Electrical Lines Geospatial Data

I recently created a GIS-based lesson focused on performing a site suitability analysis for wind turbines for the purposes of strengthening GIS skills as well as a demonstration for why GIS actually matters.  The geospatial data I used in this lesson includes wind power data from the National Renewable Energy Lab, population data from the US Census Bureau, and other data layers, including electrical lines data.   Wait, electrical lines?  Yes.  The electric lines or grid was, along with other utility-related data, the most difficult data set to gather.  This is due to reasons of security (including the passage of the post-9-11 Critical Energy Infrastructure (CEII) Act https://www.ferc.gov/legal/ceii-foia/ceii.asp) and the fact that this data is largely generated and maintained by private companies.  Making up my own electrical data for the purpose of this learning activity would have been OK, but I wanted to use real data for the entire lesson.   The breakthrough came courtesy of the company S&P Global Platts, who generously gave me some generalized data for Colorado that I used in this lesson.

A colleague of mine pointed out that a more updated and lower voltage (below 69 kV) data is available from the U.S. Department of Homeland Security (DOHS)  publicly available Homeland Foundation-Level Data (HFLD), which includes transmission lines:  https://hifld-geoplatform.opendata.arcgis.com/.   I therefore decided to investigate this resource.  It is indeed easy to use, with many options to view and download, with a wide variety of data sets.

Homeland Foundation-Level Data (HFLD) interface after searching for electrical lines data. 

On the subject of data, let’s talk briefly about the wind power data I am using in the lesson.  It was developed by AWS Truepower https://www.awstruepower.com/ and validated by the NREL Wind Resource Assessment (WRA) team in 2003; hence, the data I am using is over 15 years old.  One might say, “yes, but does wind change that much over time?”, which is a valid argument, but one should also ask, “What were the inputs to this wind data?”   The quality of this 50 m data was quite good for its time as the NREL analysts had access to about 6,000 ground measurement sites across the USA with wind measurement data, and their validations put about 80% of the modeled grid cell values within the ground measurement point values.  They made heavy use of Grid focal functions to process the data.  Fifteen years later, 80 m and 100 m data from AWS Truepower has now been generated, which can be seen on the U.S. Department of Energy (DOE) WINDExchange website https://windexchange.energy.gov/maps-data .  But, due to differences in the contractual relationship between NREL/DOE and AWS Truepower, the 80 and 100 m data is not available to the public for free.

Another set of infrastructure data can be found via the U.S. Energy Information Administration (EIA) https://www.eia.gov/  and the U.S. Energy Mapping System https://www.eia.gov/state/maps.php.  Biodiesel, coal fields, pipelines, and much more can be obtained via https://www.eia.gov/maps/layer_info-m.php.

Several key takeaways, I think, arise from this discussion:  (1)  As we pointed out in this public health related essay, sometimes the only way of obtaining the sort of “deeper metadata” that I describe above is to have an old-fashioned phone call, Skype, or email exchange with those responsible for creating the data.  This deeper metadata is often not found in typical metadata documents.  (2) The results of your geospatial analysis is highly dependent on the quality of the data (its resolution, scale, date, processes used, and so on) that goes into it.  Even in my educational activity, the results will be different if different wind power data or different electrical data are used.

Map from my wind farm site selection activity described in this essay.

Electrical infrastructure.  Photograph in The Netherlands by Joseph Kerski.