Pilot project of the electric utility of the city of Zagreb - experiences
Snježana Blagajac, Slavko Krajcar, Davor Škrlec
Introduction
The planning and the design of the electrical supply system are everyday tasks for engineers in the electric utilities. The goal of power distribution system planning is to satisfy the growing and changing system load demand during the planning period within operational constraints and with minimal costs. The planning process comprises several phases, and one of the most important is the optimization of the electric distribution network. The network optimization is considered a hard combinatorial optimization problem due to a number of limitations (network voltage level, network structure, amounts and locations of loads, routes and types of feeders, voltage drops, etc.). An additional complexity is imposed by the geographically referenced data. In this process it is important to have on time accurate relevant (related) data and information on the electric distribution system and its assets, and possibly to have data from other utilities. Computerization and development of various geographic information systems have opened new horizons for all decision-making processes as well as for manipulation and dissemination of information. Therefore, the electric utility of the city of Zagreb has started the GIS pilot project. This paper presents the development and experiences of the pilot GIS project of the city of Zagreb’s electric utility ELEKTRA.
The background
Zagreb is the capital of Croatia and it is the largest city in the country with more than a million inhabitants. The electric utility ELEKTRA as a part of the national power company HEP is responsible for distribution of electric energy at medium and low voltage level, covering the urban area of the Zagreb and several small satellite cities in Zagreb’s vicinity. ELEKTRA serves more then 400,000 customers spreading over the area of about 2,550 sq. km. The utility is organized into several operating distribution districts: Zagreb – the main and central operating district, Sveta Klara, Velika Gorica, Sveti Ivan Zelina, Dugo Selo, Zaprešić and Samobor. Its distribution network consists of 39 supply points (substations 110/10(20) kV and 30(35)/10 kV), 3180 distribution substations (10(20)/0.4 kV) and more then 13000 km of cables and overhead (aerial) lines.
The whole distribution network area is presented on maps of several different scales (1:5000, 1:2000, 1:1000 and 1:500). They contain considerably different level of details. The maps of scale 1:5000 contain limited set of data and information on medium voltage (MV) electric distribution network:
Each substation symbol is labeled with the unique name of the substation and each cable/line route is marked with the unique cable/line number. These maps also show the connectivity between the elements of MV network. The maps of scale 1:2000 similarly contain only the basic elements and information about low voltage (LV) cable and overhead electric distribution networks. Unfortunately, these maps are not complete for the whole town area and that is the major problem in their usage. On the other hand, maps of scale 1:1000 and 1:500 are surveying maps with accurate geographical positions, elevations and connectivity of all MV and LV networks’ elements. Therefore, they contain much more details and represent the best source of data and information. All these maps are created, periodically updated and managed by the utility’s technical documentation department. It has been obvious that utility needs a modern information system and a tool to manage everyday tasks related to operating, maintaining, planning, design and actual expansion of the networks [1,2].
Before GIS Pilot …
Many electric utilities have built technical information systems, usually based on relational database management system, to cover their requirements considering maintenance management and documentation, everyday operation and network extension planning. Our experiences from ELEKTRA point out the fact that their data related to distribution network planning are held within the utility in different relational databases and files, using different operating systems and user interfaces. These ”systems” are overlapped but not synchronized and not even regularly updated. Other problems related to them are: holding the large amount of data, the data structure, relations among data and the great effort required to collect, input and verify the data.
Usage of alphanumeric data as input for network planning is usually accompanied with the lack of information on existing network topology, elements connectivity, geographical locations of the network elements, possible new cable routes and the complete lack of possibility to perform spatial queries and analyses. This proves the relational databases without some improvements and additions are not very well suitable for network planning. In the late 1980s, in collaboration with Faculty of Electrical Engineering and Computing (FER), the utility developed the distribution network planning application named CADDiN. It uses system of ASCII files and relational databases upgraded with AutoCAD. The role of CAD system in such hybrid environment encompasses graphical displaying of network elements with their basic data, graphical displaying planning solutions and partially it is used for input and editing. The main drawback of this application is still missing the capability of spatially related querying and spatial analyses.
The scope of the pilot project
To successfully define, design and implement an AM/FM/GIS project it is necessary to well define objectives that specify the expected results. In addition, the projects should have well defined methodology to achieve those results [3,4]. When these ”ingredients” are present a project can be defined, designed and implemented in several years (let us say three to five years) for a single utility model. Nevertheless, it should not be forgotten that large number of projects are managed and implemented by people with little or no experience in AM/FM/GIS. To successfully manage and implement such projects ”in-the-trenches” experience is often required.
The GIS project has been originally initiated from the utility’s distribution network planning department. The goal of the initiative has been development of GIS based planning tool that will support decisions regarding planning and the design of MV and LV distribution networks. The project has been intended as the joint project of FER and ELEKTRA. It has been decided that the technical documentation department of ELEKTRA will create the GIS database of utility’s network assets and will be responsible for its maintenance. The group from the FER has to develop algorithms for optimization and planning the MV network. Following the completion of the GIS database the group will also connect the algorithms with GIS.
At the very beginning the utility’s participants knew little about GIS technology, but in time they have been educated through a serial of courses. Due to the limited finance and time, bearing in mind the aforementioned environment, the scope of the project has been restricted to the pilot type. The following objectives of the pilot project have been determined:
The spatial extent of the pilot project includes only the 9 maps of scale 1:5000 from the center of Zagreb. Because of the network structure and density this extent covers approximately 35% of the MV network. The project has started during 1995. The database model and data capturing should have been done in one year.
Database model and data structure
Table 1 shows the structure of the data model, objects with their topologic type and associated attributes.
Table 1. Data structure of MV network model
Object: substation 110/x kV |
Topologic type: node |
|
Attributes: |
name, address, description, medium voltage, installed capacity, designed capacity, the year of construction, the year of reconstruction, voltage level of equipment, no. of HV connections, no. of MV connections, no. of free MV connections |
|
Object: substation 30(35)/10 kV |
Topologic type: node |
|
Attributes: |
name, address, description, high voltage, installed capacity, designed capacity, the year of construction, the year of reconstruction, voltage level of equipment, no. of HV connections, no. of MV connections, no. of free MV connections |
|
Object: secondary substation |
Topologic type: node |
|
Attributes: |
name, address, description, installed capacity, type of substation, proprietary, transformer_1, power rate of transformer_1, transformer_2, power rate of transformer_2, the year of construction, the year of reconstruction, no. of MV connections, no. of reserved MV connections, no. of free MV connections, no. of LV connections, no. of reserved LV connections, no. of free LV connections |
|
Object: new substation |
Topologic type: node |
|
Attributes: |
name, designed capacity |
|
Object: auxiliary node |
Topologic type: node |
|
Attributes: |
name |
|
Object: cable |
Topologic type: line, comprises segment and joints |
|
Attributes: |
number, length, operating voltage, no. of segments, name of the first substation, name of the second substation |
|
Object: line |
Topologic type: line |
|
Attributes: |
number, length, construction type, material, cross section, year of mounting on the poles, construction voltage level, no. of cores |
|
Object: cable segment |
Topologic type: line |
|
Attributes: |
number, length, construction type, material, cross section, year of laying in the ground, construction voltage level, no. of cores, calculated length |
|
Object: cable joint |
Topologic type: node |
|
Attributes: |
type of joint |
|
Object: centerline |
Topologic type: line |
|
Attributes: |
length (=calculated 2D length) |
The basic set of MV network data has been digitized from scanned maps in the background, and checked against the digital maps from the Cadastre of Zagreb. Besides the spatial and attributes data the network topology has been established. The maps of scale 1:5000 are chosen for the following reasons:
The rest of the information on network assets has been extracted from existing files and documents locally distributed over different departments of ELEKTRA, and loaded into GIS database by specially developed procedures [5]. Figure 1 shows a part of the MV network from GIS database.
Figure 1. A part of the Zagreb’s MV distribution network
… And After the GIS Pilot
GIS database provides the source of sufficiently accurate data on medium-voltage distribution network assets. Its spatial data model comprises topological information on network elements (connectivity, sharing, adjacency, proximity, overlapping, etc.), providing the foundation for spatial analyses and reasoning. The accurate positions and lengths of the existing cables and all their segments in the cable route between two substations are defined. The exact locations of existing and proposed locations of future-planned substations, as a result of load growth analyses and predictions, are known. The possible routes for new cables in keeping with urban zoning plans, ecological and aesthetic constraints are also known from database spatial queries and graphic presentation of observed geographic area. The knowledge of precise cable lengths (i.e. segments of the cables between substations) and other corresponding attributes (e.g. the year the cable segment was laid down, construction type, material, cross section, operating voltage) for the existing cables and selected set of possible new cable routes enables the better electrical analyses and planning of the network.
GIS mechanisms of spatial and non-spatial querying enable us to reach the relevant data and information for the process of optimization and planning the distribution network. The same mechanisms are applied to analyses of interim or final planning results. The GIS is also used as graphical editor for displaying and taking interactive actions during the whole planning procedure. Finally, GIS is used to display and, after quality assurance from planning management, to permanently store the obtained results in the database.
Functionally the process of optimization the MV network in the new CADDiN-GIS application consists of several phases where GIS plays distinct role:
It encompasses the optimization of loop and link structures of distribution network, and it is equally used in both cases. The actual optimization of the network and its connectivity costs prepared by GIS is done employing the genetic algorithms (GA) [6,7]. The GA based optimization of loop and link structured networks is fully described in [8,9,10,11].
Conclusion
Many AM/FM/GIS projects take too much time to define, design and implement because many of them don’t start out with well-defined objectives that specify the expected results. The projects may also not have well defined methodology to achieve the results. On the opposite, organizations with projects that have well defined goals often expect to accomplish too much. There are limits to how much can be achieved within a reasonable amount of time with the knowledge that the participants of the project have.
The authors’ experiences derived from the pilot project presented here are following:
In spite of the mentioned difficulties GIS is a necessity and natural tool for every electric utility starting from technical documentation and maintenance, operating and managing the network to the future network planning. It has been decided to develop and manage the enterprise-wide interoperable GIS project in ELEKTRA Zagreb following the world GIS trends and the achieved results. Fortunately, the overall situation and GIS awareness in Zagreb are now more progressive then several years ago.
References