Such system processes the decision-making parameters and the signals representative of the individual components technical conditions (indicators of degradation). Such parameters automatically come from the network remote control system, from on-line local sensors installed on the critical power HV equipment and as an outcome of the controls locally carried out on the components. Through engineering models and decision logics defined by Terna long experience, the system shows to the technicians the appropriate actions to be carried out at the right time, by highlighting their priority based on technical-economic risk analysis. In this way the system allows to switch from a time-based maintenance policy to a condition-based maintenance and at the same time to standardize the behavior of the technicians.
This tool helps the transmission company to translate their asset performance targets and budget constraints into practical maintenance policies and plans. The set of tasks implementing the maintenance process are carefully modelled in MBI. A common practice supporting each task has been developed and implemented in the system. It is based on a set of technical/economic criteria established by the management. The adoption of MBI is expected to improve both the economic results of the company and the quality of service provided to customers. At the same time the maintenance policy adopted will increase the asset value. Another important goal pursued by MBI is the standardization of practices and unification of the maintenance practice within the whole organization. This goal is particularly important for a grid company whose asset is spread over a wide area (even the whole national territory) and non-standardized maintenance practices may be locally still in use.
The MBI maintenance process complies with the company objectives by successive refinements, where objectives are aimed at by applying consolidated practices, possibly amended at each cycle to get closer to the objectives. The effects of the maintenance policy are measured according to the main guidelines of reliability/availability and maintenance costs. The measurements obtained are used as basis for the revision of the maintenance process and in the end the analysis and revision cycle allows the needed maintenance modifications to be implemented at the same time preserving the safety requirements.
In detail, MBI covers the following tasks:
Inspection planning: the frequency of in-field inspection visits and the list of monitoring actions (e.g. diagnostic investigations) to be carried out during each visit are two crucial issues of the maintenance policy. Component type and age, operating and environmental conditions are some of the elements taken into account by the inspection planning criteria. MBI takes care of planning inspection visits at component aggregate level (e.g. bay), supports the operator in the inspection scheduling and highlights the delayed inspections.
Proposing maintenance and repair actions : a major objective of MBI is to propose maintenance and repair actions on the basis of the current technical conditions of the grid components. The information comes from inspections, on-line monitoring systems and other available data sources (e.g. operating and environmental conditions). A set of maintenance models that implement the technical/economic rules leading to the maintenance decisions are included in MBI.
Component replacement and system renewal : MBI manages the whole lifecycle of the grid components and component aggregates as well; in this respect it makes recommendations about both single component replacement and the complete renewal of component aggregates, e.g. a substation.
In order to suggest the renewal of the whole substation MBI considers in detail both the replacement actions involving the substation components and the component status. A substation renewal is proposed whenever MBI realizes that the number and quality of component replacement exceeds a predefined threshold.
Monitoring of maintenance process effectiveness: the effectiveness of the maintenance process is monitored through a set of performance indicators computed by MBI that account for the asset performance. For instance, some of the defined indicators correlate component failures with the age and the component type, or with the amount of inspection/maintenance actions performed in the past on that component.
Update and revision of maintenance policy. The maintenance performance indicators provide an important feedback to improve the company maintenance policy implemented in MBI. A comprehensive reporting function is included in MBI to support the company management in revising the maintenance policy in order to achieve new objectives. Both technical and economic data are taken into account when evaluating the effectiveness of the adopted maintenance policy and considering possible improvements.
In order to effectively tackle the tasks listed above, MBI functions are structured in two levels:
the operation support level, that includes all the functions to support the work of the staff responsible for the maintenance process at local level (on the territory);
the decision support level, that includes the tools allowing the organization to evaluate the effectiveness of the overall maintenance policy.
As an example, the next section will describe the function of ” Proposing maintenance activities (i.e. maintenance and repair actions, inspection visits and technical investigations) to be carried out on the grid components”, which is included in the operation support level.
Proposing maintenance actions
This module implements the condition based maintenance criteria that are the core of the maintenance policy endorsed by the company. The module includes a set of maintenance models implementing the technical/economic rules leading to the maintenance decisions. A maintenance model has been developed for each grid component type. The component types considered by the MBI-substation system are the following: circuit breaker, transformer, disconnector, surge arrester, CT (Current Transformer), VT (Voltage Transformer), condenser, protection devices, auxiliary systems, SF6 bay. The model output consists of a list of suggested maintenance and/or monitoring actions to be carried out on the grid component. Each action is associated with the proposed deadline. The figure below represents in graphic form the information flows involving the maintenance models.
In the central part of the picture the different pieces of input information considered by the diagnostic model are shown, namely:
- The grid component status (see the left hand side boxes), i.e. all the maintenance related data for each specific component collected during its lifetime (e.g., component age and life spent in operation, findings collected during surveillance controls or through on-line condition monitoring systems, number and date of the overhauls, malfunctions/faults detected). The component status is continuously updated with the input data considered by MBI. The component status is the main source of information for the diagnostic models; nevertheless the models can be executed also when the component status in incomplete: in this case the maintenance model results will be less accurate but still available.
- The relation between the component and the grid functionality. This information accounts for the impact of a component failure on the grid service.
- The environmental conditions of the area where the component operates (e.g., industrial pollution and salt from the sea).
The maintenance model produces its results by comparing the component status with a set of reference (target) values, that depends on the component characteristics (see the box in the lowest part of the figure). The reference values are defined taking also into account the information given by the component manufacturer. Thus, all the specimens of a given component model share the same set of target values, according to the concept that maintenance criteria must be as far possible customized on the technical characteristics of the component. A peculiar characteristic of MBI is that both status and maintenance policy of a component also depend on other components. In fact, MBI utilizes diagnostics data of some components to evaluate and carry out diagnostics of other components having different monitoring parameters.
Each maintenance model is implemented as a set of rules that represent the relations between component malfunctions and their underlying causes. The rules also define the maintenance/repair actions to overcome a given malfunction and/or to fully restore the component functionality. As an example, in Figure 2 some rules taken from the maintenance model of the circuit breaker are shown.
In the left hand side of the figure some degradation findings are reported. Some of them are detected during visual inspections (e.g.” seriously oxidized points in the control cabinet”, “insulator degradation: surface discharges”), while others can be sensed by an on-line condition monitoring system (e.g. “abnormal compressor loading time”). Each arrow represents a rule that relates the findings to the maintenance /repair/replacement actions (right hand side). For instance, rule n° 1 states that the breaker control cabinet has to be overhauled when some seriously oxidized parts in the cabinet are detected. Rule n° 4 suggests to carry out an aggregated maintenance action (“breaker major overhaul”) instead of three separated maintenance actions (“control cabinet revision”, “air compressor revision”, “breaker pole replacement”) resulting from the previous rules.
The diagnostic model for a specific component is executed in the following situations:
- whenever a visual inspection or a diagnostic test make available new findings about the component;
- whenever an on-line condition monitoring system detects a specific event that has to be analyzed by the diagnostic model rules;
- at a fixed time (e.g. monthly) in order to periodically check some cumulated indicators accounting for component degradation processes or possible incipient failures. Should the indicator value exceeds a predefined threshold, MBI schedules the relevant condition-based maintenance actions.
Finally Figure 1 also includes the inspection planner, i.e. the module in charge of planning the inspection visits. It has to be noticed that the planning rules implemented in this module refer to the same pieces of information about the component (e.g., the component status) taken also into account by the maintenance models.
The maintenance models have been implemented by resorting to the expert system technology that proved to be very effective for this aim. For each grid component type a specific knowledge base implementing the relevant maintenance model has been developed; the inference mechanism that executes the knowledge base is instead the same for all the models. The developed knowledge bases turned out to be easily maintainable and self-explaining. The achieved result can be considered as a significant step forward in the formalization, management and capitalization of knowledge about grid maintenance, a typical “intangible” company asset.
In order to avoid data duplication and limit the costs for the implementation of the input data flows, MBI is strictly integrated with the IT systems already available in the organization. In particular, specific on-line asset monitoring systems installed on critical grid components compute and make available to MBI a set of diagnostic indicators that are evaluated by the maintenance models. The increasing availability of data automatically collected by on-line condition monitoring systems enables a large adoption of the condition based maintenance practice.
The large-scale adoption of the MBI system and the condition-based maintenance practice fostered by the system resulted in the following benefits:
creation of a centralized repository containing the distinctive company knowledge that governs the maintenance process applied to the whole grid;
reduction of the frequency of periodic inspections
, without compromising the overall availability level of the network;
adoption of standardized, certified and uniform maintenance criteria within the whole company;
optimal decisions for asset management for the renewal/substitution of critical components and/or subsystems (e.g., one or more bays, a substation) identified by MBI on the basis of the continuous evaluation of their technical condition.
After the system extension in 2004, two different implementations of MBI are now in use: MBI-Substation, aimed at substation maintenance, that takes care of more than 450 substations (including more than 5000 bays and about 50.000 primary equipment), and MBI-Lines, that is applied to more than 60.000 km of HV overhead lines.
This article is linked to the following knowledge articles: