7-improving unstable processes
TRANSCRIPT
-
8/8/2019 7-Improving Unstable Processes
1/65
7-1
Investigate & FixSpecial Causes
TheQualityImprovementModel
Use SPC toMaintainCurrentProcess
Collect &Interpret
Data
SelectMeasures
DefineProcess
Is
ProcessCapable
?
ImproveProcessCapability
IsProcessStable
?
Investigate &Fix Special
Causes
No
Yes
No
Yes
Investigate & Fix Special Causes
Purpose:Assure stability of the
measures.
Establish permanency ofsolutions to recurring specialcauses.
Improve procedures andtraining.
-
8/8/2019 7-Improving Unstable Processes
2/65
7-2
Investigate & FixSpecial Causes
Unstable Process
Time
Quality Characteristic
Control Chart
UCL
CL
LCL
How dowe improve an unstable process?
-
8/8/2019 7-Improving Unstable Processes
3/65
7-3
Investigate & FixSpecial Causes
Potential Special Causes
Lack of standardization
Lack of consistent process conditions
Uncontrolled process variables
Unknown process variables
Fluctuating business environment
Equipment malfunctions
Unstable measurement process
Changes in process inputs
-
8/8/2019 7-Improving Unstable Processes
4/65
7-4
Investigate & FixSpecial Causes
The Improvement Process
Study the Symptoms
Theorize Causes Establish True
Causes
Propose AlternativeSolutions
Select & ApplySolution
MaintainImprovements
We must take the time to identify the causebefore attempting to solve the problem.
Diagnose Solve
-
8/8/2019 7-Improving Unstable Processes
5/65
7-5
Investigate & FixSpecial Causes
Special Cause Action Plan
(S.C.A.P.)
A planned sequence of activities for investigating andtaking action on a problem which has been detected.
Provides a prioritized listing of potential special causes.
The S.C.A.P. should be updated whenever necessary.
Actions1) Investigation
2) Compensation3) Documentation
4) Correction
5) Prevention
Note: This is the most important part in implementing S.P.C.
-
8/8/2019 7-Improving Unstable Processes
6/65
7-6
Investigate & FixSpecial Causes
Investigative Action
Systematic detective work to determine the rootsource of the special cause. The investigativeactions should be directed toward determining:
What When Why Where
Unless the urgency of special cause dictates otheraction, investigative action should precede any otheractions. Regardless of the urgency, investigative actionshould always be taken at some point.
Note: This is the most important part of the S.C.A.P..
-
8/8/2019 7-Improving Unstable Processes
7/65
7-7
Investigate & FixSpecial CausesCompensating Action
Adjustment made to the process after a nonconformity,
defect, or other undesirable process situation in an attemptto restore the process to the desired state, withoutaddressing the underlying cause of the situation. (ISOdefinition)
Taking action on the process in the form of adjustments to theprocess. These actions do not remove the cause but are knownto adjust the process in a favorable direction. (Adjusting)
OR
T
aking action on the process to remove the symptom of thespecial cause. This may not prevent the same problem fromrecurring in the future. (Fixing)
Adjustments may help one part of the process but could bedetrimental to others.
-
8/8/2019 7-Improving Unstable Processes
8/65
7-8
Investigate & FixSpecial Causes
Documentation
Recording the suspected cause of the problem, whatactions were taken and how effective the action was.
Benefits:
-
8/8/2019 7-Improving Unstable Processes
9/65
7-9
Investigate & FixSpecial Causes
Cause
D
Cause
G
Cause
A
Cause
B
Cause N All
others
Frequency
ofOccurrence
Pareto of Causes
Note: Putting resources to eliminating Cause D will result in most significantImpact on process improvement.
-
8/8/2019 7-Improving Unstable Processes
10/65
7-10
Investigate & FixSpecial Causes
Corrective Action
Actions taken to eliminate the causes of existingnonconformity, defect, or other undesirable situation inorder to prevent recurrence (ISO definition).
Permanent removal of the root source of recurringspecial causes that have been identified.
Focus is on preventing major recurring special causes.
Implementation results in variability reduction.
-
8/8/2019 7-Improving Unstable Processes
11/65
7-11
Investigate & FixSpecial Causes
Preventive Action
Action taken to eliminate the causes of a potentialnonconformity, defect, or other undesirable situationin order to prevent occurrence.(ISO definition)
Permanent removal of the root source of apotential special cause that could affect theprocess.
May involve using learning's from one process onanother process where the special cause might alsooccur.
-
8/8/2019 7-Improving Unstable Processes
12/65
7-12
Investigate & FixSpecial Causes
A standard way to perform a task.
A standard range that a key process variable should
be within.
Standards should be audited.
Control To A Standard
Note: S.O.P.'sNOT being followed can be a major reason for instabilities oncontrol charts!
-
8/8/2019 7-Improving Unstable Processes
13/65
7-13
Investigate & FixSpecial Causes
Control "Charts" will NOT improve the process!
S.C.A.P's IMPROVE the process
Provide a procedure for operations to follow in
identifying and removing the special cause symptom.
Promote a standardized approach to investigatingand compensating for special causes.
Provide a vehicle for documenting causes found and
actions taken to aid in future problem solving.
Why S.C.A.P.'s?
-
8/8/2019 7-Improving Unstable Processes
14/65
7-14
Investigate & FixSpecial Causes
Rating of
Importance to
Customer
10 1 9 9 5 8
1 2 3 4 5 7
Bea%
MeltTe
p
BottleOutput
Clait
Weight
Wall
Vaiation
Total
Process Step Process Input
Melt Resin Resin 9 9 3 9 9 9 324
Blow Bottle Mold Design 9 0 9 9 0 9 324
Ext ude Pa i on P og a ing 3 3 9 0 9 9 231Ext ude Pa i on Die Tip Te p 3 3 3 9 3 9 228
Blow BottleMold Water
Temp9 0 9 3 0 0 198
Blow Bottle
Water
Volume/Cooli
n Rate
9 0 9 3 0 0 198
Ext ude Pa i on Head De ign 3 9 3 3 3 9 180
Ext ude Pa i on Tooling 3 3 3 3 3 9 174
Blow Bottle Pinch Design 9 0 9 0 0 0 171
Melt Re in Ba ell Te p 3 9 9 1 3 3 168
Melt Re in S ew Speed 3 9 9 1 3 3 168
Melt Re in S ew De ign 3 9 9 1 1 1 142
Blow Bottle
# of Mold
Cooling
Zones
9 0 3 0 0 0 117
Tail Detab
Time from
extraction to
Detab
9 0 3 0 0 0 117
Re e be The C&E Mat ix
Note: C&E is a good place to start building control strategies. This information
will be used to develop a Failure Modes Effect Analysis (FMEA)
-
8/8/2019 7-Improving Unstable Processes
15/65
7-15
Investigate & FixSpecial Causes
Failure Modes and Effects AnalysisFailure Modes and Effects Analysis
A Tool forDeveloping SCAP's
-
8/8/2019 7-Improving Unstable Processes
16/65
7-16
Investigate & FixSpecial Causes
A structured approach to:A structured approach to:
Assist in developing SCAP's
identifying the ways in which a product or processcan fail
estimating the risk associated with specific causes
prioritizing the actions that should be taken toreduce the risk
evaluating the design validation plan (product) orthe current control plan (process)
Primary DirectivePrimary Directive:: Identify ways the product orprocess can fail and eliminate or reduce the risk offailure
FMEA Definition
-
8/8/2019 7-Improving Unstable Processes
17/65
7-17
Investigate & FixSpecial Causes
Vague
Workmanship
S andards
Poor Con rolPlans & SOPs
Raw Ma erial
Varia ion
Poorly DevelopedSpeci ica ion
Limi sMeasuremen
Varia ion (Online
and QC)
Machine
Reliabili y
Po en ial
Sa e yHazardsUnclear Cus omer
Expec a ions
D. H. Stamatis, FMEA:FMEA from Theoryto Practice, Quality Press, 1995
Poor Process
Capabili y
Cumula ive RiskCumula ive Risk
Where Do Risks Come From?
-
8/8/2019 7-Improving Unstable Processes
18/65
7-18
Investigate & FixSpecial Causes
Key tool of process team to improve the process in apreemptivepreemptive manner (before failures occur)
Used toprioritizeprioritize resources to insure process
improvement efforts are beneficial to customer
Used to documentdocumentcompletion of projects
Should be a dynamicdynamicdocument, continually reviewed,amended, updated
Role of FMEA
-
8/8/2019 7-Improving Unstable Processes
19/65
7-19
Investigate & FixSpecial Causes
AnalyzesAnalyzes new processes
IdentifiesIdentifies deficiencies in the process control plan
EstablishesEstablishes the priority of actions
EvaluatesEvaluates the risk of process changes
IdentifiesIdentifies potential variables to consider in E ploratoryData Analyses (EDA) and Design of E periments(DOE) studies
GuidesGuides the development of new processes
HelpsHelps set the stage for breakthrough
Purposes of FMEA
-
8/8/2019 7-Improving Unstable Processes
20/65
7-20
Investigate & FixSpecial CausesFMEA Inputs and Outputs
InputsInputs Process map
C&E matri
Process history
Process technical procedures
OutputsOutputs
List of actions to prevent causes or to detect
failure modes History of actions taken
-
8/8/2019 7-Improving Unstable Processes
21/65
7-21
Investigate & FixSpecial CausesFMEA Team
Team approach is necessary
Recommended representatives:
Design
Practitioners / Operators / Supervisors Quality
Reliability
Maintenance
Materials
Testing
Supplier
-
8/8/2019 7-Improving Unstable Processes
22/65
7-22
Investigate & FixSpecial Causes
Failure Mode
Effect
Cause
Current Controls
Severity, Occurrence, Detection
Risk Priority Number (RPN)
FMEA Terms
Next, Going to define each of these terms
-
8/8/2019 7-Improving Unstable Processes
23/65
7-23
Investigate & FixSpecial CausesFMEA E ercise
InpuFailure
ModeE ec s Causes Con rols
As we go through the different definitions and terms,complete the form below for your process
-
8/8/2019 7-Improving Unstable Processes
24/65
7-24
Investigate & FixSpecial CausesFailure Mode - Definition
Failure ModeFailure Mode - the way in which a specific process input fails -
if not detected and either corrected or removed, will cause theeffect to occur
Can be associated with a defect (in discrete manufacturing) or aprocess input variable that goes outside of specification
o Anything that an operator can see thats wrong isconsidered a failure mode
E amplesE amples
o Incorrect PO number
o Sample Size too small
o Dropped call (customer service)o Temperature too high
o Surface contamination
o Paint too thin
-
8/8/2019 7-Improving Unstable Processes
25/65
7-25
Investigate & FixSpecial CausesEffect - Definition
EffectEffect - impact on customer requirements
Generally e ternal customer focus, but can also includedownstream processes
E amplesE amples
Incorrect PO number: Accounts receivable traceability
errors Dropped call: Customer dissatisfaction
Temperature too high: Paint cracks
Surface contamination: Poor adhesion
Paint too thin: Poor coverage
-
8/8/2019 7-Improving Unstable Processes
26/65
7-26
Investigate & FixSpecial CausesCause - Definition
CauseCause
Sources of process variation that causes the failure mode tooccur
Identification of causes should start with failure modesassociated with the highest severity ratings
E amplesE amples Incorrect PO number: Typographical error
Dropped call: Insufficient number of CS representatives
Temperature too high: Thermocouple out of calibration
Surface contamination: Overhead hoist systems Paint too thin: High solvent content
-
8/8/2019 7-Improving Unstable Processes
27/65
7-27
Investigate & FixSpecial CausesCurrent Controls - Definition
Current ControlsCurrent Controls
Systematized methods/devices in place to prevent ordetect failure modes orcauses (before causingeffects)
Prevention consists of mistake proofing, automatedcontrol and set-up verifications
Controls consist of audits, checklists, inspection,laboratory testing, training, SOPs, preventivemaintenance, etc.
-
8/8/2019 7-Improving Unstable Processes
28/65
7-28
Investigate & FixSpecial Causes
Failure Mode 1
Failure Mode 2E ec 1
Note that the relationship between the failure mode and the effect is not
always 1-to-1
E ec 1
E ec 2
Failure Mode 1
Linking Failure Modes to Effects
-
8/8/2019 7-Improving Unstable Processes
29/65
7-29
Investigate & FixSpecial CausesRisk Priority Number (RPN)
The output of an FMEA is the Risk Priority Number
The RPN is a calculated number based oninformation you provide regarding
o the potential failure modes,
o the effects, and
o the current ability of the process to detect thefailures before reaching the customer
It is calculated as the product of three quantitativeratings, each one related to the effects, causes, andcontrols:
Effects Causes Controls
RPN = Severi y X Occurrence X De ec ion
-
8/8/2019 7-Improving Unstable Processes
30/65
7-30
Investigate & FixSpecial CausesDefinition of RPN Terms
Severity (of Effect)Severity (of Effect)-- importance of effect on customer
requirements - could also be concerned with safety andother risks if failure occurs (1=Not Severe, 10=VerySevere)
Occurrence (of Cause)Occurrence (of Cause)-- frequency with which a givencause occurs and creates a failure mode. Can
sometimes refer to the frequency of a failure mode(1=Not Likely, 10=Very Likely)
Detection (capability of Current Controls)Detection (capability of Current Controls) -- ability ofcurrent control scheme to detect or prevent:
the causes before creating failure mode
the failure modes before causing effect
1=Likely to Detect, 10=Not Likely at all to Detect
-
8/8/2019 7-Improving Unstable Processes
31/65
7-31
Investigate & FixSpecial Causes"E ample" Rating Scale
Ra ing Severi y o E ec Likelihood o Occurrence Abili y o De ec
10 Hazardous without warningVery high:
Can not detect
9 Hazardous with warningFailure is almost inevitable
Very remote chance of detection
8 Loss of primary functionHigh:
Remote chance of detection
7
Reduced primary function
performance
Repeated failures
Very low chance of detection
6 Loss of secondary functionModerate:
Low chance of detection
5Reduced secondary function
performance
Occasional failuresModerate chance of detection
4Minor defect noticed by most
customersModerately high chance of detection
3Minor defect noticed by some
customers Low:High chance of detection
2Minor defect noticed by
discriminating customers
Relatively few failuresVery high chance of detection
1 No effect Remote: Failure is unlikely Almost certain detection
De ec ion is ypically
assumed o imply
ac ion can be aken
-
8/8/2019 7-Improving Unstable Processes
32/65
7-32
Investigate & FixSpecial CausesQuiz!
Sev Occ Det Result Acti
Take 5 mi utesto ill i the results a d what action
should betaken. Be prepared to discuss results.
-
8/8/2019 7-Improving Unstable Processes
33/65
7-33
Investigate & FixSpecial CausesFMEA Form - Initial Assessment
Pr
cess
te /I tPote tial ail re Mode Pote tial ail re
ects
Pote tial a ses
O
rre t ontrols
P
N
Actions
ecommended
What is the
pro
ess step/
Input under
in
estigation?
In
hat
a
s does the Ke
Input go
rong?
What is the impa
t on the Ke
Output
ariables (
ustomer
equirements) or internal
requirements?
o
! "
e#
ere
isthe
effe
$
t
tothe
$uso
tmer?What
auses the Ke
Input to
go
rong?
o
! oftendoes
$ause
orFMo
$
$ur?What are the existing
ontrols and
pro
edures (inspe
tion and test)
that pre
ent eith the
ause or the
Failure Mode? Should include an
SOPnumber.
o
! !ell
$
a
n
%ou
dete
$t
$auseo
rFM?
What are the a
tions
for redu
ing the
o
urran
e of the
ause, or impro
ing
dete
tion? Should
have actions only on
hi & h PN's or easy
ixes.
0
0
0
-
8/8/2019 7-Improving Unstable Processes
34/65
-
8/8/2019 7-Improving Unstable Processes
35/65
7-35
Investigate & FixSpecial CausesFMEA Methodology
Two major approaches:
Starting with Cause & E ect Matrix
Starting with FMEA directly from the Process Map
We will e plain the approach using the C&E matri ,
though both approaches are very similar
Spreadsheet tools have been prepared to assist youin the preparation of the FMEA
-
8/8/2019 7-Improving Unstable Processes
36/65
7-36
Investigate & FixSpecial CausesFMEA Methodology - Starting with
C&E Matri
AdvantageAdvantage:: The Cause & Effect Matri assists theteam in defining the important issues that the FMEAshould address by helping to prioritize importantcustomer requirements
Process inputs that could potentially impact theserequirements
Prioritizing the Key Process Inputs according totheir impact on the Output variables (We want tofocus on Inputs that highly impact a large number
of Outputs first The C&E Matri also provides quantitative output
that can be used in the determination of thespecific severity ratings for the ne t stage of theFMEA process
-
8/8/2019 7-Improving Unstable Processes
37/65
7-37
Investigate & FixSpecial CausesFMEA - Step by Step
1. For each process input, determine the ways in which the
input can go wrong (failure modes)2. For each failure mode associated with the inputs, determine
effects of the failures on the customer
Remember the internal customers!
3. Identify potential causes of each failure mode4. List the Current Controls for each cause or failure mode
5. Create Severity, Occurrence, and Detection rating scales
6. Assign Severity, Occurrence and Detection ratings to eachcause
7. Calculate RPNs for each cause
8. Determine recommended actions to reduce high RPNs
9. Take appropriate actions and recalculate RPNs
-
8/8/2019 7-Improving Unstable Processes
38/65
7-38
Investigate & FixSpecial CausesProcess Mapping E amples
Manu acturingManu acturing
Outputs Outputs
TV of Mi -QualityCheckaround
e trusion
TV of Mi -QualityCheckarounde trusion
Surface Area
Pore Volume
Appearance
pH Temp Specific Gravity Clarity
Appearance(Color,Wetness)
Appearance(Color)
PhysicalChemicalProperties
Cycle Time(plugging)
Zone 3 Temp
Mixing
Water andMetals
Preparation
Extrusion
E trudethrough die
Dryer
Calciner
Materialdrying, surface
area issues
NADM
SolutionPreparation
Impregnation
Metalsaddition
Fluid Bed
DryerFinal
preparation /appearance
issues
Total Rate CMi er Speed CAl2O3 Qual URec Comp U
E t Rate C
Die Wear U
E t RPMs C
Die Change U
Temperature C
Rot speed C
Draft C
Feed Rate C
Drying air U
Inputs Types Inputs Types
Nitric Acid CH2O2 CWater CADM CHold Time U
Agitation C
Phos Acid C
Nozzle Type C
Spray Time C
Water C
Moly C
Nickel C
Base C
Temperature C
Rate C
Air Flow C
Res Time C
-
8/8/2019 7-Improving Unstable Processes
39/65
7-39
Investigate & FixSpecial CausesC&E Matri E ample
This is the C&E Matri for the previous Manufacturing
Process Map with the Key Inputs sorted by the Total ScoreRating of Importance to ' ustomer 8 8 4 9 10 4 8 8
1 2 3 4 5 6 7 8
P
lugging
C
ycleTime
Length
D
istribution
P
orosity
D
istribution
C
olor
C
atalyst
P
erformanc
M
oly
D
istribution
P
ellett(withi
&
between)
D
ust/Fines
C
ontent
%
Moly&%
N
ickel
Total
Process tep Process Input
Impregnation Moly 9 0 0 3 9 3 9 9 345
NADM ADM 9 0 0 1 9 3 9 9 327
Impregnation Two Step / One Step Process 9 0 0 3 9 9 9 3 321
Impregnation Base 3 9 9 0 9 3 3 1 266
Impregnation Order of Addition 3 0 0 3 9 9 3 3 225
Impregnation Water 3 0 0 9 0 9 9 0 213
Impregnation Nickel 1 0 0 9 3 0 0 9 191NADM Nitric Acid 9 0 0 0 3 9 3 1 170
NADM Hydrogen Peroxide 9 0 0 0 3 9 3 1 170
NADM Water 9 0 0 3 1 3 3 3 169
NADM Temperature (Steam Inj) 9 0 0 3 1 3 3 3 169
Fluid Bed Dryer Temp 3 0 0 9 3 0 1 0 143
Mixing Al2O3 Quality 1 9 9 0 0 0 3 0 140
Mixing Recycle Rate 1 1 1 0 1 3 3 9 138
-
8/8/2019 7-Improving Unstable Processes
40/65
7-40
Investigate & FixSpecial Causes
FMEA Worksheet
Be ore we move to the example lets look at an FMEA worksheet
The in ormation on this sheet is trans erred directly to the FMEA orm
The purpose o this worksheet is to ocus the team on the FMEA inputs
and not on scoring
The scoring should be done after the basic inputs have been madeThe scoring should be done after the basic inputs have been made
FMEA Worksheet.XLSFMEA Worksheet.XLS
Process
Step KeyProcess Input
Failure odes -hatcan o
rong? Effects Causes CurrentControls
-
8/8/2019 7-Improving Unstable Processes
41/65
7-41
Investigate & FixSpecial Causes
1. For Each Process Input, Determine theWays in Which the Input Can Go Wrong(Failure Modes)
We will first deal with the Moly Flow Rate input variable.
ro ess
tep nput
Failure ModesWha t an o
wron ? ffe ts auses
urrent
ontrols
Impregnation
Moly Flo rate
Moly Flo rate is too
high
Moly Flo rate is too
lo
FMEA Step 1
-
8/8/2019 7-Improving Unstable Processes
42/65
7-42
Investigate & FixSpecial Causes
2. For Each Failure Mode Associated withthe Inputs, Determine Effects
These effects are internal requirements forthe next process
and/orto the final customer
rocess
tep/ nput
Failure odeshat can go
rong ffects Causes
Current
Controls
Impregnation
Mol Flowrate
Mol Flowrate is too
highff pec Material
lugs dr er
Mol Flowrate is too
lowff pec Material
FMEA tep 2
-
8/8/2019 7-Improving Unstable Processes
43/65
7-43
Investigate & FixSpecial Causes
3. Identify Potential Causes of Each FailureMode
In most cases, there will be more than one Cause for a Failure
Mode but well keep it simple forthis exercise
Process
Step Input
Failure Modeshat can go
wrong Effects Causes
Current
Controls
I pregnation
Moly Flo rate
Moly Flo rate is too
highff Spec Material eigh Cell Failure
Plugs dryer eigh Cell Failure
Plugs dryer perator Error
Moly Flo rate is too
loff Spec Material eigh Cell Failure
ff Spec Material perator Error
May elect to list both effects on a single line since they
relate to a single cause, and reduce a line in the table.
FMEA Step 3
-
8/8/2019 7-Improving Unstable Processes
44/65
-
8/8/2019 7-Improving Unstable Processes
45/65
7-45
Investigate & FixSpecial Causes
5. Create Severity, Occurrence, and Detection
Rating ScalesExample Rating ScaleRating Severity ofEffect Likelihood ofOccurrence Ability to Detect
10 Hazardous without warningVer hi h:
Can not detect
9 Hazardous with warningFailure is a lmost inevitable
Very remote chance of detection
8 Loss of primary functionHi h:
Remote chance of detection
7Reduced pr imary function
performance
Repeated failuresVery low chance of detection
6 Loss of secondary functionModerate:
Low chance o f detection
5Reduced secondary function
performance
Occasional failuresModerate chance of detection
4Minor defect noticed by most
customersModerately high chance of detection
3Minor defect noticed by some
customers Low:High chance o f detection
2Minor defect noticed by
discriminating customers
Relatively few failuresVery high chance of detection
1 No effect Remote: Failure is unlikely Almost certain detection
FMEA Step 5
-
8/8/2019 7-Improving Unstable Processes
46/65
7-46
Investigate & FixSpecial Causes
6. Assign Severity, Occurrence and Detection
Ratings to Each Cause
We are now ready to transfer the worksheet input to theFMEA form
Copy and paste the worksheet columns into the appropriate
FMEA form columns
The team then starts scoring each row to compute the RPNvalues
Notes:
You will only use one Severity value
Determine which effect has the highest associated Severity anduse that SEV value forALLALL causes for the related failure mode(Worst Case)
o When combining effects that have the same cause Ne t Slide
FMEA Step 6
-
8/8/2019 7-Improving Unstable Processes
47/65
-
8/8/2019 7-Improving Unstable Processes
48/65
I ti t & Fi
-
8/8/2019 7-Improving Unstable Processes
49/65
7-49
Investigate & FixSpecial CausesRPN Review
Once you calculate the RPN for each failure mode /cause / controls combination, review the results andlook for insights
Do the gut check - does the Pareto of items makesense?
If not, maybe the ratings given are varying
Determine potential ne t steps:
Data collection
E periments
Process improvements
Process control implementations
-
8/8/2019 7-Improving Unstable Processes
50/65
Investigate & Fix
-
8/8/2019 7-Improving Unstable Processes
51/65
7-51
Investigate & FixSpecial Causes
9. Take Appropriate Actions and
Recalculate RPNs
We have recorded the action taken and the impact on the RPN
Notice thatthis is a nice way to track past activities
The FMEA should be re-evaluated by the group as new
recommended actions are identified, completed and recorded
Process
tep nput
Potential
Failure M ode
R
P
N
Actions
RecommendedResp Actions Taken E E
T
R
P
N
( )pre
0
nation1
2
step 3 step
3 step4
solution
sli ) es on
sur5
ace6
7 8 8
3 . 9 e5
ine causes.2
.@
Aat is on t
Ae
B aseC
D
.E
a ) ple t A eF
et
catalG
st5
or sli ) e.H
.E
i ) ulate t A e
) ec A anical e5 5
ects
o5
plant A andlin0
.
I A P Q R
SAR
I Q
E
1
T QR
S R
Q NT Q R
( )pre
0
nation1
2
step 3 step
3 step4
solution
sli ) es on
sur5
ace6
7 8
( )pre
0
nation1
2
step 3 step
3 step4
solution
sli ) es on
sur5
ace6
D 2 8
( ) pre0
nation1
2
step 3 step
2
E
tep appearsas 3
E
tep
D 2
T
o t A rou0
A ti ) in0
on one c
G
cle o
5
tA
eS one
G F
ell
pro0
ra )
R
U RT (E
V AW
FM A tep 9
Investigate & Fix
-
8/8/2019 7-Improving Unstable Processes
52/65
7-52
Investigate & FixSpecial CausesB-326DI FMEA E ample
Prepared by: Glen P. Wieger
FMEA Date (Orig) 5 September, 2002 (Rev) 1
Process Step/Input Potential Failure Mode Potential Failure Effects
S
E
V
Potential Causes
O
C
C
Current Controls
D
E
T
R
P
N
Item
Number
What is the process
step/ Input under
investigation?
In what ways does the
Key Input go wrong?
What is the impact on the
Key Output Variables
(Customer Requirements)
or internal requirements?
How
Severeisthe
effecttothecusotmer?
What causes the Key Input
to go wrong?
How
oftendoescauseor
FMo
ccur?
What are the existing controls
and procedures (inspection and
test) that prevent either the
cause or the Failure Mode?
Should include an SOP number.
How
wellc
anyoudetect
orpreventcauseorFM?
272
Cation Production -
Filter H2O Flow Flow Too Low
before regen required forspacing, results in
production loss 9
No flow equilization between
vessels 10 None 8 720
283
Cation Production -
Filter H2O Flow Flow Too Low
u
before regen required for
spacing, results in
production loss 9
Lack of DCS control logic to
control flow 10 None 8 720
291
Cation Production -
Filter H2O Flow Flow Too High
u g
run-lengths, early
regeneration resulting in
production loss 6
No flow equilization between
vessels 10 None 8 480
301
Cation Production -
Filter H2O Flow Flow Too High
nequa spac ng o vesse
run-lengths, early
regeneration resulting in
production loss 6
Lack of DCS control logic to
control flow 10 None 8 480
305
Cation Production -
Filter H2O Flow Flow Too High
un engt s s ortene ue
to bottom layer of resin
breakthrough, resulting in
production loss 6
No flow equilization between
vessels 10 None 8 480
315
Cation Production -
Filter H2O Flow Flow Too High
u g u
to bottom layer of resin
breakthrough, resulting in
production loss 6
Lack of DCS control logic to
control flow 10 None 8 480
111
Cation Production -
Resin Condition
Insufficient exchange
capacity
oss o exc ange capac ty -results in short run lengths,
loss of production, higher
regen costs 7
Improper regeneration (over
time) 8 DCS control 7 392
Building 326 DI PlantUtilities Division Six Sigma Team
Process or Product Name:
Responsible:
-
8/8/2019 7-Improving Unstable Processes
53/65
Investigate & Fix
-
8/8/2019 7-Improving Unstable Processes
54/65
7-54
Investigate & FixSpecial CausesApproaches to FMEA
Approach One (C&E Matri Focus)Approach One (C&E Matri Focus)
Start with key inputs with the highest scores from the C&EMatri analysis
Fill out the FMEA worksheet for those Inputs
Calculate RPNs and develop recommended actions for thehighest RPNs
Complete the Process FMEA for other Inputs over time
Approach Two (Customer Focused)Approach Two (Customer Focused)
Fill out the failure mode and effects columns of the worksheet.Copy to FMEA form and rate Severity.
For High Severity Ratings, List causes and rate Occurrencefor each Cause
For the highest Severity * Occurrence Ratings, evaluatecurrent controls
For Highest RPNs develop recommended actions
Investigate & Fix
-
8/8/2019 7-Improving Unstable Processes
55/65
7-55
Investigate & FixSpecial CausesApproaches to FMEA Continued
Approach Three (Comprehensive)Approach Three (Comprehensive)
Good approach for small processes Fill out the FMEA worksheet beginning with the
first process step and ending with the last
Score SEV, OCC and DET for all causes
Develop recommended actions for highest RPNs
Approach Four (Super Focused)Approach Four (Super Focused)
Pick the top Pareto defect item (Damaged
Components) or Failure Mode (Variability inTemperature)
Focus the FMEA process on only that defect orfailure mode
Purpose: To kill that failure mode
Investigate & Fix
FMEA Overview
-
8/8/2019 7-Improving Unstable Processes
56/65
7-56
Investigate & FixSpecial Causes
Process
Step/InputPotential Failure Mode Potential Failure Effects
S
E
V
Potential Causes
O
C
C
Current Controls
D
E
T
R
P
N
Actions
Recommended
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
What
is the
Input
WhatWhat
can gocan go
wrongwrong
with thewith the
InputInput
What canWhat can
be donebe done
What isWhat is
the Effectthe Effect
on theon the
OutputsOutputs
What areWhat are
thethe
CausesCauses
How canHow can
these bethese be
found orfound or
preventedprevented
HowHow
BadBad
HowHow
OftenOften
HowHow
wellwell
FMEA Overview
Investigate & Fix
S C A P C id ti
-
8/8/2019 7-Improving Unstable Processes
57/65
7-57
Investigate & FixSpecial CausesS.C.A.P. Considerations
1) A single joint is above UCL or below LCL2) A run of 8 above or below CL
Target
Time
2) Chronic Problems
Bad
Better
Time
1) Sporadic Problems
Note: S.C.A.P. may need to be specific to the type of instability
Investigate & Fix
A Typical S C A P
-
8/8/2019 7-Improving Unstable Processes
58/65
7-58
est gate & xSpecial Causes
Plot Data on Control Chart
Recheck Lab Analysis
Control
Chart
Signal?
RecheckValue InsideControl Chart
Limits?
Adjust Process Variable (Knob)
Wait for Another Sample
Ne tSample ValueInside ControlChart Limits?
Delete Original
Value
Replace With
New Value
No
Yes
Yes
No
YesNo
A Typical S.C.A.P.
Investigate & Fix
A S t d S C A P
-
8/8/2019 7-Improving Unstable Processes
59/65
7-59
gSpecial CausesA Suggested S.C.A.P.
Plot data on Control Chart
Control
ChartSignal?
Is
Adjustment
Process
Variable
Appropriate?
Adjust Process
Variable and Collect
Another Sample as
Soon as Appropriate.
Stability
Restored?
Continue
to
Monitor
Is
Another
Adjustment
Called for?
Document
Actions
&
Results
Process Operating Conditions
Equipment Failure
Raw Material Change
Procedure Deviation
All OK?
Correct Process and
Collect Another
Sample as Soon as
Appropriate
Stability
Restored?
Continue
to
Monitor
Last
Resort
Adjustments
Document
Actions
&
Results
Continue
to
Monitor
Document all
Actions, Results
and
Additional
Observations
Preventive
Action
Documentation
Documentation
Corrective Action
Investigative Action
Documentation
Compensating Action
No
YesYes
aboratory
ontrol
trategy
Yes Yes
Yes
NoNo
No
No
No
Yes
Yes
Production
Control
Strategy
-
8/8/2019 7-Improving Unstable Processes
60/65
Investigate & Fix
SC
-
8/8/2019 7-Improving Unstable Processes
61/65
7-61
gSpecial Causes
%9835 Control Strategy for PM-9877-C Process
Reason for low % 9835, if found:
Action taken:
Out-of-ControlCondition
PossibleCauses
Actionto Take
InvestigativeResults
InvestigatedBy
%9835 abov e uppercontrol limit for 1
1. Incorrectcalculations
1. Recheck all calculations1. Calculations Correct? ____Yes ____No_________
batch or above target
f or 8 consecutiv ebatches
2. Not enoughPM-939 added.
2. a. Check the batchrecords.
b. Check the scalesf or accuracy .
c. Check valves incharging line.
d. Check packinggland on Milroy alpump for leaks.
e. Check packing glandon autoclave agitator.
2. a. Amount of PM-939 added ______lbs.
b. Scales are accurate? ____Yes ____No
c. Any v alves leaking? _____Yes ____No
d. Packing gland leaking? ____Yes ____No
e. Agitator packing gland leaking?
_______ Yes _______ No
_________
_________
_________
_________
_________
Batch Number
__________
3. Overcharge ofPM-9835.
3. a. Check the batchrecords.
b. Check the scalesf or accuracy .
3. a. Amount of PM-9835 charged?_____lbs
b. Scales are accurate? ____Yes ____No
_________
_________
4. Assay ofPM-9835 is lowor high.
4. a. Check the assayof the PM-9835used in computer.
b. Was thisfirst batchaf ter changing lotsof PM-9835?
4. a. PM-9835 assay _____% (in computer)
PM-9835 assay _____% (f rom vendor)
b. New lot of PM-9835? ____Yes ____No
_________
5. Change inPM-1530cataly st quality.
5. a. Check if catalys tdrum was changedf or this batch.
b. Sample cataly st.
5. a. New cataly st drum? ____Yes ____No
b. 4-ounce sample. Hold f or tech staf f ._________
6. Incorrecttemperatureand/or pressureconditions.
6. Check the batchrecords and charts.
6. a. Maximum temperature _______C
b. Maximum pressure ________psig
c. Addition time ______hrs ______min
_________
_________
7. Recorders out of calibration.
7. Hav e the recorderschecked f or propercalibration.
7. Were the recorders properly calibrated?
_______ Yes _______ No_________
8. Laboratoryerror.
8. a. Ask lab whetheranaly sis wasrepeatable.
b. Re-sample batchand hold sample
f or tech staff .
8. a. Laboratory t est run in duplicate?
_______ Yes _______ NoIfno, request that sample be re-run.
Recordbothv alues on batch record.b. 4-ounce sample. Hold f or tech staf f .
_________
S.C.A.P.CheckSheet
Example
Investigate & Fix
K l d /S t R i t
-
8/8/2019 7-Improving Unstable Processes
62/65
7-62
gSpecial CausesKnowledge/Systems Requirements
for a S.C.A.P.Compensating Action
Criteria for judging if compensating action is appropriate. What process variable is to be adjusted?
Required effect of process variable adjustment.
Amount of adjustment to obtain the required effect.
Criteria for judging if stability has been restored.
Investigative Action Standard (Required) Process Operating Conditions.
Equipment failures signals.
Raw material analysis techniques.
Well-defined procedures.
Corrective Action Well-defined procedures for correcting any deviation
detected in the Investigative Action.
Last Resort adjustment procedures.
Criteria for judging if stability has been restored.
DocumentationProcedures/Systems for documenting actions, results
and additional observations.
Procedures, Lists, charts for accumulating actions,
results and additional observations.
Investigate & Fixl
-
8/8/2019 7-Improving Unstable Processes
63/65
7-63
Special Causes
Control Charts provide information, I.e., when to take
action and when not to take action.
People must know what action to take. Actions need to
be standardized for all people.
People control the process, not the control charts.
Investigation of Assignable Causes and the correction/
prevention of problems should never end.
Control Charts and the S.C.A.P. need to be
reviewed periodically and updated as needed.
Process Control is the responsibility of Operations.
Control Charts need to be in the hands of the people
who control the process.
S.C.A.P.
Investigate & FixS i l C
-
8/8/2019 7-Improving Unstable Processes
64/65
7-64
Special Causes
StatisticalProcess
Control
Use SPC toMaintainCurrentProcess
Collect &Interpret
Data
SelectMeasures
DefineProcess
IsProcess
Capable?
Improve
ProcessCapability
IsProcessStable
?
Investigate &Fix Special
Causes
No
Yes
No
Yes
Statistical Process Control (SPC)is a collection of activities:
Selection of appropriate processmeasures
Collection of process dataGraphical analysis of dataAnalysis of process stabilityUse of data to investigate and fix
special causes in a continuousimprovement cycle
Investigate & FixS i l C
-
8/8/2019 7-Improving Unstable Processes
65/65
Special Causes
E ercises
1.) Your Catapult Team should complete page 10 of theCatapult Process handout.
Limit yourselves to 30 minutes for this e ercise.