Solving Manufacturing Problems
Solving Manufacturing Problems
5-Whys TapRooT
• You cannot defeat your enemy until you intimately know your enemy
• Otherwise you don’t know what you are fighting and your Root Cause
Analysis will be led down the wrong garden path
• Defect characterization allows you to gain this knowledge systematically
• Characterization describes and quantifies the size, shape, depth, location,
color, frequency of occurrence, etc of the defect in order to point to the cause
of the defect
• This extends to the microscopic level including chemical identification of
foreign matter
• Let us consider the case of a metal part subassembly to understand the defect
characterization process
Frank Anderson, the production superintendent, who is typically a soft-spoken man unless he is speaking
about his beloved Tennessee Titans, capitalizes on a lull in the discussion to intervene “You all are talking
about this defect as if it were a phantom, with phantom causes. But I don’t see that an example part with
this defect has been brought into this room and I don’t see that a Defect Characterization Form has been
completed for this issue. Let’s grab a few parts from the production line and characterize this defect
together.”
The drawing was not a work of art. It didn’t need to be. It merely needed to serve as a communication
tool. Drawings, sketches, diagrams are the universal language. As the defect characterization form was
projected on the screen Tim Westlake, the maintenance manager, commented “Wait a minute … I have
seen that hexagonal pattern before. It looks like the base metal on the gripper pads on the pick and place.
We had to replace the air cylinder last week.”
As it turned out, the air cylinder pressure on the pick and place was set too high causing the rubber gripper
pads to wear out exposing the bare case-hardened steel base. This was the source of the scratches. This
simple example demonstrates the importance of characterizing defects to identify root causes. If, for
example, the hexagonal pattern of the gripper base was not known, analysis of the particulates at the
bottom of the scratches might have pointed to the deteriorating rubber gripper pads. Today, we have a
multitude of instrument analysis techniques available which can be used to characterize both inorganic
and organic contaminants.
• The size of the RCA Team depends on the scope of the failure incident
• If it is a simple component failure with limited scope the RCA team could consist
of just one person
Don't
Yes No Know # Root Cause Analysis Question Root Cause Implication PM Strategy Implication
1 Was the failure preceded by a process interruption? Investigate details of process Modify startup procedure & checklist
interruption
2 Was the failure preceded by a process spike (eg Investigate details of process spike Consider additional process controls
temperature, pressure, flow rate, concentration, etc)? and/or alarms
3 Did the component fail during the time frame of another Investigate which failure has
component failure? occurred first
4 Was the component operated outside of process Investigate reason for non-standard
specifications? operation
5 Was there a change in utilities (air, steam, water, electricity, Investigate details of utility change Consider additional utility controls
etc) prior to component failure? and/or alarms
6 Was there a dramatic change in the ambient environment Investigate details of ambient Consider insulating component from
(eg temp fell below freezing, thunderstorm, high humidity, environment change ambient environment
high temp) prior to component failure?
7 Is the component brand new? Infant mortality, warranty claim
8 Has this same component failed in the last 6 months? Component under-specified, start-up,
shut-down and/or operation
procedure inappropriate
9 Was the component recently rebuilt? Rebuild procedure, install procedure Depends on autopsy
10 Did the component fail suddenly? Fatigue failure, contamination, Depends on autopsy
thermal overload
11 Did the component performance degrade over time? Dirt accumulation, component Depends on autopsy
deterioration, lubrication failure
12 Is the component performance requirement at or above its Component under-specified
design limit?
13 Did the component exhibit any external signs prior to Use external sign to focus root cause
failure (eg vibration, temperature build-up, leaking, noise, detection during autopsy
odor, etc)?
Don't
Yes No Know # Root Cause Analysis Question Root Cause Implication PM Strategy Implication
14 Is there evidence of external damage to component? Component may have been Consider installing protective
accidentally damaged shielding
15 Does autopsy indicate component corrosion as failure Material selection inappropriate for Increase frequency of wall thickness
mode? process fluid & corrosion check
16 Does autopsy indicate incorrect assembly or missing parts? Warranty claim or inadequate
rebuild/install procedure
17 Does autopsy indicate excessive wear for the service time of Material selection inappropriate, Consider adding periodic lubricant
the component? lubrication insufficient analysis
18 Does autopsy indicate external particulate contamination? Component requires particulate Consider installing protective
contamination protection shielding and/or filtration
19 Does autopsy indicate presence of foreign liquid? Identify source of foreign liquid and
eliminate
20 Does autopsy indicate seal failure? Seal material/design, seal fluid Add seal fluid check to Operator
system Round Sheets
21 Does autopsy indicate electrical system failure? Identify failed component and Consider adding electric power
potential sources of failure supply controls/conditioning
22 Does autopsy indicate loose electrical connections or Connector and/or shielding design
shielding failure? and thickness
23 Does autopsy indicate wear parts have fallen below their Material selection inappropriate or
acceptable tolerance? component under-specified
24 Does autopsy indicate damaged internal parts? Identify source of internal damage
25 Does autopsy indicate jammed or slow-moving internal Identify source of internal friction or
parts? jam
26 Does autopsy indicate nothing wrong? Intermittent failure - conduct bench
top stress tests
27
stopping
Flammable
Liquid
Security Guard
Conditional noticed strong odor
Cause
Evidence Oxygen
Restricting
Primary Effect Orifice Plugged
Operator Statement
Low pressure,
alternate water source
to cool thrust bearing
is most effective
solution
Background
Pump CP4826 is a 25 Hp centrifugal pump used in the Waste Water Treatment Plant (WWTP) to recirculate the
contents of the primary digestion tank T3759. CP4826 recirculates the process fluid through eductor nozzles at the
bottom of tank T3759 to promote aeration of the tank and to prevent solids settling. It is considered a critical
component since extended periods of time without tank recirculation will cause not only tank fouling with solids
but more importantly, oxygen-deficient stratification layers within the tank causing depletion of the aerobic bacteria
necessary for the digestion process.
The mechanical seal on pump CP4826 has been a constant headache for the maintenance department. The seal has
failed three times in the last 12 months and had just been replaced 30 days ago. This chronic seal failure has
prompted the Maintenance Manager, Jim Phillips to proclaim “those darn John Crane seals are no good ... I’ve
always had better luck with Chesterton seals!” The seal fluid, as recommended by the manufacturer, is a 50/50
mixture of propylene glycol and water.
Incident Description
On Tuesday morning, at 5:15 am, Kevin Walters, the WWTP Operator, was just completing his hourly rounds and
had returned to the WWTP control room. Kevin noticed a critical alarm on the alarm management screen of his
DCS monitor. A high pressure indication was being registered by pump CP4826 within the tandem seal buffer
cavity. Kevin went out to physically inspect the pump and when he arrived at the pump location he described the
scene as such “there was seal fluid everywhere, and the pump was making a high pitch squeal like metal grinding
on metal!”. Kevin immediately shut down the pump and alerted his supervisor.
Don't
Yes No Know # Root Cause Analysis Question Comments Root Cause Implication PM Strategy Implication
X 1 Was the failure preceded by a process interruption? Not really - but there was an Investigate details of process Modify startup procedure & checklist
electrical outage at 2:00 am interruption
X 2 Was the failure preceded by a process spike (eg Investigate details of process spike Consider additional process controls
temperature, pressure, flow rate, concentration, etc)? and/or alarms
X 3 Did the component fail during the time frame of another Investigate which failure has
component failure? occurred first
X 4 Was the component operated outside of process Investigate reason for non-standard
specifications? operation
X 5 Was there a change in utilities (air, steam, water, electricity, Investigate details of utility change Consider additional utility controls
etc) prior to component failure? and/or alarms
X 6 Was there a dramatic change in the ambient environment Yes - temperature in the Investigate details of ambient Consider insulating component from
(eg temp fell below freezing, thunderstorm, high humidity, pump house fell below environment change ambient environment
high temp) prior to component failure? freezing during the night
X 7 Is the component brand new? Infant mortality, warranty claim
X 8 Has this same component failed in the last 6 months? Yes - CP4826 pump seal has Component under-specified, start-up,
failed 3 times in the last 12 shut-down and/or operation
months procedure inappropriate
X 9 Was the component recently rebuilt? Seal was replaced 12/14/09 Rebuild procedure, install procedure Depends on autopsy
X 10 Did the component fail suddenly? Seal should last longer than Fatigue failure, contamination, Depends on autopsy
30 days thermal overload
X 11 Did the component performance degrade over time? Dirt accumulation, component Depends on autopsy
deterioration, lubrication failure
X 12 Is the component performance requirement at or above its Component under-specified
design limit?
X 13 Did the component exhibit any external signs prior to Pump seemed to be more Use external sign to focus root cause
failure (eg vibration, temperature build-up, leaking, noise, noisy and vibrating more than detection during autopsy
odor, etc)? usual during hourly rounds
Don't
Yes No Know # Root Cause Analysis Question Comments Root Cause Implication PM Strategy Implication
X 14 Is there evidence of external damage to component? Component may have been Consider installing protective
accidentally damaged shielding
X 15 Does autopsy indicate component corrosion as failure Material selection inappropriate for Increase frequency of wall thickness
mode? process fluid & corrosion check
X 16 Does autopsy indicate incorrect assembly or missing parts? Warranty claim or inadequate
rebuild/install procedure
X 17 Does autopsy indicate excessive wear for the service time of Seal had oval wear pattern. Material selection inappropriate, Consider adding periodic lubricant
the component? Bearing was totally seized. lubrication insufficient analysis
X 18 Does autopsy indicate external particulate contamination? Bearing has debris inside Component requires particulate Consider installing protective
from the bearing material contamination protection shielding and/or filtration
itself.
X 19 Does autopsy indicate presence of foreign liquid? Identify source of foreign liquid and
eliminate
X 20 Does autopsy indicate seal failure? Primary seal has failed Seal material/design, seal fluid Add seal fluid check to Operator
system Round Sheets
X 21 Does autopsy indicate electrical system failure? Identify failed component and Consider adding electric power
potential sources of failure supply controls/conditioning
X 22 Does autopsy indicate loose electrical connections or Connector and/or shielding design
shielding failure? and thickness
X 23 Does autopsy indicate wear parts have fallen below their Unable to tell because bearing Material selection inappropriate or
acceptable tolerance? and seal have been badly torn component under-specified
up
X 24 Does autopsy indicate damaged internal parts? Main pump bearing has Identify source of internal damage
seized causing scoring to the
drive shaft. Metallurgical
report from SKF indicates
that bearing has fatigue
failure which could be caused
by high temperature and high
vibration conditions
Don't
Yes No Know # Root Cause Analysis Question Comments Root Cause Implication PM Strategy Implication
X 25 Does autopsy indicate jammed or slow-moving internal Bearing has debris inside Identify source of internal friction or
parts? which could have jammed jam
balls
X 26 Does autopsy indicate nothing wrong? Intermittent failure - conduct bench
top stress tests
X 27 When was last time drive motor or pump were replaced on Drive motor was replaced in
CP4826? Dec 2008 since it had burnt
out
X 28 Were any alignment tools used during motor installation to There are no tools available.
ensure that the drive shaft and pump shaft were in There is no procedure for
alignment? drive/pump shaft alignment.
Next Steps:
• Create Reality Chart
• Add Proposed Solutions
• Rank Solution Alternatives using PICK Chart
Problem Statement:
Centrifugal Pump CP4826 had to be
shut down due to loss of seal fluid
Solutions 1, 2 & 3
or 4 are best
alternatives
Date/Time Description
1/12/10 12:00 AM Round Check OK
1/12/10 1:00 AM Round Check OK
1/12/10 2:00 AM Plant-wide power outage for approx. 30 seconds.
1/12/10 3:00 AM Round Check OK Recovery from power outage OK.
1/12/10 4:00 AM Round Check OK CP4826 running rough (noisy and warm to touch).
1/12/10 5:00 AM Round Check OK
1/12/10 5:15 AM Low pressure alarm for CP4826 seal on DCS panel.
1/12/10 5:17 AM K. Walters inspected CP4826 condition.
1/12/10 5:18 AM K. Walters observed high pitch squeal and seal fluid leak around CP4826.
1/12/10 5:20 AM CP4826 manually shut down at local panel.
1/12/10 5:25 AM CP4826 locked and tagged-out
1/12/10 5:35 AM Seal fluid spill contained and cleaned-up
1/12/10 8:00 AM Root Cause Investigation Team formed.
Gano, Dean L., Apollo Root Cause Analysis, Third Edition, Apollonian
Publications, Richland, WA, 2007