Uncovering Power Quality Culprits

Apr 1, 2010 12:00 PM, Roger Lawrence, P.E., CEM, FIEEE, RGL Solutions

There’s more to determining responsibility for power quality problems than meets the eye

Every unexpected interruption in an electrical power system is accompanied not only by frustration but also by serious consequences, such as production downtime, damaged product, failed components, increased costs, compromised personnel safety, and a lack of earning power. Typically, when a disruption occurs, the first question that comes to mind deals with assigning responsibility. As evidenced in the following two case studies, power quality investigations can reveal surprising results — oftentimes where the assumed culprit is eventually exonerated.

Case history No. 1

A U.S. production plant was identified by a local electric utility as the undisputed cause of a problem that was presenting itself in the system of a remote end-user. After an initial investigation, electric utility personnel believed the plant was at fault.

The plant operated using numerous adjustable-speed drives. As material was processed and moved through the plant, the electrical system reacted and caused problems for a connected customer 10 miles away. Upon visiting the remote site of the disturbance, interference could be heard in the electric utility's pole-top transformer. This interference was taking place at the same time the production plant drives were operating. When the drives were not operating, however, there was no interference heard coming from the transformer. The natural conclusion was that the drives were causing the problem.

Action was taken to investigate the problem. Power quality monitors were placed at the plant's substation as well as at the site of the remote connected customer. Measurements taken showed a classical 12-pulse current waveform and indicated no timing or instability problems. There was an elevated level of 23rd and 25th harmonics present on the system (click here to see Fig. 1).

Current measurements taken at the plant's substation provided insight into the elevated high order harmonic currents. These are clearly seen superimposed on the fundamental waveform in Fig. 2 (click here to see Fig. 2). The voltage measured at the plant's substation confirmed that the high levels of harmonics were emanating from the electric utility substation (click here to see Fig. 3). Voltage measurements at the end-user's remote site revealed an amplification of the higher order harmonics, including a double zero crossing (click here to see Fig. 4). It was this amplification that had caused the end-user's problem.

Further investigation of the power system revealed that the electric utility had recently installed an underground distribution cable to replace the original overhead power line. Calculations of the natural resonance between the shunt capacitance in the cable and the leakage inductance of the transformer confirmed a natural resonance above the 20th harmonic. In other words, the investigation revealed the cause of the problem was on the electric utility system. Once the electric utility reconfigured its system, the problem disappeared — and the end-user was exonerated.

Case history No. 2

A U.S. offshore textile mill was convinced that the bizarre performance of its electrical apparatus was caused by the electric utility's power system. This end-user had extensive knowledge and experience with the equipment — and it had performed reliably at a previous location.

Power delivery at the new location was known to be unreliable. Neighboring electric utility customers and weather-related events caused the power to fail and recover frequently. When power delivery resumed, the textile machines needed to be restarted. At this stage — and on a random basis — some machines would run at half speed while others would suffer complete memory loss and need to be reprogrammed. Although the electric utility provided an enormous amount of information during the investigation, there was no correlation between the electric utility delivery failures and the behavior of the machines.

A site visit confirmed that the textile machines were well maintained. The batteries used by the plant to provide ride-through for equipment electronics were clean and erosion free. The power distribution system within the plant tested well; current harmonics were high yet the voltage distortion level was low. During the times when the utility electrical supply was running, it exhibited no abnormalities. Based on this investigation, it appeared that both plant and utility were performing well, with the exception of the reported symptoms demonstrated by the machines. A new approach toward solving the problem was needed — the next step was repeatedly recreating the problem under controlled conditions.

The Nature of the Power Quality Beast

Power quality events can be broadly divided into two groups — those that are evident over many cycles of the electrical supply and those that may be present only under certain combinations of events — and only for a few isolated cycles of the supply. Many times, it’s difficult to find a solution because the problem is somewhat of a moving target, characterized by elusive symptoms. In addition, complicating the complexity of the electrical system is the human reaction factor — it’s human nature to assume the problem has been created by someone or something else.

A series of rapid power transients was created (the machines were switched on and off automatically multiple times) in order to test the theory that this was a power recovery issue and to observe machine behavior under these conditions. The machines frequently started successfully and without incident. Eventually, there was a half speed startup. After still more testing, there was eventually a memory failure. The problem was demonstrated to be a deficiency within the machine. Although the end-user was not to blame, he bore responsibility for mitigation. The electric utility was not the cause of the problem, and was exonerated.

The end-user implemented a cost-effective mitigation procedure to manage and control the switching of the machines. When weather conditions threatened to interrupt power, the end-user chose to merely switch off the machines before the electric utility supply failed. When utility power was judged to be healthy, the machines were restarted. Although other solutions were considered (i.e., the installation of power mitigation devices or running a new backup power feed), they were ignored due to limited funds.

The case histories outlined in this article demonstrate that the responsibility for providing reliable power systems is a shared responsibility. When problems occur, electric utilities, manufacturers, and end-users must cooperate to find a solution. Careful analysis can define the problem, reveal a solution, and identify responsibility. But remember, don’t jump to conclusions too quickly — the entity or party who initially seems at fault may be the farthest conclusion from the truth.

Editor’s Note: The original version of this article ran in “Energy Engineering,” Vol. 106, No. 6, published by the Journal of the Association of Energy Engineers, copyright 2009 Association of Energy Engineers, www.aeecenter.org.

Lawrence is an independent consultant and sole proprietor of RGL Solutions, Raleigh, N.C. He can be reached at rogerlawrence@rglsolutions.com.