Since the onset of deregulation, market forces have placed utilities in the uncomfortable position of reducing expenditures for many activities that previously were routine with respect to construction and maintenance. In the face of these reductions, demand for reliable power delivery has increased dramatically, requiring the continuous monitoring of assets to ensure maximum reliability in a cost-effective manner.

To exacerbate this scenario, operations and maintenance managers have such a plethora of computer-generated data that they cannot sort through all of the information to make proper use of the material they have in front of them. Paradoxically, they often are faced with too little information for a variety of devices and assets because they cannot collect and filter data into manageable and useful information. Under these circumstances, certain conditions can go undetected until the situation becomes irreversible.

The Cost of Failures

A recent survey estimated that power outages cost each of the roughly two million industrial and digital-economy electric customers more than US$23,000 annually.

PacifiCorp (Portland, Oregon, U.S.), like other utilities, is interested in safely maximizing the use of its assets, minimizing customer costs by providing uninterrupted power delivery and protecting the shareholders who have invested in the company. An important, if not crucial, element for achieving these goals it to optimize maintenance and system loading and to extend the life of the key transformers on the system. Transformer management is particularly challenging because:

  • Transformers are frequently expected to operate at or above nameplate rating.

  • Peak loads are exceeding design limits, reducing safety margins.

  • Equipment replacement lead times can be more than a year.

  • Transformer average age continues to increase.

  • Spot market replacement power costs are unpredictable and continue to increase.

The common practice is to monitor transformers by collecting dissolved gas analysis (DGA) data on the coolant oil. DGA data are used to observe, diagnose and repair problems. Savings realized from proper diagnostics are significant not only because additional investment may be postponed, but also because of avoided costs due to unplanned outages (replacement energy, environmental cleanup, customer and collateral damage and increased insurance).

Avoiding the Cost of Failure

The main cause of failure is rarely associated with the age of the transformer. Rather, failures are a result of electrical disturbance (29%), deterioration of insulation (13%), lightning (16%), inadequate maintenance (13%), loose connections (13%), moisture (7%) and overloading (2%). The key to optimizing the use of assets is to minimize the risk of failures and their collateral effects on the system. Since DGA can detect impending transformer failures, the utility can decide when a transformer should be refurbished or replaced. Using DGA as a decision-making tool has has become a common practice at PacifiCorp.

The monitoring costs associated with extending the life of a critical transformer are easily recovered based on replacement cost alone. A simplified ROI analysis, based solely on delaying the replacement of a $1 million transformer for five years, and the U.S. list prices for the monitoring equipment and services used show that payback of these expenditures occurs in just over a year. A complete return on investment (ROI) analysis is subject to not only replacement costs but other costs that might include environmental cleanup of about $0.5 million, spot market purchases of as much as $1.5 million per day and insurance deductibles, which can be as high as $1 million. These numbers are estimates for the failure consequences of a single-phase 500 MVA generator step up (GSU) failure, whose costs can quickly dwarf the replacement cost of the transformer, especially those requiring spot market purchases to replace lost production.

Web-based on-line monitoring data are not designed for the direct use by the utility control center but for personnel who are responsible for reliability centered maintenance (RCM), to improve reliability, minimize downtime and reduce the cost of operations. RCM personnel are spread throughout the company as plant managers and equipment experts. In the pursuit of making RCM work, these personnel now have direct access to detailed real-time information.

Simplified Return on Investment Calculation
Values Instructions
Transformer Capital Cost $1,000,000 Cost to replace or rebuild the transformer today
Annual Interest Rate 4.000% Interest rate earned on money not spent to replace or rebuild transformer
Expected Life Extension 5 years Number of years expecting to delay replacement of transformer, 1 to 10.
TrueGas Capital Cost $31,240 Capital cost of TrueGas monitor and accessories, onetime fee
TrueGas Installation Cost $3355 Install cost of TrueGas monitor and accessories, onetime fee
TrueGas Monitoring Service $800 per month Serveron Web-based on-line monitoring services (SRC), monthly fee
Payback Period 1.1 years Number of years to recover cost of TrueGas+Monitoring Service fees
Return on Investment @ new End-of-Life 33% ROI as calculated against “Expected Life Extension”
Savings @ new End-of-Life $117,405 Total savings as calculated against “Expected Life Extension”
Year Number Cost of Monitoring System Extended Life Savings Net Cash Flow Cumulative Net Cash Flow Running ROI
0 $34,595 $34,595 $34,595 0%
1 $9600 $40,000 $30,400 $4195 0%
2 $9600 $40,000 $30,400 $26,205 6%
3 $9600 $40,000 $30,400 $56,605 23%
4 $9600 $40,000 $30,400 $87,005 30%
5 $9600 $40,000 $30,400 $117,405 33%

Outsourcing as a Solution

To achieve a continuous mode of monitoring assets, PacifiCorp elected to use the services of outside experts to check the status of key transformers. By knowing the real-time operating conditions of these assets, the company is provided with information on the necessity for providing immediate maintenance, if necessary. Outsourcing the surveillance allows PacifiCorp to focus on its core competencies involving generation, transmission and distribution. In the process, it has been demonstrated that there is a positive ROI through both a reduction in the probability of an outage and increased savings in operating expenses.

Outsourcing asset monitoring can enable utilities to improve profitability while maintaining reliable power delivery. In this connection, PacifiCorp has enlisted the services of the Serveron Corp. (Hillsboro, Oregon) to monitor six GSU transformers and six large transmission and distribution transformers using Web-based data. An eight-gas, on-line dissolved gas analyzer attached to each transformer automatically collects DGA data six times each day, on a regular 4-hour schedule. Authorized personnel can view the DGA data at any time from any location with an Internet connection using a secure Web browser. Serveron's Response Center (SRC), which consists of a Network Operations Center (NOC) and an Emergency Response Team (ERT), manages the data associated with these assets around the clock. The NOC is designed to ensure that data are collected and monitored continuously. The ERT informs PacifiCorp personnel of any unusual changes in the data, as determined automatically by software and manually through regular inspection of the data. Recognizing that DGA data are a valuable tool for monitoring transformers to maximize their availability for continuous delivery of energy, PacifiCorp initiated a trial Web-based online monitoring project in conjunction with Serveron in 2002.


Shrinking maintenance workforces, changes in regulatory rules and increased competition have forced significant changes in how utilities must now operate their systems. To maximize their assets, utilities must have access to useful information for their personnel, in the right form and at the right time. Like the supervisory control and data acquisition (SCADA) systems most utilities use, Web-based on-line monitoring involves real-time data collection. Unlike those SCADA systems, Web-based data are not for the direct control of power-delivery systems but are most useful for implementing RCM processes that affect business-planning decisions such as maintenance schedules and load guidelines. SCADA systems that feed into a control center are designed and used to provide critical alarms for immediate attention in connection with control actions. Those systems are not intended to provide the detailed information regarding events that led to triggering the alarm. A separate mechanism allows local experts to obtain data independently and in greater detail to make informed decisions for diagnosing or predicting potential problems. In this connection, the experts are in a position to specify appropriate guidelines for control center personnel to follow as part of their daily decision-making processes.

PacifiCorp has demonstrated that using outsourced monitoring has had a positive affect on maintaining critical asset reliability that results in improved profitability and a higher degree of service reliability.

Robert Augenstein received a BSEE (power) degree in 1965 from the University of Colorado and an MS degree in 1969 from the University of Southern California. Starting his career with the Los Angeles Department of Water and Power, he has worked in the utility industry for 36 years as an engineer with Bechtel, where he concentrated on power plants, and as an engineer with PacifiCorp. In the 10 years he has worked at PacifiCorp, Augenstein has been project manager for the Y2K Project and is now responsible for developing and implementing a quality-assurance program for key transformers on the company's system.