System design and protection standards coupled with the organizational structure are key to the fault-management performance of a transmission and distribution utility. As the center of all restoration actions, the System Control Center (SCC) applies and evaluates the merits and efficiency of system-management equipment and procedures. New Zealand comprises two major islands without having the benefit of international interconnections. For that reason, distribution companies take the SCC into consideration when designing and operating the transmission and distribution network.

POWER SYSTEM CHARACTERISTICS

The national transmission system operates at 220 kV and 110 kV, and is divided into two sections supplying the north and south islands. The two islands are linked electrically via a ±350-kV high-voltage direct-current submarine cable. Transmission substations, the points of supply for medium-voltage networks known as grid exit points, supply the sub-transmission (ST) network that operates mainly at 33 kV, with some sections of 22-kV underground cable.

The three-phase ST network comprises mainly underground cable in urban areas, a mix of underground cable and overhead line in semi-urban areas, and overhead lines in rural areas. The underground ST system is interconnected with zone transformers sharing common differential protection. The 11-kV feeder network allows load sharing in the event of transformer overloading. Overhead line ST networks are operated radially.

Zone substations have ST bus bars, which increase flexibility in the event of single-component element failures. These substations generally have one to four transformers with capacities ranging from 5 MVA to 30 MVA, and the units are equipped with on-load tap changers. The most common group is delta-wye grounded (DYn) and in some parts of the country, wye-wye grounded (YYn). Zone substations supply up to 20 radial feeders, each capable of being loaded up to 6 MVA. Urban feeders are shorter, mostly underground and almost completely interconnected with several backfeed options available. This is a necessary design feature because utilities use the backfeeds to solve ST network problems. The majority of rural feeders have some level of interconnection, unless they're supplying remote and isolated areas.

The feeder or distribution network operates at 11 kV with small sections still operating at 6.6 kV, but will be due for voltage upgrading to 11 kV in the near future. Wherever the fault level permits, the 11-kV bus bars are solid. The three-phase distribution system supplies secondary substations equipped with DYn transformers.

SYSTEM FAULT MANAGEMENT

The restoration of supply switching includes fault isolation, full and partial supply restoration where appropriate, network reconfiguration where required and full normalization after the event. Automatic switching systems are the first to respond if installed and subsequent to automatic switching, the SCC is responsible for coordination actions using supervisory control and data acquisition (SCADA) or other remotely controlled switchgear and field operators.

• Subtransmission

  ST networks have relatively simple configurations with a high level of redundancy. Protection is provided for each component, so a faulty component element is usually identified and isolated instantly by protection. The main concern is supply restoration following a fault outage. Load transfer and network reconfiguration are used to optimize overall security levels. These actions provide a prompt and efficient response to outages involving a large number of customers.

  A desirable feature of the after-fault network reconfiguration is the optimal sequence of load transfer in contingency conditions. Generally, feeders with major customers, high loadings and multiple customers have first priority, but the locations of field operators and traffic conditions must be assessed during emergency situations to minimize supply restoration time.

  Circulating current between substations is often higher in unusual or fault situations on ST systems. While this can cause additional problems, it also can be an advantage because it can ensure circuit-breaker tripping. Incoming feeders that remain overloaded following the ST system trip require a quick response to avoid a cascading loss of all remaining circuits.

  SCADA instantly recognizes and identifies feeder network faults and most of mid-feeder faults. Field crew(s) can then be dispatched to start the fault-location process initially driven by information provided by SCADA/DMS/OMS (distribution management systems/outage management system), which is continually updated from field crew information, fault-passage-indicator readings and reports from the customer call center. Fig. 1 shows the screen presentation of the data provided by SCADA/DMS/OMS.

  The method of performing trial switching for fault location depends on the network type, fault level and so forth. Once the fault has been located and the network isolated, fault repairs can begin. Fault-passage indicators are a simple and relatively inexpensive means of assisting in fault location and minimizing circuit-outage times.

  Effective coordination of appropriate field activities is a key driver in speeding up the process of fault location and isolation. In Fig. 2, a helicopter is used for high-voltage pole replacement. Utilities regularly employ remotely controlled switchgear, remote indication and measurements. The degree of remote control installed should be the subject of continual cost-benefit analysis. Vector created a network automation program involving the installation of 200 fully automated devices expected to produce a 10% reduction in the System Average Interruption Duration Index (SAIDI). The equipment now being installed includes G&W Electric Co. (Blue Island, Illinois, U.S.) viper reclosers (Fig. 3) and G&W rotary puffer SF₆ switches (Fig. 4) with Schweitzer Engineering Laboratories Inc. (SEL; Pullman, Washington, U.S.) relays and general packet radio system communications.

  Modern DMSs can analyze real-time data and provide topology information. If fully integrated with SCADA, geographic information system, protection and automation, they form a powerful management tool. Vector uses Spectrum Power TG (Siemens) SCADA and is now implementing SPL (Oracle) OMS/DMS. Linked together with various other databases, they will provide complete and efficient control and dispatch function.

• Distribution

  Low-voltage networks are now designed to comply with strict power-quality standards. To satisfy customers' high expectations, utilities need to use every available resource. Traditionally, low-voltage fault management relied on skilled, fully equipped and available fault-repair teams, but now to remain competitive, utilities have to strike a balance between operating costs, customer-service requirements, customer call center and field crew management. Low-voltage network fault investigation in progress is shown in Fig. 5.

Professional, efficient and customer-orientated call centers are a crucial and logical first step in the process and ensure a continuous two-way flow of necessary information. Supported by DMS prediction functions, the centers save time, relieve pressure and reduce the number of fault repair teams required.

The main field crew management task is to optimize field staff assignments geographically; therefore, real-time location (via GPS), communication facilities, OMS/DMS (that supplies the data to all relevant parties) form vital components of the fault-management system.

Vector has one control center for all areas. The company outsources the call center and divides the dispatch function between central dispatch, and separate respective subcontractors' own dispatches depending upon the time of day. A call center agent immediately answers calls from customers experiencing individual low-voltage faults (Fig. 6). However, in the event of widespread fault, upfront messaging updates customers without the need for an agent. Vector has also adapted Siebel (Oracle) customer-management system to integrate with DMS. This allows agents to have access to current two-way flow of information from the field within one software package.

Some of Vector's customers have no direct contact with the company in the event of a fault due to their energy retailer contracts. However, in these areas, the customers call their energy retailer, which forwards the jobs to Vector electronically. Vector has developed and installed these systems to allow for the two-way transfer of status information. Currently, however, not all retailers have integrated the information flow back into their systems.

NETWORK OPERATING CONDITIONS

The classic objective function, namely to minimize technical losses for network via operational optimization, is a traditional starting point. Voltage upgrading, coupled with reliability analysis, can cut costs, but it also imposes conditions on how the system operates particularly in an abnormal state. Therefore, quality checking of off-loading paths should be incorporated into the management system software. The relocation of open points can improve the load distribution for normal operation and, in particular, in the established contingency plans if the load-transfer capacity of each feeder does not exceed the maximum rating.

After reaching the satisfactory solution via the normal iterative process, normal pragmatic constrains would apply. These include easy and permanent accessibility for the switching, no major or sensitive customers connected to a high-fault-risk feeder section and not a part of an automated feeder. The company also closely monitors loading on older cable with a history of faults.

PERFORMANCE INDICATORS

To obtain a correct and useful overview of the present state, and the outcome of applied improvements and strategies, system outages need to be statistically analyzed according to the nature and duration. Planned outages need to be extracted from the faults, which then need to be sub-divided into transient faults (less than a few seconds) or permanent faults.

The internationally recognized basic system performance indicators for non-transient outages are interruption, duration and frequency presented in the form of values for SAIDI, System Average Interruption Frequency Index (SAIFI) and Customer Average Interruption Duration Index (CAIDI). Depending upon the characteristics of the particular network and the utility, further analyses are necessary to identify the type and cause of the fault, most affected feeders and so forth. Further targeted analyses should follow various implemented system improvements, changes in the maintenance pattern and all load and nonload-related investment on assets specifically installed to improve system security and reliability.

Fig. 7 shows the effect of the performance improvement in terms of the reduction in SAIDI minutes on the customer satisfaction survey score from July 2004 to July 2006.

Advanced fault-management applications directly benefit the customers and the distribution utility through reductions in outage costs, operating costs and savings in the level of investment required to improve overall system performance. Improvement of power quality is a significant benefit that improves the utility's position in the energy-trading market, irrespective of the prevailing rules and conditions.

ACKNOWLEDGEMENT

The author wishes to acknowledge the help and assistance given by John Moore and the Vector team in the preparation of this article.

---

Nenad Puljic who earned B.Sc. and M.Sc. degrees from the Faculty of Electrical Engineering and Computing in Zagreb, Croatia served as an operations engineer with Vector Limited (Auckland, New Zealand) for 11 years. Puljic has more than 20 years of experience in transmission system and distribution network operations. He was recently appointed control centre manager with WEL Networks Limited (Hamilton, New Zealand).
nenad.puljic@wel.co.nz