Skip to content

feat(fpb): add performance depth rules - #55

Merged
jscott3201 merged 2 commits into
mainfrom
faults/fpb-performance-depth
Aug 20, 2026
Merged

feat(fpb): add performance depth rules#55
jscott3201 merged 2 commits into
mainfrom
faults/fpb-performance-depth

Conversation

@jscott3201

@jscott3201 jscott3201 commented Aug 20, 2026

Copy link
Copy Markdown
Owner

Summary

  • add FPB-0004 terminal fan airflow degradation with a positive expected-flow baseline
  • add FPB-0005 ambiguous primary-airflow sensor disagreement against an independent reference
  • add FPB-0006 hydronic reheat heat-transfer degradation at near-full command
  • extend FPB points, telemetry tiers, reciprocal relationships, and the family playbook
  • use denominator-free cross-multiplied residuals and document all host-derived model obligations

Implementation decisions

Rule Evidence role Numeric policy Recorded deviation
FPB-0004 LBNL PFPU/SFPU restricted-flow class is a future replay target; PMP-0006 is the positive-baseline graph precedent 50 L/s is NO_PORTABLE_DEFAULT; 20% and 900 s are ADOPTED_TUNABLE cross-multiplication replaces Divide; signature is degradation, not a root-cause claim
FPB-0005 LBNL airflow-bias class is a future replay target; SYS-0005/0006 supply sensor-health precedent 50 L/s is NO_PORTABLE_DEFAULT; 15% and 900 s are ADOPTED_TUNABLE ambiguous adjudication does not automatically invalidate primary_airflow
FPB-0006 LBNL coil-fouling classes are future replay targets; PMP-0006 and coil-performance cards supply graph precedent 3 K is NO_PORTABLE_DEFAULT; 90%, 30%, and 900 s are ADOPTED_TUNABLE cross-multiplication replaces Divide; observable heat-transfer degradation is not called fouling

No numeric default is claimed as transcribed source guidance. PR11 owns external-data replay, so this PR makes no empirical FPR or TPR claim.

Validation

  • targeted verifier: FPB-0004 through FPB-0006, 50 scenarios
  • full verifier: 134 rules, all scenarios passed
  • registry lint: 134 rules OK
  • SVG fitment: 135 SVGs clean
  • simharness unit tests: 6 passed
  • mdBook: 134 fault pages, 21 playbooks, 13 point dictionaries
  • git diff --check

Source and ontology posture

  • LBNL simulated PFPU/SFPU data is mapped as a future replay source; this PR makes no empirical FPR or TPR claim
  • Brick 1.4.4, ASHRAE 223 PPR2.1, and QUDT 3.1.4 terms retain the family artifact pins
  • derived baseline/reference points use Function hasOutput QuantifiableProperty and do not invent expected or derived ontology aspects

Deferred

PR11 will add the external LBNL dataset adapter and replay manifest without committing the dataset.

@jscott3201
jscott3201 merged commit 4e12cb4 into main Aug 20, 2026
3 checks passed
@jscott3201
jscott3201 deleted the faults/fpb-performance-depth branch August 22, 2026 18:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant