By Edward R. Furlong & Louie J. Powell, GE Digital Energy
The increased dependence of modern industry on reliable power has shifted the site selection criteria from “available capacity” to “capacity and reliability.” As such, industrial concerns have begun to evaluate energy suppliers on their ability to meet their reliability needs. This creates a unique difficulty for a distribution utility that is eager to expand its market base but unable to invest in additional facilities to meet the needs of a single customer (or a small group of customers). At the same time, some new industrial facilities requiring premium power use batch processes or computer loads that either have a low average load, or a variable load for which the average value is small. This makes their demands for improved reliability more difficult to justify, and the solutions more finely tuned.
Power still critical
During the late 1990s, the rise of Internet-based companies-those heavily dependent on servers and other computer loads-resulted in an explosion in demand for electrical infrastructure capable of protecting delicate processes from service-interruptions as short as 1.5 cycles in duration. While some of these industries have suffered a recent downturn, the need for conditioned power continues due to the increased dependence on computer-controlled equipment in related fields such as medicine, biotechnology, and semiconductor manufacturing. Even old economy processes like coating, painting, and machining now have a single computer controlling the finishing operation on millions of dollars worth of finished goods.
As these businesses expand and select new sites to build, they must investigate the implications of local regulations, taxes, labor rates, and available infrastructure.
Traditionally, industrial operations that have needed higher reliability than could be supplied from the grid have had to take responsibility for improving the reliability of power supplied to their critical loads. Uninterruptible power supplies have found their way into many of these systems as a practical means of supporting process continuity in the face of practical service availability from commercial energy suppliers. As these industries focus on their business goals and profitability, the details of designing, maintaining, and operating this aspect of their facility are emerging as distractions. So these companies have started looking for ways to transfer responsibility for electrical reliability to their respective energy supplier.
Power distribution systems are exposed to a variety of risks that impact the reliability of service as perceived by the end consumer. The pie chart provides a historical perspective on service reliability based on actual interruption statistics for distribution customers in one region of the U.S. for the four-year period ending in 1999.
Note that two of these categories (lightning and tree contact) have a clear common denominator-weather, and that a third (accidents) is strongly influenced by weather. This is not surprising; it is a well-known fact that weather is the single most significant factor controlling the reliability of electric service from commercial power grids.
The other dominant cause depicted in these data is equipment failures. An important conclusion from that statistic is that commercial distribution systems employ physical equipment, and that there is a wearout/failure mechanism that affects the life expectancy of these components.
There are two ways that electrical distribution systems are configured: radial and network. The vast majority of distribution systems outside major metropolitan areas use a radial design. These systems have a defined source, with a tree-like structure that delivers energy to a collection of defined loads. A failure or interruption upstream of a load inexorability results in curtailment of service to that load. (The statistics shown in the chart are for radial systems.)
Radial systems are typically configured to provide for alternative connections to allow maintenance, but even under the best of circumstances the time required to reconfigure a radial system following an interruption is measured in seconds; typically, delays of minutes to hours are more common, especially in rural or suburban areas.
An alternative form of distribution is employed in some metropolitan areas. Network systems connect multiple sources of power to multiple loads through multiple parallel paths. These systems are equipped with protective accessories that detect when individual paths have failed and automatically switch them out of service while continuing to serve loads. Network systems almost always utilize underground cable, so weather is much less of a factor. The result is that the availability of service to the consumer is significantly higher than with radial systems. For economic reasons, application of network systems is limited to very high load-density metropolitan areas.
However, interruptions do occur on network systems. If a short circuit occurs on a network cable, that cable will be automatically switched out without disconnecting loads or reconfiguring the system. But the short circuit itself causes a reduction in system voltage for the few cycles of time required to detect and clear the fault. These “fault induced voltage dips” have the same effect on critical loads as physical interruptions-the only difference is that they don’t last as long.
This consideration suggests two important points. First, the statistical parameter presented in the chart is the “System Average Interruption Frequency Index,” or SAIFI. SAIFI is a standardized measure of electrical system reliability used by regulatory agencies to gage the performance of distribution companies. But because fault-induced voltage dip will create conditions that simulate interruptions, and because various control schemes can be used to either automatically restore service (automatic reclosing), or reconfigure the distribution system around a failed component, and because any of these events can be triggered by weather or other “acts of God”, the industry has agreed that interruption events shorter than some arbitrary duration will not be included in the SAIFI statistics that may be the basis for some punitive action directed toward the distribution company. The fact that short duration events (sometimes called momentaries) can affect critical loads means that looking only at SAIFI does not tell the owner of a critical process everything that he or she needs to know about the reliability of electric supply.
The second point is that regardless of what is done to enhance the reliability of the distribution system that supplies critical loads, interruptions (or voltage dips) can still occur if failures occur on the load side of the remediation. Thus, the ideal solution is to address reliability directly at the critical load.
Options for distribution utilities
Regulated public utilities often have difficulty providing differentiated services. The theory behind the establishment of rates based on return of investment is that all “customers” benefit equally from those investments, and therefore a uniform rate schedule can be applied to all customers. It is certainly possible that a differentiated rate schedule could be devised that would enable energy suppliers to offer higher-reliability service, utilizing whatever technologies are needed to achieve the reliability targets of their customers and passing the cost of those enhancements on to the customers who benefit. However, the reality is that the final decision on rates resides with state regulatory agencies.
So far, there has been little interest on the part of the various public service commissions to recognize that some customers have needs that are more stringent than others.
On the surface, it might appear that deregulation of the utility industry might open the door to creative rate structures that would address reliability concerns. As a practical matter, the distribution “wires” companies will remain regulated monopolies. So, a creative approach that does appear feasible is for a non-regulated service supplier (possibly a parallel subsidiary of the holding company that owns the regulated wires company) to offer a service of receiving power from the grid, and enhancing the reliability of that power through the use of the kinds of innovative technologies discussed below. In the simplest case, this reliability-enhancement service could be sold as a standalone service, or the reliability enhancement company could purchase power at the regulated rate from the distribution company, pass it through a reliability enhancement process, and then resell it (at a higher rate) to the end consumer.
For situations where there is the potential to offer industrial concerns reliability-dependent rates, local reliability can be improved through careful analysis and hardware selection. First, load-flow software is used to identify customer susceptibility and grid improvement potential by simulating events at varying distances from the customer or at nearby facilities. These events can include faults, large motors starting, smelters, and distributed generation resources coming on-line or tripping. The congestion points are identified and hardware solutions, such as voltage compensators, energy storage systems, or advanced power systems, are selected. The customers then benefit from the improved power reliability, without the distraction of operating equipment outside of their core competency and business goals.
Furlong, new product introduction leader with GE Digital Energy, can be reached at firstname.lastname@example.org. Powell, manager of T&D consulting with GE Power System’s power systems energy consulting division, can be reached at email@example.com. For more information visit the website at www.gedigitalenergy.com.
Causes of service interruptions