ALL UTILITIES FACE PRESSURES to increase efficiency while meeting higher expectations for service and reliability. Many are examining how new maintenance strategies can increase reliability and avoid unnecessary costs.

The traditional “brute force” method of equipment-intrusive, time-based periodic maintenance is a manpower-intensive approach. In addition to the inherent costs of this methodology, the unnecessary periodic maintenance can itself contribute to future failures because of the opportunity to create additional problems during operational switching, outage hold offs and planned maintenance.

For these reasons, many utilities have been modifying their maintenance strategies, first moving to reliability-centered maintenance, then subsequently to sensor-enabled, condition-based maintenance in order to meet the simultaneous needs of increasing reliability and efficiency.

KEY ATTRIBUTES FOR CONDITION MONITORING

Monitoring trigger maintenance is effective only if:

• Cost-effective sensors are available.

• An appropriate algorithm exists to trigger necessary action.

• A communication medium exists to deliver information to maintenance and operations decision makers.

• Appropriate filtering, prioritization and information-delivery mechanisms can reduce the human burden in deciding which data must be acted upon (for large-scale, broad-based condition-monitoring programs).

  Additional key attributes should be considered and may include:

• Sensing for anomalies versus absolute fixed values. Having an algorithm track and trend differences between similar equipment can often be more effective than trying to dynamically adjust alarm triggers that might be effected by seasonal or loading changes.

• Using system-level data versus component-level data where possible to avoid expensive monitoring at the device level.

KEY ATTRIBUTES FOR SOFTWARE TOOLS

Since device data collected at the local or system level may be a valuable input to multiple algorithms with different purposes and functions, software tools are needed to collect, integrate and prioritize data and to allow algorithm triggers to share information, communication and decision logic as appropriate.

Key attributes of software tools that implement condition-based maintenance include open architecture, interoperability, modularity and are user configurable with common interface architectures and languages.

LINKING CONDITION MONITORING TO WORK ORDERS

The choice of which element of condition monitoring should be implemented next should be based on the incremental value it provides. An early key to successful broad deployment of condition-based monitoring is linkage to the work-order system, so that successive deployments are providing value by effectively facilitating resource application where the highest number of dollars or the largest number of man-hours are currently being spent on an annual basis.

An additional value that can be achieved by a two-way linkage is the automation of work processes. Consolidated Edison of New York is currently using software tools to allow operators to assist substation personnel in completing job plans (rounds) and to improve data collection input into the Computerized Maintenance Management System (CMMS), referred to as MAXIMO.

This software is designed to support field data collection via a handheld computer (HHC) as well as desktop computers. Using an unhardened HHC device, substation personnel can download their periodic compliance work orders via a LAN connection (including readings taken to date and current completion status), accept any additional round entries entered into the HHC, and subsequently resynchronize to the LAN to upload the changes. This software automatically tracks and changes the status of these completed work orders. Maintenance Management Workstation (MMW) is then used to report the data, track it, send e-mails (if required) and publish the results on the company website for review by authorized personnel.

The company is expanding the system to include periodic maintenance activities as well. More urgent alarm-trigger functions can automatically trigger BlackBerry e-mail or pager messages to communicate with the operator even when in the field doing switch operations or traveling between stations.

Once you identify your biggest business concerns based on annual dollar and manpower expenditures, determine if any algorithms or alerts can help mitigate the annual cost impacts by making better, more timely decisions and moving toward more precise “just-in-time” maintenance that avoids catastrophic failures and the potential for associated ancillary damage.

CIRCUIT-BREAKER MONITORING CASE STUDY

As just one example of what can be done with focused data mining, our Substations Operations (SSO) group reviewed circuit-breaker operational data for routine maintenance hold off and automatic breaker operations as evidence of a completed operational test. This data identified three breakers that hadn't had an operation within a year. Upon inspection, one of the three breakers would not have operated if called upon, enabling Con Ed to prevent a potential over trip.

This is just one example of why, right after transformers, circuit breakers are the most significant burden to both maintenance and testing in substations, and a major contributing element to operational reliability. Their importance in maintaining normal paths of service is further increased due to their vital roles in outage hold offs, operational switching and protective operations to isolate faults in response to relay protection demands. All of the diverse service roles in which circuit breakers are used increase the different ways in which circuit breakers could potentially fail.

Among the critical functions is operational timing of the circuit breaker, which is the largest contributing element to overall fault clearing time. The ability of the system to take “timing shots” as they occur during normal breaker operations reduces what would be a major separate test burden and, depending on the particular degrading condition, might allow for earlier identification of the symptoms of incipient failures.

FIELD DEMONSTRATION

To this end, the Electric Power Research Institute (EPRI; Palo Alto, California, U.S.) and Con Edison, working together with a vendor, developed and deployed a field demonstration of a wireless circuit-breaker monitoring system on four 138-kV circuit breakers at one of Con Edison's urban substations.

A wireless technology was chosen to avoid the costs associated with extensive wire routing. The obstacles to a successful wireless design were anticipated to potentially include physical obstructions within the station preventing direct line-of-sight communications, background levels of interference from external urban electronic traffic such as radios and mobile-phone networks, and the constant electric fields associated with high-voltage equipment. Potential interference from high-voltage switching was also recognized as an even more substantial potential interference, but not critical for a condition-based monitoring application where occasional short interruptions of data transfer are not a problem.

To mitigate potential problems, Con Edison chose an architecture where each individual circuit-breaker monitor communication assembly could act as both a transmitter and receiver (wireless sensor networks [WSN] that utilize ad hoc, redundant, automatic message-routing protocols implemented on low-cost hardware including data-acquisition circuitry), effectively networking multiple communication paths.

A PC was connected to the wireless network and ran software that collected data from the sensors and saved it in an appropriate format. Initial field trials were conducted that compared two alternatives: Amplitude Shift Keying (ASK) and Discrete Sequence Spread Spectrum (DSSS) WSNs.

The DSSS system showed greater resilience in Con Edison's demonstration trials. Laboratory tests on elements of prototype systems also took place in the vendors' high voltage and environmental laboratories. These as well as field trials at the substation proved successful, and a tentative specification was developed for a practical post-demonstration system.

WIRELESS DATA TRANSFER

Since loss of signal (LOS) problems were seen in the lab setting, several WSNs were investigated. WSNs are new technologies that take the place of universal wireless transmitters. They offer a solution to the LOS problem by providing message relaying between devices, thereby allowing data to be routed via any nearby device and to automatically find a route to its destination. By placing devices around the substation, the LOS problem is readily solved. Provided that the cost is acceptable, the WSN approach should play a prominent role in the final system architecture.

Unfortunately, the communication equipment and technology installed at the substation used early prototypes and are no longer commercially available. The proprietary protocol used in the project has evolved into IEEE 802.15.4. Con Edison learned some valuable lessons that can now be incorporated into a future wireless monitoring project. For example, we successfully demonstrated that we can get a signal from any location in the yard back to the control room and that we can take meaningful measurements at a reasonable cost.

SEEKING PARTICIPATION

Con Edison and EPRI have decided to initiate a new project to resolve the communication technology issues that prevented follow-up deployment of the original prototype. We welcome participation and broader-based specification and design inputs from other utilities, to help assure design acceptability for more potential users and contribute to market demand for successful commercialization.

DEMONSTRATION PROJECT WITH WIRELESS SENSORS

Another demonstration project launched at one of our substations is deploying prototype wireless universal pressure and temperature sensors on a circuit switcher. For the trial, Con Edison installed wireless temperature and pressure sensors, a remote transceiver, a local receiver, a cell modem and a data-retrieval computer. There are three temperature sensors connected. Each of these sensors replaced a washer on the bus of each phase of the circuit switcher. An ambient temperature sensor was also installed locally. A pressure sensor is connected to the common manifold to measure the switcher's SF₆ gas pressure. Signals are sent out periodically with an average time between data packets of about 20 minutes. There is a remote transceiver mounted at ground level near the transmitters that receives the wireless transmitter signals from the circuit switcher and sends them to the local receiver set up in the LAN room in Con Ed's Hell Gate reporting center. Software is installed on a computer at Hell Gate that allows remote retrieval of the signals received.

The data itself is easily viewed using Microsoft Excel. This equipment can be used to record and trend gas pressures and excessive surface temperatures caused by hot spots.

FUTURE EFFORTS

EPRI's MMW is being used to support data base compilation and analysis of cable tests and forensics at the ECTN Cable Center for Excellence. Con Edison intends to start an industry failure database initially focusing on organizing the company's substation equipment data for internal purposes. Con Edison also plans to automate reliability data input on an outage-by-outage process basis, so that its transmission PRA model will be up to date and meaningful on a continuing basis. With EPRI, Con Edison has developed a conceptual asset management model that crosses organizational silos (e.g. life-cycle management of equipment) and plans to implement applications in a bang-for-the-buck, step-by-step, value-added sequence, linking applications together to increase their power and functionality.

In the longer term, as the utility increases the sophistication of decision making and moves the asset management program from being data driven to model driven, the models (e.g. configuration management of system topography) will be shared across different applications for ease of configuration management. EPRI continues to research the use of wireless communication links in substations in its Integrated Monitoring and Diagnostics project.

THE ULTIMATE GOAL IN ASSET MANAGEMENT

Asset management — more recently referred to as resource optimization — is an attempt to take a step toward broader-based enterprisewide decision support. Asset management will provide tools and methods to support the ability to make informed decisions regarding substation assets and manage costs, benefits and risk. There are many utility decisions that depend on information within traditionally separate silos of operation.

EDITOR'S NOTE

The results of this effort were presented earlier this year at EPRI's annual Substation Equipment Diagnostics Conference. The next event will be held in San Diego on July 17 — 19, 2006.

---

Pat Duggan is an R&D program manager at Consolidated Edison Company of New York, where he is responsible for engineering-, operations- and maintenance-related R&D projects for various substation, transmission and system operations organizations. Since joining Con Edison in 1968, Duggan has held engineering and management roles in nuclear power instrumentation and controls, project management, construction and electrical engineering, generation and controls. He earned his BSEE degree from Manhattan College in Riverdale, New York. dugganp@coned.com

Theo Theodorou is a manager with Consolidated Edison Company of New York in the Substation Operations Department. He is responsible for the computer-management systems and procedures that are used in all construction, maintenance and testing activities within Con Edison's transmission and area stations. Since joining Con Edison in 1988, he has held various management positions in substations, system operation and construction. Theodorou earned his BSEE degree from Polytechnic University in Brooklyn, New York.

Philip DiScala is a senior specialist with Consolidated Edison Company of New York in the Substation Operations — Maximo/Asset Management Department, where he is responsible for supervising and administrating computer management applications. Since joining Con Edison in 1974, DiScala has held various position in demand side management, energy services and commercial services. He earned his BS degree Mercy College in Dobbs Ferry, New York. discalap@coned.com

Patrick Di Lillo is a technical specialist with Consolidated Edison Company of New York in the Equipment and Field Engineering Department. He has extensive experience in the specifying, evaluating, installation, maintenance and root-cause failure analysis of high-voltage circuit breakers, disconnect switches and related substation equipment. Di Lillo earned his BSEE degree from Manhattan College in Riverdale, New York.

APPLYING CONDITION-BASED MONITORING TO HANDLING SF₆ GAS PROGRAM

The value proposition needed to support a growing condition-based maintenance program needs to create incremental value on a step-by-step basis. As just one possible example, the progression at Con Edison was generally as follows.

The Substation Operations (SSO) group started with a problem it had as a result of the company's early aggressive implementation of SF₆ equipment. Con Edison's first vintage equipment had seal problems, so to properly focus the company's continuing work on SF₆ recovery gas carts, SF₆ camera development and now improved SF₆ sealing methods, Con Edison had to locate the worst problems.

By monitoring “gas calls” and inventory additions, Con Edison was able to identify the worst leakers by area, station and specific equipment, and direct its programmatic efforts and maintenance, replacement, redesign (i.e., conversions to single pressure) and overhaul activities to be the most effective in managing its SF₆.

At the same time, this first application allowed the company to build expertise in using all the features available in the Maintenance Management Workstation (MMW). Since Central Engineering's root-cause evaluations identified a lack of LTC contact wipe (crossing the neutral tap) as a significant contributor to LTC coking buildup that could lead to failures, SSO started a program of routine data mining, identified LTCs that had not crossed neutral within a certain period of time, and scheduled exercising of tap travel through neutral on the off shifts to mitigate this failure contributor.

Working with the software vendor to develop an Autotour interface for MMW allowed Con Edison to use LAN synchronization of handheld devices to more easily assign and track completion of compliance inspection reports. Linking to the work-order system clearly identified the company's most important business problems (where it spent the most dollars and people).

Con Edison also developed a Planning and Resource Optimizer to deal with the “stutter-step” realities of “on again/off again,” interrupted and “switch-to-return” outages in the substation environment, allowing us to do schedule and resource “what-if” optimization.

As the latest step, Con Edison participated with EPRI and other utilities in the initial Asset Management project, which has developed a detailed overall functional model for asset management decision making. Con Edison is now working with EPRI and other utilities to identify and develop elements of this model that would be of most common benefit to participating utilities in the form of software tools.