Over the years, American electric power has constructed thousands of circuit miles of underground cables throughout its 11-state territory. These underground cables provide electric service to American Electric Power's (AEP's; Columbus, Ohio, U.S.) residential, commercial and industrial customers as well as the integrated underground network systems. Through the years, the utility's underground cable systems have followed industry trends, with the installation of paper-insulated, lead-covered cables early on to the cross-linked polyethylene and ethylene propylene rubber cables of today. Environmental issues, line mechanic skill levels, infrastructure-expansion limitations and the increasing load-carrying requirements have all had influencing roles in driving underground cable trends.

In today's climate of doing more with less, the industry is increasingly challenged with regard to the underground cables of its distribution systems. As with many utilities, AEP is now faced with supporting an aging infrastructure — with an aging workforce. This infrastructure will require a significant amount of attention in order to meet the ever-growing reliability requirements of today and the energy demands of tomorrow. Recently, AEP has been placing increased emphasis on cost control, including operation and maintenance expenses as well as technical and physical workforce human resources. New tools and technologies are needed to compensate so that AEP can obtain the efficiencies needed to stay competitive in this ever-changing industry.


The most significant of AEP's underground systems in terms of cost, complexity and load served are the duct and manhole cable installations. Duct and manhole systems are prevalent in high-load-density areas and are designed to facilitate the installation and maintenance of underground cables. It is not uncommon to have several distribution and/or subtransmission feeders of various voltages and sizes passing through a common duct and manhole system. This presents a special challenge when trying to evaluate a feeder's load-carrying limitations. The amount of load that can be placed on a particular feeder is highly dependent on the loading of the other feeders that occupy the same duct system. Thus, to evaluate a particular feeder in a duct system, one must evaluate all the feeders in that system. The common question of “How much load can be placed on a specific feeder?” depends entirely on the loading of the other feeders common to that duct system and their physical relationship to the feeder in question.

As costs for constructing duct and manhole systems increase and right-of-way availability becomes more difficult to obtain, more detailed analytical approaches are needed to assist the engineers in determining more precisely what, where and when new or additional facilities will be needed.


By analyzing the transient response of duct installations, AEP engineers can take advantage of the time lag phenomena that exists between the temperature rise of the cable and the temperature rise of the surrounding duct system. Since not all the feeders in a typical duct installation peak at the same time, higher current ratings can be applied to cables because of the differing load-cycle characteristics with respect to time of the feeders involved. Transient analysis gives engineers a much broader perspective of a cable's load/temperature cycle. Where previously an engineer would have looked at making improvements for cables that exceeded their normal or emergency operating temperatures, the engineer can now consider deferring such improvements and, consequently, very costly expenditures. The transient analysis is not only beneficial in determining operating temperature violations, but also the duration of those violations. From this, the engineer can then assume an acceptable amount of risk in deferring expenditures if, for example, the operating temperature violation is only a few hours in duration for a peak loading period.

Manually conducting a transient analysis for a duct installation can require a significant amount of engineering resource time. However, the benefits of doing transient analysis are far too valuable to dismiss because of resource constraints. AEP found it needed to develop a system for engineers to evaluate duct installations quickly, accurately and consistently. The system needed to be automated and make use of information and infrastructures already in place. In addition, the system needed to allow AEP to prioritize and optimize its underground cable installations with regard to reliability and costs by effectively managing risk. And of course, the system needed to be inexpensive to implement.


In 2005, AEP solicited the resources of CYME International to develop the Cable Historical Operating Temperature Estimator (CHOTE) system. CHOTE is based on the field-proven cable thermal-analysis program CYMCAP. The CHOTE system was designed to automate the processes involved in conducting the transient cable analysis. To augment the CHOTE system, AEP developed the Dynamic Cable Ampacity Function (DCAF). This is a front end for CHOTE that scrubs the AEP load data to ensure its integrity and to format the data for input to CHOTE. The AEP load data presently exists in two forms: in 15-minute to 1-hour intervals collected digitally and in a single peak-demand reading for a given month. The interval load-data-collection system, which started during the mid-1990s is available for approximately 80% of underground feeders in the Columbus area. (At the time of this writing, this is the largest concentration of underground duct installations in the AEP system.) For feeders that do not yet have interval load data available, the other function of DCAF is to create a load-cycle profile from similar feeders that do have interval data available. DCAF scales a time-based load profile from the single peak monthly load value to create interval load data where there is none.

The CHOTE system then reads the historical load data and the AEP CYMCAP database, which contains the duct installation data, and — using the transient analysis function of CYMCAP — calculates the operating temperatures for the cables. A report on which cables have exceeded their respective normal and emergency operating temperatures, along with the duration, is then generated.

The CHOTE system keeps track of all the calculated cable operating temperatures with respect to time and keeps a continuous tally of the amount of time a cable exceeds its normal or emergency operating temperature for as long as the cable remains in the CHOTE system. Simply stated: Time interval load values in amps go into CHOTE, and the computation of cable temperatures with time come out. Initially, DCAF will be processed and input to CHOTE on a monthly basis, by which CHOTE will perform its calculations and process its reports on the DCAF monthly schedule. In addition, the CHOTE system will perform some basic what-if scenarios. For example, if an engineer wants to see the effects of adding or removing a cable to a duct installation or wants to see the effects of increasing or decreasing the load on a feeder or feeders, the CHOTE system can provide useful insight for those scenarios. The CHOTE system can even provide insight as to when, in the future, cables will reach their operating temperature parameters based on a given load growth value.

The benefits of implementing the CHOTE system are:

  • Cost savings for engineering resources
  • Informed engineering decisions to better manage risk
  • Better prioritization of capital investments
  • Deferment or redirection of capital investments.


During the initial phase of this project, AEP is using CHOTE to model its station exits that are in underground duct installations. The CHOTE reports will be ported out to an intranet website where engineers and dispatch centers will have universal access to the information. A near-term objective is to expand the scope of CHOTE to include the distribution station transformer tie cables and subtransmission cable routes. A longer-term goal is to have CYME incorporate a thermal-aging model for cables. This will further enhance the decision-making process as to when and where to enact system improvements for cable systems. Another goal is to enhance CHOTE to keep track of duct installation changes so that if, for example, an existing feeder has changed cable sizes, CHOTE knows to reset the time clock for that feeder aging while keeping tracking of the other feeders in the installation.

As data acquisition systems improve and expand, another development will be to conduct real-time cable analysis. This will be assisting engineers and dispatch centers to make faster and better-informed decisions concerning underground cable installations during emergency switching conditions.


The author would like to acknowledge the contributions of several AEP employees in IT Applications Support: William Cameron, manager, and Jaquie Weyers, Francis Strawler and Mark Kaschner. A special acknowledgment goes to Robert Zigmund for all his project management efforts.

The author also would like to thank several CYME International T&D employees for their assistance with this article: Christian Beauregard, director of business development, Patrick St-Roch, senior programmer, and Dr. Francisco de Leon, research and development.

Scott Konkus is a senior engineer, Network Systems Engineering, at American Electric Power. During his 27 years with AEP, Konkus has worked in Distribution Planning at Appalachian Power Co. (Roanoke, Virginia) and more recently in Network Systems Engineering at AEP Service Corp. (Columbus, Ohio). His present duties include the planning and design for the AEP underground networks in Columbus, Roanoke and Wheeling, West Virginia. Konkus is the lead engineer on the AEP Network Design Standards Committee. He holds a BSEE degree from Virginia Military Institute and a master's of administration from Lynchburg College. sjkonkus@aep.com