Optimising gas turbine maintenance
Maintenance is all about increasing the life of the equipment by planned and unplanned activities that verify component functionality and detect faults preferably before they occur. Maintaining high levels of availability and reliability are essential objectives for many industries, especially those that are subject to high costs due to shutdowns of critical systems and components, such as gas turbines.
To utilise these systems as effectively as possible, preventive maintenance must be optimised. Determining what is optimal is, however, a multi-variable task requiring detailed knowledge about the components in the system and their different damage mechanisms. Traditional maintenance planning consisted of maintenance packages comprising a set of maintenance activities for a set of components, in other words, it was predetermined, but with the introduction of lifetime assessment methods it has become possible to estimate and measure correctly for individual components during operation.
In general, downtime of a production unit due to failure of one of its components can result in high costs. However, executing maintenance activities too frequently can be costly as well. To minimize life cycle costs and maximize earnings, it is important to optimize maintenance according to a customer’s specific conditions. Achieving the optimal maintenance plan, which minimizes the total cost, depends on the availability and coexistence of diagnostics, prognostics and maintenance planning technologies.
Condition-Based Maintenance (CBM) is a term for methods and methodology that, based on the actual condition and predicted future use, in theory allows maintenance to be performed at the best possible date for each component. CBM tasks are normally applied to components of a system that fail over a period of time and give some early indication of an impending failure, enabling the user to avoid the consequences of an unexpected breakdown.
Siemens defines a fully implemented CBM strategy as a process consisting of prognostics, diagnostics and optimization tools, where all pieces are equally essential. We call this process CAMP, Customer Adapted Maintenance Plan. The real benefit of CAMP lies in the ability to precisely adjust the maintenance timing and level to the actual condition of the turbine and its components. The maintenance strategy of the CAMP approach is Retirement For Cause (RFC), where components are not replaced until a potential failure has been detected. RFC is an approach that allows each component in an engine to be used for the full extent of its safe life.
Due to the more exact knowledge of the maintenance needs of the gas turbine, maintenance intervals no longer remain fixed, but instead vary depending on the condition of several components and a variety of other factors eg actual load profile, quality of fuel, ambient temperature, particle levels etc. This information will be combined with the customer’s opportunities and demands for the maintenance to be carried out as effectively as possible. Thus, a CBM business model will ensure that the potential for short-term profits will be evaluated in an overriding life cycle cost perspective.
To manage all the information in our CBM strategy, we have developed a Preventive Maintenance optimisation tool called PM-opt. PM-opt will plan preventive maintenance for any complex technical system and maximise earnings for a system operator. It is done by the use of an advanced prognosis process and input from an operator regarding operation profile, ambient conditions and financial data such as production value and standstill costs. This information will be processed in PM-opt, generating an optimised preventive maintenance schedule adapted to an operation-unique situation and hence maximising profit. The process is supported by an advanced diagnostic tool to further increase reliability and availability.
The goal is to provide operating conditions that will increase availability with predictable scheduled maintenance, based on condition monitoring assessment with little or no downtime during deployments. Any changes in e.g. operation profile will instantly affect the preventive maintenance, of course. Also, if an unplanned opportunity occurs, maintenance can be re-scheduled if it is proved profitable to use this ‘slot’. PM-opt can deal with these situations and re-optimize maintenance if this is financially justifiable for the operator. PM-Opt uses advanced diagnostics and prognostics tools to estimate when in the future a component in an engine must be maintained. This needs to be sensitively handled in combination with a customer’s business system.
With a CAMP approach, flexibility depends on a lot of different factors such as risk willingness, condition of components, value of production and future operation profile. Through intelligent processing, and integration with other parameters, valuable information can be acquired including actual life consumed, life remaining, and the condition of the gas turbine relating to its operation profile and ambient conditions. Applications include calculating the risk involved in extending the life of components.
It is a well known fact that components in a gas turbine face different wear, depending on parameters such as environment, load, events, fuel type etc. This means that two identical gas turbines with different operators can present significant differences in wear. In order to make an optimisation, every component in a gas turbine must be monitored and the accumulated equivalent operating hours (EOH) and equivalent operating cycles (EOC) must be considered. In addition, predictions based on an estimate of the expected future wear should be available in order to compute expected maintenance deadlines.
If components face unequal wear, they must be replaced with great care. Failure to do so may result in sub-optimisation, causing increased cost for a customer due to higher maintenance frequency as a function of uneven component wear. This can be avoided by the use of a highly detailed components database with component traceability, ensuring an effective replacement schedule throughout the gas turbine lifetime.
When creating and managing a customer-adapted maintenance plan, a lifetime prediction tool (LPT) is essential. The LPT will keep track of a number of damage locations for each component in a gas turbine, eg creep, fatigue, erosion, oxidation and corrosion. The prognostic tool calculates the residual lifetime depending on a customer’s operation profile.
The goal of a maintenance strategy should be to reach a Retirement for cause (RFC) condition, where components are not replaced until a potential failure has been detected. Further, the inspection interval should be large enough to allow spare parts to be ordered and delivered during the time between failure detection and failure, with sufficient safety margins. This requires measurement techniques that can monitor how the turbine is operated, prognostic capabilities that foresee maintenance needs, and test methods that can determine the state of a component during maintenance events. We can also show how eg changes in operation profile will affect the future maintenance needs. If a customer changes, for example, from a base load application to a peak load application, we can re-calculate lifetime of critical components and create a new maintenance plan, building on operation history and future operation profile. This will give the operator new flexibility in terms of planning his preventive maintenance. With a CBM contract, each critical component within the gas turbine is condition-monitored and any influences from, for instance, a changed operation profile or from unforeseen events will result in a new, opti¬mised maintenance plan.
In general there are two ways to determine the amount of damage a gas turbine component has been subjected to – calculations and examination of components exposed to service. Optimum results should be obtained by using calculations as a basis and continuously reviewing/modifying their interpretation and the underlying damage models using best available experience. This Siemens process is called Maximum Utilisation of Parts Process (MUPP) – a process for systematic testing of used components: by this process, data can be turned into modifications to a gas turbine’s maintenance plan with same or decreased risks, and the implications for equipment operator as well as maintenance provider.
Siemens’ approach to CBM is a maintenance philosophy and methodology that will allow us to meet high demands from the market regarding life cycle profit. To be successful requires advanced technologies and a close relationship between maintenance supplier and operator. Full and adequate implementation will require CBM-enabling of health monitoring and assessment of systems having diagnostic, prognostic and optimisation. The systems must also have the capability to fully utilize the possibility to gather data and make use of remote communication technologies in order to transmit real time data between a site and a logistic centre with experienced OEM personnel available. This enables more accurate trending and projection analyses on the life cycle management.
The Siemens approach not only uses advances diagnostic and prognostic tools, we can also use the information to create optimised maintenance plans adapted to a customer’s operation profile, technical specifics and financial situation. This approach is built on our core competency in understanding the diagnostic and prognostics of engine health and is used as input into PM-opt software in order to provide a way to maximize life cycle profit of gas turbine ownership.
For more information, visit www.siemens.com
