BMS Data Analytics + Fault Detection for Indian Commercial — ASHRAE GL 36 + ISO 50001

A 25,000 m² office leaks 15-25 % above design EUI within 24 months of commissioning. ASHRAE Guideline 36 + ISO 50001 + ECBC 2024 push Continuous FDD as the cure. A 10-rule library — simultaneous heat-cool, economiser stuck, static-SP reset, CHW ΔT degradation, schedule overrides — saves 18-25 % annually. Capex ₹18 lakh; year-1 saving ₹32 lakh; simple payback 7 months. Three deployment errors that kill ROI: BACnet naming chaos blocking rule scale, 15-min trend default missing valve hunting, no ML fault prioritisation drowning FM teams in alerts.

Why FDD matters now in Indian commercial

Building energy waste is rarely a design problem. After commissioning, schedule overrides, dirty sensors, stuck dampers, and chiller fouling accumulate silently until utility bills creep 15-25 % above design. ASHRAE Guideline 36 + ISO 50001 + ECBC 2024 commissioning section all push toward continuous Fault Detection + Diagnostics. The Indian market for cloud BMS analytics has matured — Vendor-agnostic platforms (Bueno, ICONICS, Switch Automation, ABB Ability, JCI OpenBlue Connect) ingest BACnet/Modbus + apply rule engines + ML anomaly detection.

FDD rule library — 25,000 m² office

Three deployment mistakes Indian projects keep making

1. BACnet object naming chaos — every system integrator names points differently (AHU01-SAT vs ahu_01_supply_air_temp vs SAT_AHU_01). FDD platforms cannot apply rules at scale without normalised naming. Specify Project Haystack or Brick Schema tagging in tender BoQ — not as an afterthought.
2. 5-min trend interval too slow — FDD needs 1-min trends for valve hunting + simultaneous heat-cool detection. Most Indian Tridium / Niagara setups default to 15-min. Specify 60-second trend intervals in commissioning protocol.
3. No fault prioritisation — out-of-the-box FDD generates 200-400 alerts/day on a typical 25,000 m² office. Without ML-based prioritisation + comfort-impact weighting, FM teams burn out + ignore the system. Buy a platform with cost-and-comfort impact scoring built in.