You are here

Transformer Load Analysis at MTEMC
w/ArcGIS Online

Transformer RepairEditor's Note: Make sure not to miss the video demo of the Transformer Loading ArcGIS Online iPad app near the bottom of the article!

Every electric utility regularly deals with transformer replacements. Transformers encounter issues related to everything from animals, theft, lightning, voltage changes, fires, and, of course, sizing issues. Transformers are sized based on the expected customer load.

The rated KVA of a transformer or a bank of transformers should be equal to or greater than the load placed on them by the customer consumption at any given time. As with any device under constant load, the consumption will vary based on how much electricity is being used at that moment.

We often see summertime peaks during very hot days when everyone is cranking the AC or in the middle of the winter in locations where electricity is the predominant source of heat.

In both cases, engineers often use historical and predictive analysis to determine the most effective size for a transformer when it is installed. And there are many third party engineering analysis tools that provide software to assist with the sizing.

But what type of analysis is usually done after a transformer is installed? And what about when a transformer is replaced as part of an outage? These are both questions that Middle Tennessee Electric (MTEMC) asked to determine how they could improve their decision making process around installing and replacing these crucial assets.

MTEMC is very focused on reliability and historically they have oversized most of their transformer installations based on the expected customer load. And when transformer outages occur, MTEMC would typically upsize the transformer to ensure that sizing was not the cause of the problem.

While this has resulted in a very stable network, it has not always resulted in the most effective use of the assets in the field. Transformers are expensive assets and when a transformer is rated significantly higher than the load it supplies there is an opportunity cost associated with that mismatch in capacity vs. consumption.

To address this issue, MTEMC engineers were looking for a better way to visualize and analyze transformer consumption data against transformer capacity. The raw consumption data was available via their customer information system (CIS) and the engineers had the required calculations to compare the peak consumption data against a transformer’s rated KVA.

The goal was to find a way to bring this data together to drive more effective decision making. And what better place to do this than GIS?

Last year, MTEMC joined forces with SSP Innovations on a project to improve this process. The project was broken across three interrelated goals:

  1. TXfr Relationshiphe first goal was to automatically load the customer consumption data into GIS on a monthly basis from the SAP CIS (MTEMC is working toward a new AMI implementation and more real time data will be available in the future, but for now the monthly peaks will be loaded). The raw data is then aggregated to determine peak usage at the transformer level for both winter and summer seasons as well as month over month peak usage for larger transformers. The aggregation is also used to apply the engineering calculations to convert the peak usage into a meaningful value that can be compared against a transformer’s rated KVA to expose a utilization statistic.

    Winter Algorithm:
    Utilization = (kWh / 360) / RatedKva


    Summer Algorithm:
    Utilization = (kWh / 380) / RatedKva

    An existing relationship was used to match the CIS consumption data to a GIS service location. But the system was missing an explicit relationship between the service location and a transformer asset. The Esri geometric network provided connectivity between a transformer bank and the service location. But we needed to delve deeper to utilize electric phase to relate individual transformer unit/asset records to the service locations. New custom AutoUpdaters were put in place to maintain this data and a mass data update was applied to initialize the data (applied at the database/SQL level to ensure there was no versioning performance impact).

    The result is that every service location is related to one or more transformer asset records via the asset/company number (a three phase customer could be related to three individual overhead transformers). This allows for the consumption data to be easily and accurately aggregated for each transformer asset.
  2. Next, a transformer loading report was created allowing engineers to query the aggregated data for either overloaded or under-loaded transformers based on the utilization. The report allows engineers to input a utilization threshold as a percentage value to find all transformers above or below the mark along with detail on the season or month to be analyzed:
    Threshold

    Any matching transformer assets can be expanded to view the underlying aggregation and raw CIS consumption data that make up the utilization:
    Xfr Report

    The regular aggregation of this data allows this report to be run very quickly based on the cached statistics. The report was built using Microsoft SQL Server Reporting Services which consumes a combination of Esri Multi-Version Views (MVV) and the aggregated and raw CIS data to render the report via a web page.

    The report can be easily exported to PDF, printed for internal distribution, or loaded into Excel for further analysis and planning. This report enables the engineers to quickly query the data to proactively target specific assets for possible right-sizing thus driving a more efficient and reliable network.
  3. The final goal was to visualize the aggregated data points in GIS via a color-coded thematic map. The engineers first established colors to use based on pre-determined utilization thresholds in the aggregated data.
    Loading Layer

    But the GIS natively only shows a single transformer bank on the map whereas the data was aggregated at the transformer unit level. To make the unique unit data available on the map, the aggregated data points are regularly replicated into a read only analysis feature class based on the transformer bank location on the map. Multi-asset locations use an offset algorithm to create multiple points.  This allows for individual transformer assets to be visualized by phase – i.e. in an overhead scenario, the transformer utilization could be different for A, B, and C phases and they must therefore be symbolized individually.
    1 Phase Xfr    3 Phase Xfr

    The last challenge was to make the data available to the troubleshooters and other field personnel responsible for making decisions around transformer replacements. To accomplish this, the feature data was exposed to the field via an enterprise implementation of ArcGIS Online.
    WebMap

    Troubleshooters can now use Collector for ArcGIS to locate themselves on the map and to view the most recent load profile for any transformer showing the utilization along with other related information for the peak value. This targeted usage of ArcGIS Online is a great example of utilizing the Esri platform to enable new areas of the business to utilize GIS to empower decision making.

The result of these efforts has enabled MTEMC to drive better decisions around transformer installations and replacements very quickly. It’s a great example of using GIS to bridge the gap between two disparate data sources using spatial and network to develop data relationships that otherwise would not have existed.

SSP was excited to work with MTEMC on this project because of the strong nature of utilizing the significant strengths of their existing ArcFM™ GIS investment alongside the new Esri platform technologies to make this data available to new GIS users on any device anywhere in their service territory. The initial reporting solution went live in November 2014 with the ArcGIS Online capability following it along with MTEMC's 10.2.1 upgrade in early 2015.

An-Louise De Klerk, MTEMC GIS Coordinator, said “Working with SSP is one of the best decisions MTEMC GIS has made. Their exceptional solution design and excellent services are not only refreshing, but truly have transformed GIS at MTEMC.