ = 110839t0.696
$ 16,000,000
Announced
TPM Program
$ 14,000,000
Started TPM
Efforts Jan 2002
$ 12,000,000
Lambda Beta r^2
3361 1.192 0.998
110839 0.696 0.995
C$ = 3361t1.192
$ 10,000,000
800
900
1000
1200
1400
1600
1800
2000
11
Table 4
Safety Record--Major Chemical Plant Incidents
Cum
Days
Cum
Incidents
Cum
Days
Cum
Incidents
Cum
Days
Cum
Incidents
Cum
Days
Cum
Incidents
1
8
23
47
53
58
65
67
72
78
94
105
106
108
124
149
226
228
248
285
288
289
310
312
1
2
3
4
5
6
7
8
9
10
12
13
14
15
16
17
18
19
20
21
22
23
24
25
367
368
429
526
553
585
598
599
600
632
635
660
677
690
719
759
773
830
878
1009
1018
1031
1040
1044
26
27
28
29
30
31
32
33
34
36
37
39
40
41
42
44
45
46
47
48
49
50
51
52
1046
1096
1184
1195
1291
1345
1397
1565
1591
1598
1624
1626
1634
1655
1670
1692
1711
1753
1759
1990
2186
2430
2472
2509
53
54
55
56
57
58
59
60
61
62
63
74
75
76
77
78
79
81
82
83
84
85
86
87
2622
2742
2754
2825
2846
2851
2888
2922
2969
2984
3099
3106
88
89
90
92
93
94
95
96
97
99
100
101
8.5 years
Safety Incidents
Are Failures!
Long term
average MTBF
is 3106/101 =
30.7 days/incident
12
Figure 5
Safety Incidents
Jan '94-Aug '02
300
Forecast
Crow/IEC1164
Safety Incidents
(Total Cumulative)
100
2-Aug-02
19-Sep-02
7-Nov-02
26-Dec-02
13-Feb-03
4-Apr-03
24-May-03
13-Jul-03
2-Sep-03
22-Oct-03
12-Dec-03
2-Feb-04
24-Mar-04
15-May-04
Unfavorable Trends
10
Lambda
0.6177
1
10
Actual
101
102
103
104
105
106
107
108
109
110
111
112
113
114
Forecast
Beta
0.634
1000
100
Time (Total Cumulative Days)
10000
13
Crow/AMSAA Data
Spill Date
Days
Between
Spill
Spill
Events
Cum.
Days
Cum.
Spills
11/18/1995
1/31/1996
5/8/1996
5/22/1996
7/29/1996
8/23/1996
8/25/1996
6/20/1997
2/22/1998
2/10/1999
35
74
98
14
68
25
2
299
247
353
1
1
2
1
1
1
1
1
1
1
35
109
207
221
289
314
316
615
862
1215
1
2
4
5
6
7
8
9
10
11
Forecasts
Failures
New
Predicted
Method
By Old
Savings
Method
8 Failures In
316 Days
This Is The
Datum
18
27
41
9
17
30
14
Figure 6
100
Spills
Crow/IEC1164
Old Way
Spill Reduction
10
Improvement
10
100
1000
Time
(Total
Cumulative
Days)
Barringer & Associates, Inc. 2003
10000
15
Table 6
Raw Data
Spill Date
11/18/1995
1/31/1996
5/8/1996
5/22/1996
7/29/1996
8/23/1996
8/25/1996
6/20/1997
2/22/1998
2/10/1999
8/16/1999
11/7/1999
2/12/2000
4/29/2000
11/16/2000
12/25/2000
3/25/2001
8/1/2001
10/28/2001
7/10/2002
7/25/2002
9/6/2002
2/18/2003
Crow/AMSAA Data
Days
Between
Spill
Spill
Events
Cum.
Days
Cum.
Spills
35
74
98
14
68
25
2
299
247
353
187
83
97
77
201
39
90
129
88
255
15
43
165
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
35
109
207
221
289
314
316
615
862
1215
1402
1485
1582
1659
1860
1899
1989
2118
2206
2461
2476
2519
2684
1
2
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Forecasted Failures
Failures
Missed
New
Predicted
Opportunities
Method
By Old
From
Savings
Method
Relapse
18
27
41
9
17
30
Bad News!
Failures
Coming
Faster!
Good
News!
1
2
2
3
4
5
5
6
7
9
9
9
11
16
Figure 7
Spills
100
Crow/IEC1164
Relapse
Re
tre
nc
h
ct
io
n
Re
du
ill
Sp
Old Way
Weve
Lost
The
Way To
Success!
10
Improvement
Lambda
0.0074011
1.989
0.0091392
1
10
Beta
1.214
0.241
0.997
Fit-P%
<10!
>=10
>=10
100
1000
Time (Total Cumulative Days)
10000
17
Table 7
All Outages
Date
Event
Outage
Event Description
2/1/1999
2/20/1999
2/24/1999
5/22/1999
7/9/1999
8/9/1999
9/13/1999
10/13/1999
11/3/1999
11/6/1999
11/10/1999
1/3/2000
6/12/2000
6/21/2000
9/11/2000
11/7/2000
12/2/2000
12/12/2000
4/11/2001
4/12/2001
4/19/2001
6/7/2001
8/22/2001
9/13/2001
9/16/2001
10/6/2001
10/12/2001
10/31/2001
12/1/2001
1/1/2002
4/15/2002
4/18/2002
9/27/2002
12/6/2002
1/3/2003
Planned
Planned
Forced
Forced
Planned
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Planned
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Tie In
Tie In
Gas Line Outage
Animal Contact
Interconnect Energized
Switching Error
Tie Wrap Failure
Lightning Strike
Static Wire Short
Switch Failed
Not Logged
Cable Bond Fault
Underground Cable Fault
Bird Contact
Lightning Strike
Animal Contact
Animal Contact
High Winds
Not Logged
Not Logged
Tie In
Not Logged
Pole Damage
Interconnect Opened
Supplemental Power Out
Power Dip
Control Tripped
Power Dip
Power Dip
Steam Outage
Switching Error
Load Shedding Error
Water In Switch Gear
Generator Air Intake Frozen
UPS Failure
Days
Between
Event
0
19
4
87
48
31
35
30
21
3
4
54
161
9
82
57
25
10
120
1
7
49
76
22
3
20
6
19
31
31
104
3
162
70
28
Cum.
Days
Cum
Failures
0
19
23
110
158
189
224
254
275
278
282
336
497
506
588
645
670
680
800
801
808
857
933
955
958
978
984
1003
1034
1065
1169
1172
1334
1404
1432
0
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
Forced Outages
Cum.
Days
Cum
Failures
23
110
1
2
189
224
254
275
278
282
336
497
506
588
645
670
680
800
801
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
857
933
955
958
978
984
1003
1034
1065
1169
1172
1334
1404
1432
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Co-gen
Failure Log:
31 Forced
Outages In
1432 Days
= 46.5 days
per failure
18
Figure 8
100
Crow/IEC1164
Forced Outages
10
1000
100
Time (Total Cumulative Days)
10
Event
Outage
2/1/1999
2/20/1999
2/24/1999
5/22/1999
7/9/1999
8/9/1999
9/13/1999
10/13/1999
11/3/1999
11/6/1999
9/27/2002
12/6/2002
1/3/2003
2/23/2003
4/11/2003
5/27/2003
7/13/2003
8/28/2003
10/14/2003
11/29/2003
1/15/2004
3/2/2004
4/17/2004
6/3/2004
Planned
Planned
Forced
Forced
Planned
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
Forced
All Outages
Event Description
Tie In
Tie In
Gas Line Outage
Animal Contact
Interconnect Energized
Switching Error
Tie Wrap Failure
Lightning Strike
Static Wire Short
Switch Failed
Water In Switch Gear
Generator Air Intake Frozen
UPS Failure
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Predicted
Days
Between
Event
0
19
4
87
48
31
35
30
21
3
162
70
28
19
Forced Outages
Cum.
Days
Cum
Failures
Cum.
Days
Cum
Failures
0
19
23
110
158
189
224
254
275
278
1334
1404
1432
0
1
2
3
4
5
6
7
8
9
32
33
34
23
110
1
2
189
224
254
275
278
1334
1404
1432
1483
1530
1577
1623
1670
1716
1763
1809
1856
1903
1949
3
4
5
6
7
29
30
31
32
33
34
35
36
37
38
39
40
41
42
More Failures
Predicted Into
The Future From
Lack Of An
Improvement
Program ( 1)!
5000
No
Change
20
10
Summary
Five actual datasets demonstrate straight
lines of cum failures vs cum time
Crow/AMSAA technology allows failure
forecast of the mixed failure modes
The forecast objective is to start decisive
corrective action and prevent predicted
failures
Preventing failures requires proactive effort
Preventing future failures saves $!
Barringer & Associates, Inc. 2003
21
11