Following the Trail of Destruction

Equipment with blown fuses, charred resistors, fried semiconductors, and vaporized circuit traces are on the workbenches of repair shops everywhere. Such obvious damage can be easily — perhaps too quickly — attributed to power surges.

When claims of equipment damage caused by faulty electricity cross the customer service desks of utilities, questions about the original cause border on the forensic. Was the damage due to old age? What about faulty components pushed over the edge by a common electrical disturbance? Was lightning to blame? These questions are never easy to answer, but definitive surge testing can reveal true stories about good components gone bad.

Utilities aren't the only companies interested in the cause-and-effect of equipment damage. Insurance companies, which shell out millions of dollars every year to owners of damaged equipment, are highly motivated to exploit standardized surge testing that reveals original cause.

Consider data collected from a single regional insurance company. Between 1998 and 2001, one insurance company recorded 2,017 payments to policyholders with claims of damaged equipment. Of those 2,017 payments, 889 — or 44% — were attributed to either lightning or surges, for an average of $2,700 per claim. Over those four years, the company paid out more than $2.4 million because of electrical excursions, an average of $600,000 per year. Extrapolating this figure to the worldwide population of property-insurance companies yields the possibility of a significant amount of surge-inflicted equipment damage.

Developing test protocols, investigative procedures, and a catalog of cause-and-effect associated with surge damage will require an outlay of time and resources that cannot be born by a single industry. A collaborative effort between utilities, insurance companies, manufacturers, and end users promises to shed some light on the destructive nature of power surges. The EPRI project, called Analysis of Failure Mechanisms for Electrical and Electronic Equipment, began this year with some exploratory testing in the laboratory.

Surge Testing Equipment

Surges can impinge upon equipment at the power port, communication port, or both. Because surge voltage can be coupled from the power port to a communication port, power quality investigators are generally interested in both ports. Surge testing the power port of equipment requires a surge generator capable of simulating the types of surges that impinge upon equipment in the real world. For example, “Zeus,” the surge generator used at the authors' laboratory, includes plug-in modules for the ANSI/IEEE Standard C62.41 Ring Wave and Combination Wave.

Impulsive Surges Caused by Lightning

Based upon field experience and testing, Standard C62.41 proposes testing with a waveform that represents a surge resulting from lightning directly impinging upon or coupled into the power system (called a “combination wave”). Fig. 1, on page 62, shows the waveform of the open-circuit voltage and short-circuit current. Three characteristics specify the waveform: rise time (or front time, in the case of short-circuit current), duration, and amplitude.

During testing, the rise time and duration are held constant according to the ANSI/IEEE definition. On the other hand, varying the amplitude by ramping it up in 500V increments for each trial can cause equipment failure.

Applying Surges in the Laboratory

Whereas traditional surge testing is intended to reveal the withstand limits of equipment without deliberately damaging the equipment, the surge testing described here was intended to cause damage so investigators can correlate observable and measurable effects with a known cause. Fig. 2 shows a simplified diagram of the setup used to surge test equipment. Zeus is composed of a surge generator to supply the surge voltage and current and a back filter/coupler to prevent surge current from feeding back into the power source and to couple the surge into the equipment under test. A digital signal analyzer monitors and records the voltage and current at the equipment power port.

Trial by Fire and Smoke

During the first stage of the project, about a dozen motor-and-control systems were tested. Each system was composed of a motor, an adjustable-speed drive (ASD), and a programmable logic controller (PLC), as shown in Fig. 3. The PLC was connected to three-phase voltage via a step-down transformer, which supplied single-phase 120V power.

Fig. 4, below, and Fig. 5, on page 66, show examples of recorded waveforms during surge testing a motor-and-control system. Fig. 4 shows the results of a trial where the current peaked at 1,200A. MOVs inside the ASD were able to clamp the voltage at about 600V. In this particular trial, the equipment survived the surge.

Fig. 5 tells a different story. During this trial, the current peaked at 1,400A. As shown in the voltage trace, the MOV clamped the voltage, but an unidentified component failed just after the peak current.

Fig. 6 shows one of the ASDs that failed during surge testing. The catastrophic failure of a filter capacitor resulted in an explosion followed by smoke billowing from the ASD vents. The failure of MOVs and filter capacitors is not an unexpected consequence of a strong surge applied to equipment with these components in the front end. Note the soot remaining on the components (near the bottom left) resulting from the explosion of the filter capacitor. Also note the burned circuit trace resulting from excessive current — revealing evidence about the mode of failure.

Failure Analysis

The next stage of the project — to be conducted later this year — is the meticulous analyzing of each tested ASD and PLC to determine modes of failure. This analysis will go beyond obvious damage detected during visual inspection. To construct a comprehensive catalog of failed components and equipment parts, it is necessary to restore each piece of equipment to working condition.

By troubleshooting and repairing surge-damaged equipment, you can develop a knowledge base of surge-related failure for classes of equipment: industrial equipment, such as ASDs and PLCs; commercial equipment, such as video cameras and card readers; and residential appliances, such as security systems and entertainment systems.

Understanding common failure modes among classes of equipment will enable manufacturers to improve the surge-withstand capabilities of their equipment, resulting in fewer claims of damaged equipment for utilities and insurance companies, as well as reduced inconvenience for end users.

Next Month

In next month's article, the authors will discuss surge testing the communication port of equipment and reveal some preliminary findings of their investigation.

Brad Connatser is the publications manager at EPRI PEAC Corp. in Knoxville, Tenn. You can reach him at [email protected]. Doni Nastasi is a power quality engineer at EPRI PEAC Corp. You can reach him at [email protected]. Kermit Phipps is a power quality and EMC specialist at EPRI PEAC Corp. You can reach him at [email protected].