New MV products can transfer in milliseconds while others sense sags and boost voltages quickly.

Although we usually don't associate the term "power quality" with higher voltage levels, the concept is just as important for medium- and high-voltage systems as it is for 600V systems. In fact, electric utilities and their research consortiums have devoted time and funding to improve service technology at the primary voltage level. The end result has become known as "custom power."

Customer-directed developments

A recent study, the Distribution Power Quality Monitoring Project, conducted by the Electric Power Research Institute (EPRI) with the cooperation of 24 utilities, involved the monitoring of 277 sites. It produced information on voltage sags and swells, transient disturbances, momentary disruptions, and harmonic resonance conditions. Fig. 1 shows a portion of the data.

As you can see, there's a rather large percentage of sags and momentary outages (the data found at time durations of 1 sec or less). These short-term power system disruptions weren't as troublesome when analog devices were the norm for end users. Now, almost every major voltage sag and momentary breaker operation causes loss of system control or a complete production shutdown.

Faced with increased end user equipment sensitivity, utilities redoubled their efforts to develop technologies dealing with sags and momentary outages. One result of these efforts is the static series voltage regulator (SSVR), as shown in Fig. 2. Although designed by an eastern utility to fit its customers' needs, the product has been brought to market with the help of a major electrical equipment manufacturer.

When a phase-to-ground fault occurs on an MV power system, feeders nearby the faulted area experience phase voltage distortions. If these distortions last for several cycles (which seems to happen frequently, as the EPRI data indicates), sensitive loads are negatively affected. With an SSVR, if phase voltages drop below a preset threshold, it will "boost" the voltage to support continuity of the correct voltage to a critical load area. And, with proper sizing, an SSVR will support other voltage drops, such as those resulting from momentary breaker operation. We're not talking about protection from a complete loss of power (as with a UPS) but rather protection against short duration problems.

Another product is the result of one utility's survey of its customer base. Those customers served by two separate primary power feeders were asked if they would be interested in a method of "upgrading" the slow automatic transfer (2 to 10 sec) equipment. More than half surveyed said they would be interested, especially since the new technology could provide uninterruptibility without a UPS and batteries. The result: An MV subcycle transfer switch that can be used in both new and retrofitted switchgear. After study by EPRI, the product was found to be worthy, and is now on order with four more utilities.

Applications of custom power strategies

Let's look at a few typical customer power applications, concentrating on the 1.5kV and 25kV distribution voltage levels. First, we'll look at the MV subcycle transfer switch. These transfer switch applications are derived from field studies made at various customer locations. They demonstrate the diversity of available solutions, even when the technology seems to be limited to a narrow range of sites, namely those having two independent power sources.

Application No. 1. A semiconductor manufacturer with 14,000kVA of load was faced with a UPS budget of more than $4 million. Even when all the sensitive loads were concentrated, the same amount of battery support protection was needed. Since the facility had the required two independent power sources, the company was able to reduce the cost expectations, through the use of a subcycle MV transfer switch, down to below $600,000. Yet it achieved similar protection.

Application No. 2. An industrial facility was facing yearly losses of $700,000 from process downtime due simply to circuit breaker operations. As shown in Fig. 3 (on page 23), the company already owned an automatic transfer switching scheme of the "slow" variety (shown with the dotted line) that used motorized circuit breakers (designated "M3" and "M4"). The recommended solution was to use "silicon" technology transfer switches (designated "SS1" and "SS2") as the normal (high-speed transfer) circuit between Lines 1 and 2. The existing motorized breakers would serve as the backup and bypass. A one-year payback is estimated for this installation.

Any danger areas?

Looking at the above examples, you may suggest that there's still a danger in the applications discussed above, in that an upstream disturbance on the utility system might possibly affect both independent power sources at the same instant of time. After all, how far back into the utility grid can you maintain "independency." Surely, these lines must come together somewhere.

This possibility already has been anticipated. The result is a combination of the "boost" concept [ILLUSTRATION FOR FIGURE 2 OMITTED] and the high-speed switching concept discussed above. This works very well in providing fast transfer between lines while maintaining correct voltage should a common disturbance on the upstream circuits (in this case, a single exposed transmission line) occur.

Multiple/redundant services

Many facilities, such as hospitals and health care centers, have existing dual feed services. Another alternative for reliable power is a scheme called multiple/redundant services. As shown in Fig. 4, a new fast-transfer device is installed between two existing substations to keep the facility on-line in the normal manner. If an incoming line is interrupted, this transfer switch transfers to the alternate feeder in less than 4 milliseconds, thus keeping the load on-line. The existing engine-generator set is brought on-line (with a 1-sec to 2-sec time delay to avoid false starting), synchronized, and made available to substitute for the lost incoming line. The net result: A restoration of the two-source capability within several seconds. The possibility of a cascading failure of both lines is avoided by bringing the third source (the gen-set) on-line after the loss of the first source.