ReV tutorials

update package names from `NREL` to `nlr`

fix broken links

concepts

Geospatial Data Science using Geographic Information Systems

Energy Planning and Production

resource

• the input utilized for energy generation by a generator

resource potential

• theoretical availability of renewable energy across an area
• quantifies the energy constant of a resource at specific location

NLR

\NREL

technical potential

• the maximum amount of capacity and generation possible in an area
  • given physical, technological, and regulatory constraints
• the measure of developable resource potential based on user assumptioms of the area
  • e.g. user may constrain developmented based on what land-use intersects/overlaps

technical potential establishes an upper boundary estimate of development potential

oil and gas

availability potential

• will be represented as a combination of available gas supply and deliverable gas volumes
• each potential power plant will have access to deliverable capacity of the nearest service line

deliverable potential

• maximum amout of deliverable oil or gas
  • based on delivery infrastructure and assumptions of available resources

reV model

• simulates renewable energy generation and cost potential across a landscape
• simulates discrete generation plants with spatiotemporal constraints
• serves a spatial coordinator for SAM, connecting energy plants to transmission network

Levelized Cost of Energy

• average revenue per unit of electricity generated
  • that would be required to recover the costs of building and operating a generating plant during an assumed financial life

technical potential

• represented as a function of available land after applying…
  • exclusions
  • net capacity factor (esimtated by generation module)
  • user-defined power density appropriate for specific technology
    • power density value represents the maximum potential capacity for a given unit of area \((MW/km^2)\)
• the LCOE output provides insight into decision-making for economics, performance, and regionalized differences
  • LCOE also used in supply curves

installing reV

pip install NREL-reV
# for retrieving remote resource data from NREL servres
pip install NREL-reV[hsds]

pip install NREL-reV[hsds] (Windows)

error: Microsoft Visual C++ 14.0 or greater is required

`conda create -n rev python=3.12`

`conda activate rev`

`python -m pip install NREL-reV`

alt

`uv venv –python 3.12`

`source .venv/Scripts/activate`

`uv pip install NREL-reV`

reV inputs

full model pipeline of reV requires:

• SAM configuration
  • used to simulate energy generation and costs
• Resource HDF5 dataset
  • spatial timeseries of energy resources
• project points csv
  • where to run SAM configuration
• exclusion HDF5 dataset
  • used to model available land and capacity given technologies
• transmission connection csv
  • table of supply curve points representing potential power generation sites
    • and their potential interconnection tie-line distances and costs

SAM Configurations

• SAM C/C++ code simulates performance and economics for a suite of energy technologies
  • reV accesses PySAM which wraps SAM C API

reV is a spatial-wise application of SAM

• runs SAM at given location(s) to simulate given technology(s)

Building Configuration

• use SAM GUI

• write JSON from scratch

  example with a minimum set of parameters for a PV system and LCOE calculator

{
    "array_type": 2,
    "azimuth": 180,
    "capital_cost": 39767200,
    "dc_ac_ratio": 1.3,
    "fixed_charge_rate": 0.096,
    "fixed_operating_cost": 260000,
    "losses": 14.07566,
    "module_type": 0,
    "system_capacity": 20000,
    "tilt": 0,
    "variable_operating_cost": 0
}

Resource HDF5 dataset

• 2D array for each data variable (GHI, windspeed, air pressure, etc.)
• time index on the y axis
  • site index on the x axis
• `meta` stored as a struct array
  • containing `coordinate`,`timezone`, `elevation`, and other (associated with x axis)
• 1D `time<sub>index</sub>` vector that contains datetime strings (associated with y axis)
• `scale<sub>factor</sub>` attribute on each variable that is used to translate integers back into floats where scaling is used for data storage
• `units` attribute that stores the units for each variable

remote access using `rex` and `hsds`

• seamlessly integrate remote resource access into reV modeling pipeline

Project Points

points that subset the data that pulled from resource datasets for a given geographic region

• project points files must have two columns
  • `gid`
    • index position of x axis of resource array
      • or row index of the resource meta data table
  • `config`
    • key associated with SAM configuration in `config<sub>gen</sub>.json`
      • e.g. for wind it is `onshore`

to run make sure to have an editable install

• `pip install -e .`

import os
import h5py
import pandas as pd
from rex import Resource
from rev_tutorial import DATA


# Path to the HDF5 resource file
RES_FILE = DATA.joinpath("ri_100_nsrdb_2012.h5")


def main():
    """Read, format, and write a project points file."""
    # Use rex's Resource class to pull out a formatted resource meta dataframe
    with Resource(RES_FILE) as res:
        meta = res.meta

    # The "gid" column is the original index and tell reV where to run
    meta.loc[:, "gid"] = meta.index

    # Now you can subset, as long as the gid refers to the original index
    meta = meta[meta["population"] < 10_000]

    # The config column tell reV which SAM config to run
    # Here we are assuming a single 'default' config will be run every where
    meta.loc[:, "config"] = "default"

    # All reV needs is config and gid, but you can include other fields
    project_points = meta[["latitude", "longitude", "gid", "config"]]

    # Save this to a project points file
    fname = "project_points.csv"
    dst = HOME.joinpath(f"tutorial_3_input_builds/make_project_points/{fname}")
    project_points.to_csv(dst, index=False)


if __name__ == "__main__":
    main()

Exclusion data

• rev-compatible high resolution land use hdf5 datasets used for exclusions and characterizations

  • e.g. values within the arrays indicate whether a grid cell is excluded based on specific exclusion layer such as protected land area

• metadata includes description, sources, and more to identify goal of each type of exclusion

  exclusion.py provides steps to upload existing geotiffs into a new or existing reV-compatible hdf5 file using reVX

  • exclusion datasets in `*.tif` format
  • file paths in `LAYER<sub>FPATHS</sub>`
  • descriptions in `LAYER<sub>DESCS</sub>`
  • first function `geotiffs<sub>toh5</sub>()` uploads geotiffs to a new/existing h5 file
  • second function `h5<sub>togeotiffs</sub>()` pulls existing h5 datasets out as a geotiff
    • e.g.
      • protected areas: `padus.tif`
      • ReEDS: `reeds<sub>regions</sub>.tif`
      • human settlement: `smod.tif`
      • srtm-derived slope `srtm<sub>slope</sub>.tif`

Transmission data

• a lookup table with costs and distances associated with all available connections for a reV supply-curve aggregation table

• supply curve algorithm uses four types of electricity infrastructure features:

  • transmissions lines, substations, load centers and synthetic features that can be defined by the user or based on the specific modeling application

reV generation

• controls what and where SAM runs
  • using a JSON configuration

for a basic run

`config<sub>generation</sub>.json`

• `technology`
  • SAM generation module string available in reV
    • e.g. 'windpower', 'geothermal', 'Pvsamv1'
• `project<sub>points</sub>`
  • csv table of locations to run simulations on
• `sam<sub>files</sub>`
  • dict of paths to SAM config file(s)
• `resource<sub>file</sub>`
  • path(s) to resource h5 file(s)
  • multiyear resources are stored in a single string with a set of `{}`
• `output<sub>request</sub>`
  • list of datasets you want in output hdf5 file
  • availability of outputs depends SAM config used
  • some outputs have names specific to reV
    • e.g. `capacity<sub>factor</sub>` SAM variable is `cf<sub>mean</sub>` in reV

      • full capacity factor timeseries is `cf<sub>profile</sub>` in reV

        // Below is the sample output request with links to descriptions or comments for each variable:
        
        "output_request": [
            "cf_mean",  # reV-derived, see [[https://nrel.github.io/reV/_autosummary/reV.SAM.generation.PvWattsv8.html#reV.SAM.generation.PvWattsv8.cf_mean][description]]
            "cf_profile",  # reV-derived, see [[https://nrel.github.io/reV/_autosummary/reV.SAM.generation.PvWattsv8.html#reV.SAM.generation.PvWattsv8.cf_profile][description]]
            "lcoe_fcr",  # reV-derived, see [[https://nrel.github.io/reV/_autosummary/reV.SAM.econ.LCOE.html#reV.SAM.econ.LCOE.lcoe_fcr][description]]
            "gh",  # Directly from SAM, see [[https://nrel-pysam.readthedocs.io/en/latest/modules/Pvwattsv8.html#PySAM.Pvwattsv8.Pvwattsv8.Outputs.gh][description]]
            "ghi_mean",  # reV-derived, see [[https://github.com/NREL/reV/blob/0f71e9e97cc320a085c519819750f3a5a6889f5f/reV/SAM/generation.py#L161][method]]
            "capital_cost"  # Passed through from SAM inputs, see [[https://nrel-pysam.readthedocs.io/en/latest/modules/Pvwattsv8.html#PySAM.Pvwattsv8.Pvwattsv8.Outputs.gh][description]]
        ]
        

• `execution<sub>control</sub>`
  • dict for execution control arguments
  • can be left out for small, local runs
  • entry includes configuring:
    • `max<sub>workers</sub>` (# of cores)
    • `sites<sub>perworker</sub>` (project points being processed per core)
    • `memory<sub>utilizationlimit</sub>` (max memory allocation before dumping data)
• `log<sub>directory</sub>`
  • path to logs directory
  • optional
  • each log file will be named after containing folder
    • and the process ID of the job
• `log<sub>level</sub>`
  • string of desired logger verbosity

running the model

reV generation -c config_generation.json

• command is `reV`
  • submodule is `generation`
  • `-c` points to rev generation config json file