By: Stephen Rowbotham, Quest Integrity NZL Ltd.
As seen in the Jan-March 2014 issue of Energy Generation Magazine. View the PDF version.
The topic of critical spares for industrial plant is an important aspect in the successful operation of a plant. Minimising any delays in restarting following a forced outage due to component failure greatly benefits real-world availability. A review of the required critical spares can optimise the capital cost implications of carrying the required parts.
To most people facility stores contain consumable items, the nuts and bolts of everyday maintenance. However, lurking in the dark corners are critical or insurance spare parts: large, expensive crates that reside in the storage facility, never appearing to move and often requiring maintenance or replacement due to a fixed shelf life.
As a plant ages and personnel changes, the existence of critical spares can become a mystery, especially if systems and policies are not in place to monitor and maintain the parts.
To help clear up
the mystery let us answer what is a critical spare part? There are several
definitions but the one that covers most bases is the following: "A part that, if not
available when required, results in the loss of production, power generation or
endangers people’s lives or seriously damages the environment”. The latter two
points have always been important but have become more important following a series
of high profile failures. The operation
of any industrial facility requires various types of spares to be carried:
Figure 1 - What's in that box?
Each is used over the life cycle of the unit as represented by a bath tub curve in Figure 2. The critical and insurance parts are predominately components with long lead times.
So once we define a critical part, the question will still be asked: why do we need to tie up capital in parts that we may never need if the prescribed maintenance schedule is followed? The cost of holding the required critical spares are not only the acquisition costs but the inventory holding costs (typically 20 to 40 % of the purchase price per year).
The answer is not straight forward. Even with excellent maintenance regimes, failures will happen and can be predicted if you know or can estimate typical failure rates. Implementing a critical spares assessment will identify the likelihood and consequences of the occurrence in relation to the plant so that sound financial and engineering decisions can be made.
Figure 2 - Spares required over the life of plant
The critical spares assessment will be affected by the operational strategy for risk which is in turn affected by the contracts and penalties in force for loss of generation or production. Other factors include maintenance systems, skills within the organisation and dependency on external suppliers. The two extremes of risk strategy are conservative operations or run until it drops. Most operational strategies will fall somewhere in between.
financial or IT institutions were the risk to operation is so great that
duplication of the entire systems is justified. Most businesses or industries
however do not have this luxury. Assessments
must be made to determine the parts of a system that are critical to operation,
should a failure occur.
Figure 3 - Extremes of risk strategy
So where do you start? The first stage of the critical parts assessment is to complete the following review:
The first filter
in the system review is to identify redundancy, as all parts of a system will
be critical unless redundancy is present. This is not to say that the parts
should not be carried for a system which does have redundancy, as you need to
assess how long the system can be exposed to operation with no redundancy.
Fortunately, most power plants have considerable redundancy built into the
system with several but not all gas turbine power stations being dual fuel. For
gas turbines power stations in particular, criticality only appears after the last
fuel filter/pump skid when systems rely on single components and failure will
result in the loss of production.
Once the component or areas at risk have been identified, the detailed risk assessments of likelihood and consequences of a failure can be undertaken. Some typical questions that could be asked to determine the likelihood of failure are:
Typical questions for the assessment of consequences are:
Figure 4 - Risk for mechanical systems in a gas turbine power plant
A critical parts list can now be generated. The next step is to determine how many components are required to achieve the desired availability.
Many component manufacturers have failure rates for their components. To complete the assessment, some statistical analysis taking into account the failure rate is required. In this we have to make some assumptions, one being a constant failure rate. The failure rate (Î») reciprocal is the mean time between failures (MTBF). The reliability can then be determined using:
R(t) = exp (-Î»t)
For example, let’s look at an electric motor, using the data available, a failure rate of 14.3 x10”‘6per hour could be expected. The reliability over one operating year (approximately 8000 hours) is approximately 89%. The survival probability or exchange rate (tr) can then be determined using:
tr= 1/ðœ† x ln(1/R(tr ))
which for the motor is 2 years.
Table 1 - Parts required to meet desired availability
The spares required or stock level to cover the failure rate and survival probability is based on a Poisson probability distribution.
P(x) = (Î»x)(e-Î»)/x!
The above can be simplified to the following formula.
P = KâˆšÎ»
K being a factor depending on the service level or availability required. As Table 1 shows, the failure rate of insurance critical parts is low and the analysis would indicate only one part to be held.
One final question is, how long does the assessment take? The on-site portion of both the mechanical and electrical systems is usually completed in 20 personnel days (5 days on site with a team of 4 engineers). The report is normally issued 6 to 8 weeks after demobilisation from site.
In conclusion, the critical part assessment is an integral part of a facilities equipment integrity management system, and as such helps the plant operators understand, develop and manage the following:
Finally, remember that the critical parts assessment is not a static document, SAE JA1011 recommends that the assessment should be revisited every two years.