IN 1998, EESTI ENERGIA (EE), WHICH OWNED THE EXISTING FIVE ELECTRICITY DISTRIBUTION COMPANIES IN ESTONIA, was restructured and reorganized. The Estonian distribution network company was combined with the transmission system company, which served seven regions, and together they now supply 91% of the country's population living in 97% of Estonia's geographical area. Figure 1 shows the distribution companies before and after the 1998 restructuring. Currently, the key statistics for the EE distribution network are as follows:

• Maximum load: 1200 MW
• Annual increase in energy consumption: 3%
• Distribution network voltages: 35 kV, 20 kV, 15 kV, 10 kV, 6 kV and 0.4 kV
• Distribution substations medium/low voltage (MV/LV): 18,442
• Length of MV overhead network: 18,480 km (11,485 miles)
• Length of MV underground network: 3096 km (1924 miles)
• Length of LV overhead network: 33,825 km (21,022 miles)
• Length of LV underground network: 4750 km (2592 miles).

To ensure that the new organization was unified and efficient, a geographical network information system (NIS) was needed. Prior to the industry restructuring, practically all of the network records were retained in paper format and the use of computerized systems was minimal. Therefore, in the autumn of 1998, EE put out a bid for the supply and implementation of an integrated distribution NIS for network documentation, load dispatch, planning and maintenance processes.

NETWORK INFORMATION SYSTEM SELECTION

As a result of the procurement process, EE selected Xpower NIS, marketed by Tekla (Espoo, Finland). Tekla offered EE two packages:

• The first package, SmallWorld GIS, was to establish a geographical information system (GIS) environment where the majority of the necessary functions and modules could be developed in-house or by a third party.

• The second package, Tekla Xpower, was a readily available geographical-based NIS that required some configuration. The advantage of this package was Tekla's knowledge, long-term experience and the customer base that would be built into the data model and applications. After assessing that the risk of investing in a rigid system was lower than the risk of delaying the project due to system specification problems, EE selected this package.

THE NIS IMPLEMENTATION PROCESS

Implementation of the Xpower NIS started in 2000. The first tasks were to define the data specification (objects and data fields to be collected) and start the data acquisition subproject. Little reliable data was available for direct conversion into the new system, and it was necessary to undertake a visible inspection of practically the entire distribution network. Approximately 50% of the fieldwork inspections were outsourced to mapping companies, retired personnel and students. The information and basic location data were collected and recorded in tables and on maps, and the Estonian Mapping Center provided network location data. Missing records on cable locations were acquired using cable detectors. These data sources and data input were combined and processed in seven regional centers.

One of the project's key objectives was to integrate customer data with geographical network data. This was done by address geo-coding and customer-point location measurement using global positioning satellite (GPS).

NIS FEATURES IN OPERATION

Now, six years after the decision to invest in Xpower, EE's systematic data acquisition is almost complete; however, data quality improvement and error correction will continue for a few more years. To date, 95% of the customer-point or service records are already connected to the network model.

Xpower is now fully operational for the following system management processes:

• Network documentation, load-dispatching distribution management system (DMS), network planning and calculations. Pilot studies are underway that include network construction, planning applications and a mobile solution for maintenance data collection (Fig. 2).

• DMS is integrated with supervisory control and data acquisition (SCADA) and, therefore, is able to obtain state information for more than 5800 circuit breakers.

• The DMS application is currently the most critical application because it culls EE's customer interruption statistics. This is invaluable because in 2005 EE began compensating customers who experience power outages that extend beyond 24 hours. Also, DMS is used in 14 dispatch centers to manage all MV and LV network switching operations for planned and unplanned (fault) outages.

• The network planning process helps analyze the existing network and provides comparisons with alternative development solutions. Because resources are tight and the need for investment is high, the key success factors are investment timing and priorities.

• The network calculation module evaluates MV/LV short-circuit and load-flow conditions using customer-point yearly consumption data in combination with the standard load curves for estimated load levels. The calculated results are compared with the feeder load-current measurements available from high-voltage (HV)/MV substations to assess their accuracy.

Soon DMS will be used to list the customers affected by EE's planned outages. This summer, development trials will be underway in one of EE's regions, whereby customers will receive a postcard stating the time and period of the planned supply interruption. Also, to further expand the system usage and ensure data quality, EE is testing the Xpower mobile application, which could be used for programming network condition inspections and for basic network data checks.

NIS PROJECT INVESTMENT AND BENEFITS

To date, EE estimates it has invested 5 million euros (US$5.92 million) on this NIS project. The largest component of this investment was the cost of data collection, input and user training. Internally, there has been constant pressure by the management to see tangible results, particularly when it appeared that the project was taking longer than originally planned. Initially following the decision to invest in this project, EE selected a team comprising of management and end-user specialist staff who were able to foresee the benefits of this project, which introduced several changes to existing work practices.

Xpower has given EE improved network data quality and availability, which has resulted in improved investment decisions. Planned and unplanned distribution network outages can be managed more effectively, reducing the outage time and frequency occurrences. Furthermore, EE now compensates customers who experience outages lasting more than 24 hours.

EE has been able to benefit from some downsizing of staff attributable to restructuring the control of the distribution network, as the number of load dispatch centers was reduced from 17 to 11. In the next five years, further rationalization is planned that will leave EE with seven regional load dispatch centers and one central dispatch center for the distribution network.

During the past five to six years, EE has reduced the number of field operatives as a result of outsourcing construction, maintenance and fault-repair activities. Although EE is experiencing a rapid growth of new connections, it anticipates that further modules of Xpower NIS will provide the tool to execute the growth in planning and improve service quality without the need for additional staff.

The electricity distribution market is regulated, and EE now has a computerized process for analyzing network losses, which are subject to strict monitoring in order to comply with the levels agreed upon with the Estonian Energy Market Regulator (Fig. 3). For example, the measured 10.7% losses in 2005 must be reduced to 8% in 2009.

---

Jaanus Kivistu earned his BSEE degree from Tallinn Technical University in Estonia in 1998, just prior to joining Eesti Energia. Kivistu has served as NIS project manager, system support and system manager in the network development department. gismo.mail@mail.ee