Contact us
Kansas State University
Electrical and Computer Engineering
2061 Rathbone Hall
Manhattan, KS 66506
Phone: (785) 532-5600
Fax: (785) 532-1188
Hours: 8 am-12pm, 1pm-5pm M-F

Chapter 3

Outage Management


Outage management in a distribution system includes activities associated with location of outages and restoration of service to the customers who have lost power due to the outage. In an automated environment, outage management requires some form of data from the system. Many utilities depend on customers to report an outage via telephone to a trouble call analysis system. The customers are asked to report their address and telephone number. The collected data is then used to determine a probable location of problem based on heuristic or fuzzy reasoning. Reporting of outage could be automated by installing a communication device at customer-ends. Such data also could be obtained in conjunction with automated meter reading, which is being contemplated by some utilities. The task of fault location becomes easier and more precise if time of outage at customer-ends can be obtained. In this chapter we will discuss three different methods, which use time-of-outage data, for location of outages. From theoretical point of view these methods work but from practical point of view there are several questions. First of all, availability of timing devices needed is questionable. Secondly, there is a possibility of congestion on the communication channel if many devices start communication simultaneously with the central control system immediately following the outage. These issues will have to be resolved before the methods discussed in these papers can be implemented.

Restoration activities following the location of outages depend on the type of outage. If it is a single outage in a part of the distribution system, then usually a few switching operations are required to restore power to the customers. However, a few customers may still be without power until repairs on the affected circuits are made. The objective of the restoration scheme in such cases is to use a switching sequence which can be performed quickly without overloading the facilities. If the outges are widespread such as those due to severe weather conditions, it takes much longer to repair the circuits and restore the system. In those cases the utilities must take into consideration the impacts of cold load pickup (CLPU) , which is a condition created by saturation of thermostatically-controlled devices. In this chapter we will discuss theses impacts and discuss strategies to overcome the adverse effects of cold load pickup.

 


3.1 Location of Outages


Refer to the following papers:

  1. R. Balakrishnan and A. Pahwa,"A Computer Assisted Intelligent Storm Outage Evaluator for Power Distribution Systems," IEEE Transactions on Power Delivery, July 1990, pp. 1591-1597.
  2. P. Deepal Rodrigo and Anil Pahwa, "Location of Outages in Distribution Systems Based on Statistical Hypotheses Testing," IEEE Transactions on Power Delivery, January 1996, pp.546-551.
  3. P. Deepal Rodrigo and Anil Pahwa, "A Modified Least-Mean-Square-Error Algorithm for Location of Outages in Distribution Systems," submitted to Electric Machines and Power Systems.

3.2 Restoration Following Extended Outages


The load upon restoration following an extended outage is generally much larger than the normal load due to loss of diversity amongst loads, particularly thermostatically controlled loads. We will use the term diversified load for the load in normal state and undiversified load for the load upon restoration. Because the undiversified load of the system is larger than the diversified load, restoration problems may occur when there is not enough reserve transformer capacity in the substation. Presently, most of the utilities provide enough transformer capacity in the distribution systems; therefore, large undiversified load does not cause any problems. However, in the future, with increase in automation of distribution systems, the reserve margin of substation capacity will decrease. The decreased margin will require different ways of restoration following extended and widespread outages. Although such occurrences are not common, the effects could be severe and may last for a long time. Therefore, it is important to explore alternatives for restoration of the system as fast as possible to improve reliability and customer satisfaction.

One possible approach is to install remotely operated sectionalizer switches on the main feeder to split the system into several sections. Then during restoration the sections can be restored in steps. Load behavior of sections and restoration sequence of these sections play an important role in the restoration procedure. Based on the load dynamics of each section, the restoration procedure should be chosen in such a way that some restoration objectives are met. For example, one of these objectives is to minimize customer interruption duration because it directly affects the reliability of the system. The shorter the customer interruption is, the more reliable the system will be.

3.2.1 Residential Load Behavior After an Interruption

Individual loads on a residential feeder can be categorized into two different groups: thermostatically-controlled and manually-controlled. In general, thermostatically-controlled devices such as air-conditioners, heaters, and heat pumps provide the largest contribution to the total load in a typical house. Manually-controlled loads are switched on and off by occupants of the house in an undetermined fashion. The life-style of the occupants of the house has a significant influence on the contribution of these loads to the total load of the house.

During normal conditions, diversity among loads is present, and therefore, the aggregate load of a number of houses is less than the connected load. If an abnormal condition such as when an extended outage occurs in a distribution system, some or all thermostatically-controlled devices will be on as soon as the power is restored. Similarly, the aggregate load of manually-controlled devices will be higher than normal upon restoration because more people may want to use different devices. If an outage involves a large number of customers and has a long duration, it may result in excessive load during restoration. Restoring power to a circuit under such conditions is called cold load pickup.

CLPU currents can be categorized into four phases according to the current levels and durations. These phases are inrush, motor starting, motor running and enduring current phases. The first three phases last approximately less than 15 seconds and the current may reach 5 to 15 times of the pre-outage current [1-4]. The enduring current phase follows the third phase and continues until the normal diversity amongst the loads is re-established. The load in this phase may vary from 2 to 5 times of the diversified load level. This phase may last for several hours depending on outage time and outside temperature. The magnitude and duration of load during CLPU will depend on the following factors: outside temperature, duration of outage, the type and ratings of devices.

CLPU appeared first in the literature in 1940's as a problem related to high inrush currents that last a few seconds and prevent the circuit from being re-energized after extended outages. Application of very inverse characteristic relays or sectionalizing the distribution systems were some of the solutions engineers used to overcome this problem. Since then penetration of thermostatically-controlled devices in distribution systems has increased [1, 5]. These types of loads may cause restoration problems during CLPU before they cause serious overload problems in normal operation. Therefore, sustained load after restoration becomes an important issue for loading limitations of distribution equipment. Although enduring component was not a problem in 1940's and 1950's, Oliver Ramaur mentioned in his paper in 1952 that the enduring component might limit the amount of load that could be picked up at once [4] .

Utilities do not encounter CLPU problems often. Therefore, an increase in thermostatically-controlled devices in a distribution system may go unnoticed; but during CLPU this type of loads may exceed system capacity. In that case, one of the traditional ways of dealing with the CLPU problem is sectionalizing the system and restoring power to the sections using manual switches. For example, manually-operated sectionalizers have been used successfully to deal with the enduring portion of CLPU [6]. The operator has to coordinate the field personnel using radio communication to effectively complete switching operations. With remote control more precise and timely switching can be exercised. Some utilities have tested remote control of sectionalizers as a part of demonstration projects dealing with distribution automation. In these studies, time of operation of sectionalizers during an actual restoration is based on trial and error. If closing of a sectionalizer results in excessive load, it is opened and reclosed after a time delay. The operator can read meters to obtain field data as well as operate sectionalizers remotely from the control center.

3.2.2 A Review of CLPU Dynamics

Enduring current of CLPU, which is a result of the loss of diversity, did not capture much attention in the 1940's and 1950's. Partially, this happened because this current was not as high as inrush and motor starting currents. Also, long duration of enduring current did not force the thermal limits of distribution equipment because of large margins between substation capacity and system load. Large margins were necessary for high system reliability since the distribution systems were in infancy state and load supply from other substations was either very limited or did not exist.

It was 1970's when further investigation on the enduring component of CLPU was initiated by a few researchers. In 1979, McDonald, Bruning, and Mahieu [1] monitored electrically heated homes to predict the magnitude and the duration of the peak demand following a power outage in cold weather. Their work was one of the first attempts to predict experimentally magnitude and duration of peak demand as a function of outside temperature and outage duration. The results of this study were very limited and could only be applied to the systems with similar type of loads. Miller, Serhal, and Morris have used a modified version of this model in their cold load pickup study[7].

Also, in the late 1970's pioneering work on physically-based modeling of loads was initiated by many researchers under sponsorship of the Department of Energy. The idea was to develop a physically-based load model for individual loads as a function of weather and human use patterns. This was followed by aggregation of these loads to find the total demand. Aggregation was generally was based on a model based on stochastic theory [8-20]. Mortensen and Haggerty [21]present and discuss five such mathematical models that have been studied in the literature. Aggregated load behavior, when a large number of customers are considered, can be determined based on these models either by use of numerical techniques to solve partial differential equations derived from individual load model or by Monte Carlo simulation based on the stochastic difference equation. Some on these models will be discussed in the following sections.

On the other hand, instead of stochastic and detailed models, simplified models have been used to find the effect of CLPU on substation and distribution transformers. Because thermal response of a transformer to a load is slow, simplified models are sufficient to analyze loading capabilities. Wilde [22]investigated the effects of CLPU on the substation transformer using a model in which post-outage load is constant for some time and then decreases linearly from undiversified load to diversified load. J. Aubin, R. Bergeron, and R. Morin did a similar study using a piece-wise linear CLPU model to find the overloading capability of distribution transformers [23].


The rest of Chapter 3 is available as a postscript document: click here to download

Chapter3-2

References

  1. J.E. McDonald, A.M. Bruning, and W.R. Mahieu, "Cold load pickup," IEEE Transactions on Power Apparatus and Systems, vol. PAS-98, pp. 1384-1386, July/August 1979.
  2. C.E. Hartay and C.J. Couy, "Diversity: A new problem in feeder pickup," Electric Light and Power, pp. 142-146, October 1952.
  3. R.S. Smithley, "Normal relay settings handle cold load," Electrical World, pp. 52-54, June 15, 1959.
  4. Oliver Ramsaur, "A new approach to cold load restoration," Electrical World, pp. 101-103, October 6, 1952.
  5. W.W. Lang, M.D. Anderson, and D.R. Fannin, "An analytical method for quantifying the electrical space heating component of a cold load pick up," IEEE Transactions on Power Apparatus and Systems, vol. PAS-101, pp. 924-932, April 1982.
  6. D.W. Butts, Winter cold load pickup study, Technical report, Planning Department - Illinois Power Company, Decatur, Illinois, Feb. 1979.
  7. William H. Miller, Ali S. Serhal, and Emanuel Morris, "Cold load prediction in electrically heated homes," Proceedings of the American Power Conference, vol. 48, pp. 495-500. Chicago, Illinois Institute of Technology, April 16, 1986.
  8. C.Y. Chong and A.S. Debs, "Statistical synthesis of power system functional load models," Proc. IEEE Conf. Decision Control, pp. 264-269, Fort Lauderdale, FL., 1979, session WP-4.
  9. S. Ihara and F.C. Schweppe, "Physically based modeling of cold load pickup," IEEE Transactions on Power Apparatus and Systems, vol. PAS-100, pp. 4142-4150, September 1981.
  10. C.Y. Chong and R.P. Malhamé‚ "Statistical synthesis of physically based load models with applications to cold load pickup," IEEE Transactions on Power Apparatus and Systems, vol. PAS-103, pp. 1612-1628, July 1984.
  11. R. Malhamé‚ and C.Y. Chong, "Electric load model synthesis by diffusion approximation of a high-order hybrid-state stochastic system," IEEE Transactions on Automatic Control, vol. AC-30,pp. 854-860, September 1985.
  12. R.E. Mortensen and K.P. Haggerty, "A stochastic computer model for heating and cooling loads," IEEE Transactions on Power Systems, vol. 3, pp. 1213-1219, August 1988.
  13. C. Alvarez, R. Malhamé‚ and A. Gabaldón, "A class of models for load management application and evaluation revisited," IEEE Transactions on Power Systems, vol. 7, pp. 1435-1443, November 1992.
  14. J.C. Laurent and R.P. Malhamé, " A physically-based computer model of aggregate electric water heating loads," Paper no. 93 SM 496-0 PWRS, IEEE PES Summer Meeting, Vancouver, B.C., Canada, 1993.
  15. C.W. III Brice and S.K. Jones, Physically-based load modeling," Technical Report for DOE Contract EC 77-5-01-5057, Electrical Engineering Department, Texas A&M University, 1978.
  16. T.M. Calloway and C.W. Brice, "Physically-based model of demand with applications to load management assessment and load forecasting," IEEE Transactions on Power Apparatus and Systems, vol. PAS-101, pp. 4625-4631, December 1982.
  17. A. Pahwa, Physical Stochastic Modeling of Power System Loads: Modeling and System Identification of Residential Air Conditioning System, Ph.D. thesis, Texas A&M University, Dec. 1983.
  18. A. Pahwa and C.W. Brice, "Modeling and system identification of residential air conditioning load," IEEE Transactions on Power Apparatus and Systems, vol. PAS-104, pp. 1418-1425, June 1985.
  19. T. Roy, "A diffusion approximation approach to stochastic modeling of air conditioner type loads," Masters Thesis, Texas A&M University, 1981.
  20. Irvin C. Schick, Patrick B. Usoro, Michael F. Ruane, and Fred C. Schweppe, "Modeling and weather-normalization of whole-house metered data for residential end-use load shape estimation," IEEE Transaction on Power Systems, vol. 3, pp. 213-219, February 1998.
  21. R.E. Mortensen and K.P. Haggerty, "Dynamics of heating and cooling loads: Models, simulation, and actual utility data," IEEE Transactions on Power Systems, vol. 5, pp. 243-249, February 1990.
  22. R.L. Wilde, "Effects of cold load pickup at the distribution substation transformer," IEEE Transactions on Power Apparatus and Systems, vol. PAS-104, pp. 704-710, March 1985.
  23. J. Aubin, R. Bergeron and R. Morin, "Distribution transformer overloading capability under cold-load pickup conditions," IEEE Transactions on Power Delivery, vol. 5, pp. 1883-1891, November 1990.
  24. M.L. Chan, G. Ackerman, et. al., "Simulation-based load synthesis methodology for evaluating load-management programs," IEEE Transactions on Power Apparatus and Systems, vol. PAS-100, pp. 1771-1778, April 1981.
  25. M.N. Nehrir, P.S. Dolan, V. Gerez, and W.J. Jameson, "Development and validation of a physically-based computer model for predicting winter electric heating loads," Paper No. 94 WM 228-7 PWRS,IEEE PES Winter Meeting, New York, NY, 1994.
  26. D. Athow and J. Law, "Development and applications of a random variable model for cold pickup," Paper No. 94 WM 073-7 PWRD, IEEE PES Winter Meeting, New York, NY, 1994.
  27. P.S. Dolan and M.H. Nehrir, "Development of a residential electric water heater model using energy flow analysis techniques," The Proceedings of the 24th Annual North American Power Symposium, pp. 272-277, Reno, Nevada, October 5-6, 1992.
  28. J. Woodard, Electric Load Meeting, Ph.D. theses, M.I.T., 1974.