Metropolitan Kansas City Has Enjoyed Consistent Growth in New Customer Loads over the years, highlighted most recently by a resurgence of the downtown area. In the past couple of years, new, large downtown loads, such as a new sports arena and Federal Reserve and Internal Revenue Service buildings, have come on-line.

Kansas City Power & Light (KCP&L; Kansas City, Missouri, U.S.) serves the metropolitan area, which includes customers in both Kansas and Missouri. The utility also has seen significant growth in other residential and commercial loads, both downtown and in the surrounding service area. At the same time major infrastructure additions and modifications were going on to accommodate this new load, commodity prices — and, therefore, material prices — were escalating 15% a year on average, driving up construction costs. In order to use capital in the most efficient manner possible and maximize the value of the existing distribution assets, the distribution engineering group decided to take another look at how it does capacity planning and additions.

A NEW APPROACH

In the past, KCP&L designed and built a system based on N-1 contingency planning design criteria that could handle the loss of any one piece of equipment at a given time. The circuit loads used in this type of planning were seasonal peak loads, which were not always coincident peak loads. The highest load a circuit reached in a given year was assumed to be coincident to a neighboring circuit's highest load.

This presumed coincident peak load was used to determine how much load could be shifted to another circuit during failure conditions. If a cable rating was violated during this simulated load-shifting procedure, capital improvements were designed and built to resolve the switching issue. On one hand, this could be as simple as installing a feeder-class switch to sectionalize a circuit differently and break apart the load onto neighboring circuits. On the other hand, it could be as complex as installing a new substation with new distribution circuits to cut over loads that would otherwise overload neighboring circuits in failure conditions. In any case, the coincidence or non-coincidence of these loads was not examined.

The original non-coincident contingency method should still be used as a first screening of which circuits might be problematic. If a non-coincident contingency switching order works, there is no need for further examination. If it doesn't work, the engineer should look more closely at those days of the year where there could be overload. During that examination, the engineer would be able to tell how many hours and by how much a circuit is overloaded — both key pieces of information when making decisions based on risk and for use with energy-efficiency and demand-response programs.

FEEDER DATA AND PLANNING PROCESS

There are four levels in which to examine the behavior of a given feeder or feeders: multiple-year seasonal demand, single-year daily demand, multiple-day hourly demand and single-day hourly demand. All these may be based on hourly amp-flow at a feeder breaker from a distribution supervisory control and data acquisition (SCADA) system, which can be sorted to find a 3-phase daily peak kilovolt-amp. Each level of examination serves a different purpose in distribution planning. For example, a multiple-year seasonal demand would be used to determine a circuit's future growth rate.

A multiple-day hourly demand between several feeders would be used to determine if distribution ties would be overloaded if one feeder failed. Generally speaking, if a feeder has five distribution ties and equal peak demand to its tie circuits, then each circuit in the area should have at least one-fifth of its capacity reserved for contingency switching. Another use of multiple-day hourly demand between feeders that serve a given geographic area is to analyze a geographic area's power usage, which is valuable when determining a neighborhood's growth rate.

By looking at the behavior of multiple feeders simultaneously, it is possible to determine if the distribution ties between the circuits will be overloaded during failure conditions on a seasonal, daily or hourly basis, also called an N-1 coincident contingency analysis. There is a method to reduce the level of analysis required for such a study.

First, perform a seasonal peak contingency study. By taking the apparent peak demand of several circuits during a given summer, determine what switching operations are necessary to restore all customers if one of the feeders were to fail. Then determine what effect that switching operation will have to see which tie circuits are overloaded. Identify the demand level at which the tie circuit would not be overloaded when switched around and set a desired threshold.

Once the threshold is determined, examine those days in the summer where the circuit's demand was higher than the threshold. By looking at this closely for a circuit, the number of hours of risk is apparent (and a necessary piece of information needed for decision-making). The number of hours curtailment needed in a summer is also required for curtailment programs, also called energy efficiency and demand response (EE/DR). A logical step-by-step process for this analysis is outlined in the nearby sidebar.