Add FCU fault family: 5 engine-verified rules — base library complete at 55 (batch 10) (#19)

FCU-FC-001 excessive operating state changes (change-counter with
in-graph yWindowFull warm-up gate; count_scale 20.0 at 180 s tick,
legal band 57.15 s <= dt <= 450 s), FCU-FC-002/003 SAT low/high in
full heating/cooling (saturation conjunct > 99%, exposed as
yHtgFullOk/yClgFullOk), FCU-FC-004/005 inactive coil leak pair
(inactive_coil_threshold 3.0 C composed per G36 FC#3 sensor bands,
fan-heat asymmetry documented).

Closeout: AHU-FC-004 card's MovingAverage ring bound corrected
(57.15 s, n+1 checkpoints — was 56.25 s); FCU-FC-001 added to CLU-01
(card proposal accepted); FCU README rows verified; root README
refreshed to 55 rules / seven families.

All 55 rules replay green at ENGINE_PIN e2ff2f8; lint ok.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

Author: jscott3201

README.md

@@ -34,10 +34,10 @@ Here is a real one — AHU-FC-050, simultaneous heating and cooling:
 
 ## What's inside
 
-- **43 verified fault rules** (and counting) across air handlers
-  (`faults/ahu/`, all 31 reference faults), rooftop units (`faults/rtu/`), and
-  VAV terminal boxes (`faults/vav/`) — heat pump, fan coil, energy recovery,
-  and VFD families in progress.
+- **55 verified fault rules** across seven equipment families: air handlers
+  (`faults/ahu/`, all 31 reference faults), rooftop units (6), VAV terminal
+  boxes (6), fan coil units (5), heat pumps (3), energy recovery ventilators
+  (2), and VFDs (2) — every base fault in the reference, engine-verified.
 - **Point dictionaries** (`points/`) grounding every canonical point name in
   Brick 1.4.4 and ASHRAE 223P, so binding a rule to a real building is
   mechanical.

clusters/clusters.json

@@ -6,7 +6,14 @@
       "id": "CLU-01",
       "name": "Simultaneous Heating & Cooling",
       "trigger": "AHU-FC-050",
-      "members": ["AHU-FC-004", "AHU-FC-005", "AHU-FC-012", "AHU-FC-059", "AHU-FC-063"],
+      "members": [
+        "AHU-FC-004",
+        "AHU-FC-005",
+        "AHU-FC-012",
+        "AHU-FC-059",
+        "AHU-FC-063",
+        "FCU-FC-001"
+      ],
       "playbook": "simultaneous-hc",
       "prevalence": "common",
       "energy_impact": "15-40% of AHU energy"
@@ -15,7 +22,11 @@
       "id": "CLU-02",
       "name": "Missing Reset Strategy",
       "trigger": "AHU-FC-057",
-      "members": ["AHU-FC-058", "AHU-FC-053", "AHU-FC-065"],
+      "members": [
+        "AHU-FC-058",
+        "AHU-FC-053",
+        "AHU-FC-065"
+      ],
       "playbook": "missing-reset",
       "prevalence": "74% of buildings (PNNL 151-building study)",
       "energy_impact": "5-20% of site energy"
@@ -24,7 +35,11 @@
       "id": "CLU-03",
       "name": "Economizer Failure",
       "trigger": "AHU-FC-051",
-      "members": ["AHU-FC-009", "AHU-FC-011", "RTU-FC-053"],
+      "members": [
+        "AHU-FC-009",
+        "AHU-FC-011",
+        "RTU-FC-053"
+      ],
       "playbook": "economizer-failure",
       "prevalence": "54% of RTUs (Cowan 2004)",
       "energy_impact": "5-20% of cooling energy"
@@ -33,7 +48,11 @@
       "id": "CLU-04",
       "name": "After-Hours Operation",
       "trigger": "AHU-FC-052",
-      "members": ["SYS-FC-052", "SYS-FC-053", "AHU-FC-060"],
+      "members": [
+        "SYS-FC-052",
+        "SYS-FC-053",
+        "AHU-FC-060"
+      ],
       "playbook": "after-hours-operation",
       "prevalence": "~15%",
       "energy_impact": "3-16% of site energy"
@@ -42,7 +61,10 @@
       "id": "CLU-05",
       "name": "Zone Heating & Cooling Conflict",
       "trigger": "VAV-FC-055",
-      "members": ["VAV-FC-052", "SYS-FC-056"],
+      "members": [
+        "VAV-FC-052",
+        "SYS-FC-056"
+      ],
       "playbook": "vav-min-flow-reheat",
       "prevalence": "common",
       "energy_impact": "5-20% of zone energy"
@@ -51,7 +73,11 @@
       "id": "CLU-06",
       "name": "Chilled Water Plant Inefficiency",
       "trigger": "CHW-FC-050",
-      "members": ["CHW-FC-051", "CHW-FC-052", "CHW-FC-053"],
+      "members": [
+        "CHW-FC-051",
+        "CHW-FC-052",
+        "CHW-FC-053"
+      ],
       "playbook": "chiller-efficiency",
       "prevalence": null,
       "energy_impact": "5-15% of chiller energy"
@@ -60,7 +86,9 @@
       "id": "CLU-07",
       "name": "Unnecessary Plant Operation",
       "trigger": "SYS-FC-050",
-      "members": ["SYS-FC-051"],
+      "members": [
+        "SYS-FC-051"
+      ],
       "playbook": "unnecessary-plant-operation",
       "prevalence": null,
       "energy_impact": "100% of plant energy while active"
@@ -69,7 +97,10 @@
       "id": "CLU-08",
       "name": "Schedule Dysfunction",
       "trigger": "AHU-FC-052",
-      "members": ["SYS-FC-057", "SYS-FC-053"],
+      "members": [
+        "SYS-FC-057",
+        "SYS-FC-053"
+      ],
       "playbook": "after-hours-operation",
       "prevalence": "~15%",
       "energy_impact": "3-16% of site energy"
@@ -78,7 +109,12 @@
       "id": "CLU-09",
       "name": "Sensor Integrity Failure",
       "trigger": "AHU-FC-062",
-      "members": ["AHU-FC-002", "AHU-FC-003", "SYS-FC-054", "SYS-FC-055"],
+      "members": [
+        "AHU-FC-002",
+        "AHU-FC-003",
+        "SYS-FC-054",
+        "SYS-FC-055"
+      ],
       "playbook": "sensor-drift",
       "prevalence": "~15%",
       "energy_impact": "prevents false alarms; sensor-dependent"

faults/ahu/AHU-FC-004/card.md

@@ -178,10 +178,12 @@ basis: N/A.
   quietly — a mis-set `count_scale` produces a plausible-looking number.
 - **Minimum tick interval, from the same block.** Each `MovingAverage` instance
   keeps a fixed 64-checkpoint ring and silently drops the oldest in-window
-  sample past that (one warning per instance), so the window must not hold more
-  than 64 ticks: `dt ≥ count_window/64` = 3600/64 = **56.25 s**. Combined with
-  the scale rule, a legal deployment has `dt ≥ 56.25 s` and
-  `count_scale = 3600/dt ≤ 64`. At the default 300 s tick the window holds 12
+  sample past that (one warning per instance). A window spanning n ticks
+  retains n + 1 checkpoints (one sits at or before the trailing edge), so the
+  window may span at most 63 ticks: `dt ≥ count_window/63` = 3600/63 =
+  **57.15 s** (an earlier revision of this card said 56.25 s, dividing by 64).
+  Combined with the scale rule, a legal deployment has `dt ≥ 57.15 s` and
+  `count_scale = 3600/dt ≤ 63`. At the default 300 s tick the window holds 12
   samples, well inside the ring.
 - **Startup artifact (a): a spurious first-tick pulse, which costs nothing
   here.** `Integers.Change` compares against `pre_u_start` on the first tick, so