chapter 4 - results & analysis
DESCRIPTION
Maths V - StatisticsTRANSCRIPT
CHAPTER 4
RESULTS AND ANALYSIS
4.1 Overview
Analysis is the principal tool for obtaining information from the data.
Statistical theories and methods are used to analyse the information
contained in its data. The data is then tabulated and calculated using
Microsoft Excel. Then, graphs for respective pumps are plotted using
Microsoft Excel. Finally, interpretations and comments are doing
throughout the process of analysis.
4.2 Results and Analysis
The basic data (running hours) for each pump has collected from SOGT
and then interpret it through statistical theory. The mean and standard
deviation for crude oil transfer pump P802, P803 and P804 are calculated
by using Microsoft Excel. After that, the data is tabulated in Table 4.1,
Table 4.2 and Table 4.3. The graph is then plotted in order to further
present the data more clearly.
4.2.1 Data Simulation for Crude Oil Transfer Pump P802, P803 and
P804
The data is collected, then tabulated and calculated by using Microsoft
Excel.
Table 4.1: Running Hours for Crude Oil Transfer Pump P802 from
Year 2011 to Year 2012 (Simulate Using Excel)
Month Running Hours
for P802
(Hours)
Cumulative
Running
Hours
Deviation,
x−X
¿
Jan-11 744 744 304.833 92923.361
Feb-11 332 1076 -107.167 11484.694
Mar-11 744 1820 304.833 92923.361
Apr-11 707 2527 267.833 71734.694
May-11 744 3271 304.833 92923.361
Jun-11 178 3449 -261.167 68208.028
Jul-11 197 3646 -242.167 58644.694
Aug-11 594 4240 154.833 23973.361
Sep-11 0 4240 -439.167 192867.361
Oct-11 504 4744 64.833 4203.361
Nov-11 0 4744 -439.167 192867.361
Dec-11 0 4744 -439.167 192867.361
Jan-12 86 4830 -353.167 124726.694
Feb-12 23 4853 -416.167 173194.694
Mar-12 88 4941 -351.167 123318.028
Apr-12 263 5204 -176.167 31034.694
May-12 744 5948 304.833 92923.361
Jun-12 694 6642 254.833 64940.028
Jul-12 735 7377 295.833 87517.361
Aug-12 704 8081 264.833 70136.694
Sep-12 273 8354 -166.167 27611.361
Oct-12 736 9090 296.833 88110.028
Nov-12 706 9796 266.833 71200.028
Dec-12 744 10540 304.833 92923.361
Sum,∑❑ 10540 0.000 2143257.33
3
Count, n 24
Average
(mean, X )=1054024
=439.167
Variance, s2 ¿∑ ¿¿¿
Standard
Deviation, s
¿√¿¿¿
Table 4.2: Running Hours for Crude Oil Transfer Pump P803 from
Year 2011 to Year 2012 (Simulate Using Excel)
Month Running Hours
for P803
(Hours)
Cumulative
Running
Hours
Deviation,
x−X
¿
Jan-11 744 744 283.667 80466.778
Feb-11 176 920 -284.333 80845.444
Mar-11 0 920 -460.333 211906.778
Apr-11 0 920 -460.333 211906.778
May-11 0 920 -460.333 211906.778
Jun-11 475 1395 14.667 215.111
Jul-11 744 2139 283.667 80466.778
Aug-11 734 2873 273.667 74893.444
Sep-11 692 3565 231.667 53669.444
Oct-11 720 4285 259.667 67426.778
Nov-11 720 5005 259.667 67426.778
Dec-11 744 5749 283.667 80466.778
Jan-12 712 6461 251.667 63336.111
Feb-12 627 7088 166.667 27777.778
Mar-12 728 7816 267.667 71645.444
Apr-12 458 8274 -2.333 5.444
May-12 0 8274 -460.333 211906.778
Jun-12 0 8274 -460.333 211906.778
Jul-12 0 8274 -460.333 211906.778
Aug-12 40 8314 -420.333 176680.111
Sep-12 542 8856 81.667 6669.444
Oct-12 742 9598 281.667 79336.111
Nov-12 706 10304 245.667 60352.111
Dec-12 744 11048 283.667 80466.778
Sum,∑❑ 11048 0.000 2423587.33
3
Count, n 24
Average
(mean, X )=11104824
=439.167
Variance, s2 ¿∑ ¿¿¿
Standard
Deviation, s
¿√¿¿¿
Table 4.3: Running Hours for Crude Oil Transfer Pump P804 from
Year 2011 to Year 2012 (Simulate Using Excel)
Month Running Hours
for P804
(Hours)
Cumulative
Running
Hours
Deviation,
x−X
¿
Jan-11 744 744 84.250 7098.0625
Feb-11 670 1414 10.250 105.0625
Mar-11 742 2156 82.250 6765.0625
Apr-11 709 2865 49.250 2425.5625
May-11 744 3609 84.250 7098.0625
Jun-11 576 4185 -83.750 7014.0625
Jul-11 706 4891 46.250 2139.0625
Aug-11 119 5010 -540.750 292410.5625
Sep-11 720 5730 60.250 3630.0625
Oct-11 744 6474 84.250 7098.0625
Nov-11 714 7188 54.250 2943.0625
Dec-11 744 7932 84.250 7098.0625
Jan-12 730 8662 70.250 4935.0625
Feb-12 627 9289 -32.750 1072.5625
Mar-12 667 9956 7.250 52.5625
Apr-12 720 10676 60.250 3630.0625
May-12 744 11420 84.250 7098.0625
Jun-12 720 12140 60.250 3630.0625
Jul-12 647 12787 -12.750 162.5625
Aug-12 743 13530 83.250 6930.5625
Sep-12 434 13964 -225.750 50963.0625
Oct-12 420 14384 -239.750 57480.0625
Nov-12 706 15090 46.250 2139.0625
Dec-12 744 15834 84.250 7098.0625
Sum,∑❑ 15834 0.000 491016.500
Count, n 24
Average
(mean, X )=1583424
=659.750
Variance,
s2¿∑ ¿¿¿
Standard
Deviation,
s
¿√¿¿¿
Table 4.4: Mean Time between Failure (MTBF) and Cumulative
Number of failure for each pump P804 from Year 2011 to Year
2012 (Simulate Using Excel)
Month Cumulative No.
of Failures
Mean Time Between
Failure (MTBF)
Failure Rate / year
P80
2
P80
3
P80
4
P802 P803 P804 P802 P803 P804
Jan-11 0 0 0 - - - - - -
Feb-11 0 1 1 - 1.278 1.964 - 9.522 6.195
Mar-
11
0 1 2 - 1.278 1.497 - 9.522 8.126
Apr-11 0 1 2 - 1.278 1.990 - 9.522 6.115
May-
11
0 1 2 - 1.278 2.506 - 9.522 4.855
Jun-11 0 1 2 - 1.938 2.906 - 6.280 4.186
Jul-11 0 1 2 - 2.971 3.397 - 4.095 3.582
Aug-
11
1 4 2 5.889 0.998 3.479 2.066 12.196 3.497
Sep-
11
2 4 2 2.944 1.238 3.979 4.132 9.829 3.058
Oct-11 3 4 2 2.196 1.488 4.496 5.540 8.177 2.706
Nov-
11
3 4 3 2.196 1.738 3.328 5.540 7.001 3.656
Dec-
11
3 4 3 2.196 1.996 3.672 5.540 6.095 3.313
Jan-12 4 4 3 1.677 2.243 4.010 7.255 5.423 3.034
Feb-12 4 4 3 1.685 2.461 4.300 7.220 4.944 2.829
Mar-
12
4 4 3 1.716 2.714 4.609 7.092 4.483 2.640
Apr-12 4 4 3 1.807 2.873 4.943 6.733 4.235 2.462
May-
12
4 5 3 2.065 2.298 5.287 5.891 5.294 2.301
Jun-12 5 5 3 1.845 2.298 5.620 6.594 5.294 2.165
Jul-12 5 5 4 2.049 2.298 4.440 5.937 5.294 2.740
Aug-
12
5 5 4 2.245 2.309 4.698 5.420 5.268 2.590
Sep-
12
5 5 4 2.321 2.460 4.849 5.243 4.946 2.509
Oct-12 5 5 4 2.525 2.666 4.994 4.818 4.563 2.436
Nov-
12
5 5 4 2.721 2.862 5.240 4.471 4.251 2.322
Dec-
12
5 5 4 2.928 3.069 5.498 4.156 3.965 2.213
4.2.2 Graphs of the Running Hour for Crude Oil Transfer Pump
P802, P803 and P804 from Year 2011 to Year 2012
Figure (): The graph of cumulative running hours against time for pump
P803 from year 2011 to year 2012.
Figure (): The graph of running hour against time for pump P802 from year
2011 to year 2012.
Figure (): The graph of running hour against time for pump P803 from year
2011 to year 2012.
Figure (): The graph of running hour against time for pump P804 from year
2011 to year 2012.
Figure (): The graph number of failure for respective pump from year 2011
to year 2012.
Figure (): The graph number of Mean Time between Failure (MTBF) for
each pump from year 2011 to year 2012.
Figure (): The graph number of failure rate per 2 years for each pump from
year 2011 to year 2012.
4.3 Discussion and Analysis
4.3.1 Interpretation and Comparison for each pump
From Table (), pump P804 has the highest cumulative running hours
compare with pump P802 and P803 from year 2011 to year 2012. Pump
P804 has the cumulative running hour of 15834 hours whereas pump P802
and P803 have the cumulative running hours of 10540 hours and 11048
hours. Besides, pump P804 also achieves the highest average (mean)
running hours of 659.750 hours from year 2011 to year 2012 compare
with pump P802 and P803. Meanwhile, from table (), pump P802 has the
lowest average running hours of 439.167 hours compare with pump P803
and P804.
From the table (), it can be observed that all the crude oil delivery
pumps didn’t work at its optimum running hours throughout two years.
The maximum running hours can be 24 x 2 x 365 = 17520 hours
throughout the two years’ time. The minor reasons due to they can’t work
at the optimum running hours are vibrations or cavitation, gearbox issues,
engines issues, misalignment between the pump and the driver and
leakage happen. However, the problems proposed by engineers from
SOGT are pump P804 more likely act as the duty pump meanwhile pump
P802 and pump P803 act as standby pump. In a clearer meaning is the
standby pump shall continue to operate until the duty pump is placed
back into service.
From the table (), it is showed that pump P804 has the lower
standard deviation compare with pump P802 and P803. Pump P804 has
achieved the standard deviation value of 146.111 hours compare with
pump P802 of 305.262 hours and pump P803 of 324.613 hours. From the
measurement of uncertainty, the lesser the standard deviation, the lesser
the uncertainty is. Thus, the more confidence the pump and higher the
reliability of the pump (Siddharth Kalla, 2009). So, it can be concluded that
pump P804 has the higher reliability than pump P802 and P803.
4.3.2 Number of Failure, MTBF and Fail Rate for each pump
From table (), pump P802 has the highest MTBF which is 5.889 on month
of August 2011. Pump P802 can achieve the highest MTBF without 7
months of failure. However, pump P803 has the lowest MTBF on month of
August 2011 because there are 3 times failures occur within a month
period. From the above interpretation, a relationship can be obtained
where MTBF somehow proportional to number of failures occur within that
period. This prove that the justification for MTBF is somehow correct.
From the table (), the MTBF values are calculated based on the
period for one month. Initial uses of MTBF can provide us with the average
time between failures2 for a given time period, and that this can then be
manipulated to give a failure rate3 for any specified period of time. The
following calculation example can provides more understanding about
MTBF.
For one measurement of MTBF on Month of August 2011 for P802
MTBF of the primary function = 5.889
Probability of a failure for P802 in a month = 1
5.889=0.169=16.9%
Probability of a failure for P802 in a day = 1
5.889×30=5.66×10−3=0.566%
Likelihood of a failure for P802 in an hour =
15.889×30×24
=2.36×10−4=0.024%
There are several ways to improve the MTBF for crude oil delivery
pump, including all their component parts, the consensus is that the best
option is to operate as near as possible to their best efficiency point, BEP.
Besides, there are other recommended methods to help out in prolonging
the MTBF. For example, reduce impeller diameter, install by-passes, have
speed-controlled drivers, control flow rates with automatically operated
valves, put holes through impellers to equalize pressures are other ways
to improve the MTBF.
Instead of using MTBF, there is another way to be suggested by
using MTBR, or known as mean time between repairs (Robert X. Perez,
2014). Robert (2014) define MTBR as the mean number of life units
between repair activities required to bring all parts of the item back to
within their specified limits, during a particular measurements interval
under stated conditions. MTBR is similar to MTBF, but uses repair events
instead of failure events. The following equation is used to determine
MTBR:
MTBR=NR
Where N is the number of machines in the populations
R are the number of repairs in the measurement
period.
Of these three metrics, MTBR is probably the most widely used for
evaluating pump reliability. However, MTBR also encounter with
limitations and caveats. First of all, it includes the mean pump life along
with the mean time for the organization to identify, plan, and repair the
pump, which tends to inflate the value of MTBR (Robert X. Perez, 2014).
The MTBR metric is an amalgamation of repair data for all pumps, running
and idle, that are included in the population tends to greatly increase its
value compared to the true mean time to failure (Robert X. Perez, 2014).
Besides, the repair data supplied for the calculation is subject to
interpretation, therefore it is prone to errors and inconsistency (Robert X.
Perez, 2014).