BACKGROUND

This repo supports two-level grid planning given a shapefile of points. It creates MV and LV .shp files and also returns relevant connectivity metrics : Average LV wire length (m) per structure, Average MV wire length (m) per structure, Average Number of Structures per Transformer andAverage Cost per structure ($USD)

@article{fobi2021scalable, title={A scalable framework to measure the impact of spatial heterogeneity on electrification}, author={Fobi, Simone and Kocaman, Ayse Selin and Taneja, Jay and Modi, Vijay}, journal={Energy for Sustainable Development}, volume={60}, pages={67--81}, year={2021}, publisher={Elsevier} }

Directory Structure

├── README.md               <- The top-level README for developers using this project.
├── data
│   └── raw                 <- Sample point locations to create a network design (i.e. low voltage, medium voltage and transfor)
│
├── reports                 <- This directory contains the published paper for this work
│
├── requirements.txt        <- The requirements file for reproducing the analysis environment
│
├── src                     <- Source code for use in this project.
    │
    ├── dataprep            <- Scripts to preprocess data
    │   └── split_data.py   <- Recursively split points to obtain smaller coverage areas
    │
    ├── TLND                <- Scripts to run network design model and reconstruct MV for larger areas
        ├── TwoLevelNetworkDesign.py
        └── mvReconstruction.py
    

Getting started.

• Fork this repo, DO NOT clone it as it will not support pushing by noncollaborators.
• Create a python 3 environment (conda create -n network python=3.6)
• Install gdal with working numpy (conda install gdal numpy pandas)
• Run the requirements file to install other packages (pip install -r requirements)

Running the Code

• Case 1 (Small number of nodes e.g. <1000): If the shapefile has a small number of points (< 1000), the TLND can be run without further splitting.
  • Run python src/TLND/TwoLevelNetworkDesign.py)
• Case 2 (Large number of nodes): When there are many nodes to consider within a shapefile, first split the shapefile to smaller files, then run the TLND and after reconstruct the medium voltage network.
  • Run python src/dataprep/split_data.py
  • Next for each generated sub-file run python src/TLND/TwoLevelNetworkDesign.py
  • After all the networks are generated run mvReconstruction.py to obtain the final network.

Things to Note

• The provided shapefiles have to be in a CRS that uses meters. There is a function (convert_to_utm_shp) in TwoLevelNetworkDesign.py that supports projection from WGS84 to desired crs. The function takes in a csv and outputs a shapefile in the crs.