【正文】 of air, gas, steam, and vacuum. Ultrasonic noise can be emitted as a result of friction between moving parts. ? Partial Discharge Detection This technology employs an electrical sensor to detect the initial insulation breakdown in electrical equipment such as insulators and terminators. Partial discharge detection is used to detect incipient failures before significant damage occurs. ? Transformer GasinOil Analysis This is needed to keep the transformer online as much as possible. One indicator of abnormalities is the dissolved gas content in the transformer oil. Certain gas levels can indicate aging, the need for maintenance, or potential failure. ? Infrared Thermography Thermography surveys involving the use of an infrared camera to detect hot spots in large motors used in power plants. ? Sound Intensity Measurement Sound Intensity Meter is needed to identify potential problems in equipment and record the historical changes in sound and output of equipment For transmission and distribution systems, sensors such as transformer fault gas analyzer might prove to be beneficial. This device provides realtime measurement of the four key gases associated with fault currents in transformer: carbon monoxide, hydrogen, acetylene, and ethylene. The next step is to incorporate an additional sensor to detect the presence of moisture which can reduce dielectric strength and lead to failure. This will be used with another device that measures the transformer loading so that the evolution of key gases and moisture can be characterized as a function of the load. With the moisture sensor and load current monitor we can develop accurate criteria for loading transformers under stressful conditions rather than having to rely on the overly conservative ratings now provided. INTEGRATED APPROACH An integrated approach for transmission and distribution systems would ensure that equipment or subsystems leading to serve a particular load would receive uniform and consistent level of maintenance in all departments, thus enhancing and optimizing the maintenance process. In order to establish a maintenance program, the RCM process needs to be the driving point. Figure 2. Shows different strategies required for a maintenance optimization program. The classical 13 RCM process involves identifying the systems to be studied, their functions, functional failures, failure modes, failure causes, and the maintenance task selection The most critical pieces of equipment which affect the overall function of the system need to be identified. In other words, we need to identify equipment with severe consequences when failed. Also, critical customers and the equipment leading to their loads need to be identified. All equipment affecting these customers need to be analyzed, and the most critical pieces of equipment should be determined. We also need to understand the customer needs in terms of reliability, safety, power quality, cost, etc. These attributes must be weighted to determine the optimum maintenance policy to deal with these customers. In order to optimize the utilities’ maintenance program, all relevant information must be used to most effectively initiate, schedule, track, record, and analyze maintenance tasks. An open munication protocol that enables various monitoring devices (regardless of the manufacturer) to talk to each other, with utilityoffices, and with control centers would best suit the needs of electric utilities by integrating various data sources and other software modules. Currently, many large utility organizations perform some equipment diagnostic tests. Many have puterized their maintenance work management functions including retaining maintenance histories, logging significant operational activities, and maintaining a library of design information for equipment. The proper munication, integration and analysis of all of this information would result in more accurate remendations concerning when to perform maintenance and/or how to operate a specific piece of equipment. A typical problem in the industry is that, at times, a large amount of data is collected but it is not integrated and processed expediently for a quick evaluation. This results in the inability to make immediate Oamp。M) costs are top priorities for electric utilities. As system equipment continue to age and gradually deteriorate, the probability of service interruption due to ponent failure increases. An effective maintenance strategy is essential in delivering safe and reliable electric power to customers economically. The objective of this paper is to provide a framework for