ELECTRICAL MAINTENANCE

Refining Assest Management by Integrating Condition Monitoring and Work Management

Faced with deregulation and growing competition, many utilities (and industries) are seeking to maximize the value of their existing assets by leveraging new technologies to optimize Operations & Maintenance activities. One of the most successful maintenance strategies is a Condition-Based approach to performing maintenance, which utilizes data collected from periodic inspections, testing, and Predictive Maintenance (PDM) technologies to determine the optimum maintenance strategy. Contrary to the traditional time-based maintenance approach, Condition-Based Maintenance (CBM) is a process that utilizes monitoring and diagnostic data to drive the maintenance decision process. The process of collecting equipment condition data, converting that data to information, and taking action based on that information is a vital part of implementing an effective asset management program. This article will define the role of condition monitoring in today's asset management strategy, and how data management techniques can be integrated with work management to facilitate the implementation of a Condition Based Maintenance program.

Redefining Asset Management
Historically, asset management has typically focused on planning, scheduling, and executing maintenance tasks using a Computerized Maintenance Management System (CMMS). Although this approach has proven successful in how managing preventive and corrective maintenance work gets done, it is limited in determining what maintenance needs to be performed, and when it needs to occur. In order to effectively determine what maintenance needs to be done, utilities have implemented more advanced maintenance strategies like Predictive and Proactive Maintenance which collect and analyze monitoring and diagnostic data to assess equipment condition. By integrating these condition monitoring techniques with the Work Management process, a more effective asset management strategy can be executed, a strategy where condition data drives maintenance decisions. This process is defined as Condition Based Maintenance (CBM).

Understanding Condition Based Maintenance (CBM)
Condition Based Maintenance integrates and utilizes inspection, testing, and PDM diagnostic data, operator logs, and maintenance history to assess equipment condition and make more effective decisions regarding how equipment is operated and maintained. The goals of CBM are to increase equipment reliability, reduce operating and maintenance costs, and maximize equipment performance and production. It is important to realize that CBM is a process, not the application of advanced technologies. The CBM process involves efficiently collecting data, converting that data into useful information, and taking action (where needed) to ensure business goals are achieved.

In a typical power plant, there are many different data collection/equipment monitoring processes which are being employed to assess equipment condition: operator rounds, periodic surveillance and performance testing, PDM diagnostics (vibration, oil, and infrared analysis), electrical testing (resistance, power factor, etc.), instrument calibrations, the use of online monitoring systems (DCS), and many others. Unfortunately, many of the these data processes are manual and utilize paper-based inspection and testing forms to collect information; others are electronic, but do not make the data readily available to the people that need it to make operating and maintenance decisions; and some processes are not documented at all. At best, this equipment condition information resides in islands of information scattered throughout the enterprise in independent databases and spreadsheets. In order to effectively utilize this data to determine maintenance needs, these data sources need to be integrated in a central data warehouse, with standardized business processes for managing the data and ensuring this information is integrated into the work management process.

Data Management vs. Work Management
Work Management Systems (CMMS) have been utilized for years in the utility industry. These systems focus on the effective planning, scheduling, and executing of work orders to perform Preventive and Corrective Maintenance activities. They track labor hours, parts, purchasing information, maintenance costs, the results of maintenance activities, and other work management information. Inherent in these work management systems are business rules and processes to effectively execute how maintenance is performed.

In contrast, Data Management Systems collect, analyze, trend, and manage Condition Monitoring data including process data (pressures, temperatures, flow), inspection and log data, testing data (resistance readings, calibration data, performance data), and PDM diagnostic data (vibration, oil analysis, infrared thermography, ultrasonic). Data Management Systems rely on data mining tools like statistical analysis, data integration techniques, and advanced analysis tools to take raw data and convert it into useful information. They also have the capability to automatically notify personnel when certain conditions exist, and allow plant personnel to document analysis results and decisions made regarding actions taken.

Essentially, Data Management Systems are designed to assess and manage equipment condition and determine what maintenance is required and when. In a CBM environment, the Data Management System is a front-end system that drives the Work Management System; a process where data drives decisions; a process where work orders are triggered based on condition data (see Figure 2). By integrating the data management process with the work management process, an improved asset management strategy can be implemented; a strategy that supports the CBM process.

Automating the CBM Process
There are many different monitoring and diagnostic activities that are typically performed in a power plant to assess equipment condition. Many of these processes are manual, where the condition data is recorded on hardcopy inspection sheets. For example, most utilities have operator rounds that collect process variables from existing field instrumentation, check oil levels, and test the operation of equipment. Typically, operators either utilize paper inspection sheets to record the data/results of their rounds, or they do not record the data at all. Other work groups perform periodic testing of equipment like 'meggering' motors, utilizing preconfigured forms to record the actual test results. In some cases, the CMMS system may be used to document labor hours and the overall results of the inspection or test, but the actual testing data is usually stored on hardcopy inspection/testing forms. In both of these cases, the condition data is typically filed away, rarely to be used again for trending, analysis, or comparison with other condition monitoring information.

One technology which has proven effective in automating these processes and maximizing the utilization of this data is the use of Automated Data Collection (ADC) techniques. ADC uses mobile data collection devices to perform periodic inspections and testing, electronically recording the condition monitoring data in the field at the point of collection.

Once data is collected in the field, it can be uploaded into the central Data Management System for alarming, trending, analysis, and integration with other condition monitoring data. The use of bar-coding technology can also be used to streamline the data collection process and eliminate the need for plant personnel to manually type in descriptive comments when problems are identified in the field. The benefits of utilizing the ADC technology can be significant: increases in productivity of 10-15%, reduced data errors, improved data accuracy and integrity, and an increased utilization of data for maintenance decisions.

Of course, the use of ADC techniques alone does not result in an effective CBM process; it only helps to streamline the first step of the process, that being efficient data collection. Once the data is obtained electronically, analysis and data mining tools are required to make information out of the raw data. One key requirement for effective data analysis is the ability to notify plant personnel when adverse conditions exist. In a typical CBM program, there are literally millions of data points that may be collected every year to assess equipment condition. Most utilities do not have the resources required to review all of this data and find potential problems. Thus an automated alarming and notification strategy is required to "identify and push" potential problems to the people that need to know. Using advanced tools like statistical analysis, multi-point cross trending, threshold alarming, and knowledge-based decision rules, helps take the data and convert it into useful information. Using this approach, plant personnel do not have to review all of the data to identify degraded trends or conditions. The system will automatically analyze the data and notify them via email or paging when certain conditions exist. This automated analysis capability minimizes human intervention and helps focus resources on the problems that really require investigation. Less time is spent trying to find potential problems, and more time is spent resolving the issues that warrant attention.

Integrating Data Management with Work Management
Condition Based Maintenance is a process of collecting data, converting that data into useful information, and taking action based on the information. In order to effectively implement a CBM strategy, the Data Management process must be integrated with the Work Management process, to create the Data-to-Information-to-Action model.

Ultimately, these two processes can be tied together, so that work orders can be triggered based on the analysis results of the condition data. For example, assume an operator records a bearing temperature that exceeds some pre-determined alarm limit during his normal equipment inspection. At the initiating stage, this data point is essentially an "unanalyzed" alarm condition; it is data, not information. In order to effectively analyze this condition and determine what action (if any) is required, plant personnel may:

1. Analyze data trends and statistical parameters;
2. Review other open issues or alarms that may exist on the equipment (from other monitoring and diagnostic activities);
3. Review equipment fault history records (alarm history, maintenance history);
4. Compare data from the equipment in alarm to a similar piece of equipment (i.e. same manufacturer, same application);
5. Perform cost benefit analysis to assess financial/risk factors associated with various action scenarios.

Instead, the data management process should be linked with the work process such that work orders are only triggered after the condition has been adequately analyzed.

In cases where the analysis of condition data indicates maintenance action is required, work orders can be triggered automatically by passing critical data/information from the data management system to the CMMS. Once the work order is initiated, the work management process typically takes over; the work order is planned, scheduled, and executed as part of the normal work managementlCMMS process. Once the maintenance is performed and the work order is closed, information from the CMMS (i.e. what maintenance was performed) can be transferred back to the Data Management system.

Internet Utilization for Asset Management
The emergence of the Internet technology has opened up new methods for utilizing and sharing information within the utility and industries alike. Today, using Internet-based applications, plants can share reliability and diagnostic information, network plant resources, communicate experiences and knowledge, and maximize the utilization of information throughout the enterprise. In some utilities, corporate resources have actually been consolidated into central "diagnostic" teams that provide technical analysis and decision support for multiple plants in a given geographic region. Internet technology has also opened up new deployment models for business applications, which has significantly lowered the total cost of ownership for these highly advanced information systems; systems that are the backbone of today's business infrastructure.

The emergence of Application Service Providers (ASP) has provided utilities with low-cost methods for utilizing advanced technology without having to make a significant investment in upgrading their infrastructure. ASP's essentially "rent" their applications to utility customers and use the Internet to deploy it. No hardware, software, or IT support is required at the utility level, which significantly lowers the cost of deployment. Access to the system and all of the information it manages, is obtained anytime, from anywhere, using a standard Internet browser.

Under an ASP model, this requires a minimal monthly fee for application hosting and data warehousing. Best of all, the application can be deployed within a couple of weeks, not a couple of years like traditional client/servers applications.

In the CBM environment, the ASP model allows utilities to integrate their plants and share condition monitoring information and maintenance experiences, to help achieve operational and maintenance excellence.

Conclusion
As deregulation and increased competition drives utilities and industries to become more performing, business processes must be critically evaluated to identify the value-added components that help the organization succeed. Condition Based Maintenance is a proven process that utilizes monitoring and diagnostic data to optimize maintenance, and improve the operational effectiveness of plant assets. The process of collecting data, converting it into useful information, and taking actions based on the information, is the foundation of the CBM approach.

When implemented effectively, significant benefits can be realized from CBM:

• Reduced O&M Costs
• Increased Equipment Reliability
• Improved Process for Work Identification
• Optimization of Maintenance Tasks
• Reduced Failures
• Increased Utilization of Data

As a result, it is not something that is implemented once. It is a continuous process of improvement; a living program that never ends.

Jean-Marc Demers is with ABB Inc. He can be reached at jean-marc.m.demers@ca.abb.com ET