ArcGIS Utility Data Modeling: Electric Distribution Conductors
This post is the first in a series in which we’ll explore specifics of modeling utility infrastructure components within ArcGIS/ArcFM. To start we’ll address electric distribution conductors. There are two standard ArcGIS data models typically used within North America for electric distribution: the ArcGIS Electric Model and the ArcFM Multispeak model. While similar in many respects, the Multispeak model is closer to “ready-to-use-out-of-the-box,” so our discussion will begin there.
Electric Conductor Model
The following diagram illustrates the classes that describe electric circuit segments in the Multispeak model. All features are sub-classed off of the ArcObjects ComplexEdge class. (Note that this can easily be changed to SimpleEdge if needed.) Meta-data is defined in the ArcFMUserMetaData_ComplexEdge class and all properties common to all types of conductor, of which there are many, are defined in the ElectricLineSegment class.
While the standard model defines all conductor classes sub-classed from ComplexEdge, an implementation may choose to exclude secondaries from the network altogether, in which these would be defined as the type SimpleFeature. Reasons an implementation may choose to exclude secondary from the network include:
- There are no secondaries present in the company’s source data; not an unusual condition.
- While secondaries are represented in the company’s source data, its positional accuracy is not deemed sufficient for creating correct network connectivity from transformer to customer service location.
- The amount of effort the company has been able to expend on the creation and maintenance of secondary lines has lagged the effort applied to primaries, and as a result overall confidence in secondary accuracy is small.
If secondaries are to be included in the network, another option to be considered by the implementer is to define these as simple edges rather than complex edges. Unlike primary lines where there may be value in using complex edges to support connection of transformers without splitting the line, secondaries are typically connected to devices or to one another only at their end-points.
Circuit segments describe a span of conductor that includes all phase and neutral wires carrying electricity along the same circuit at the same voltage. The model divides the set of all distribution circuit segments by voltage separating lines carrying primary voltage (over 600 volts) from secondary voltage (less than 600 volts), and further separating overhead lines from underground.
In implementation the company can determine what the extent of a “span” means. Though this question does not strictly impact the data model, it is important to establish as a convention for the model’s implementation. Typically for overhead conductors this is either from pole to pole or from switchable device to switchable device. The choice often depends on how data is defined in the company’s legacy mapping and/or asset management systems as well as planned uses within the GIS. Note that either approach can neb sued to achieve network connectivity and is fully supported by ArcFM and Feeder Manager. Here are a few pros and cons for each approach:
- Defining Overhead Conductor Segments Pole-to-Pole. This approach involves beginning and ending each overhead conductor span at a pole, and could be supported using either ArcGIS simple edges or complex edges, since there would typically be no case where a device would split a line between poles. (An exception being the case of flying taps, or mid-span taps – where one conductor is tapped from another in the middle of a span. Some companies consider the case of a flying tap to be equivalent to a pole for asset management purposes.)
- A company’s work and/or asset management system may track circuit segments as assets that run from pole to pole. For consistency and/or to support an interface between the systems the GIS would follow the same approach.
- If there is a requirement that the company track specifically what conductors are supported by what poles, then the conductor-to-pole relationship is more straightforward to track and validate using a pole-to-pole mapping convention. [Note that the out-of-the-box ArcFM MultiSpeak model would need to be extended for an implementation with this requirement.]
- There will be more conductor features in the database compared to conductors defined from device to device. An overhead feeder may have from a few hundred to a few thousand pole-to-pole conductor spans. If the conductor is split only at switches and protective devices (and not transformers) that number might be reduced by a factor of between 5 and ten. So, for example, a feeder with 2,000 overhead pole-to-pole spans could be represented with from 200 to 400 device-to-device spans, and a 70,000 customer utility with 50,000 overhead pole-to-pole spans could be represented with from 5,000 to 10,000 spans. Further, since a given span may have one to four (if neutrals are modeled as ConductorInfo) related records in the ConductorInfo table, which means 100,000 pole-to-pole ConductorInfo records could be represented using 10,000 to 20,000 rows
- More rows in a database impact query performance; including tracing queries (though neither 5,000 nor 50,000 rows are large numbers for any reasonable database management system.) It’s also the case that more rows require more effort in maintenance and quality assurance.
- There are no out-of-the box processes to enable, enforce or validate the pole-to-pole mapping convention.
- Defining Overhead Conductor Segments Switchable Device to Switchable Device. This approach involves beginning and ending overhead conductor spans only at switching or protective devices. Transformers can be connected to conductors at mid-segment snapped to vertices along the ArcGIS complex edge. The same is true for primary tap points, flying or otherwise. The approach requires use of complex edges.
- The same set of overhead conductors can be represented with a smaller number of GIS features and ConductorInfo table rows. The opposite of the “con” described for the pole-to-pole convention described above.
- When a new protective device is introduced along a device-to-device span it can be automatically split and ConductorInfo records duplicated by out-of-the-box ArcFM software
- Other external systems supported by the GIS may not represent conductors in the same way. For example, an asset or work management system may recognize conductor spans only from pole to pole. Most engineering analysis products do not recognize the concept of a complex edge and multi-vertex spans may need to be “exploded” into vertex-to-vertex segments for export to an application like SynerGEE, Cyme, Windmil or other analysis product. The same may also be true of non-ArcGIS based outage management or distribution management applications. The conversion, or “explosion” of complex edges to multiple segments is performed automatically by ArcFM Network Adapter and other GIS export products, so this is not an issue. What is lost however is the direct one-to-one correspondence between a span in the GIS to a span in the analysis or OMS/DMS product which can be useful for data validation.
This post has provided summary level information about modeling electric conductors in an ArcGIS/ArcFM solution Geodatabase. Clearly even within this “standard” model there are options from which you can choose to best support the needs of your implementation.