Escolar Documentos
Profissional Documentos
Cultura Documentos
“Your system is perfectly designed to give you the results that you get.”
W. Edwards Deming PhD
H ow good is your organization at identifying failures? Of course you see failures when they occur, but
can you identify when recurring failures are creating serious equipment reliability issues? Most compa-
nies begin applying RCA or RCFA to “high value failures”. While this is not wrong, I prefer to either not
see the failure in the first place, or at the least, to reduce the failures to a controllable level.
Failure Reporting Analysis and Corrective Action System role in unreliability. Here are a few examples of those
(FRACAS) is an excellent process that can be used to con- losses:
trol or eliminate failures. This is a process in which you
identify any reports from your CMMS/EAM or a special- 1. Equipment Breakdown (total functional failure)
ized Reliability Software that can help you to eliminate, A. Causes of Equipment Breakdown
mitigate or control failures. These reports could include 1. No Repeatable Effective Repair, Preventive
cost variance, Mean Time Between Failure, Mean Time Be- Maintenance, Lubrication, or Predictive
tween Repair, dominant failure patterns in your operation, Maintenance Procedure
common threads between failures such as “lack of lubrica- 2. No one following effective procedures
tion” (perhaps due to lubricator not using known industry 2 Equipment not running to rate (partial
standards). One poll was conducted recently covering 80 functional failure)
large companies. Shockingly, none of these companies A. Causes of Equipment not Running to Rate
were capturing the data required to understand and con- 1. Operator not having an effective
trol equipment failures. procedure to follow
2. Operator not trained to operate or
Answer the following questions honestly before you go troubleshoot equipment
any further to see if you have any problems with identify- 3. Management thinking this is the best rate
ing failures and effectively eliminating or mitigating their at which the equipment can operate because
effects on total process and asset reliability. of age or condition
3. Off-Quality Product that is identified as “first pass
1. Can you identify the top 10 assets which had the quality” (could be a partial or total functional failure)
most losses due to a partial or total functional A. Causes of Quality Issues
failure by running a report on your maintenance 1. Acceptance by management that “first pass
software? quality” is not a loss because the product can
2. Can you identify the total losses in your organization be recycled
and separate them into process and asset losses for 4. Premature Equipment Breakdown
the past 365 days? A. Ineffective or no commissioning procedures. We
3. Can you identify components with a common thread are talking about maintenance replacement of
due to a specific failure pattern, such as the one parts or equipment and engineering/contractor
shown in Figure 1? that fails prematurely because no one has iden-
tified if a defect is present after the equipment
has been installed, repaired, serviced, etc. See
Infant Mortality
Figures 2 and 3.
(If you have ever seen equipment break down or not run-
ning to rate immediately after a shutdown, you know
what we are talking about.)
Figure 1 - Failure Pattern from Nowlan
and Heap Study
The Proactive Workflow Model – Eliminating unreliability
Many times, the cost of unreliability remains unknown be- is a continuous improvement process much like the Proac-
cause the causes of unreliability are so many. Whether tive Work Flow Model in Figure 4. The Proactive Work-
you want to point the finger at maintenance, production flow Model illustrates the steps required in order to move
(operations) or engineering, each functional area plays a from a reactive to a proactive maintenance program.
44 june/july 2010
the FRACAS provides con-
tinuous improvement for Failure Modes
your maintenance strate- Analysis
Asset Catalog
ProActive Work Flow Model ure reporting or FRACAS (Failure Reporting,
Analysis and Corrective Action System). In
Asset Criticality Analysis
order to eliminate failures, one needs to en-
sure this is a successful first step. Figure 6 (on
Failure Modes Analysis Predictive & Preventative Work
80% of Total Work M.O.C. Procedures the following page) displays the findings from
C.A.S. LOOP
F.R.A. a plant with 32 total “Part – Bearing” failures
Equipment Maintenance from different size electric motors (“Part” is
Plan (EMP) PdM/PM Inspections Results of PdM/PM Equipment Reliability
Improvements identified from a CMMS/EAM Codes drop down
screen). One type “Defect – Wear” occurred
Proactive Work Flow
M.O.C. Procedures Planning M.O.C. Procedures in 85% of the failures (“Defect” is identified
Failure Modes
Analysis
Job Plan
Daily Coordination Improvements Work Order
”Inadequate Lubrication”. Now it is time to
perform a Root Cause Failure Analysis on this
Root Cause
History
www.uptimemagazine.com 45
The purpose of the Asset Criticality Analysis
32 Bearing Failures Accross the Plant is to identify which equipment has the most
Wear Defect for 85% of Bearing Failures serious potential consequences on business
98% of Defects Caused by Lack of Lubrication performance, if it fails. Consequences on the
business can include:
46 june/july 2010
cept of part-defect-cause, and go through the
Criticality Severity Order basic equipment types in your facility such as
High 1 1st
centrifugal pumps, piston pumps, gearboxes,
High High 2 2nd
motors, etc.. You will find that a relatively
Medium 1 3rd
small number of failure codes will cover a lot
Medium 2 4th
High 3 5th
of failure modes in your facility. The failure
modes developed during this exercise can later
Medium Low 1 6th
Medium 3 7th
become the basis for the failure modes, effects,
Intercept
Ranking Low 2 8th and criticality analysis that takes place during
Criticality
High 5 16th
If, over a period of one year, the dominant
©2009 GPAllied
Medium 5 17th
failure mode is “wear” for bearings caused by
5 4 3 2 1 Not 4 18th
www.uptimemagazine.com 47
pect about Asset Health. Once this change is
Asset Health Summary underway, Asset Health, as a metric, becomes
From: Jun 2006 To: June 2007
View: Overall
what most maintenance managers and plant
managers have wanted for a long time — a
1,400 leading indicator of maintenance costs and
Plot Legend
Asset Count Green business risk.
Yellow
Red
0 2. Mean Time Between Failures and Mean Time
Between Repairs – reported by maintenance
Jun 2006
Jul 2006
Aug 2006
Sep 2006
Oct 2006
Nov 2006
Dec 2006
Jan 2007
Feb 2007
Mar 2007
Apr 2007
May 2007
Jun 2007
or reliability engineers on a monthly basis on
Month
the top 5-20% of critical equipment. The report
to management should include recommenda-
Figure 9 – Percent of Assets with No Identifiable Defect tions to improve both metrics and should be
measured and posted on a line graph for all to
in an existing condition, or to keep, preserve, Strategy, Maintenance Planning consisting of see.
protect, while Strategy is development of a repeatable procedures, work scheduled based
prescriptive plan toward a specific goal. on equipment criticality, work executed using 3. Cost Variance by area of the plant – report-
precision techniques, proper commissioning of ed by maintenance and production supervisor
So, a Maintenance Strategy is a prescriptive equipment when a new part or equipment is area of responsibility. Cost variance must be
plan to keep, preserve, or protect an asset or installed, and quality control using Predictive reported to maintenance and production man-
assets. Keep in mind that one specific type of Maintenance Technologies to ensure no de- agement on a monthly basis. The report should
maintenance strategy is “run to failure” (RTF). fects are present after this event occurs. The not be acceptable without a known cause of
However, RTF is used only if, based on thor- very last part of your maintenance strategy is the variance and a plan to bring it in compli-
ough analysis, it is identified as the best solu- FRACAS, because it drives the continuous im- ance.
tion for specific equipment to optimize reliabil- provement portion of this strategy.
ity at optimal cost. Less invasive maintenance 4. Most Frequent Part-Defect-Cause Report –
is preferred to more invasive maintenance. Failure Reporting reported monthly by maintenance or reliability
This is one of the fundamental concepts of any engineers. If you do not have maintenance or
well-defined maintenance strategy. Specific Failure reporting can come in many forms. reliability engineers, you may need to appoint
maintenance strategies are designed to miti- The key is to have a disciplined plan to review a couple of your best maintenance technicians
gate the consequences of each failure mode. failure reports over a specific time period, and as “Reliability Engineering” Technicians, even if
As a result, maintenance is viewed as a reliabil- then to develop actions to eliminate failure. unofficially, and train them to be a key player
ity function instead of a repair function. Saying Following are a few Failure Report examples, in this failure elimination process. This one re-
this means Predictive Maintenance or Condi- which should be included as part of your FRA- port can identify common failure threads with-
tion Monitoring is the best solution because it CAS Continuous Improvement and Defect Elim- in your operation which, when resolved, can
is mainly noninvasive. ination Process. make a quick impact to failure elimination.
Knowing that both systemic problems and op- 1. Asset Health or Percent of Assets with No There are many more reports that can be used
erating envelope problems produce the same Identifiable Defect – reported by maintenance effectively, but will not fit in the space of this
type of defects, a maintenance strategy that management to plant and production manage- article. You will be able to find more reports
merely attempts to discover the defects and ment on a monthly basis at least (see Figure 9). in the book on “FRACAS” written by Ricky and
correct them will never be able to reach a pro- An asset that has an identifiable defect is said Bill, which will be published by mid July.
active state. Technicians will be too busy fixing to be in a condition RED. An asset that does
the symptoms of problems instead of address- not have an identifiable defect is said to be in Bill Keeter is currently a Senior Techni-
ing the root cause. To reach a truly proactive condition GREEN. That is it. It is that simple. cal Advisor for Allied Reliability. Bill joined
state, the root cause of the defects will need There are no other “but ifs”, “what ifs” or “if Allied in 2006 after serving as President
to be identified and eliminated. Maintenance then”. If there is an identifiable defect the asset of BK Reliability Engineers, Inc. where he
strategies that accomplish this are able to is in condition RED. If there is no identifiable provided training and facilitation services
achieve a step change in performance and defect, it is GREEN. The percentage of ma- to help facilities improve asset performance
achieve incredible cost savings. Maintenance chines that are in condition GREEN is the Asset using Weibull Analysis, Reliability Centered
strategies that do not attempt to address the Health (as a percentage) for that plant or area. Maintenance, Availability Simulation, and Life
root cause of defects will continue to see lack- Cycle Cost Analysis. Bill has over 30 years of
luster results and struggle with financial per- The definition for defect is: an abnormality in a experience in Maintenance Engineering and
formance. part which leads to equipment or asset failure Management. He has successfully imple-
if not corrected in time. mented maintenance improvement programs
A Maintenance Strategy involves all elements in a variety of manufacturing and production
that aim the prescriptive plan toward a com- Example: the plant has 1,000 pieces of equip- facilities. Bill’s experience includes mainte-
mon goal. Key parts of a maintenance strategy ment. Of that number, 750 of them have no nance leadership positions in the US Military,
include Preventive and Predictive Maintenance identifiable defects. The plant is said to have the nuclear industry, chemicals, paper con-
based on a solid Failure Mode Elimination 75% Asset Health. There is an interesting as- verting, and plastic film manufacturing. He
48 june/july 2010
has provided training and reliability consult-
ing services to petroleum, process, mining,
and defense industries in the United States, Get more
done
Mid-East, and Europe. Bill has developed
competency maps for Reliability, Availabil-
ity, and Maintainability Engineering for the
Petroleum Industry’s PetroSkills® program.