Development of site and route selections in Eskom using GIS

May 10th, 2016, Published in Articles: PositionIT


This article discusses the education of top management by showing the dependency of the electricity value chain on the earth with its man-made and natural features. Some of the benefits that the geoscience functions can bring to the business environment are defined and explained with a short discussion of a business case on a site selection process. Finally, some of the future success criteria are briefly mentioned.

Geoscience and geo-information systems are functions and tools which, if applied correctly, may enhance the whole business process and make it far more effective and efficient. The purpose of this paper is to educate the business from top management to operational level about the potential of this combination. The close relation between land and the Eskom value chain is highlighted, as well as the three obvious benefits of the geo-information function and tool, namely visualisation, operational improvement and enhanced decision making (especially multi-criteria decision-making). The Delphi technique handles uncertainty and meagre data, as will be discussed. This is followed by a case study to show the practical applications and the perceived challenges of the technology.

Fig. 1: Limit of coal deposits.

Fig. 1: Limit of coal deposits.

Value chain of electricity

The vertically integrated Eskom system produces 95% of the electricity energy in South Africa. The energy sources are coal, nuclear and hydro; with reasonable generating capacity now coming from renewable energy sources, including wind turbines, PV, and concentrated solar energy (with and without storage capacity).

These new resources of electricity also adhere to the value chain for electricity which involves the delivery of electricity to customers. To cater for this, redistributors are geographically spread. Electricity has to be delivered to these customers and this customer point is always connected to a geographical point on earth. Distribution channels may be used to distribute the surplus electricity of the customer to the distributors for re-sale. The supply of electricity has many characteristics, one of which is a geographical area that is fed from a substation with a certain capacity. A number of these substations are fed by larger substations that are interconnected to ensure that the electricity supply is balanced. These main or master substations are connected to the power generating plant. Another characteristic is the irreversible linking with land with its man-made and natural features. This relationship with land is the de facto or primary asset which makes these interactions possible.

More formally, this utility can be described as a system with major processes that contains subsystems with their own processes. All of these are driven by inputs through processes with the aim of transforming inputs through other subsystems into something of value that satisfies a need. The system and its subsystems require the acquiring and consumption of large amounts of resources – money, management, administration, labour, land, buildings, fuel and technology. The way these activities or operations are carried out determine cost, profits and ultimately the sustainability of the utility.

Electricity is provided through a physical value chain with real activities performed in order to deliver the electricity. These activities:

  • Have evolved over time by the workforce from the business operations and other activities
  • Need to be sustainable and/or profitable
  • Include time as a variable

All these activities and variables are related to geographical positions and this means that one of the dimensions of the system should be the whole applicable geographical area. The best option to depict this operation area with its features is a map. The most common map is the geographical definition of the earth through latitude, longitude and elevation. This model defines the earth space time. The latitude, longitude and elevation may represent other earth space-time quantified systems such as a survey beacon and the elevation at a benchmark as a height above mean sea level. The aim of this system is to have all earth space-time references related to each other and to a real physical location. This relation of references through accurate and timeous spatial information brings a large amount of real world, projected past and future data to the forefront for analysis, depiction, and complex investigation, applying some of the results and predicting some future situations. The integration of different data systems is no longer a dream, and geospatial data are stored only once for use by the organisation.

Geoscience to better understand the value chain and improve it

Geoscience has immense benefits for the whole community, not only for specialists.

GIS enhances communication

Visualisation is the presentation of geospatial data in various groups of layers to display the feature with its attributes and relations, convey the results of operations to the wider community, to understand the result of the operations and to analyse it better. The real operational results can be shared against the significances of different potential events from the predictions. The specific site and/or route of electrical infrastructure can be captured and stored with the applicable attributes for future use. Through land cover, the influence of that infrastructure on the natural and man-made environment can be monitored. (The concept of earth space-time).

Fig. 2: Water footprint based on distance from sufficiently available water.

Fig. 2: Water footprint based on distance from sufficiently available water.

Today cartography is done with the applicable software on super desktop computers from one central database. To produce a physical product, various layers can be combined from several standard and geospatial datasets to convey specific messages. The product remains digitally filed with the data/information in a central place. Paper copies and web images are only produced when there is a specific need to promote a concept. The experienced geoscience or subject matter user has the last say over the visualisation of the geospatial data.

Websites such as Google Maps, Bing Maps, Openlayers and OpenStreetMap give the public access to huge amounts of geographic data. It is no longer a few professionals that collect geospatial data but the potential of crowdsourcing geodata is a reality. Customer-specific products are produced and all of the information can instantly be available for the whole world.

GIS improves the operational and maintenance activities of an entity

Communication or visualisation through the cartographic processes translates into improvement of various operational and maintenance processes.

This planning information serves as input into the decision-making process and is saved as a record of decision. Operators produce plans or scenarios of situations on the screen or on a physical medium to convey the results of analysis and instructions. This information is produced for the people who have to perform various tasks. The assets and other functions’ information sets are all integrated and are stored only once.

A series of images of a veldfire is produced, stored and serves as a permanent physical record of earth space-time. This serves as indisputable evidence of the incident for future reference. Display of pollution areas of any kind is a further application. The affected area can be shown in earth space-time and then tracked over time to monitor the realisation or non-realisation of the moderation action. The position of faults on the line-specific classes of customers, working teams, and assets with their attributes are used in managing these incidents and assets through various integrated processes.

The decision maker, end user or specific audience can visualise the situation and understand the state of affairs and can get an immediate reply if certain criteria is changed. It visualises the concept of earth space-time, and improves communications.

Web map servers facilitate the distribution of generated maps through web browsers and servers. Information technology has made it possible for nearly every one to have access to certain imagery using various web-based applications.

The geospatial datasets can be manipulated and integrated with various other data just by adding the geographical coordinate or combine different geospatial and non-geospatial datasets. Value can also added to a geospatial dataset by re-classifying the geospatial data to satisfy the customer need. In the Eskom situation it could identify a constrained network or show the proximity of a settlement to a power station.

The possibilities of the modern graphic display techniques may combine different datasets, stored only once, of the same area for different users without the fear of scale or projection distortions. Modern technology has made it possible for a topographical map to be tailor-made for its user within a reasonable time and cost.

Geospatial analyses are done with the various features hosted in GIS databases. Physical relations can be determined between features.

Geometric networks are linear networks of objects representing an interconnected feature. It may simulate the operations of that entity, allow special spatial analysis, modelling and even manage the network. Such a network is composed of sources, which are connected at nodes, to represent the real world. A plant slot is an entity in the electric database that represents a “thing or concept” in the form of a position with its own attributes. An asset can be placed in it, allowing the asset to be managed in a different application independent from the geometric network. Eskom Distribution has gained experience with electrical networks from the conceptual, physical planning, design and construction of the networks. The relationship does not stop here as the operations, maintenance, de-commissioning and even dismantling of the networks are also conducted through the GIS. Datasets (both geospatial and alpha numeric) are stored once, assisting in data security, storage and maintenance management.

GIS enables decision-making

With all these geospatial datasets available it becomes an option to use multiple criteria decision-making.

Complex problems don’t have unique optimal solutions and it is necessary to decrease the bias from a single or group of decision makers. Therefore the problem should be structured so that multiple criteria decision-making could be used for more informed and better decisions.

A whole array of possible alternatives may be developed that can include the “best” alternative from a set of available alternatives or a small set of a “good” alternatives, or alternatives can be grouped together. Trade-offs have to be made between various criteria as the very challenge comes from having more than one criterion.

The classical site selection process starts with a detailed assessment of the project needs which are then equated against the merits of potential locations. Site selection plays a significant role in pre-planning discussions and normally starts early in the capital development process.

Fig. 3: Overlapping inclusion zone.

Fig. 3: Overlapping inclusion zone.

Projects are different of course, but this article provides a short overview of the processes and philosophies at ESI-GIS.

The site selection begins as a pre-conceptual project where the confidentiality is one of the customer’s primary requirements. The project needs, along with matching criteria, are carefully designed, tested and re-tested before acceptance. These criteria, are ranked, ordered and weighted by subject matters, usually with the support of a professional facilitator. The project team is then able to delete certain geographical areas out of the comparison, and rank order the alternative criteria to take emotion out of the evaluation process. The approach must also prevent placing too much emphasis on any one criterion.

Typical information requested and debated during the opening discussions of the site selection process include, but are not restricted, to:

  • Man-made features such as roads, railways, other transport options, airfields, dams, irrigation, agriculture, residential, commercial, industrial, heritage, heavy industrial and heritage sites, nature reserves, and sanctuaries.
  • Natural features such as topography, mountain slopes, soil, minerals, waterways, wet lands, natural habitat, virgin land, forestry, natural sanctuaries and coastal challenges.
  • Legal and financial incentives, and cost off-sets considerations such as water rights/usage.
  • Land requirements for initial project and future expansions.
  • Plant requirements (water, sewer, electricity, transport, storm water and housing), current and future, dismantling, decommissioning, and customer closeness.
  • Plant output energy pollution (noise, air, atmosphere and soil).
  • Demographic requirements such as semi-skilled, skilled, technical and professional labour force with associated training strategies, housing, schools and hospitals.

A number of possible sites are selected by Eskom for submission to the Department of Environmental Affairs environmental analysis processes and it may expand criteria or lengthen the time required to ensure proper consultation.

Assessments are done by independent consultants, and Eskom is then provided with a number of options or alternatives for final selection. Some of the work can be performed by internal staff while others may require external advisors.

Case study in selection of pre-conceptual sites

ESI-GIS was originally requested to assist with the creation of a shelf of potential pre-conceptual sites for power stations using different energy sources. Over the last ten years the section has developed the basic process to select such a site or sites for similar uses. The original process has been redefined and is still adjusted and enhanced to satisfy the need of the end-customers. ESI-GIS now has a number of pre-conceptual sites that may be offered as possible contenders to Eskom, the Departments of Energy, and the Department of Public Enterprises for further considerations for possible site collections.

The process must be working as the Eskom end-customers are returning on a regular basis for new pre-conceptual sites or for including ESI-GIS in the final selection process by requesting increasingly more information, reports and involvement in the processes.

The following were some of the assumptions and criteria that were given by the end user to determine a number of possible pre-conceptual sites. These criteria and assumptions are not a true reflection of the concluding assumptions and criteria as some were adjusted or excluded for various reasons.

Fig. 4: Protected areas.

Fig. 4: Protected areas.

The plant must be:

  • Able to produce 3600 MW.
  • Dry cooled with a maximum water requirement of 7-million m³ per annum.
  • Within 50 km of coal reserves in excess of 480 Mt proven for 40 years and with specific characteristics.
  • Within 200 km of limestone or dolomite sorbent sources.
  • Further than 10 km from urban residential areas.
  • Closer than 35 km from urban centres (skills base and load centres).
  • Closer than 35 km from road and rail infrastructure.

Plant cannot be sited in the following areas:

  • Protected areas.
  • Prime agricultural land.
  • Biodiversity endangered areas.
  • Potential areas for tourism investment initiatives.
  • Dams and lakes.
  • Areas with high levels of air pollution.
  • Areas earmarked for other strategic infrastructure projects.

Geospatial information processing involves the production of two master information sets visualised through an inclusion dataset and an exclusion dataset. These datasets are visualised as maps to enhance understanding, and their information combined to show the possible pre-conceptual sites.

Inclusion areas

The inclusion information was based on localities which met the criteria for distance from the mentioned criteria.

Only the coal and water areas are shown. The coal deposits were obtained from the Council for Geoscience and enhanced by using information obtained from consultants. The basic dataset for the coal is not in a very user-friendly format and was accepted even though it was “old”. The final overlapping inclusion zone was manipulated by using all of the applicable geospatial data/information sets that were developed for each criterion. This is perceived as sufficient work to have the indication of where a coal powered station of this size could be situated. It cannot be stressed enough that the map is just a tool to assist in understanding and visualising the information.

Exclusion areas

The exclusion dataset was based on the combination of the previously mentioned criteria.

Only the agriculture potential and protected areas were shown as a visual picture of the geospatial information sets. The scale and age of the raw geospatial data has created some challenges as they were on relatively small scales and the metadata was not always completed and dated. Subject matter experts were used to enhance the geospatial information to get to the final exclusion zones. The information from the other criteria was included to deliver the final exclusion zone that could be used for a pre-conceptual sites. Again it must be stressed that the power of the GIS is not the maps but the underlying data or information.

The first pass of pre-conceptual sites

By matching the information from the inclusion areas with the exclusion areas and subtracting the last mentioned, a number of areas could be identified as pre-conceptual sites that could be used for future investigation.

The process was to take a criterion and find the data and then manipulate the data to become information or knowledge. This part of the exercise is time consuming, but is the most important. Datasets were investigated and validated based on their metadata, but are these datasets also acknowledged by the specific subject experts?

Fig. 5: Exclusion zones.

Fig. 5: Exclusion zones.

Rather than pointing out a “right” decision, decision makers must find the alternative that helps with their understanding of the decision and the alternative that best suits their goal in the long term and in a sustainable way. The decision making process must provide a comprehensive and rational framework for structuring the challenge, identifying and quantifying its clashes and collaborations, highlighting the main alternatives and the well thoughtout reasoning behind them. This is applicable where there is multiple decision criteria involved and a similar answer is needed in a future exercise.

With the site selection, it is not the selection of one alternative from a given set of alternatives, but rather defining the relation between the value of a set of alternatives.

The Delphi technique was originally the preferred decision-making process because of the lack of structured information. A facilitator communicates the responses of the panel of experts in conjunction with the project owner and his in-house subject expert. These panel members are subject matter experts, who were selected because they hold knowledge and/or wisdom on the decisions to be made. The participants (subject matter experts) remain anonymous “forever”. This approach facilitates:

  • Admission of errors when revising earlier judgments.
  • Free expression of opinions.
  • Open critique.
  • Even playing fields between authority, personality and/or reputation.
  • A process that avoids the negative effects of face-to-face panel discussions, and solves the usual problems of group dynamics.

The panel members in this case were chosen through brainstorming to ensure that the “best subject matter experts” were identified, their participation requested by the project owner. The anonymous rule was explained and only people who accepted it become part of the panel.

The facilitator sends out questionnaires, developed jointly with in-house experts, and the answers are recorded and examined, then common and conflicting standpoints are identified and shared for further discussion with the panel. Participants comment on their own forecasts, the responses of others and on the progress of the panel as a whole. At any moment they can revise their earlier statements. This process is repeated to gradually build consensus.

This site selection after the first stage has a limited number of alternatives, openly known in the beginning of the solution process. Each alternative is represented by its performance in multiple criteria.

The process for the final site may be as follows using the Delphi technique:

  • Criteria development.
  • Determination of weightings per criteria (spatial GIS layer).
  • Determination of scores per criterion based on GIS proximity modelling.
  • Suitability geo-information creation based on MCE and GIS modelling techniques.
  • Visualisation of the information as maps to communicate each option.

Once the possible sites were selected and presented to the top decision makers, these potential sites are stored for further use as and when required.

Fig. 6: Matching inclusion areas less exclusion areas.

Fig. 6: Matching inclusion areas less exclusion areas.

Some of the conclusions that were reached

  • Some people in Eskom’s top management are very scared of the process and do not appreciate the connections between the value chain and its geographical environment. When the wrong decision is made the difficult task of educating these managers is nearly always successful.
  • Even some of the subject matter experts need GIS education, no matter how good they are in their field. It is very difficult to relate or comprehend what can be done with geospatial data and information. It is important to understand that Google commercial data is not suitable for engineering applications. The consequences of the activities are not appreciated.
  • Eskom has many fine and well-educated people to do GIS, building data collection and capturing into the system and they are not always given the recognition and credit that they deserve. On the other hand, many subject matter experts have theoretically burnt their fingers when facing the challenges mentioned, or have not checked the cadastral data in the GIS with the real data in the surveyor-general’s office.
  • Formal education and facing challenges is the only way to teach people the right way.
  • There is a craving for instant satisfaction that wants all the results now and people have trouble accepting that certain activities do take time.
  • South Africa doesn’t have the detailed geospatial data for site and route selection. In the case discussed, the team was happy with geological data at a scale of 1:50 000 and even 1:250 000. There is no problem in presenting the merger of the exclusion and inclusion maps, but if further detail is required it becomes difficult due to the lack of skilled people and appropriate data. The scale, age and completeness of the geospatial datasets is not meeting the requirements.
  • Areas that are covered by air pollution are not fully covered or demarcated. It has to be accepted that there is air pollution, but it could not be built into the models.
  • South Africa doesn’t have a road map coming from any organ of state that is available to the public.
  • Statistics South Africa is doing its best to provide sound data, but ten-year-old demarcation data is not suitable.

Recommendations to improve the situation

  • Geomatics Professional Act (Act 19 of 2013): The Geomatics Act will force the registration of professional people and will be a way forward, but it has to be applied carefully as there is no way to stop someone from downloading data and software from the internet. The Act has to be changed to make provision for a category that caters for people with qualifications other than mathematics and cartography. Previous work experience and sound education have to be given more credit.
  • Spatial Data Infrastructure Act (Act 54 of 2003): The appointment of an official geospatial data custodian for a base geospatial data or information set in terms of the Act is needed. The committee of spatial information has to determine these core or fundamental geospatial datasets. Political will must be in government to enforce these decisions.
  • Education of the whole of South Africa in geosciences: The profession must educate their political leaders. Each one of us has to be committed to make it work.
  • Collaboration by middle managers between organs who carry the load to implement GIS must take place.


This paper was presented at Geomatics Indaba 2015 and is republished here with permission.

Contact Adri de la Rey, Eskom ESI-GIS, Tel 011 651-6908,

Related Articles

  • Latest GIS further supports electric and gas networks
  • EE Publishers 2020 Year Planners
  • To readers, customers, suppliers, staff and other stakeholders of PositionIT
  • Mozambique’s land administration modernisation
  • Spatial database receives major update