Risk-based Asset Management identifies the most appropriate and sustainable maintenance, operating, and replacement strategies based on the actual risk to the business posed by a crane asset in its current operating context.
It
is important to recognize that cranes operate as part of a system; risk is
usually a function of the entire system’s performance, rather than an
individual unit’s. Therefore if funds and resources are limited, it is best to
optimize an individual crane’s performance only when it contributes
significantly to the overall risk.
In other words, deal
with the weakest links in the chain first. As one link is strengthened, begin
work on the next weakest and so on.
To
implement an RBAM approach, four major steps are required:
(1)
Identify priorities. If
the organization has a number of different cranesystems, it is important to
work out which system should be analyzed first.
(2)
Understand the risks posed
by an operating system, and identify its weak points. These may vary for
different types of risk. One crane may be contributing the most to cost, while
another poses the greatest environmental threat.
(3)
Identify strategies for
reducing the risk, or risks, posed by that specific crane/s.
(4)
Manage corporate knowledge to ensure the business can adequately implement and
sustain the initiatives developed in steps 2 and 3.
Risk
= probability x consequence x confidence
Undertaking
a RBAM analysis and implementation process takes time and commitment. With a
large and disparate asset base that is made up of lots of different cranes, it
is impossible to solve all problems in one pass. It is also important to get
some big wins early. So determining the order in which cranes will be analyzed
is an important first step in applying RBAM.
Technically speaking,
highest risk cranes should be tackled first. However, these may be the hardest
to change. So a more flexible strategy is generally warranted.
A basic ranking
exercise, based on a set of agreed risk criteria, is the easiest way to get
started. This ranking should be performed in a group, which includes
representatives from the organization that manage the business as well as the
cranes themselves, in addition to experienced risk analysts to facilitate the
process.
Many
crane owners have a standard risk matrix that can be applied; however if no
appropriate approach exists, ranking criteria might include:
•Safety
and environmental risks posed by failure of the crane
•Operational
criticality of the crane
•Historical
cost of maintenance (in the current configuration)
•Operating
costs
•Number
of similar cranes within the entire asset base
•Ability
to effect the required changes (e.g. unionized site, highly regulated,
prevalence of old-school attitudes or openness to new ideas.
All
unique configurations are then ranked according to the combined total of all
risk criteria. This determines the order in which cranes should be evaluated
Once
the system has been modeled, individual components that contribute
significantly to overall system risk and/or cost are easily identified. More
comprehensive analysis and/or improvement strategies can then be implemented.
Reducing risk is
usually achieved by reducing the number of unexpected failures in critical
system components. Reducing whole of life cost however, is a little more
complicated. As reducing unexpected failures also reduces cost (unplanned
failures can be 10 times more expensive than planned maintenance), these
factors will be considered together.
Get
Smarter About Maintenance
To reduce risk and
minimize whole-of-life costs, cranes should be maintained in accordance with
their combined risk/consequence and failure profiles. There are four options:
(1) Run to failure is
appropriate for low consequence scenarios where the cost of failure is
approximately the same as the cost of replacement/refurbishment. It is also
appropriate when the risk of failure is the highest immediately after
installation and reduces with age. Electronic
components typically fall into
this category.
(2) Preventive
replacement/refurbishment is appropriate when assets fail predictably with
time-dependent failure modes. Age (e.g. Run-hours, days, cycles, etc) is then
monitored and parts/cranes are replaced or refurbished when this age reaches a
predefined level. Applying this regime when failure is not time-dependent does
nothing to reduce the risk of failure. It is also called fixed-term
maintenance.
(3) Condition-based
maintenance should be considered when cranes have no strong time- dependent
failure modes (i.e. failure age is random), or when consequences of failure are
significant and must be avoided at any (reasonable) cost. Most large cranes
generally fit into this category, which is also called predictive maintenance.
This approach requires suitable condition indicators that are able to warn of
incipient failure.
CBM
assumes that the risk of failure of a crane is a function of its age, and its
response to its operating environment.
However CBM implies
much more than just collecting information about the crane’s integrity; it
assumes that this condition data will then be used intelligently to decide what
and when particular maintenance tasks should be performed. Unlike condition monitoring,
CBM is a maintenance management philosophy, not an inspection process.
The
major components of a CBM approach are illustrated in Figure 3. Associated work
tasks include:
•Sourcing,
acquiring, cleansing and formatting data (crane’s failure history and condition
monitoring or process data).
•Developing
models that relate failure events to CM data.
•Evaluating
data shortfalls and auditing data quality.
•Determining
appropriate on-condition maintenance tasks.
•Implementing
FMEA (failure mode effects analysis) to determine if failure modes that have
not occurred historically may still pose a significant risk to the business.
•Disseminating
information appropriately.
•Changing
maintenance philosophies when current strategies are inappropriate.
So
why do CBM?
Several tangible
benefits to an organization are expected from the introduction of CBM
practices. Specific benefits depend on the extent to which CBM is implemented,
however, these generally include:
1)
Cost reductions via:
•Reduction
of unnecessary maintenance & replacement
•Establishment
of more appropriate maintenance requirements for individual cranes and crane
types
•Reduced
reactive maintenance
•Less
collateral damage from failed equipment
•Fewer
surprises during major overhauls
•Reduced
sparing
•Informed
decision-making
2)
Increased availability through:
•Improved
planning and decision making
•Reduced
number of breakdowns and less collateral damage
•Improved
understanding of machinery and maintenance limitations
•Improved
troubleshooting
3)
Improved data capture and
exploitation
4)
Improved knowledge transfer between personnel
collecting, analyzing and using data, resulting in improved diligence in
performing maintenance tasks and data collection.
5)
Reduced risks of crane failure and consequent loss of availability.
For
CBM to be useful a particular failure mode must produce one or more measurable
warning indicators prior to failure, and these indicators must provide
sufficient time for systems and/or personnel to react appropriately.
If a crane has
multiple failure modes, then multiple indicators usually need to be monitored.
For defining suitable maintenance each failure mode must be distinguishable by
a unique set of parameter values.
With a large number
of failure modes and CM indicators, this mapping can not be achieved manually.
We use a program which analyses the
relationships between historical failure data and available condition
monitoring data. It also:
•Identifies
which condition monitoring parameters are useful in warning of incipient
failure, and which are not; these can be safely removed from the inspection
program
•Predicts
remaining useful life (time to repair/replacement)
•Quantifies
the current risk of failure
•Estimates
the risk of failure at a specified time in the future
•Determines
how much is saved by implementing the recommended maintenance policy, as
compared to the current policy and/or run-to-failure.
A typical output from
the software is shown in Figure 3.
Figure 3: output
graph that can help asset managers identify if items need to be repaired/replaced on the basis of
condition monitoring result
The
Life-Cycle Cost (LCC) Equation makes the assumption that the crane is only
purchased once in the plant’s lifecycle.
In practice, it may
be feasible to replace a crane at periodic intervals due to escalating costs of
operation and maintenance. Therefore to minimize the overall lifecycle costs,
the optimal time to replace should be determined and the LCC equation adjusted
accordingly. Common reasons for considering replacing functioning cranes with
new models include:
•Cheaper
to replace than repair.
•Requirements
have changed from the time of original installation, resulting in the equipment
operating under sub-optimal conditions, increasing maintenance and/or operating
costs.
•No
longer possible to restore the crane to an as-new state resulting in increased
maintenance and/or operating costs.
Newer, technologically-improved models are available with lower
operating and maintenance costs.
The
quantities of spare parts carried by an organization typically reflect the
variety of the equipment to be maintained, the
Spares management is
something that should be considered once initiatives specific to the needs of a
crane and the systems it works in have been put in place and are well proven.
Thereafter, programs that can often yield value include:
Rationalizing
crane types –
This allows an organization to share spares
Establishing
supply agreements –
These become easier to arrange when work is preplanned, parts have been
rationalized and spares are no longer held on site as suppliers are able to
better estimate their revenue and manage their stock holdings. Their reduction
in risk can be returned to the asset owner in reduced parts costs.
Vendor
management of on-site spares – This moves the parts management and carrying cost of
stocking appropriate spares to the supply vendor. They are the people best
situated to provide the correct inventory mix to match the equipment mix for
the site. The vendor is most aware of their manufacturing and supply capability
and can make the more optimal stocking decisions.
Creating
Knowledge Workers
Analysis tools have
fundamental data requirements and data outputs. This data output needs to find
its way to whomever needs that data in order to make better business decisions.
Be it an operator, maintenance planner, or reliability engineer in multiple
locations, with each needing the information presented in a slightly different
way. This is the essence of a “knowledge worker”. Good business decisions
therefore require good quality data and an efficient delivery mechanism. These
factors are not independent.
Experience
tells us that maintenance related data held by most organizations is not fit
for purpose. Ultimately, this is because it is not being used. Reasons for data
not being used include:
•Data
is difficult to access and/or combine in a meaningful way (required information
may be held in disparate systems)
•Data
fields are incomplete or not suitable for the task at hand (e.g. failure mode
descriptions are inconsistent);
•Wrong
data is being captured
•No-one
is responsible for ensuring data integrity.
Knowledge
management (KM) help
organizations overcome these issues, and underpin the savings and risk
reduction strategies discussed, by reducing the time taken to collate, analyze
and present data
As interoperability between disparate
systems is improved, data can be accessed efficiently and effectively. Thus
analysis of data becomes a more convenient and trusted input to decision
making.
The
greatest benefit for an organization that RBAM offers is that it
Although some upfront
investment is required to develop an implementation plan and install tools that
facilitate appropriate data mining and interpretation, each system analysis and
subsequent implementation of its recommendations can be justified against the
savings on offer.
These are discrete,
specific and objective. Performance of the program, as well as consultants
brought in to help in its implementation, can therefore be closely monitored.
Costs are controlled and further initiatives become much easier to justify.
References
[1] Carson, J.S.,
(2005) “Introduction to Modeling and Simulation”, in Proceedings of the 2005
Winter Simulation Conference, pp. 16-23
[2] Barlog, R.J., (2004)
“Reliability Review and Reliability/Availability Model: Whiting Refinery Waste
Water Treating Plant”, BP Report, pp.1-23
[3] Mackenzie, M. and
Briggs, R. (2006) MODELLING AND SIMULATION - A PROFITABLE TOOL FOR ALL PHASES
OF THE LIFECYCLE, in Proceedings of 1PstP WCEAM (submitted for publication).
[4] (2001) “Pump
Lifecycle Costs: A guide for LCC Analysis for Pumping Systems”, Europump and Hydraulic
Institute, USA.
[5] Nowlan, F. S., and Heap,
H.F. (1978) “Reliability-centred Maintenance”, Report
No. A066-579, National Technical Information Service.
[6] Collins, A.
(2004) “Vibration Analysis of Pumps – Course Notes”, Blakers Pump Engineers.
[7] ETSU and AEAT PLC
(2001) “Study on improving the energy efficiency of pumps”, European
Commission.
[8] Jantunen, E., Miettinen, J. and Ollola, A. (1993)
"Maintenance and downtime costs of centrifugal pumps in Finnish
Industry" in Proceedings of The 13th International Pump Technical
Conference, Pumps for a Safer Future, London, UK, pp.9-20.
[9] Yates, M.A.,
(1996) “Pump Performance Monitoring, Proceedings of the 5th International
Conference on Profitable Condition Monitoring”, BHR Group, pp.207-228.
[10] Sikorska, J.Z., and Hodkiewicz, M. (2005) “Flow
monitoring of a double-suction pump”, Proceedings of the Comadem, pp. 192-202.
[11] Hodkiewicz, M., Kelly, P., Sikorska, J. and Gouws, L. (2005), “A
Framework to Assess Data Quality for Reliability Variables”, in Proceedings of
1st WCEAM (submitted for publication).
