T&D World Magazine
Optimized Maintenance Edinor

Edenor crews perform major on-site repair tasks on a 300-MVA transformer.

Optimized Maintenance

Argentine utility reduces power transformer outages and increases availability with a proactive strategy.

The development of electricity distribution networks is influenced by several factors, including increasing demand, a reduced level of redundancy and operating restrictions limiting maintenance outages. Regulatory frameworks require strict technical requirements and severe fines are imposed for noncompliance, so utility revenue is linked to network availability.

Utilities operating in this environment regard power transformers as critical physical and capital-intensive assets. They have a strategic role in the operation of distribution networks because of the huge overall investment in the capital, operational and maintenance costs of these assets during their operational life span.

Edenor is the largest electric distribution utility in Argentina in terms of customer numbers (2.7 million) and annual energy sales (20,769 GWh), with some 2,930 employees. The utility operates a network that extends to 36,462 km (22,657 miles) and includes 73 high-voltage transformer substations with more than 200 high-voltage power transformers, having a total installed transformer capacity of 15,000 MVA. The voltage levels of the transformer population range from 132 kV to 500 kV.

The average age of Edenor's transformer population is 25 years, with an annual failure rate of 2% based on major failures that require either replacing the unit or making major repairs in the field. The use of proactive techniques becomes crucial for comprehensive maintenance management of the power transformer population in any distribution utility. This requires focusing on all available resources and management tools that will help to minimize unplanned outages and reduce the frequency of network downtime for maintenance to optimize network performance and reduce operating as well as maintenance costs.

Optimized maintenance, Edenor
Key drivers of the data management model.

Proactive Strategy Approach

Edenor bases the maintenance strategies it applies to power transformers on several key drivers:

• Intensive use of periodic checklist inspections

• Development of suitable diagnostic determinations

• Systematic monitoring of results

• Performing, when necessary, additional or major maintenance tasks.

The utility has placed a strong emphasis on the condition assessment of its oil-paper insulation system, tank and main accessories. High-voltage bushing and on-load-tap-changer (OLTC) failures are responsible for a large number of major transformer outages, so the utility pays close attention to the condition of these components. The condition of the tank, oil leakages, and proper functioning of OLTCs and cooling systems are also evaluated from periodic technician inspections.

To assign the most appropriate maintenance activities, Edenor defined different work programs:

• Execution of tasks

• Predictive analysis

• Measure with the definition of acceptable thresholds.

The results obtained are used as a basis for assessing the paper-oil insulation system condition and determining the need for major maintenance tasks. When necessary, such determinations are complemented by the execution of additional electrical measures to delve further into condition diagnosis.

Internally developed by Edenor personnel, this data management model is strongly supported by a number of assessment and monitoring tools. They represent a tool kit, essential for the results of the data management model developed, and are considered a core factor in the success of a transformer’s condition assessment for defining further maintenance activities.

Insulating Oil Analysis

The insulating oil samples can be obtained from an on-line transformer in a relatively short time with reduced resources. Evaluation of this sample provides a general profile of the transformer’s condition, detecting incipient faults or just monitoring the condition of the unit. The state of the insulating oil is widely considered a witness to assess the transformer’s internal condition.

Therefore, the oil analysis forms part of the routine maintenance activities executed. At Edenor, more than 1,000 oil samples are taken from power transformers annually to perform physical-chemical analysis (PCA) and dissolved-gas analysis (DGA), considering regular predictive maintenance activities and control of units under special monitoring.

Periodic PCA (dielectric strength, water content, dissipation factor, neutralization number, interfacial tension) identifies the evolution and degradation processes in the oil-paper insulation system and their evolution over time. The use of DGA enables the detection of electrical, mechanical and thermal faults. Furan analysis provides additional helpful information about the aging process in the paper insulation; where possible, this information is supported with degree-of-polymerization values.

The use of DGA in the field by means of portable equipment is a valuable tool to obtain test results on site in a few minutes. This is useful for control units in critical condition or to evaluate emergency situations more quickly, without having to take the sample to a laboratory.

Optimized maintenance, Edenor
Physical-chemical analysis of insulating oil in the laboratory.

Data Management Process

The appropriate management of information is critical for the success of any maintenance strategy. Thereby, the maintenance management and decision-making processes are supported by an IT tool, whose core is a database where all equipment is inventoried. This software provides workflow functions, work order issuing and data sto­rage tracking, which help with the planning and programming of maintenance activities. Through this decision-support system, information such as the evolution of electrical measures and critical parameters can be determined from the data.

Oil PCA and DGA play a key role in this IT tool. An application shows a register of the oil physical-chemical parameters of the transformer fleet in addition to a detailed record of the different dissolved gases in oil. It is also possible to obtain a DGA transformer diagnosis automatically through different methods. Because the best use of oil diagnostics data is to determine trends together with all the other test data taken during the life of the transformer, this tool also provides an easy-to-access detailed record of all oil tests and events.

The water-in-oil content for different operating temperatures and loading conditions can be obtained, and using different methods, the probable level of water-in-paper content can be estimated. In a dynamic query linked to the main database, frequencies and priorities for the oil analysis are defined according to the different conditions and critical level assigned.

Other outputs are used in the decision-making process, extracting trends and patterns from the data collected as well as knowledge-based rules. This monitoring identifies ab­normalities, enabling corrective actions to be prioritized.

A working group of transformer specialists can then assess the internal health of the transformers, quantify their criticality and detect weak components, labeling their condition. With the supporting IT tool, the results obtained are evaluated systematically.

Optimized maintenance, Edenor
Technical risk index for a number of transformers.

Health Indices Development

Applying a relationship analysis tool, the dependencies of the evaluated parameters are calculated. As a result, an algorithm was developed, performing a correlation analysis among the obtained results and the transformer condition, to grade the state of every transformer, weighting such parameters to qualify the condition of the units. The algorithm ponders the condition of every unit, evaluating different variables and grading the condition of the transformers between 0 and 100 to qualify them from the worst to the best condition, to create health indices.

Such an algorithm considers the oil analysis, state of key components (high-voltage bushings, OLTC) and tank (including oil leakages), weighing each variable in a different way, according to their relative importance, as assigned by the expert opinion of the utility’s maintenance specialists, and complemented, when needed, by additional electrical measurements and tests. Therefore, different condition levels are established.

The numeric classification (0 to 100) is finally ranked in three critical levels: low, medium and high. This allows the utility to identify the technical risk of its transformer population.

Because of the wide variety of types, models and ages of units in service, different patterns of abnormalities and failures are identified, defining probable aging models. In transformers of the same design and age, by the relative benchmarking of their indices, typical aging patterns are outlined. From the information obtained, a risk index of the whole transformer fleet can be represented in a 2-D graph with the health indices and the number of transformers on the two axes. This tool helps to identify, at first sight, the general condition of the transformer fleet.

The overall results allow labeling the condition of the transformer population in terms of defining the need for additional or major maintenance actions to reduce levels of risk.

Optimized maintenance. Edenor
Risk matrix of the transformer fleet.

Risk Matrix

A risk consists of two different aspects, the probability an event will occur during a time interval and the consequence of such occurrence. Edenor defined its risk matrix as a product of the technical risk (criticality) and the relative importance (consequences) of the physical assets in its high-voltage grid.

Considering the level of criticality of the transformers and their relative importance in the grid, a map of risks of the high-voltage network was developed, identifying levels of criticality of the installations. Therefore, priorities for transformer maintenance are defined and activities addressed considering the risk assessment of the high-voltage network as a whole.

Tasks to reduce levels of risk include refurbishments, replacement of components with a history or pattern of failures detected (for example, high-voltage bushings), technological upgrades (for example, OLTCs), treatment processes (drying out, oil filtering, degasifying or regeneration) and so on.

Also, the relocation of units is performed, placing transformers considered not reliable enough in areas of lower criticality in the network. Units with higher critical levels are replaced and moved to areas of reduced risk levels, minimizing the risk of faults and increasing the reliability and availability of the installations.

Such a dynamic matrix of risk, which undergoes periodic revision and updating, makes it possible to identify potential areas of high criticality quickly in order to define major maintenance decisions with a rational allocation of resources.

Optimized maintenance, Edenor
Edenor technicians are in the process of replacing a 230-kV power transformer bushing.

Development of Personnel Skills

A central issue for the success of this management model is to count on highly qualified man­power as well as the expert knowledge of specialists in this matter. Therefore, Edenor focused on preserving the required expertise concerning power transformer maintenance tasks. As a result, a strong emphasis was put on the train­ing of very specialized in-house working teams. By periodically reviewing and updating knowledge and skills, the best maintenance practices are highlighted for continuous improvement.

Evaluation of the condition of the transformer population relies on the knowledge base of an expert working group. This includes broad experience and up-to-date knowledge on state-of-the-art transformer design, operation and maintenance, and tests to strongly support the evaluation of results for making maintenance decisions.

Thus, a best value was added to the maintenance activities, not only in a profitable way but also through seeking the additional advantage in terms of quality, performance and service improvement. Overall, the use of this model has resulted in the more effective use of Edenor’s resources.

Positive Results

The aim of Edenor’s data management model was focused on improving the maintenance management of power transformers using staff, technical and economic resources to increase simultaneously the reliability and availability of such physical assets. This model, which has now been in operation for more than 10 years, is supported by a strong tendency toward the intensive execution of product data management (on-line and off-line) activities to assess the condition of the equipment and the use of risk-based maintenance techniques.

The extended use of this data management approach has led to positive results regarding the improvement of Edenor’s maintenance management, reducing transformer outages and increasing the availability of the network, making it possible for the utility to reduce its maintenance costs.

A survey of several key performance indicators (for example, number of unplanned interruptions, programmed outages, maintenance costs, quality of supply) confirms the success achieved by the model’s use in the efficiency of maintenance management as a whole.

José Luis Martinez ([email protected]) is a senior electrical engineer at Edenor, where he manages the maintenance of high-voltage substations. Martinez has more than 25 years of experience in the maintenance of high-voltage power equipment and is a specialist in the operation and maintenance of high-voltage power transformers. In his position, he also acts as a consultant, providing technical support on these subjects to different transmission and distribution utilities. He actively participates in CIRED and other international conferences on electricity distribution.

Mentioned in this article:

Edenor | www.edenor.com.ar


Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.