Home / Docs / GAT – Grid Analysis Toolkit / Reference & Theory / Reliability Metrics
2 min read

Analytics & Reliability Metrics

Analytics commands extract grid insights: power transfer distribution, reliability indices, and interconnection limits.

Key Metrics

PTDF — Power Transfer Distribution Factors

  • Sensitivity: 1 MW injection at bus A → how much MW flows on each branch?
  • Use case: Determine congestion risks for renewable injection
  • Linear analysis (post-contingency or steady-state)

Reliability Metrics

  • ENS — Energy Not Served (MWh unserved per year)
  • LOLE — Loss-of-Load Expectancy (hours/year at risk)
  • CAIFI — Customer Average Interruption Frequency Index
  • Based on outage scenarios (N-1, N-2) and restoration times

ELCC — Effective Load Carrying Capability

  • How much load can be served with a new resource (wind/solar/battery)?
  • Incorporates weather, demand, and existing resources

Deliverability — Transmission hosting capacity

  • How much renewable energy can be delivered to load?
  • Limited by transmission thermal ratings and voltage stability

DS — Demand Served

  • Fraction of load met after N-1 screening
  • Input to reliability index calculations

Usage Examples

Compute PTDF for a source-sink pair

gat analytics ptdf \
  --grid transmission_network.arrow \
  --source bus_1 \
  --sink bus_2 \
  --transfer 1.0 \
  --solver gauss \
  --out ptdf_1_2.parquet

Output:

  • Branch ID, flow (MW), PTDF (fraction per MW)
  • Summary: max PTDF, min PTDF, branches above threshold

Reliability analysis

gat analytics reliability \
  --grid network.arrow \
  --outages contingency_scenarios.yaml \
  --restoration-times outage_mttr.csv \
  --demand demand_profile.csv \
  --out reliability_indices.parquet

Output:

  • Outage ID
  • Peak unserved load (MW)
  • Energy not served (MWh)
  • Estimated LOLE, CAIFI contributions

ELCC for solar

gat analytics elcc \
  --grid network.arrow \
  --weather solar_irradiance.csv \
  --demand load_profile.csv \
  --existing-resources existing_gens.yaml \
  --candidate-capacity 50 \
  --candidate-type solar \
  --out elcc_solar.parquet

Output:

  • Effective load carrying capability (MW)
  • Margin above nameplate (if any)
  • Sensitivity to weather data

Deliverability screening

gat analytics deliverability \
  --grid network.arrow \
  --injection-point feeder_123 \
  --injection-ramp 100 \
  --max-penetration 30 \
  --out deliverability.parquet

Integration with Planning Workflows

Use analytics in batch studies:

# Scenario-based reliability
gat scenarios materialize --spec scenarios.yaml --grid-file network.arrow --out-dir runs/scenarios
gat batch opf --manifest runs/scenarios/scenario_manifest.json --out runs/batch/opf_results

# Then compute reliability from batch results
gat analytics reliability \
  --grid network.arrow \
  --batch-root runs/batch/opf_results \
  --out reliability_summary.parquet

References

  • CLI: gat analytics --help
  • Schemas: docs/schemas/analytics_*.json
  • Examples: test_data/analytics/