This repo contains code to produce and interactively visualize publicly available geospatial data to support trucking fleets in navigating the transition to alternative energy carriers. The tool uses data from the "freight analysis framework" (FAF5) database and other public data sources.

• python3

git clone [email protected]:mcsc-impact-climate/FAF5-Analysis.git

Install python requirements

pip install -r requirements.txt

The script download_data.sh downloads all the data needed to run this code into the data directory. Note that it will only download a file if it doesn't already exist in the data directory. To run:

bash download_data.sh

The code in web contains all functionality to visualize the geojsons interactively on a web interface. The code can be executed as follows:

# Install python requirements if needed
install -r requirements.txt

python manage.py runserver

If that executes without issue, you should be able to view the map in your browser at http://127.0.0.1:5000/transportation. It currently looks something like this:

The script ProcessFAFHighwayData.py reads in both the FAF5 network links for the entire US and the associated highway network assignments for total trucking flows, and joins the total flows for 2022 (all commodities combined) with the FAF5 network links via their common link IDs to produce a combined shapefile.

To run:

python processFAFHighwayData.py 

This should produce a shapefile in data/highway_assignment_links.

The script ProcessGridData.py reads in the shapefile containing the borders of subregions within which eGRIDs reports grid emissions data, along with the associated eGRIDs data, and joins the shapefile with the eGRIDs data via the subregion ID to produce a combined shapefile.

To run:

python source/ProcessGridData.py 

This should produce shapefiles in data/egrid2020_subregions_merged and data/eia2020_subregions_merged.

The script ProcessPrices.py reads in the shapefile containing borders of zip codes and states, along with the associated electricity price data and demand charges, and joins the shapefiles with the electricity price data via the subregion ID to produce combined shapefiles. It also evaluates electricity price, demand charge and diesel price by state.

To run:

python source/ProcessPrices.py 

The script ProcessStateSupport.py reads in the shapefile containing borders of US states, along with CSV files containing state-level incentives relevant to trucking from the AFDC website, and joins the CSV files with the shapefile to produce a set of shapefiles with the number of incentives of each type (fuel, vehicle purchase, emissions and infrastructure) and fuel target (electrification, hydrogen, ethanol, etc.) for each state.

To run:

python source/ProcessStateSupport.py

The script PrepareInfrastructureCorridors.py reads in either a shapefile with the US highway system, or shapefiles with specific regions of planned heavy duty vehicle infrastructure corridors announced by the Biden-Harris administration. For corridors represented as subsets of the national highway system, the code produces shapefiles for each highway segment with a planned infrastructure project. For corridors represented as regions of the US, the code produces shapefiles showing the region(s) where the planned infrastructure project will take place.

To run:

python source/PrepareInfrastructureCorridors.py

This should produce shapefiles for zipcode-level and state-level electricity prices in data/electricity_rates_merged

The script AnalyzeVius.py produces distributions of GREET vehicle class, fuel type, age, and payload from the VIUS data. To run:

python source/AnalyzeVius.py

Run the script ViusTools.py to produce an output file tabulating the product of fuel efficiency (mpg) times payload for each commodity, along with the associated standard deviation:

python source/ViusTools.py

This should produce the following output file: data/VIUS_Results/mpg_times_payload.csv.

The script Point2PointFAF.py combines outputs from VIUS, GREET and FAF5 and merges it with geospatial shapefiles with the contours of FAF5 regions to associate each region with tons, ton-miles, and associated emissions of imports to and exports from each region, along with areal densities of these three quantities (i.e. divided by the surface area of the associated region). There is also functionality to evaluate these quantities for a user-specified mode, commodity, origin region, or destination region.

Before running this code, you'll need to have first run the following:

python source/ViusTools.py

To run:

python source/Point2PointFAF.py -m user_specified_mode -c "user_specified_commodity" -o user_specified_origin_ID -d user_specified_destination_ID

This should produce a csv and shapefile in data/Point2Point_outputs/mode_truck_commodity_Logs_origin_11_dest_all.[extension].

NOTE: The "" around the commodity option is important because some commodities contain spaces, and python does NOT like command line arguments with spaces...

where each argument defaults to 'all' if left unspecified. The mode is one of {all, truck, water, rail}. The available commodities can be found in the 'Commodity (SCTG2)' sheet in data/FAF5_regional_flows_origin_destination/FAF5_metadata.xlsx ('Description' column). The origin and destination region IDs can be found in the 'FAF Zone (Domestic)' sheet of the same excel file ('Numeric Label' column').

For example, to filter for logs carried by trucks from FAF5 region 11 to FAF5 region 139:

python source/Point2PointFAF.py -m truck -c Logs -o 11 -d 139

There's also a bash script in source/run_all_Point2Point.sh that can be executed to produce merged shapefiles for all combinations of modes, commodities, origins and destinations.

To run:

bash source/run_all_Point2Point.sh

WARNING: This may take several hours to run in full, and the shapefiles and csv files produced will take up ~100 GB. To reduce this, you can comment out items that you don't want in the COMMODITIES, REGIONS and MODES variables.

The script PrepareHydrogenHubs.py combines locations and information about operating and planned hydrogen production facilities and the U.S. and Canada into shapefiles located in data/hydrogen_hubs/shapefiles. To run:

python source/PrepareHydrogenHubs.py

The script IdentifyFacilitiesInRadius.py identifies truck stops and hydrogen production facilities within a user-provided radius and central location - (33N, 97W) and 600 miles by default.

The script AnalyzeTruckStopCharging.py is designed to quantify the charging demand at truck stops along U.S. interstates that are sparsified to support the specified truck range, and estimate the potential difference in infrastructure costs needed if the entire electrified trucking fleet were to electrify and either:

• The full electrified fleet shared the investment and usage of charging infrastructure, or
• The fleet was divided in half, and each half invested in and used their respective charging infrastructure separately.

The idea of this exercise is to understand the potential infrastructure savings from trucking fleets pooling infrastructure investments in charging infrastructure based on real-world freight flow data.

The methodology is detailed in MacDonell and Borrero, 2024.

To run with a specified set of options:

python source/AnalyzeTruckStopCharging.py -c [charging time (hours)] -m [max allowable wait time (hours)] -r [truck range (miles)]

To run over all options visualized in the geospatial mapping tool:

bash source/run_all_AnalyzeTruckStopCharging.sh

The script EvaluateTruckingEnergyDemand.py aggregates highway-level FAF5 commodity flows and trips to evaluate the approximate annual energy demand (in MWh) that would be placed on the grid for each state if all trucking operations were to be fully electrified. The energy demand is calculated assuming that the flows are carried by the Tesla Semi, using the mileage with respect to payload calibrated using code in this repo (link to relevant section of README). The underlying calibration is performed in this repo using data from the PepsiCo Tesla Semi pilot.

To run:

python source/EvaluateTruckingEnergyDemand.py

This produces an output shapefile in data/trucking_energy_demand containing the energy demand for each state from electrified trucking, both as an absolute value (in MWh), and as a percent of each of the following:

• The total energy generated in the state in 2022
• The theoretical total energy generation capacity for the state in 2022 (if the grid were to run at its full summer generating capacity 24/7)
• The theoretical excess energy generation capacity (i.e. theoretical - actual energy generated in 2022)

The script TT_charging_analysis.py produces a plot visualizing the demands associated with electrifying trucking with charging at 8 sites in the Texas triangle region. To run:

python source/TT_charging_analysis.py

This will produce Texas_charger_locations.png in the plots directory that compares the

The script MakeChargingLoadByZone.py produces a csv file for each ERCOT weather zone containing one or more charging sites. For each such zone, the csv file contains the daily load from each charging site in the weather zone, assuming it follows the most extreme variation found in Borlaug et al (2021) for immediate charging (see red curve in Fig. 5 in the paper).

To run:

python source/MakeChargingLoadByZone.py

This will produce a csv file daily_ev_load_[zone].csv for each zone.

The script AnalyzeErcotData.py compares the daily EV demand each charging site in a zone (along with the total combined demand) with the estimated excess capacity of the grid over the day.

To run:

python source/AnalyzeErcotData.py

This will produce a plot for each zone and month called daily_ev_load_with_excess_[zone]_[month].png in the plots directory.