GIS@development April 2002: Modelling the Electricity Distribution Networks

In almost every walk of life, it is a common practice to create models, generally as scaled down versions, to depict real-life situations. The models are created for a variety of purposes. No matter what the purpose, the models allow one to visualise a given real-life condition and analyse the impacts it would have if an aspect in which one is interested were to be introduced. For example, floor plans of an existing building may show the sizes and dimensions of rooms, placement of doors, windows etc. More detailed floor plans may even contain the electrical wiring and plumbing layouts. It is a common practice with architects, town planners, engineers with utilities etc., who are responsible for implementing any project, to create models of their projects before actual implementation commences.

Though the main subject matter of this discussion is for models of electricity distribution networks, we will first briefly review the types of models, their utility to the users, and then go over to the subject matter in greater details.

Type of Models
Models are generally created as the scaled down versions of what actually exists or is likely to exist in real-life. Models allow one to visualise the real-life situations, study them, and in some particular cases subject them to analyse the effects of real-life changes that may take place from time to time. Models fall into two distinct categories - Static Models and Dynamic Models.

Static models are created for fixed type objects, such as those for buildings, prototype of machine parts etc. They are generally created so that one can visualise, in advance, the real-life scenario, and analyse if the objective set for real-life situation would be adequately served.

Static models are created more to create the visual impact. They are very rarely used for any study or for analytical purposes.

In situations where existing or assumed conditions are expected to undergo frequent changes, it is necessary to create appropriate models of such conditions that would dynamically respond to the changes in condition that would take place from time to time. Models of this type fall into Dynamic model category. These models, though are first created to depict the initial static condition, have the ability to respond to the changes in conditions, and dynamically show the new static condition. In that respect, they can be termed as Dynamic Models.

For example, if one has to study and predict population growth in a certain region, it would be necessary to first create a static model of the existing population and then subject it to dynamic analysis that would take into account the socio-economic parameters that would govern the population growth. Similarly, if one were to study, analyse and predict the performance of a given supply network, it would be necessary to have a model of the existing network, and then analyse its performance under simulated conditions of changes that are expected to take place. In case of Dynamic models, the visual impact they create is of lesser importance. It is desirable that they rather provide excellent facilities to simulate various conditions, study them and analyse the results.

Dynamic Models
As said before, Dynamic models are generally used to simulate, study and analyse scientific or engineering environments, which normally undergo frequent changes. The scientific and engineering environments, in general, are inherently complex in nature, and are extremely sensitive to changes in the conditions under which they are operating. Few examples of situations where dynamic modeling is required are:

Performance studies of Electricity Distribution Networks, Water Supply Networks, which have to operate under constantly varying conditions of demand.
Study of agricultural yields in a region based on previous crop patterns and current data about the soil conditions and rainfall.
Prediction of Election results based on demographic data and results of previous elections.

The very complexity of the scientific or engineering environments, and the fact that they undergo frequent changes to the conditions that govern their operations, demands that, the models for such situations are preferably created by using some software specifically designed for the purpose.

To create dynamic models for scientific or engineering environments, it is essential that the software has the built-in graphics facilities to depict the prevailing conditions, and is able to quickly incorporate the changes in conditions. Therefore, it is important that software to create these models provides suitable graphical facilities and allows to incorporate as much intelligence as possible in the model so that the users can use this intelligence for performance studies and analytical purposes. For Electricity Distribution Networks, working under dynamically changing conditions of power demands, the software has to provide minimum facilities to model them with appropriate symbols, and add intelligence to them by way of parametric data attached to each of the network elements.

Electricity Distribution Network Models
Electricity Distribution Networks essentially consist of power flow control equipment such as transformers, switches, breakers etc. These equipment are connected to each other, either thorough connecting cables or bus bars. It is a common practice with electricity distribution agencies to create for their supply networks, feeder Single Line Diagrams, which are commonly referred to as SLDs. The SLDs are not drawn to any scale since they are the mere representation of the supply network. Feeders drafted on maps may or may not be to scale. It is a common practice with the utilities to create these SLDs or draft the maps, by using a symbol for each type of equipment they use in the network and different line types for the variety of cables they have. To distinguish one feeder from the other, utilities sometimes use different colours to show different feeder networks. In modern times, the feeder supply networks, for almost every Electricity Distribution Utility, have several equipment, and are spread out over a large area. The utilities, therefore, would preferably need computerised facilities to model their supply networks to get the visual effect of the electrical connectivity of their supply feeders. The utilities would also prefer to have facilities of computerised network analysis by which they can conduct load flow studies to quickly calculate the Technical Losses in their distribution network. The other facility that the utilities wish to have is to use the software facilities to conduct analytical studies that would help them to make optimal use of their feeder supply networks.

Software for Modeling Electricity Distribution Networks
The most important requirement for providing the above-mentioned facilities to the utilities, therefore, is to have the software that would help utilities to correctly model their supply networks. The software should preferably have the following facilities.

Allow users to use the equipment symbols they would prefer to model their feeders.
Give them extensive graphics facilities so that they can create feeder models and edit them from time to time, as they would require.
Should be designed intelligently so that the users are required to give minimal inputs while creating their feeder models. This can be easily accomplished by providing to the users, pull down menus for equipment and cable types.
Should preferably guide the users to the next logical step in the construction of FEEDER MODELS. For example, while creating a feeder model, whenever a user would put any equipment in it, the modeling software should force him to connect this equipment either by a cable or a bus bar only. The user should not be able to proceed with the modeling until this step of connecting the equipment is completed. With this type of built-in intelligence in the software, the models created would be logical, conforming to the engineering practices followed, and would be free of any errors.
Users should be preferably given the facilities to annotate the attributes of their choice to the equipment and cables in the feeder network they are modeling. This can be easily done by incorporating the pull down menus in the software for all the standard equipment utilities use in constructing the feeder networks. For example, users should be able to automatically annotate, in their models, the ratings of transformers by choosing them from pull down menu, which contains ratings of various types of transformers used in the utility. In case the users, for want of correct information with them, is unable to provide the rating of the transformer they are modeling, they should be given the facility, in the software, to use default values. The default value should be so set that, it is none of the common ratings used by the utility. This way, whenever the model is reviewed, the user would immediately know that the transformer rating they are seeing is not correct and needs to be edited to get the correct model of the feeder. For example, if the default value for transformers is set at 99.99 KVA, the users would immediately know that this is a default value assigned at the time of creating the model needs to be corrected to reflect the correct rating of the transformer in question. Similarly, for the cable sections, the users should get the facility to automatically annotate cable type only from the pull down menu they have. In the feeder model created by using such an intelligent software, all that the user would need to key-in would be the length of the cable section they are modeling.
Most of the Network Analysis software require that the feeder equipment, cable sections, and bus bars are assigned unique numbers ( IDs ). The modeling software, therefore, should preferably have a built-in module, which would allow users the choice to assign unique IDs to the feeder elements, as per the scheme preferred by the utilities, or as per the scheme incorporated in the modeling software. A built-in facility of this type, in the modeling software, for assigning unique IDs to the equipment in the feeder network will remove a lot of tedium for users while modeling their feeders. This facility, moreover, would allow the users to quickly switch over to Network Analysis task as soon as they complete modeling of their feeder networks.

It is equally important that the modeling software has the built-in facility to check the correctness and reliability of the models created by it.

Some of the basic and important checks to which the supply feeder models can be subjected to are:

Checks for ELECTRICAL CONNECTIVITY of all the equipment in the FEEDER MODEL
Checks for proper placement of NODES
Checks for proper placement of EQUIPMENT modeled in the feeder network
Checks to ensure that the EQUIPMENT modeled is only one of those that are used by the utility, and none other.
As said before, the utilities normally carry their feeder diagrams either as SLDs or feeder layouts drafted on the geographical maps of their supply area. Therefore, it is desirable that the modeling software allows utilities to use both these inputs and create the digital models of their supply feeder network. This facility can be easily provided in the modeling software if it allows users the option to use scanned images of their SLDs or maps, and just trace over them to create the digital models of their supply network.

These are some of the examples of how intelligent software could be offered to the Electricity Distribution Utilities so that their engineers are able to model their feeder supply networks quickly, correctly, and with complete reliability in the created model.

On the lines above, Indicos Information Technology Pvt. Ltd. has designed and developed software for modeling supply feeder networks. This software, named INDISYN, takes into account, in its design, the prevailing practices followed by most of the State Electricity Boards in India. Therefore, it is ideally suited for their use.

The software has an in-built module to transfer feeder models to automatically transfer feeder network data to SynerGEE, one of the most widely used Network Analysis software for Electricity Distribution Networks. SynerGEE software is designed and developed by Advantica Stoner of USA.

A few select screen shots taken from this software, INDISYN, will highlight its versatility, and its utility to the Electricity Supply Utilities in India.