Printed from: Quest Integrity

Print Page


Critical Spares - Who Needs Them and When?

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:

  • Commissioning
  • Initial
  • operating
  • outage and capital insurance
  • critical spare parts.

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.

Except for 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:

  1. Facility operation, and the demands and restrictions to operation. Ideally this should have been considered during the procurement and build phases, when the known operational data for various packages can be assessed before a commitment has been made. Depending on the industry, this would look at information such as fleet data, required availability/supply network and the option to have a possible central storage of critical parts - the latter being more applicable to power generation operators who can have a significant fleet of similar units.
  2. If the build is complete and the facility operational, an assessment of the whole production system using reliability centred maintenance (RCM, SAE JA1011 for example) or risk based assessment/inspection (RBA/RBI) processes should be completed. This should include the following.
    1. Review history of failed components for the individual plant and across the industry.
    2. Review OEM manuals.
    3. Interviewing the plant operational and maintenance personnel to capture and record existing institutional knowledge.
    4. Inspection of the plant’s operational and maintenance supply chain. Do you have the capability and skills to do this or is there a local network of suppliers who could hold the critical spares and deliver them within the required timeframes? This requires the supply chain to be regularly audited to ensure availability of the parts is maintained.  
    5. Perform a risk assessment for the mechanical and electrical systems within the plant.

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:

  • What is the age of the component?
  • Are there damage processes and what is the likelihood of …?
  • Is rapid or unpredictable failure likely?
  • How good are the current inspections?
  • How good is the current maintenance?
  • How well is the component currently operated?

Typical questions for the assessment of consequences are:

  • What would be the extent of damage to other equipment?
  • What is the part availability?
  • What is the length of repair time?
  • What is the safety hazard to personnel?
  • What is the potential loss of production?
  • What is the cost to repair or replace?
  • Is there potential to cause damage to the environment?
Once the RCM or RBI process has been completed, a total risk profile can be plotted (Figure4). Not surprising, the area most at risk is the within the gas turbine enclosure where redundancy is not present.



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 (λ) reciprocal is the mean time between failures (MTBF). The reliability can then be determined using:

R(t) = exp (-λt)

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/𝜆 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) = (λx)(e-λ)/x!

The above can be simplified to the following formula. 

P = K√λ

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:  

  • Identifying equipment risk through reliability based inspection.
  • Identify all failure modes for each component.
  • Detailed monitoring & inspection test plans specific to each component.
  • The most appropriate maintenance approach for particular machines, systems, and areas.
  • Suppliers, vendors, contactors, and internal personal with regard to expectations in relation to reliability and availability of critical items of plant.

Finally, remember that the critical parts assessment is not a static document, SAE JA1011 recommends that the assessment should be revisited every two years.


   Email Print

All Content Rights Reserved , Quest Integrity
Captavi - SaaS CMS Platform ©