J B Barnard
Eskom Distribution, Eastern Region
ABSTRACT
The South African government has set a target of universal access to basic electricity by the year 2012. Eskom Distribution, in conjunction with local municipalities, is responsible for the outstanding electrification predominantly in rural areas. In KwaZulu-Natal, mountainous terrain and scattered settlement patterns of communities complicate the achievement of this goal. This study was aimed at showing how GIS can be used to design the shortest networks from the grid to an electrification area.
Designing the shortest path involved the use of four factors, these being roads, land cover, household positions and slope. Each layer was reclassified, ratings applied and a combined suitability raster created in order to successfully design the final path.
Extensive research into GIS practices worldwide, resolving of data quality issues, testing of points and weighting systems, and working together on making changes to age-old system structures and processes is needed before any of the recommendations resulting from this study can be effectively implemented.
INTRODUCTION
In order to determine the shortest path for an electricity network extension different network planning concerns need to be addressed. Currently network planners work manually, using maps and aerial photographs in order to find the optimal routes for network extensions. The more factors considered, the longer it would take to evaluate the maps and determine the shortest path manually. Electricity utilities and other route planners are turning more and more to using GIS as an effective alternative (Williams et al, 2003, Glasgow et al, 2004).
Using GIS a model can be created to overlay any number of requirements into one multi-criteria map which is created by reclassifying layers using relevant scores, which are then added together spatially. In the case of an electrical network, ideally the best route would run along the least slope, avoid forests, wetlands and other ecologically sensitive areas, be routed near to roads and avoid households, while running near densely populated areas in order to easily supply them with electricity. Planners, engineers and environmental scientists need work together to determine a standard points system for each layer in such a model (Glasgow et al, 2004).
CASE STUDY AREA
A case study site in the sub place area of Shayilanga / Kamlenze was chosen from a list of possible electrification areas to demonstrate how GIS can be applied in planning an extension from the nearest network to the schools within that area. The RAPS method of scoring was used to prioritise villages . This area scored the highest number of points out of 34 short-listed electrification areas which can easily be supplied electricity from the grid, due to the number and density of households, and the number of pupils at the two schools in the area. There are very few projects currently planned for the area as per the IDP for Ingwe Local Municipality
Figure 1: Case study site of Shayilanga/Kamlenze showing relief, schools, roads, rivers, and current Eskom MV networks and transformers
METHODOLOGY
Four of the most important factors were used and these were roads, land cover, household positions, and slope (Eskom, 2000):
Roads
Roads are ideal routes along which to plan electricity networks as they are usually already leveled, at a reasonable gradient, cleared of vegetation and networks next to roads are convenient for line patrols, planned maintenance and emergency work. A distance allocation raster was created and reclassified according to distance from the road.
Land Cover
Land cover data contains areas demarcated as different types of cultivated land, degraded land, grasslands, mines, scrubland, forests, wetlands and others. The land cover data was converted to raster and reclassified as shown in figure 2.
Household Positions
A raster was created from the household positions and points allocated to ensure that networks run near to existing households.
Slope
A TIN was created from 20 metre DEM points and a slope raster derived. Slopes were allocated points according to the steepness of the slope.
Figure 2: Study sites: Determining shortest path from network to school
Combined Suitability Raster
The layers were then overlaid and the values for corresponding pixels added together. Where the resultant combined suitability raster has pixels with a combined rating equal or near to zero, this will identify the best path for an extension or new network.
Figure 3: Study site: Shayilanga / Kamlenze: Reclassification of straight line distance raster created from roads
Figure 4: Study site: Shayilanga / Kamlenze: Reclassification of raster created from land cover
Figure 5: Study site: Shayilanga / Kamlenze: Reclassification of straight line distance raster created from households
Figure 6: Study site: Shayilanga / Kamlenze: Reclassification of raster created from slope
Figure 7: Study site: Shayalinga / Kamlenze: Combined suitability raster
RESULTS
For routing a new network, four criteria were considered, these being household positions, land cover, slope, and proximity to roads. Weighting and overlaying these four layers in a combined raster enabled the calculation of a cost weighted route from the network to schools in the study sites. Initial estimations for electrification of an area must be within 65% of the final cost (Eskom, 2000) and GIS can help to make this far more accurate.
Shayilanga / Kamlenze area is closest to UBR680 on Bulwer NB2 so the most logical start point was to extend the network from that transformer. This area has two schools with a large number of pupils – namely Mashaliyanga Primary School (889 pupils) and Skofill School (342 pupils) and over 52% of the people living in the ward this area falls in are under 15 years old. The schools were therefore used as a central point for placing a transformer, but in practice, any logical central point could be used or a centroid determined from groupings of households.
The two schools and nearest existing transformer fall roughly in a triangle so different routes could be considered. Initially two paths were derived; from the transformer to the two schools and then a third, which ran between the two schools. Comparisons between the routes found that the shortest path would start at the transformer, run via Mashaliyanga Primary School and end at Skofill School. The route between the schools is shorter than running the connection along the road, an option that may not have been obvious without using GIS.
Land cover in this area is predominantly divided between unimproved grassland and temporary cultivated subsistence farming which were allocated the same number of points so land cover did not have an effect on the outcome. Most of the slope is between 0 and 10% so slope was not a factor in determining the optimal route.
Household density along the extensions is relatively high (in places around 80 houses per square kilometre) – and the network extensions were designed to avoid going through those households.
Figure 8: Study site: Shayilanga / Kamlenze: Shortest path from Bulwer NB2 to schools in Shayalinga / Kamlenze.
CONCLUSIONS
Utility companies worldwide are starting to appreciate the benefits of using GIS for purposes other than that of facilities management and automated mapping and their experience could benefit Eskom Distribution in restructuring their data model to also realize the benefits of spatial analysis. GIS is increasingly being used for managing rural electrification, ranging from basic planning models to full Information Management Systems. Ground breaking practices in using GIS and sharing information with other service providers and local communities are gaining momentum. It is strongly recommended that Eskom Distribution research some of these projects, especially in areas with similar demographics such as in other countries in Africa, with a view to deciding what type of system would be best to use.
Planning the shortest path for an extension to a network is a complex exercise when working manually with a combination of a number of maps, statistics, and constraints. Instead, all standards and current planning procedures could be used to build a network planning tool based on GIS that would standardize analysis and allow the planner to easily compare all constraints in order to accurately determine the shortest path from a selected network to a predefined point or points in a community. Using GIS for electrification and network planning would completely revise current methods used and result in viable planning for network extensions where there is still currently spare capacity. Where no spare capacity exists, GIS can also be used for identification of overloaded networks and ideal placing of new substations.
Any number of further optional criteria could be built into the network planning tool to assist in determining the shortest path for connecting new customers to existing networks and in so doing assist in improving network performance and reducing short and long-term costs. These could include conditions such as environmental impact studies which have to be assessed as part of each new network or extension (Eskom, 2003), and areas with extreme weather conditions which can prove costly in terms of future maintenance. For instance, avoiding areas with extremely damp conditions can extend the life of wood poles and save on costs of early replacement. Care also needs to be taken when incorporating rivers in network planning as only perennial rivers with no bridge near the proposed network position need to be avoided. This could be verified as part of the standard site visit and then added into the network planning tool if necessary.
Currently, electrification plans already far advanced have to be put aside on occasion, due to various reasons such as a lack of capacity, and when those constraints are resolved, the entire project has to be researched and replanned due to changes on the ground. However, if a plan created using the principles in this study has to be put aside for any reason, it would be relatively simple to rerun the analysis with new data.
In essence, electrification plans could be created now for every village, and the planners could keep those plans up to date by rerunning analysis as new developments are planned or growth in areas takes place. This would result in a transparent plan to support other service providers and the South African government in building a better life for all.
REFERENCES
(Williams et al, 2003, Glasgow et al, 2004).
Raps
(Eskom, 2003)
(Eskom, 2000):
