By Ron Farquharson and Ken Caird
The electric utility industry has seen unprecedented changes in recent years due to deregulation of the industry. Not the least effected is the management of primary equipment assets throughout the utility. In particular, maintenance of primary equipment in substations, has frequently been delayed in order to reduce O&M costs.
At the same time the majority of the primary equipment in Western countries was installed before 1965 with an operating life of approximately 30 to 40 years depending a range of factors. In North America primary equipment is ageing at the rate of 0.7 years per year net of upgrades and replacements etc. Equipment is not only aging but is increasingly operated at higher levels -- impacting equipment life and possibly reliability. In the process of reducing maintenance budgets utilities have lost a lot of knowledge and expertise through downsizing and early retirements.
In the drive for efficiency primary equipment maintenance programs have been cut or rolled back. At the same time customers, power pools, regulatory agencies (and ISOs in the future?) are demanding higher and higher levels of availability and overall reliability. The electric utility apparatus engineer, stuck in the middle, seems to be in a 'no win' situation.
There are now a wide range of new (to electric utilities) methodologies and technologies available to aid the engineering team in dealing with these conflicting interests. The technology solutions tend to fall into one of three groups:
a. off-line monitoring devices
b. on-line monitoring devices
c. on-line diagnostics systems
In order to provide the most cost effective on-line solutions, monitoring data from these Intelligent Electronic Devices (IEDs) needs to be integrated onto an existing common platform using standard protocols and LANs. The preferred common platform appears to be the substation RTU or automation system as it is designed to perform this type of device interfacing, is likely already in place or is required for many other purposes such as protection, SCADA, metering, planning. Monitoring refers to the accumulation of basic data such as temperature or pressure thresholds. Diagnostics refers to the intelligent grouping of related data including statistically significant values and trends which are then processed through an expert system to generate a more complete knowledge of the condition of the equipment and recommended actions to be taken.
This article will provide a top level overview of the current options available to utilities including the application of reliability centred maintenance (RCM) programs, base line equipment monitoring sensors, basic equipment monitoring algorithms for use at the substation, IED integration issues for sensors, substation automation LAN technology necessary for equipment monitoring, advanced sensors and new expert systems for equipment diagnostics and predictive maintenance.
Background
Traditionally, utilities have used a time-based preventative maintenance system. Periodically the utility maintenance personnel would take the equipment out of service and 'tear the equipment down' for a maintenance check.
This time based approach had a number of drawbacks:
1. The equipment had to be taken out of service in order to perform diagnostic tests.
2. After diagnostic tests and 'tear downs' often nothing was found wrong with the equipment (i.e. wasted time and resources, which translates to higher costs).
3. Failures occurred shortly after the maintenance check (i.e. equipment was not put back together properly or 'infant mortality').
4. Limited ability to respond to certain types of failure modes that progress quickly from the incipient to critical stage. An example of this scenario is transformer bushing failures.
5. Requires extensive apparatus knowledge and experience.
6. Expensive to implement and administer
As utilities started to come under deregulation in the 1990's, utilities faced pressures to reduce costs, which forced them to reevaluate maintenance intervals and restrict hiring or reduce maintenance staff. These staffing limitations also caused a backlog of maintenance jobs which in turn caused maintenance intervals to be greatly extended. As a result utilities have started to see a dramatic increase in primary equipment failure. Maintenance costs may have gone down but this is increasingly now offset by higher repair costs after failure.
Since utilities cannot afford to go back in time and go back to the old ways of doing things, utilities have started to look at new methods of maintaining their assets. The methodology most commonly used is 'Reliability Centered Maintenance' or RCM which also utilizes Failure Modes and Effects Analysis or FMEA.
An example of a substation apparatus that consumes significant maintenance efforts and impacts reliability is the transformer load tap changer. The LTC is generally viewed to be the cause of about 40 per cent of transformer failures. Figure 1 shows the temperature affects on the transformer LTC tank when contact coking results in arcing and hence overheating as the LTC is operated.
Why, When and Where to Implement Monitoring and Diagnostics
Manual inspection can trigger appropriate maintenance if a failure characteristic can be observed and a predictable pattern recognized. However, on-line condition monitoring is the only effective choice if any of the following occur:
- The failure characteristic cannot be identified by routine inspection.
- The failure development time is shorter than the inspection period.
- The failure characteristics are not predictable in advance.
Continuous on-line monitoring can provide many other benefits to the utility.
On-line condition monitoring only alerts maintenance personnel when equipment needs maintaining. This allows utilities to move from a periodic-based maintenance schedule to a just-in-time maintenance program.
At a utility in the U.S., on-line monitoring has allowed them to move from a 5-year breaker maintenance schedule to a 12-year average maintenance schedule. This utility routinely checked 87 LTCs annually. Now, they only receive about 5 LTC maintenance alarms a year from their on-line monitoring system.
There was a payback of less than 2 years for their substation diagnostic system. The utility has now implemented this system at over 60 substations. They estimate the overall savings at 40 per cent of their O&M expenses, as compared to their previous scheduled-based maintenance program, while enhancing reliability.
Summary benefits of on-line monitoring and diagnostics include the following:
- Reduce inspection costs.
- Reduce maintenance costs.
- Reduce failure-related repair or replacement costs.
- Enhance system reliability: fewer unplanned outages.
- Provide better planning for scheduled outages.
- Defer planned upgrade or replacement capital costs.
- Reduce insurance premiums.
- Retain knowledge of most skilled staff (expert system).
- Provide wide access to key knowledge.
To broadly implement continuous on-line monitoring, the economic evaluation process may involve a number of analysis tools. These tools include Fixed Charge Rate Analysis or Discounted Cash Flow Analysis. However, where sites fall into a special category, such as system critical sites for transmission, generation step-up sites, stations with troubled equipment, or where the loads are especially critical, the normal economic evaluation may not apply.
Where and When to Implement On-Line Monitoring and Diagnostics
RCM is a technique aimed at identifying the most appropriate inspection and maintenance tasks to preserve functional reliability. The technique can be aimed at components or defined systems. RCM uses FMEA and risk assessment techniques to select maintenance tasks that are directly related to high-criticality failure causes. The same methods can select monitoring that is directly related to advising condition or impending failure information.
FMEA or failure modes and effects and criticality analysis (FMECA) identify the following:
- The failure modes.
- Their possible and probable causes.
- The system and component effect(s) of that failure.
- The 'criticality' involved.
Criticality is viewed in the context of risk. Therefore, risk identification is an integral part of RCM. Risk can be defined as the product of the chance of an event occurring and the impact of such an event.
Using a ranking system, risk or criticality can be ranked from highest to lowest. In this case:
- Product numbers 1 through 3/4 are high risk (A).
- Numbers 4/5 through 8 are medium risk (B).
- Numbers 8/9 through 14 are low (C).
- Numbers 15 to 20 are minimal risk (D).
Risk can then be quantified using a matrix of these products. Any issue such as financial impact, customer impact, safety, environment, or legal can be ranked in the risk (criticality) assessment.
Once a utility business has a justification and objective for monitoring and diagnostics, a risk based analysis such as the one above can be useful in determining which stations and primary equipment should be monitored for which failure modes.
Furthermore, the matrix can be used to determine whether basic monitoring is sufficient or more comprehensive diagnostics are needed. Finally, a per site economic or value analysis for on-line monitoring and diagnostics must be conducted.
Reminder
When discussing these issues you should remember that the term 'monitoring' describes basic parameter measurement with threshold alarms. The term 'diagnostics' indicates the addition of sophisticated analysis, an expert system capable of providing an assessment of the equipment condition and suggested actions.
Credits
[1] 'Equipment Monitoring Selection as a part of Substation Automation' Panel Session presentation to IEEE Winter Power Meeting, February 1999 by W. Bergman, TransAlta Utilities.
[2] CEA Project 485T1049 'On-line condition monitoring of substation power quipment --- utility needs' January 1997.
Ron Farquharson and Ken Caird are with GE Harris. This article was reprinted with permission from GE Harris' Newsletter 'Synergy'. Part 2 will appear in the next issue of Electricity Today.