Test Systems / RBTS / InfraRel Validation
RBTS Substation Configurations: InfraRel Validation Protocol
Reproduce selected historical SUBREL results, explain discrepancies, then compare alternative station configurations under matched assumptions.
Research objectives
- Reconstruct the original RBTS bus-branch reference and the breaker-and-a-half and ring-substation node-breaker representations.
- Represent station-originated single faults, isolation operations, stuck breakers, and relevant two-fault combinations without energizing an uncleared fault.
- Compare the deterministic service outcomes of a selected traceable event set across legacy SUBREL and InfraRel.
- Reproduce statistical reliability indices using compatible component outage, repair and restoration assumptions.
- Separate the effect of station architecture from the effect of modeling detail and study assumptions.
Stage A — Provenance and equivalence audit
Inventory and checksum original RAW, SAV, PWB/PWD, AUX, SUBREL project, Access MDB, contingency export and historical results files. Match buses, load points, generators, line endpoints, breaker identities, normal status, protective zones, maintenance policies, component outage rates and customer damage curves. Record differences rather than adjusting historic data silently.
Stage B — Deterministic event parity
For each selected event, compare the initiating component, clearing breakers, unsuccessful breaker operations, devices isolated, surviving energized paths, available load, curtailment, restoration action sequence and outage duration. Reconcile fault/isolation/fault-plus-stuck definitions between the 484-case ring contingency export and the larger historical workbook enumeration.
Stage C — Reliability and service indices
With source-matched failure and repair data, calculate SAIFI, SAIDI, CAIDI, ASAI, expected unserved energy and outage costs by configuration. Report raw and customer-weighted denominators, cut sets and the treatment of overlapping events. Include study-scoped contingency counts and explain deviations before interpreting comparisons.
Stage D — Extensions after parity
Only after reproducing reference behavior, evaluate architecture changes, automatic and manual switching, protection settings, common-mode failures and operational/maintenance states. Distinguish verified historical assumptions from new InfraRel scenarios.
Acceptance and publication rules
| Item | Minimum evidence |
|---|---|
| Topology | Bus and equipment mapping, normally open/closed states, no inadvertent parallel source or uncleared fault |
| Event behavior | Case-by-case comparison of tripped/isolated devices, lost load and restoration time; all deviations logged |
| Numerical indices | Declared input datasets, formulas, run identifiers, reproducible files and reconciliation tolerances fixed before runs |
| Comparability | Matched outage rates and operating assumptions or explicit documented exceptions |
| Paper claims | Clearly identify legacy reported values, independently reproduced values and new InfraRel results |
Proposed paper
Impact of Substation Configurations on Power-System Reliability: A Reproducible Comparison Using RBTS, SUBREL and InfraRel. A benchmark methodology paper should report differences between physical architectures and the accuracy and interpretability of the modeling approach, not rank station schemes from inconsistent input models.