Project : Compressed Air Energy consumption reduction by 12%.
DM A I C
Presented to
: CII
Presented by : SKF Team
Category
: DMAIC - Utilities
SKF Knowledge Engineering Company
\
• 100 years of technology progress and innovations.
SKF Group
•
40,000 employees
•
104 factories
•
6.5 Bn US$ turnover
•
83 production facilities
•
1 out of 5 bearings in the world.
October 30, 2007 © SKF Group Slide 1
SKF India
• 2,000 employees
• 4 production plants (Pune,
• Bangalore, Haridwar & Ahmedabad)
• Rs. 1,600 cr. turnover.
• 1 out of 4 bearings in India
SKF Bangalore (SDGBB)
• 400+ employees
• 6 Manufacturing channels
• Rs. 315 cr. turnover.
SKF & Sustainability
Environmental Care @ SKF in India
•Environmental Care Score Card for each location
•CO2 Emission Reduction
•
•
LightTheFuture Project - Reduction in Energy Consumption (each factory location)
Number of projects focused on reduction of Energy Consumption (CO2 emission reduction) happened in
•
•
•
2007 – 1 project closed
2008 – 8 Projects closed
2009 – 21 Projects closed
All Factories Together (in Tons of Co2)
2007
2008
2009
w.r.t 2008
w.r.t 2007
29522.42
27172.40
24703.19
9.09%
16.32%
15359
13583.74
14158.93
-4.23%
7.81%
Bangalore 3S
2539.01
2377.66
2160.43
9.14%
14.91%
SKF India
47420.43
43133.81
41022.55
4.89%
13.49%
Pune
Bangalore DGBB
October 30, 2007 © SKF Group Slide 2
Bangalore Plant Energy Consumption Reduction by 12%Umbrella
Sponsor :
G.W.Chidambar Rao
Review committee
Harsha Kadam
Amar Bhat
Anil K Gupta
G.W.Chidambar Rao
Umbrella Leader
K.N.Manjunatha
Umbrella Process owner
C. Ravindranath
Compressed air BB: N
Jnanesh
Chiller
GB: C
Ravindranath
Chl Group C
GB:
Balashanmuga m Project Organization
Chl Group A
GB: Rakesha
MST
GB: M N
Satishkumar
Metrics
Base line
1580815
1391117
National Press
GB: K
Chandrashekara
Ch Group B
GB: Basavaraj
Goal
Bangalore Plant
Energy
consumption per month, kWh
Deployment Champion :
K.N.Manjunatha
October 30, 2007 © SKF Group Slide 3
Forecasted hard savings 9
MINR/Annum
CO2 Emission Reduction
1620 Tons /Annum
Factory
Lighting
GB: Bhanu
Why this Project ???
D M A I C
Pareto Chart Energy Consumption
16000000
100
12000000
80
Total
10000000
8000000
60
23% of Total Energy Consumption is by Compressed air
6000000
40
4000000
20
2000000
0
Ar ea
0
l l s
1
2
2
4
3
4
5
6 s ) c )
1
g r U ta ta
1A
1B er e e es & 3
Mc
CF
H
e he t in
To
To hil l
EL
EL ac ac
EL
EL
EL
EL
Pr
es
&
es
5 A na c i gh
Ot
te
N
N
N
N
N
N or ce
C
rn rn al
, 2 rvi c ri M vi c l4 & r L ss (1
D
AN
AN
Fu
Fu
AN
AN
AN
AN
e
Sp
er
Fu
na on e e i r S
S
(R n es te pr
CH
CH
CH
CH
CH
CH N a t
4
Fu an hi m Lin k W
Ch
oc
Co
ST
Bl
M
Total
Per cent
Cum %
3296435 1517342
1715424
750253
750033
594816
553540
550898
514471
497528
493546
438903
424168
351834
240760
238086
235696
210877
205794
203304
646434
23
23
12
35
11
45
5
50
5
56
4
60
4
64
4
67
4
71
3
74
3
78
3
81
3
84
2
86
2
88
2
90
2
91
1
93
1
94
1
96
4
100
Percent
14000000
Is this impacting Strategic Goals
• Operating Margin
• Growth
• Zero Defects
• Zero Broken Promise
• Zero Accidents
• Zero Loss making Business
• Beyond Zero
• Full Factory Per Day Energy Consumption 50000 units
• 13,500 Tons of CO2 Emission annually
• 23% of Total Energy consumption is due to Compressed air
• Approx- 1.6 Million Rs spent on Compressed air every Month
October 30, 2007 © SKF Group Slide 4
Project Charter : VOC, Scope & Metrics
Element
Problem
Statement
Process Impacted
Project Scope
Description
D M A I C
Charter
State project’s problem and background. ("What, where when, who, why, how.")
* Energy cost of DGBB Bangalore is on upward trend
* Per day consumption is about 50000 units
* Compressed air is one of the most expensive sources of energy in a plant. It is the 1st highest of existing energy usage (approximate 23
% of the Total energy)
Describe the Process where improvement opportunity exists, i.e. the process where defect occurs.
Team activities will target on minimization of
- Compressed Air Generation,
- Compressed Air Distribution
- Compressed Air Consumption
Define the part of the process (start and end point) or product(s) that will and will not be investigated. Describe how the project conditions are when it is finished, e.g. proposal, investigation, full or partly implementation, etc.
Included:
Metrics
[Choose based on Project
Y(s)]
Compressed air energy consumption per day
October 30, 2007 © SKF Group Slide 5
Compressed Air generation and control system
Excluded:
Source of supply
Investment approval for major change
Baseline;
("Y’s")
Project
Goal by
End Date
12737 kwh
/ day
11208 kwh / day Cross Functional Team
Balashanmugam, Karthikeyan,
C.Subramanian, Bhanuprakash, Ronal,
Rajendran, Krishnappa
Base line / Initial Capability
Summary for Comp - Energy
Daily Power Consumption
A nderson-D arling N ormality Test
Mean
StD ev
V ariance
Skew ness
Kurtosis
N
12400
12600
12800
13000
0.55
0.144
12737
216
46591
0.314940
0.736199
46
Minimum
1st Q uartile
Median
3rd Q uartile
Maximum
13200
1
12300
LC L=12184
15-Dec-09 20-Dec-09 26-D ec-09 31-Dec-09
12662
12852
95% C onfidence Interv al for StDev
9 5 % C onfidence Inter vals
179
272
5-Jan-10 10-Jan-10
Date
15-Jan-10 20-Jan-10
25-Jan-10
31-Jan-10
U C L=680.4
600
M ov ing Range
12802
_
X=12737
12600
95% C onfidence Interv al for M ean
12673
U C L=13291
12900
12208
12563
12722
12874
13300
95% C onfidence I nterv al for Median
Baseline mean
12737
13200
I ndiv idual Value
A -Squared
P -V alue
12200
D M A I C
450
300
__
MR=208.2
150
Mean
0
Median
12650
15-Dec-09 20-Dec-09 26-D ec-09 31-Dec-09
12700
12750
12800
12850
P Value 0.144 Data is normal
Baseline :
•Mean – 12737 units (kwh)
•Variation – 216 units (kwh)
October 30, 2007 © SKF Group Slide 6
LC L=0
5-Jan-10 10-Jan-10
Date
15-Jan-10 20-Jan-10
25-Jan-10
31-Jan-10
Sipoc , Process Map & C&E Matrix
D M A I C
Process Map
1
SIPOC
S
I
Suppliers
Providers of the required resources P
Inputs
Resources required by the process Numerical requirements on inputs O
Process
Top level description of the activity P-MAP
2
Process Name: COMPRESSED AIR - ENERGY CONSUMPTION
Process Owner: RAVINDRANATH
Outputs
Compressed Air Energy consumption reduction by 12%.
Belt:
Black
BU:
DGBB
Date: 30/12/2009 Version:
C = Controllable, C! = Controllable but critical , S = Standerd Operating Procedure , N = Noise
Class
Input
Step
KPIV
C
1
Page
Output
1
Notes
Customers
Deliverables from the process Numerical requirements on outputs Requirements
Project Title
1. Generation
Stakeholders who place the requirements on outputs
Power supply with 415 V from KEB
Power supply with 415 V from DG
Requirements
C
N
Compressed Air Generation
Power Consumption (Kwh) at Full load
Power Consumption (Kwh) at No load
ENERGY SUPPLYTO
COMPRESSOR
INLET AIR FILTER
FILER QULAITY
COMPRESSED AIR
GENERATION
CLEAN AIR
WATER COOL
COMPRESSOR
COOLING
CLEANING FREQUENCY
MOISTURE FREE AIR
REFRIGERANT
DRYIER
AIR FILERATION
AIR DRYIER
INTER COLLER
AFTER COOLING
DRY MAINTENANC
IER
E
AMIANT TEMP AIR
CIVIL DEPT
DISTRIBUTION PIPE
REQUIRED SIZE AND
LINE
LENGTH
LENGTH OF NETWORK MINIMUM REQUIRED
RESERVIOUR
AIR STORAGE
S
S
C
C
C
C!
C!
C!
C
C!
C!
C
Inlet Air temperature (natural air)
C!
Air filter change frequency (1000 hours)
C!
Electrical supply frequency variation
C!
Electrical supply fluctuations
C!
Specific energy of Compressor
C!
Motor type (Star Delta)
C
Motor Rating (160 Kw)
S
Motor Bearings (greasing once in 3 months) C!
Motor Bearing change frequency
S
Balancing of motor
S
No of times motor rewind
C!
Internal Maintenance
S
Maintenance-PM's activities
S
Condenser condition
C!
Scaling of condenser tubes (once in year /2 year)
N
Ageing effect
C
Present plant demand pattern
C
Compressore running at high pressure
C
Calibration of controller
C!
Air cooler
C!
air cooler fan terace
C!
air cooler fan motor
C!
water cooler pump
C!
245 X’s identified - P Map
Controllable
= 187
(Critical -127)
Noise
= 11
SOP
= 47
2. Distribution
COMPRESSED AIR
DISTRIBUTION
REQUIRED STORAGE
CAPACITY
MINIMUM PRESSURE
DROP
SAFE AIR
DISTRIBUTION
AIR PRESSURE @
DELIVERY
LEAK FREE
MACHINES
USAGE WITH OUT
WASTE
TO THE POINTS OF
USE
C
ONSTANT AIR
PRESSURE
AIR CONSERVIER
VARIOUS SUPPLIERS
LUBRICATING
COMPRESSED AIR
CONSUMPTION
PIPE LINE
AS FLEXIBLE AS
POSSIBLE
CONNECTORS
REGULATORS
NO OF CONNECTIONS
ON ONE POINT
FLEXLINK
CONNECTIONS
PUSHERS
@ REQUIRED TEMP
MACHINE
FUNC
TIONING
ADEQUATE PRESSURE
ACTUATORS
COOLING MEDIA
3. Consumption
3
internal exhaust fan
Six Sigma Cause and Effect Matrix
Process Step
4
8
10
13
7
9
12
11
2
5
6
1
14
3
15
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Com
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Air
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
Distribution
10
8
Air Consumption
Process Input
length of flexible pipe to end point
No of Hoses
Pipe Condition
Pipe Diameter
No of Mufflers
Pressure required
U bends
L Bends
Fluctuating demand
Connections
Fittings
Rust in GI pipeline
Threaded Joint
Flange Joints
Welded Joint
October 30, 2007 © SKF Group Slide 7
3
4
5
Total
2
Air leakage
Key Process Outputs
Screw Compressor
5
nos Atlas Copco of 838 cfm each
C
Pareto Chart of Process Input Generation
1
Rating of Importance to Customer
Moist Air at certain Pressure psig
& temp in degree C
Noise Level
Air quality (lubrication oil mix up)
Power Consumption (Kwh) at Full load
Power Consumption(Kwh) at No load
Oil type (Servo System 68)
Oil change frequency (1000 hrs)
Oil quality (120 lts)
Oil temp (80 - 120 centigrade)
Electronic controller
Oil seperators
Oil seperators change frequency (3000 hrs)
Delta P diff
Cooling water temp (26 to 29 degree)
Switchgear spares availability
Condition of screw element
PROTECT EQUIPMENT PNEUMATIC
FROM AIR BORNE
EQUIPMENTS AND
PARTICLES
COMPRESSED AIR
REQUIRED PLAC
ES
QUALITY AIR @
REQUIRED PRESSURE
Total
9
9
9
9
9
3
9
9
3
3
3
3
1
1
1
9
9
9
9
3
9
1
1
3
3
3
1
1
1
1
162
162
162
162
114
102
98
98
54
54
54
38
18
18
18
3500
3000
2500
2000
100
80
60
1500
1000
500
0
40
20
0
Process Input ff r s ) r e r r p t ) n n n ) d ) ) ) e ) n ) t r e s ) ) y ) r d i le o r r s l e r so t o m e c lta r fa io a r in rs 68 e e c r s fa lt s en l e c ie w e l it ir e
P o o at h r o ssu es o u eff e tte r dit y e w h m gr era ou st 0 m r o an i vit K r a d bi a l a th ta ir c pe r3000o nt pr em pran m r p g ta r D paooleo n / 2 r r e 000ste d ean t 0 hhau (12 eleo ntntena ct( 160 t igv ailatu r O l le n
De A il sey ( o f cigh C o r f o o ge i (S an dr c er c ea r ot oy ( 1 Sy 29 r f 100 ex l ity r e w ic c ai M 's g c ens a (na
O nc n t h o f o le r c A pe em A i n s n y s m nc rv o6 to o le y ( n al qu a sc on a l M - P atin 20 a re re ue atiog a gy r co ate r ty t d n de e i m e qu e Se (2 r coe nc ter il ofe ctr e r n nce r R - 1 sp atu q r c ti
( p i r i w to lan n O n l t a o 0 r r frea lib n in n e a
C o( on o f e fre pe te m a e qu i it io E I n te nM otp (8ge ampe o p
M nt s o g ty r fr g e C un c e nd m c h te ain se b e N a n O il at e g e a n re re cifi
M il te w it A ir e Co h tu c h o w h an
Pr
O S le t s c e ss Sp er Oi l g r r in e r c ns to
In
ol t ra mp de
C o ir fil p e Co on se
A
fc l o
Oi
ng a li
Sc
74 Input's taken into PFMEA
Total
Percent
C um %
Percent
FILTER SUPPLIER
80% of Inputs taken to PFMEA
180 168 162 162 116 114 108 1089896 72 60 5656 5656 5454 5454 54 48 3836158
168 168 162 116 114 114 108
6 5 5 5 5 5 5 4 4 4 4 4 3 3 33 3 2 2 2 2 2 2 2 2 2 2 2 2 1 1 5
6 11 16 2227 3237 41 44 4851 5558 62656871 74 76 7779 8183 8486 8889 91 93 9495100
PFMEA – snap shot
4
Process Step
Compressed Air
Generation
Input
Pressure Setting
Potential Failure Mode
Loading Setting
D M A I C
Potential Effect(s) of
Failure
S e v e r i t y High energy Consumption
5
Energy
More energy Consumption
Low Compressor Efficency
C l s a s
Potential
Cause(s)/Mechanisms of
Failure
Load / Unload Setting Manual
Error
Filter Choke
Delta P High
5
Compressed Air
Distribution
Pipe Length
Length More
High Consumption
Delta p difference due to intake air quality
Pressure Required for application High Pressure
6
105
6
252
6
216
6
108
6
180
6
180
No Control
6
Pipe Damages due to wear &
Tear
3
5
High Pressure de to Excessive design 6
High Pressure due to non availability of service
6
High Consumption
126
7
7
5
Compressed air
Consumption
120
No control
More no of Hoses due to Excessive Design
High Consumption
R
P
N
6
3
6
Pipe Damages
c t n i o
No control
Length More - Excessive Design
High Consumption
D e t e No control
4
6
Pipe Condition
More no of Hoses
Current Process Controls
Detection
3
6
No of Hoses
r r c e e n Wear & Tear
7
Filter
O c c u No Control
No Control
No control
No control
X’s taken further:
Just do it:
X1-Pipe length from Valve
X2-Pipe Diameter from Valve
X3- Air Flow with different Pressure
X4- Compressor Loading & Unloading
Setting
X5- No of pipe Connections from the same pipe
X6- Delta P Difference
X7- Pipe Condition
X8 – Compressor Efficiency
X9- Pressure Drop during Distribution
X10-Pipe Diameter to the consumption point October 30, 2007 © SKF Group Slide 8
Multi-vari Planning Sheet
Multi-Vari Organization Form
Project Y
Energy Consumption for compressed air
Hypothesis
Ho- Compressed air Consumption is the same with different pipe lengths from the Valve to Application
Y
Possible Xs
Air
Pipe length
Consumption in
CFM
Ho- Compressed air Consumption is Air
Pipe Diameter the same with different pipe Diameter Consumption in from the Valve to Application
CFM
Ho- Compressed air flow is same in
Different pressure (5 bar, 3bar, 2bar,
1.5 bar
Ho-Compressor Loading and
Unloading Setting is same on all the days below columns are applicable to continuous data for discrete data
In
Levels to control Centering
Y: Discrete or the Y
Spread (Std and Normal ?
(Mean or
Continuous?
(1, 2, 3 or
Dev)
Stable
Median)
more)
?
Mean Continuous
Yes
Yes
12737
Std Dev- 216
In
Levels to
Multicontrol
Centering
X: Discrete or the X
Spread (Std
Vari
and
Normal ?
(Mean or
P value
Continuous?
(1, 2, 3 or
Dev)
Analysis
Stable
Median) more) Tool
?
Discrete
2
No
Median
mannP value-0.000
Whitney
Discrete
Air
Air Pressure
Continuous
Consumption in
CFM
Unloading
Unloading setting Discrete
Pressure
Ho - Compressed air flow is Same with Air
No of Pipes one pipe outlet & Two pipe out put
Consumption in
CFM
D M A I C
Discrete
2
Compressed air consumption increases by
Increase in Length of Pipe from Valve
Increase in pipe daimeter increases the flow of air
Validated with no of cycles run to change
0.1m^3 Volume of air
Median
mannWhitney
No
Median
Increase in air pressure increase air flow
Validated with M^3 of air flow per minute
2
No
Median
Moods
P value-0.000 median Test
1 Sample P value-0.000
Wilcoxon
Unloading pressure isnot remainig the same on all the days
2
No
Median
mannWhitney
No difference between
1pipe or 2 pipe from the same input
Unloading was set @
78PSI and data colleced for 2 weeks daily one point
Validated with M^3 of air flow per minute with single opening &
Two Opening
X’s Taken for Validation :
X1 – Pipe Length from Value to Application
X2 – Pipe Diameter from Valve to Application
X3 – Compressed Air Pressure
X4 – Unloading Pressure of Compressor
X5 – No of connections from single line
October 30, 2007 © SKF Group Slide 9
Notes
No
More than 2 Not stable P value-0.000
Conclusion
P Value- 0.264
Validated with M^3 of air flow per minute
Validation of X1 – Pipe length from Valve
D M A I C
Boxplot of small (495),(long 2360)- pipe length- air consumption
Y- No of Stokes / Units CFM (Continuous Data)
X- Smaller & Longer Pipe (Attribute)
21
20
Data is non - Normal
Mann-Whitney test Done
20
Data
19
Ho- There is no evidence of difference in population median of Compressed air Consumption with different pipe lengths from the Valve to Application
18
17
17
16
495
2360
Ha-There is evidence of difference in population median of Compressed air Consumption with different pipe lengths from the Valve to Application
P Value-0.000
Mann-Whitney Test and CI: 495, 2360
N Median
495 32 20.000
2360 32 17.000
Point estimate for ETA1-ETA2 is 3.000
95.1 Percent CI for ETA1-ETA2 is (3.000,3.000)
W = 1552.0
Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at
0.0000
The test is significant at 0.0000 (adjusted for ties)
October 30, 2007 © SKF Group Slide 10
Conclusion:
Since p Value is 0.000 which is less than 0.05 there is sufficient evidence to Rej Ho, Hence smaller the diameter lesser the air consumption
D M A I C
Validation of X2 – Pipe Diameter from Valve
Boxplot of air consumption in pu 6 & pu 4
Y -Air Consumption - Continuous Data
X- Pipe diameter smaller & Bigger -Discrete Data
22
Data is non - Normal
Mann-Whitney test Done
CFM pu 6&4 in 3 bar
20
18
Ho- There is no evidence of difference in population median of Compressed air
Consumption with different pipe Diameter from the Valve to Application
16
14
12
10
4
6
Ha- There is evidence of difference in population median of Compressed air Consumption with different pipe Diameter from the Valve to
Application
Pu Dia 6&4- 3bar
Mann-Whitney Test and CI: 6mm dia pipe -cfm, 4mm dia pipe -cfm
N Median
6mm dia pipe -cfm 30 24.000
4mm dia pipe -cfm 30 19.700
Point estimate for ETA1-ETA2 is 4.300
95.2 Percent CI for ETA1-ETA2 is (4.200,4.400)
W = 1365.0
Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at 0.0000
The test is significant at 0.0000 (adjusted for ties)
October 30, 2007 © SKF Group Slide 11
P Value-0.000
Conclusion:
Since p Value is 0.000 which is less than 0.05 there is sufficient evidence to Rej Ho, Hence shorter the length lesser the air consumption
Validation of X3 – Compressed Air Pressure
Test for Equal Variances for CFM_pu 4
Bartlett's Test
Test Statistic
P-Value
1.5
34.69
0.000
Pressure_pu 4
Levene's Test
Test Statistic
P-Value
2.0
11.02
0.000
D M A I C
Y- Air consumption - Continuous Data
X- Different Air Pressure - Discrete Data (>2 level)
Data is non-normal & data is not stable - Test of Variance Failed
Moods median Test done
Ho- Compressed air flow is same in 5bar,3bar,2bar & 1.5bar
3.0
Ha- Compressed air flow is not same in 5bar,3bar,2bar & 1.5bar
5.0
P Value-0.000
Conclusion:
Compressed air consumption is different @ different pressure,
Hence High Air Pressure increases the air consumption.
0.050
0.075
0.100
0.125
0.150
0.175
95% Bonferroni Confidence Intervals for StDevs
Boxplot of Compressed air pressure vs consumption
20.0
17.5
CFM_pu 4
15.0
12.5
10.0
7.5
5.0
1.5
2.0
3.0
Pressure_pu 4
October 30, 2007 © SKF Group Slide 12
5.0
Mood Median Test: CFM_pu 4 versus Pressure_pu 4
Mood median test for CFM_pu 4
Chi-Square = 120.00 DF = 3 P = 0.000\
Individual 95.0% CIs
Pressure_pu 4 N Median Q3-Q1 -------+---------+---------+--------1.5
30 0 5.3 0.1 *
2.0
30 0 6.9 0.1
*
3.0
0 30 11.7 0.2
*)
5.0
0 30 19.7 0.1
*)
-------+---------+---------+--------8.0
12.0
16.0
Overall median = 9.2
X4 – Compressor Loading & Unloading
Setting
D M A I C
Y- Unloading Pressure - Continuous Data
X- Set Unloading pressure (1Level- targeted Value)
Summary for Unloading Pressure
A nderson-Darling N ormality Test
A -S quared
P -V alue <
M ean
S tD ev
V ariance
S kew ness
Kurtosis
N
78
79
80
81
2.58
0.005
79.818
1.610
2.591
0.17087
-1.42119
33
M inimum
1st Q uartile
M edian
3rd Q uartile
M aximum
82
78.000
78.000
80.000
82.000
82.000
95% C onfidence Interv al for M ean
79.247
80.389
95% C onfidence Interv al for M edian
78.000
80.000
Data is not normal
1 Sample Wilcoxon test conducted
Ho- There is no evidence of difference of
Compressor Loading and Unloading Setting is same on all the days
Ha- There is evidence of difference of Compressor
Loading and Unloading Setting is same on all the days 95% C onfidence Interv al for S tDev
9 5 % Confidence Inter vals
1.294
Mean
2.129
P Value-0.000
Median
78.0
78.5
79.0
79.5
80.0
Conclusion :
Since p Value is 0.000 which is less than 0.05 there is sufficient evidence to Rej Ho. The Unloading setting is not the same on all the days
80.5
Wilcoxon Signed Rank Test: Unloading Pressure
Test of median = 78.00 versus median not = 78.00
N for Wilcoxon
Estimated
N Test Statistic
P Median
Unloading Pressure 33 21
231.0 0.000
October 30, 2007 © SKF Group Slide 13
80.00
X5 – No of pipe Connections from the same pipe D M A I C
Boxplot of 2 pipe, 1 pipe
Y- Air consumption - Continuous Data
X- One pipe outlet & Two pipe outlet- Discrete (2Level)
24.10
24.05
Data is non - Normal
Mann-Whitney test Done
Data
24.00
23.95
Ho- There is no evidence of difference in population median of compressed air flow with one pipe outlet &
Two pipe out put
23.90
23.85
23.80
2 pipe
1 pipe
Ha- There is evidence of difference in population median of compressed air flow with one pipe outlet &
Two pipe out put
Mann-Whitney Test and CI: 2 pipe, 1 pipe
P Value-0.264
N Median
2 pipe 30 24.000
1 pipe 30 24.000
Conclusion :
Since p Value is 0.264 which is more than 0.05 there is insufficient evidence to Rej Ho, Hence there is no difference in air consumption when we are using multiple outlets from single inlet
Point estimate for ETA1-ETA2 is -0.000
95.2 Percent CI for ETA1-ETA2 is (-0.100,-0.000)
W = 839.0
Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at 0.2643
The test is significant at 0.2352 (adjusted for ties)
October 30, 2007 © SKF Group Slide 14
Input Funnel – Filtration of Xs
Define
To find the X’s
D M A I C
245 X's
Process Map
First list
Analyze
To filter the list
Reduce the X'S , C & E Analysis
Failure Modes & Effects Analysis
Improve
October 30, 2007 © SKF Group Slide 15
10 X's
To find the critical X’s
Hypothesis Tests,
Control
74 X's
To control the critical X’s
5 X's
4 X’s
Critical X’s
Measure
Critical 4 Xs for Improvement
D M A I C
• Compressed air Pressure
–
–
Consumption of air will come down when the pressure is lower
Energy Required to Generate Low Compressed air is lower than the High
Compressed air
–
–
5 Bar Pressure require 4 kwh per day for per CFM Generation
3 bar Pressure Require 2.8 kwh per day for per CFM Generation
Pressure set 5 bar
High Air Consumption
Pressure set 3 bar
Low Air Consumption
• Pipe Diameter from Valve to Application
–
–
Lower the Pipe Diameter low the Air Consumption
Lower the Pipe Diameter higher the cycle time (Less Volume of Flow /
Minute)
• Pipe Length from Valve to Application
–
Lesser the length of Pipe Lesser the Wastage of air
• Compressor Unloading Setting at Generation Stage
–
Higher the Unloading Setting Higher the Energy Consumed for
Generation
October 30, 2007 © SKF Group Slide 16
Cylinder pipe length & diameter from Valve
Valve
Improvement Plan
Sl # Xs
1
2
Compressed Air pressure Pipe Diameter
Action Plan
D M A I C
Resposibilty
Soft M/cing area convert from high pressure to low pressure
Krishnappa
compressed air
Inspection machines convert from high pressure to low pressure
Ronald
compressed air
Outter ring honning m/c High pressure to low pressure
Ronald
compressed air
Washing machine pipe diameter to
Ronald
be changed form 12mm to 9 mm
Planed on
Completed on
Status
Wk 12
Wk 12
Completed
Wk 12
Wk 12
Completed
Wk 15
Wk 15
Completed
Wk 23
Wk 23
Remarks
Completed
Critical Xs
Critical Xs
5
Spindle seal pipe diameter to be changed from 6mm to 3 mm
Pipe Length
HIT machine
Unloading pressure to be
Compresser
monitored
Unloading Settign unloading setting to be converted to digitial gauge
Delta P Difference To be monitored daily basis
6
Pipe Condition
To be Monitored on Monthly basis Rajendran
WK7
Wk8
Completed
To-do
7
Air Leakages
To be Monitored on Monthly basis Rajendran
Wk6
Wk6
Completed
To-do
8
Starting of 200 cfm compressor To optimize the load the the compressor Wk 23
Wk 23
Completed
3
4
C Subramanian
Wk 23
Wk 23
Completed
C Subramanian
Wk 20
Wk 20
Completed
Rajendran
Wk 6
Wk 6
Completed
Bhanu prakash
Wk 15
Wk 15
Completed
Rajendran
Wk6
Wk6
Completed
To-do
Critical Xs
Critical Xs
Jnanesh / C
Ravindranth
• All the activities completed on time
• Major activity planned was Starting of smaller compressor due to load reduction on compressor by converting the high pressure line to low pressure line
October 30, 2007 © SKF Group Slide 17
Improvement of the Project
Before Project Status
• 3 compressors of High Pressure
•1 Compressor of Low pressure was run to fulfill the compressed air requirement D M A I C
-2400
CFM to Line
838CFM
838CFM
838CFM
700
CFM to
Line
1100 CFM
High Pressure Compressors
One Compressor stopped completely
After Improvement Activities
• 2 + 2 compressors of High
Pressure
•1 Compressor of Low pressure will be run to fulfill the compressed air requirement
1900CFM to
Line
838CFM
838CFM
200CFM 200CFM
1100
CFM
to Line
1100 CFM
High Pressure Compressors
• One - 838 CFM Compressor energy consumption per day is 3300 units
• Two - 200 CFM Compressor energy consumption per day is 1700 units
• Total Energy Saved per day by this change is 1600 units
• Annual saving of 3 Million Rs/• 432 Tons of CO2 emission Reduced Per Year
October 30, 2007 © SKF Group Slide 18
Low Pressure Compressors
Low Pressure Compressor
Final Capability – Before & After results
D M A I C
I Chart of kwh/day by Phase
Baseline
Measure & Analyse
Improve
Control
14000
1
13000
Individual Value
12000
UCL=11705
_
X=11139
LCL=10574
11000
1
10000
1
9000
8000
1
7000
6000
1
1
22
Metrics
43
64
85
106 127
Observation
148
169
Project
Goal by
End Date
Actual
[Choose based on
Project Y(s)]
Baseline
;
("Y’s")
Compressed air energy consumption per day
12737 kwh / day
11208 kwh / day 11139 kwh / day
October 30, 2007 © SKF Group Slide 19
190
Energy consumed during Holidays
Control Plan & Replication
05- Jun
Date:
Revision:
Product:
Control Plan
Generation, Distribution & Consumption
What's
C
ontrolled?
Process Step
2
Compressed Air
Process:
D M A I C
Input or
Output?
Spec. Limits /
Requirements
Measurement
Method
C ontrol Method
Sample Size
Frequency
Who/What
Measures
Where Recorded
Decision Rule/
C
orrective Action
10700 to 11500
Meter readings
Daily Monitoring
1
Daily
IES Personnel
Input
74 Psi
Pressure Guage
Display
Daily Monitoring
1
Daily
IES Personnel
Daily Monitoring
Checklist
Input
48 Psi
Pressure Guage
Display
Daily Monitoring
1
Daily
IES Personnel
Daily Monitoring
Checklist
Compressor
Unloading Setting
Input
78 PSI
Pressure Guage
Display
Daily Monitoring
1
Daily
IES Personnel
Daily Monitoring
Checklist
Delta p Difference
Input
Less than 0.1 bar
Pressure Guage
Display
Daily Monitoring
1
Daily
IES Personnel
Daily Monitoring
Checklist
Air Leakage
Input
Zero Leakage
Manual Check
Monthly
Monitoring
1
Monthly
Leakage monitoring
Team (NJ, CR & CS)
Monthly Report
If not able to arrest the leakage inform the maintenance Engineer
No of Connections
Input
No free end connection
Manual Check
Weekly audit
1
Every Week
IES
Monthly Report
Remove all unwanted connection
No of Connections
Input
New Connections
On need
No Connection with out approval
Monthly
Maintenance
New Connection
Request
Approval of the Maintenance manager is mandatory
Pipe Diameter &
Valve Position for flexlink Input
PU3 & Close to the Cylinder
On need
No Connection with out approval
1
Monthly
Maintenance
New Connection
Request
Approval of the Maintenance manager is mandatory
Pipe Diameter of
Spindle Seal
Input
PU3
On need
No Connection with out approval
1
Monthly
Maintenance
New Connection
Request
Approval of the Maintenance manager is mandatory
Pipe Diameter of
Washing M/c
Input
PU9
On need
No Connection with out approval
1
Monthly
Maintenance
New Connection
Request
Approval of the Maintenance manager is mandatory
Low Pressure
Connections 3bar
Input
All applications which does not take load more than 5 kgs On need
No Connection with out approval
1
Monthly
Maintenance
New Connection
Request
Approval of the Maintenance manager is mandatory
Energy Consumption
Compressed air
Generation
Compressed air
Distribution
Compressed air
C
onsumption
Remarks
Energy Meter
Readings
High Pressure
Controller unit
Pressure setting
Low Pressure
Controller unit
Pressure setting
Output
If not able to set the Required parameter inform Maintenance
Engineer
12 inputs included in Control Plan
Replication
• The Improvements done thro this project has been taken to the new Lines which are being installed during the year 2010 & 2011
• Replication of this project is being taken in all the similar factories of SKF.
October 30, 2007 © SKF Group Slide 20
October 30, 2007 © SKF Group Slide 21 