Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Playbook: Proof-of-Operation Failure
Applies toAHU-0039 (supply fan), PMP-0003 (pump), HW-0009 (boiler), ERV-0004 (active recovery device), FCU-0006 (fan), FPB-0001 (terminal fan), RTU-0010 (supply fan), HP-0007 (compressor)
Fix complexityRemote (40%) · On-site (60%)
Typical time5–15 min remote; 30 min–2 h on-site
Typical cost$0 remote / $100–$800 on-site (belt, contactor, overload reset; motor or drive repairs run higher)
Energy impactDirection-dependent: fail-to-start costs delivery, not energy; unexpected-run is pure waste — a 10 kW pump left in HAND over a weekend is ~640 kWh nobody asked for, and the fan-affinity cube makes oversized fans worse

Library-authored playbook for the proof-of-operation family: command versus status, both directions (yFailToStart — commanded on, not proven; yUnexpectedRun — proven on, not commanded). Grounded in ASHRAE Guideline 36-2021’s proof definition (§5.1.6) and its alarm instances (§5.16.13.2 fans, §5.21.10.5 / §5.20.17.6 pumps and tower fans, §5.21.3 boiler prove), paraphrased.

Step 1 — Verify the fault

  1. Confirm mode, schedule, maintenance/test state, and point scope before touching equipment. Compare the final device command with independent proof; inspect interlocks and OEM alarms before assuming a mechanical failure.
  2. Identify the status device — the diagnosis depends on it. A current switch proves the motor draws amps (a broken belt can still read ON); a DP or flow switch proves air/water actually moves; a VFD status word proves only what the drive believes; an aux contact proves the starter closed and nothing downstream.
  3. For fail-to-start: confirm the command is actually reaching the equipment — read the BAS output at the controller, then the terminal. Check the HOA/HOA-equivalent switch position, the overload/reset flag, the drive’s local/remote mode and fault code, and the disconnect.
  4. For unexpected-run: look for HAND at the starter or drive, a local override, or a welded contactor. A qualified person must apply lockout/tagout and verify absence of hazardous energy before opening panels or approaching belts, fans, capacitors, contactors, or compressors.
  5. Boilers (HW-0009): a failed prove usually means the burner-management system locked out — that is the BMS doing its safety job. Read the lockout code at the boiler. Never bypass or repeatedly reset a safety lockout to clear an FDD alarm; repeated lockout-retry cycles also surface as short-cycling (HW-0001).
  6. Active recovery (ERV-0004): identify the technology before interpreting proof. Wheel speed/rotation can prove the media; motor current alone cannot prove a belt. Runaround-loop pump proof has the same current/flow distinction as PMP-0003. Passive plate cores have no proof rule.
  7. FCU/FPB/RTU fans: inspect subtype sequence, condensate, freeze, smoke, purge, post-heat, and local ownership before mechanical work. Heat pumps: inspect defrost, pump-down, pressure/oil/electrical lockouts, inverter state, and matched compressor scope. Never bypass a safety to make command and proof agree.

Step 2 — Remote fix

  1. Release BAS/software overrides on the command point; confirm the output actually changes state at the controller.
  2. Reset a tripped overload or OEM lockout only when a qualified person, the manufacturer procedure, and site policy explicitly allow one reset. Record the code first; a repeated trip requires service, not another reset.
  3. If the status point is derived (VFD word, current threshold), sanity check its configuration — a current-switch threshold set above the motor’s actual draw reports a running fan as off forever.

Step 3 — On-site service

Only qualified personnel following site LOTO and OEM procedures may open electrical/refrigerant compartments or service rotating equipment. Refrigerant work and forced compressor cycles require the appropriate licensed technician.

  1. HOA in HAND: return to AUTO and find out why someone put it there — HAND is usually a workaround for a control problem this library has a rule for.
  2. Broken or slipping belt, failed coupling: replace; check sheave alignment and tension. A current switch that kept reporting ON while the belt was broken should be replaced with a DP switch on rebuild.
  3. Welded contactor or failed starter: replace the contactor; check coil voltage and cycling rate (chronic short-cycling welds contacts — see the equipment’s short-cycling rule).
  4. Drive faults: record the fault history before clearing; recurring drive faults are a power-quality or motor problem, not a reset ritual.
  5. Boiler lockouts: service per the burner-management manual — flame sensor, pilot assembly, fuel train. FDD’s job here ends at reporting availability; the fix belongs to a qualified burner technician.

Step 4 — Confirm resolution

  1. When the OEM procedure and operating conditions permit, coordinate a safe stop→start→stop cycle from the final command and watch both directions prove within their windows. Never force past smoke, freeze, condensate, high/low-pressure, oil, electrical, or burner safeties.
  2. Both sub-condition flags should clear immediately on agreement; the fault should not reassert across at least one scheduled occupied/unoccupied transition.
  3. If the same equipment reappears here monthly, the root cause is upstream: sizing, power quality, or an operator working around a control defect with the HOA switch.