Dataset Weather-Grid GridView Operations Dataset

download Download

This dataset contains time series of output operational features from production cost simulations over historical weather conditions. Datasets are presented with respect to the Western Interconnection of the US, on a zonal scale. Operational features are obtained from production cost simulations based on the Western Electricity Coordinating Council (WECC) 2032 Anchor Data Set (ADS), over weather conditions from 1982-2019.

 

For more details, please refer to: O. Anderson, B. K. Ketineni, K. Oikonomou, C.D. Burleyson, C. Bracken, and N. Voisin, "A Dataset of Operational Characteristics Based on 38 Weather Years Impacting the U.S. Western Interconnection", under review in Scientific Data.

 

Corresponding author(s): Osten Anderson (osten.anderson@pnnl.gov) and Nathalie Voisin (nathalie.voisin@pnnl.gov)

 

Overview

Hourly WECC operations output from a GridView production-cost simulation, organized by variable. The study sweeps 38 historical weather years (1982–2019) against a single infrastructure scenario. Per-variable yearly files hold one weather year each; regional aggregation files hold all 38 weather years stacked.

  • Root: GridView/
  • Total size: ~1.1 GB
  • Total files: 248 CSVs
  • Subdirectories: 7

Subdirectory inventory

Shapes are rows × columns, excluding the header.

  • Curtailment/

    • Files: 38
    • Size: 61 M
    • File pattern: curtailment_Y{yyyy}.csv
    • Shape: 8 760 × 36 BAs
    • Resolution: Hourly, per BA
  • Import_Limits/
    • Files: 38
    • Size: 168 K
    • File pattern: import_limits_Y{yyyy}.csv
    • Shape: 36 × 1
    • Resolution: Annual scalar per BA
  • LMPs/
    • Files: 38
    • Size: 219 M
    • File pattern: lmps_Y{yyyy}.csv
    • Shape: 8 760 × 36 BAs
    • Resolution: Hourly, per BA
  • Load/
    • Files: 38
    • Size: 202 M
    • File pattern: load_Y{yyyy}.csv
    • Shape: 8 760 × 36 BAs
    • Resolution: Hourly, per BA
  • Load_Shed/
    • Files: 38
    • Size: 54 M
    • File pattern: load_shed_Y{yyyy}.csv
    • Shape: 8 760 × 36 BAs
    • Resolution: Hourly, per BA
  • Net_Imports/
    • Files: 38
    • Size: 208 M
    • File pattern: net_imports_Y{yyyy}.csv
    • Shape: 8 760 × 36 BAs
    • Resolution: Hourly, per BA
  • Regional_Aggregations/
    • Files: 20
    • Size: 384 M
    • File pattern: all_{REGION}_{variable}.csv
    • Shape: 332 880 × 1 or 7
    • Resolution: Hourly, 38 years stacked, per region

{yyyy} ∈ 1982 … 2019 (38 values). Leap days are not included, every hourly file has exactly 8 760 hourly rows (365 × 24).

Timestamp conventions

The per-variable yearly files and the regional-aggregation files use different calendar conventions. Both carry weather-year semantics; only the regional files put the weather year in the timestamp itself.

  • {var}/{var}_Y{yyyy}.csv

    • Timestamp year: constant 2030
    • Weather year source: {yyyy} in the filename
  • Regional_Aggregations/
    • Timestamp year: actual weather year (1982-01-01 … 2019-12-31)
    • Weather year source: timestamp column

Example: in Curtailment/curtailment_Y1982.csv the first row is 2030-01-01 00:00:00, not 1982-01-01 00:00:00. The year 2030 is inherited from the underlying production cost case. Code utilizing this data must read the weather year from the filename for the per-BA files, and from the timestamp for the regional files.

All timestamps are provided in MST (UTC-7) and hour-beginning. First row is … 00:00:00, last row is … 23:00:00 on Dec 31.

Regional aggregation files

all_{REGION}_{variable}.csv where:

  • {REGION} ∈ BS, CA, NW, RM, SW — WECC sub-regions (Basin, California, Northwest, Rocky Mountain, Southwest).
  • {variable} ∈ generation, lmp, load, load_shed.

Each file holds 332 880 rows = 38 weather years × 8 760 h, timestamped with the actual weather year. generation files have 7 fuel/import columns (see data dictionary); the other three have a single value column named 0.

Note: Curtailment, Net_Imports, and Import_Limits have no regional aggregation companion.

Row/column constants

  • 36 BA columns, identical set and order across every per-BA yearly file and across all years. See data dictionary for the list.
  • 7 generation-mix columns in all_{REGION}_generation.csv: Coal, Gas, Hydro, Other, Solar, Wind, Imports.

File-count check

  • Yearly files: expected 6 variables × 38 years = 228; observed 228
  • Regional files: expected 5 regions × 4 variables = 20; observed 20
  • Total: 248

Caveats

  1. The model also includes external BAs and trading hubs (AESO, BCHA, CFE, TH_Malin, TH_PV, TH_Mead) that are not included as outputs.
  2. In leap years, the leap day is dropped. All time-indexed files have 8 760 rows.

See also

  • DATA_DICTIONARY.md — column-level definitions, units, BA code reference.
  • COMPLETENESS.md — integrity and completeness audit.
English