Dataset DOI: 10.5061/dryad.dz08kpsbm

Description of the data and file structure

This repository contains input data and software required to run GridPath-India power system planning models for India from 2020–2050. The data include hourly capacity factor time series (capacity_factors.zip) for solar and wind resources simulated for FY2018–19 using ERA5 (wind) and NREL NSRDB PSM v3 (solar), with wind speeds bias-corrected using the Global Wind Atlas. Technologies represented include offshore (/offshore) and onshore wind (existing /wind_existing/, adjusted /wind_existingAdjusted, and new /wind_new) and solar PV (fixed-tilt /SolarPV_singleAxis, single-axis tracking /solarPV_singleAxis, and rooftop /solarPV_roofTop). Technology-specific assumptions are applied for turbine type, hub height, PV configuration, and losses, and adjusted wind profiles are calibrated to match reported FY2018–19 generation at the state level.

The repository also provides technology cost trajectories (technology_cost.zip) for bettery (Battery.csv), pumped storage hydro (Hydro_Pumped.csv), hydrogen (Hydrogen.csv), thermal generators (Thermal.csv), solar PV (SolarPV.csv), and wind (Wind.csv) from 2020–2050, along with complete GridPath CSV input files (gridpath_csv_files.zip). Two alternative sets of hourly load demand projections are included: (i) india_2020-2050-24_ICED, which linearly scales FY2018–19 demand profiles to future years, and (ii) india_2020-2050-24_PIERv2, which uses bottom-up, weather-synchronized FY2018–19 demand profiles from PIERv2 projected forward to 2050.

Files and variables

capacity_factors.zip

/offshore

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

Offshore wind capacity factors are simulated using ERA5 hourly wind speeds, bias-corrected so that annual mean wind speeds match the Global Wind Atlas (GWA) at each site. Generation is calculated using 135m hub-height, 7MW offshore turbines (Vestas 164). Simulations are performed for 2019 at hourly resolution.

/wind_existing

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

Existing onshore wind capacity factors are simulated using ERA5 hourly wind speeds adjusted to match GWA annual mean wind speeds. Site-specific turbine classes (low, mid, or high wind) are used based on local resource conditions. Simulated generation reflects physical turbine performance but does not include operational derates beyond standard availability factors and losses. Simulations are performed for 2019 at hourly resolution.

/wind_existingAdjusted

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

This dataset applies post-simulation adjustments to wind_existing capacity factors so that FY2018–19 generation aggregated at the state level matches reported wind generation. These adjustments account for operational constraints, curtailment, and performance factors not captured in the physical wind model.

/wind_new

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

New onshore wind capacity factors are simulated using ERA5 wind data bias-corrected with GWA, assuming newer turbine models with 120m hub height and 2MW capacity. Turbine selection is site-specific (low, mid, or high wind class), reflecting improved performance relative to existing installations. Simulations are performed for 2019 at hourly resolution.

/solarPV_fixedTilt

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

Fixed-tilt solar PV capacity factors are simulated using NREL NSRDB PSM v3 irradiance and temperature data at 4x4 km spatial resolution. Generation is calculated assuming standard monocrystalline modules with regionally appropriate tilt and standard utility-scale loss assumptions. Simulations are performed at hourly resolution for FY2018–19.

/solarPV_singleAxis

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

Single-axis tracking solar PV capacity factors are simulated using NSRDB PSM v3 data, incorporating a solar tracking mount to maximize plane-of-array irradiance incidence. Systems assume high-efficiency monocrystalline modules and standard tracking system losses. Simulations are performed at hourly resolution for FY2018–19.

/solarPV_roofTop

• time: date in format mm/dd/yyyy HH:MM for the value capacity_factor column.
• capacity_factor: ratio of power output with respect to installed capacity.
• project: project name (XX_Y), where XX is load_zone_abr and Y is the zone letter in the load_zone (alphabetic order).
• capacity_mw: maximum available capacity in the zone in MW.
• load_zone_abr: Indian states common two-letter short forms.

Rooftop solar PV capacity factors are simulated using NSRDB PSM v3, assuming standard monocrystalline modules with higher losses due to soiling and installation constraints relative to utility-scale PV. These capacity factors represent distributed rooftop systems rather than optimized ground-mounted plants. The profile for each state corresponds to the most populated city. Simulations are performed at hourly resolution for FY2018–19.

technology_cost.zip

Battery.csv

• Power cost (USD/kW): Low, mid, and high trajectories (2020–2050)
• Energy cost (USD/kWh): Low, mid, and high trajectories (2020–2050)
• Power O&M (USD/kW-yr): Low, mid, and high trajectories (2020–2050)
• Energy O&M (USD/kWh-yr): Low, mid, and high trajectories (2020–2050)

Hydro_Pumped.csv

• Power cost (USD/kW): Low, mid, and high constant trajectories
• Energy cost (USD/kWh): Low, mid, and high constant trajectories
• O&M cost (USD/kW-yr): Low, mid, and high constant trajectories

Hydrogen.csv

• Capacity costs:
  • Power (USD/kW): Electrolyzer (proton exchange membrane) input combined with fuel cell output; low, mid, and high trajectories (2020–2050)
  • Energy (USD/kWh): Tank or cavern; low, mid, and high trajectories (2020–2050)
• O&M (USD/kW-yr): Low, mid, and high trajectories (2020–2050)

Thermal.csv

• Capital cost (USD/kW) and O&M (USD/kW-yr) by technology:
  • Nuclear: mid constant trajectory
  • CT and CCGT: mid constant trajectory
  • Supercritical coal: low and high constant trajectories

SolarPV.csv

• Capital cost (USD/kW):
  • Single-axis tracking: low, mid, and high trajectories (2020–2050)
  • Fixed-tilt: low, mid, and high trajectories (2020–2050)
  • Rooftop PV: low, mid, and high trajectories (2020–2050)
• O&M (USD/kW-yr):
  • Single-axis, fixed-tilt, and rooftop PV: mid trajectory (2020–2050)

Wind.csv

• Capital cost (USD/kW):
  • Onshore and offshore wind: low, mid, and high trajectories (2020–2050)
• O&M (USD/kW-yr):
  • Onshore and offshore wind: mid trajectory (2020–2050)

gridpath_csv_files.zip

/india_2020-2050-24_ICED

This folder contains GridPath input files (gridpath_input) in which hourly load demand profiles for 2020–2050 are derived by linearly scaling FY2018–19 demand profiles. The temporal shape of demand is preserved from FY19, while annual demand levels are scaled to match projected growth in PIERv2. 

/india_2020-2050-24_PIERv2

This folder contains GridPath input files (gridpath_input) that use bottom-up hourly load demand profiles from PIERv2. These profiles are weather-synchronized with FY2018–19 and projected forward to 2020–2050.

The gridpath_input directory contains all model inputs required to reproduce the GridPath simulations used in this study. Inputs are organized by model component following the native GridPath input structure. Each top-level folder corresponds to a specific class of model assumptions:

• fuels/: fuel prices and fuel availability assumptions
• markets/: market structure and market-related constraints
• policy/: policy constraints (e.g., emissions targets, renewable requirements)
• project/: generation, storage, and transmission project definitions
• reliability/: resource adequacy and reliability constraints
• reserves/: operating reserve requirements
• solver/: optimization solver settings
• system_load/: hourly electricity demand profiles (ICED or PIERv2)
• temporal/: timepoint, period, and chronology definitions
• transmission/: inter-zonal transmission constraints
• tuning/: numerical and performance tuning parameters

Scenario selection and model configuration are controlled through the scenarios*.csv files. These files define which inputs from the above folders are activated in each model run and specify the different combinations of demand assumptions, cost trajectories, and policy settings analyzed.

Some folders are empty. However, all folders are included intentionally to document the experimental design and to allow exact reproduction of the simulations in GridPath.

GridPath-India

Pre-Setup Instructions

Before running the model, download the gridpath_input_csvs.zip folder from Dryad repository and save it to your desktop. The path should be ~/Desktop/gridpath-india/gridpath_input_csvs.

Setting up Conda

Install Anaconda

Anaconda is a package manager and environment manager for Python/R. It simplifies managing libraries and creating isolated environments. Follow the installation instructions based on your OS:

• macOS
• Windows
• Linux

Terminal Access

Use the terminal to interact with Conda:

• macOS: Open Terminal via Spotlight or under Applications > Utilities.
• Linux: Open Terminal from the applications menu or by pressing Ctrl + Alt + T.
• Windows: Open "Anaconda Prompt" from the Start menu.

Creating a Conda Environment

In the terminal, create a new environment named <myenv> with Python version 3.9 for GridPath:

conda create -n <myenv> python=3.9

Conda offers various tools to manage environments, including activating, deactivating, listing, and removing them as needed. For detailed information on managing Conda environments, refer to this guide. For example, you can view all existing environments in Conda by running:

conda env list

Setting up model requirements in terminal

Activating the Conda environment

To activate the environment by run the following in the terminal:

conda activate <myenv>

Once activated, the environment name will appear at the beginning of the terminal prompt in parenthesis - any subsequent package installations will take place within this environment.

Installing open-source solver

GridPath needs an open-source solver software (cbc) to solve the optimization problem. To install cbc:

For Mac users: run the following command to set up the cbc solver:

conda install -c conda-forge pyomo coincbc

For Windows PC users:

1. Windows PC users need to copy the cbc.exe file to the Library/bin folder of your Anaconda environment. The cbc.exe file is located under the ‘solver’ in the primary model folder (state_model).
2. In the Anaconda Prompt, run the following to get the list of conda environments in the system:

conda env list

1. Copy the path next to the environment where is installed. It would look something like this: C:\ProgramData\anaconda3\envs
2. Open File Explorer and paste the path you copied into the address bar. Navigate to the Library folder, then to the bin folder.
3. Copy the cbc.exe file from the location state_level\solver\cbc.exe and paste it into ~\Library\bin\ (the path navigated in step 5).

Git repositories

A Git repository (repo) is a version-controlled storage space for project files, tracking changes over time and allowing users to collaborate without overwriting each other's work. Repos can be hosted locally or on remote platforms like GitHub, facilitating efficient code management and collaboration. For more information, go to https://github.com/

For Mac users: Git is installed by default in the OS and no extra steps are required.

For Windows users: visit the Git for Windows website and download the latest version. After downloading, run the installer and follow the prompts.

Clone gridpath repo

GridPath is a versatile grid-analytics platform that seamlessly integrates several power-system planning approaches – including production-cost, capacity-expansion, asset-valuation, and reliability modeling – within the same software ecosystem. More information can be found here: https://gridpath.readthedocs.io/en/latest/index.html

For Mac users:

cd ~/Desktop/gridpath-india 
git clone https://github.com/blue-marble/gridpath.git --branch v0.16.1
cd ~/Desktop/gridpath-india/gridpath
pip install .

For Windows users:

cd *\Desktop\gridpath-india
git clone https://github.com/blue-marble/gridpath.git --branch v0.16.1
cd gridpath
pip install .

Access information

State-Level Model Tutorial

For demo state-level models capable of running in your local machine, please follow the instructions on this repository: https://github.com/cetlab-ucsb/gridpath-workshop-ucsb 

GridPath Documentation

GridPath is a versatile grid-analytics platform that seamlessly integrates several power-system planning approaches – including production-cost, capacity-expansion, asset-valuation, and reliability modeling – within the same software ecosystem. 

More information on Gridpath can be found here:https://gridpath.readthedocs.io/en/latest/introduction.html 

Perspectives on Indian Energy based on Rumi (PIER) 

PIER is an Indian energy systems model for India built on the Rumi modelling platform that estimates demand and cost-optimal supply options to meet the demand. Version 2.0 of PIER currently features energy demand estimation for five different sectors: residential, transport, industry, agriculture and "others" up to 2040-41

For more information on PIER v2, visit: https://energy.prayaspune.org/our-work/research-report/pier-detailed-demand-side-energy-modelling-of-residential-transport-industry-sectors-for-india-from-fy2023-24-to-fy2040-41  

Multi-criteria Analysis for Planning Renewable Energy (MapRE) 

The Multi-criteria Analysis for Planning Renewable Energy (MapRE) seeks to provide a framework for the systematic identification and valuation of areas for renewable energy development–focusing mainly on solar and wind technologies–for developing countries.

For more information, visit https://mapre.es.ucsb.edu/