DA for a Rural Distribution Company
By Timo Laine, VTT Energy, and Erkki Antila, ABB Transmit Oy
North-Carelian Power Co. is a typical Finnish rural power distribution utility which operates on a rather sparsely populated area of more than 2,1000 km2. The neutral compensated, medium-voltage network, 9,100 km of overhead lines, is divided into 179 feeders on 32 HV/MV substations. Thus, an average feeder is approximately 50 km and many of the feeders have a length of more than 100 km. Often located in forested areas, the overhead lines are vulnerable to storms and, during the long winter, to snow burdens that bend the trees on the lines. Especially in the winter the outage frequency in the utility is rather high. At the end of 1995, the utility had 72,765 customers, most of which were residential and agricultural customers. Average load per kilometer of medium voltage line is low, and therefore the network is relatively expensive to construct and maintain. Thus, cost-benefit calculations of different network investments, including investments in distribution automation, are of substantial importance.
The 32 HV/MV substations, five power plants and more than 120 of the network disconnectors of North Carelian Power Co. are controlled from a control center in the town of Joensuu. In the control center, the utility has a VMS-based MicroSCADA that communicates with the HV/MV substations and power plants via either microwave links or fiber-optic cables. In addition to the usual functions of a SCADA system, real-time graphical display of the medium-voltage network switching state on a geographical map and calculated location of short-circuit faults have been implemented.
For communications with remote controlled disconnectors, a VHF-range radio network has been built. The utility also has a radio network in the same frequency range for communication needs of working crews. Separate networks are needed in order to prevent data and speech transmissions from disturbing one another. In addition to the substation telecontrol system, the utility has a distribution management system (DMS) and a customer information system (CIS). Although about 40 remote controllable disconnector stations have already been installed in the network, an assurance of the profitability of continuing the installation program was needed, and therefore the function was included in the study. Line switch remote control makes changing the network connection state substantially easier and faster. This is especially important when isolating faulty line sections from healthy networks and when restoring supply to the healthy sections of a faulted feeder. Time is also saved in routine connections under normal network operating conditions.
Computational Fault Location
If voltages and currents of a faulted feeder are registered in the substation bay level before and during the fault, an estimate of the distance of the fault from the station can be computed using this information. This applies for short-circuit faults and earth faults with low-fault resistance. For short-circuit faults, the computational location is already in use at North Carelian Power Co. Computational location of earth faults can be introduced using mainly the same software and hardware at the control center level. Because the data transmission connection between the control center and the HV/MV substations already exists, the only additional components needed are the current and voltage registration devices for the substations on which this function is implemented. Until just recently, registration units with an adequate sampling frequency for the needs of the calculated location of earth faults have not been available. For short-circuit fault location no additional equipment is needed because the necessary current data can easily be obtained from the protective relays.
Detection of High Resistance Earth Faults
A significant portion of outage-causing earth faults in overhead line networks originate from high-resistance earth faults with a small fault current. In many cases, a developing earth fault can be detected and located before it causes a triggering action of protective relays, the result of which is an outage. If the fault is found before an unplanned outage, the repairing actions can be planned, performed and timed, thus minimizing the damage to the customers. The detection of high-resistance earth faults is based on monitoring the busbar zero voltage at the HV/MV substations. A faulty line section can be located more accurately by moving line sections between feeders from different stations. If this is not possible, the faulty feeder can be found out by opening the circuit breakers one by one at the station. When the faulty feeder has been identified, the faulty line section can then be detected in a similar manner by opening network disconnectors. The latter procedure always causes outages to some customers and the timing of the switching must be considered carefully. Even if the faulty line section is known, the exact fault location cannot always be found by working crews. This decreases the value of this function to some extent.
Automatic Voltage Control
The load flows during the peak demand can be reduced by improved voltage control and by reactive power compensation. These functions are particularly important in rural systems where voltage drop is the limiting factor for distribution capacity. Improved voltage control allows the utility to reduce active and reactive power flows during the peak demand period by temporarily lowering the supply voltage. Essential for this function is a relatively accurate knowledge of voltages in the customers` locations so that the risk of violating the voltage limits set by quality standards is minimized. To implement this function, a real-time network calculation system is needed. A separate computer and the software that combines information from the utility`s databases, load models and network measurements must be acquired. In the case of North Carelian Power Co., all the necessary communication connections already exist because the actual changing of voltage is performed by controlling the HV/MV transformers` on-load tap changers.
Network Topology Optimization
By optimizing network switching state more frequently than the current practice, transmission power and energy losses can be reduced. The effective use of this function depends on how the variation of loads between feeders can be more effectively divided in the network. The density of remote controllable line switches in the network is of vital importance. This function is based on changing the open points of the network using remote controllable line switches.
Remote Metering and Load Control by Dynamic Tariffs
In this study, load control was assumed to be based only on the use of dynamic tariffs. Direct-load control has so far been found unprofitable in the utility. Remote metering and load control were assessed together because the two functions can be implemented using the same communication system.
In the case of a rural distribution company whose customers are distributed over a geographically large area, remote reading of all customers` meters is not justified. Because of the costs of telecommunication, only the largest customers can be included in the remote reading system. In this study, two alternative telecommunication techniques suitable for the purpose were considered: a distribution line carrier system and communication using a public-switched telephone network. If implemented, using a distribution line carrier, a central unit in the control center, a communication terminal unit on every HV/MV substation and customer terminals with appropriate meters is needed. The use of a switched telephone network only requires a PC with a modem and suitable software and, at the customers` locations, customer terminals with kWh meters.
Calculation of Costs and Benefits
To be able to compare the costs and benefits of different automation functions, applications were calculated per HV/MV substation feeder over an observation period of 10 years. All the numerical values of costs and benefits presented are present values of costs and benefits generated during the observation period.
The calculation was done using numerical values of an average feeder. These values and other utility specific data were collected in North Carelian Power Co. The cost estimates of system components including installation and maintenance costs were obtained from ABB Transmit Oy and other manufacturing companies. Values for outage costs in different customer groups were collected from a Nordic study.
The costs of automation consist of the costs of computer equipment and software, data transmission systems and automation-related network equipment. The costs can be further divided into purchase, installation, operation and maintenance costs. In this study, costs of already existing system components, such as DMS and CIS, were not taken into account. As an exception, the communication costs for the remote controlled line switches were included in full.
From the results, it can be seen that many of the assessed new automation functions are characteristically low-cost computer-system level functions. The implementation of these functions requires an increase in the metered real-time data from the network. In Finland, the newly introduced free electricity market is rapidly increasing the need for network measurements, and the same measurements can be used as a source of data for various automation applications, such as voltage control and various DSM functions. However, the introduction of some functions requires installation of additional specific measurement devices in the network. Calculated location of earth faults and detection of high resistance earth faults are essentially such functions.
All the functions studied appeared to be cost effective in the conditions prevailing at North Carelian Power Co. Hence, substantial savings can be achieved in the cost of electricity distribution. The customers benefit from this in the form of lower electricity prices, reduced supply outage cost and better voltage quality. From the power company`s point of view, savings in investment and labor costs can be achieved.
If you would like to see more articles on this topic, circle R.S. 101.
For more information on this article, circle R.S. 102.
Heavy snow conditions put a strain on North Carelian Power`s (Finland) distribution network.
Timo Laine and Erkki Antila won the Best DA Project at DA/DSM Europe`s Innovation Awards.