Data Centre MEP Design & TIA-942 Engineering Services India | KVRM
TIA-942 · ASHRAE TC 9.9 · NFPA 2001 · PUE Optimisation · Immersion Cooling — Delhi NCR · Mumbai · Hyderabad · Bangalore · Pune · Chennai

Data Centre MEP Design &
TIA-942 Engineering Services India

Specialist MEP engineering for hyperscale data centres, colocation facilities, and enterprise data rooms — from Tier II to Tier IV. Critical power design, precision cooling (including immersion cooling for AI GPU racks), NFPA 2001 fire suppression, PUE optimisation, and REVIT BIM coordination. Serving data centre projects across Delhi NCR, Mumbai, Hyderabad, Bangalore, Pune, and Chennai.

Why KVRM

Data Centre Engineering Expertise

Every cooling and power system modelled and validated before construction begins — not assumed correct. TIA-942 Tier compliance from design concept, not retrofitted at the end.

TIA-942 Rated-3 / Rated-4 DesignEvery power and cooling system modelled and optimised for the required Tier level — not assumed compliant.
CFD Thermal AnalysisIdentifying hotspots before build — hot aisle / cold aisle containment validated through computational fluid dynamics.
NFPA 2001 Clean Agent SuppressionFM-200, Novec 1230, INERGEN, CO₂ — every system fully hydraulically calculated to NFPA standards.
REVIT MEP BIM CoordinationFull multi-discipline BIM — clash detection, MTO generation, IFC delivery to reduce construction RFIs.
PUE OptimisationISO/IEC 30134 reporting, economiser design, partial load performance analysis and DCIM integration strategy.
Data Centre PMCIndependent owner’s representative from pre-construction through SMSTS, CMSTS, IMTS commissioning and acceptance testing.
Our Services

What We Deliver

Six core engineering disciplines for data centre projects — each calculation-backed, code-compliant, and fully documented.

1
Critical Power System Design

Reliable, resilient power supply is the defining engineering challenge in data centre design. Every configuration — N, N+1, 2N — modelled through ETAP power system studies before construction.

  • UPS design — N, N+1, 2N redundancy configurations
  • HT/LT power distribution — substation sizing, ring main unit, transformer specification
  • Generator sizing and fuel storage — N+1 or 2N diesel/gas backup
  • Automatic transfer switching (ATS) and static transfer switch (STS)
  • PDU design and branch circuit layout for IT halls
  • Busbar trunking system (BTS) design for floor-level distribution
  • Load schedule and maximum demand calculation (ETAP)
  • Power monitoring and branch circuit monitoring (DCIM integration)
Deliverables: SLDs, load schedules, ETAP calculation reports, equipment specifications, UPS and generator data sheets, power quality analysis.
2
❄️
Precision Cooling & HVAC Design

Precision cooling keeps servers within ASHRAE TC 9.9 thermal parameters. Hydraulic analysis using AFT Fathom or PIPENET for chilled water loop sizing and pump selection — every system modelled and validated before construction begins.

  • Chilled water system — primary/secondary/tertiary loops, variable primary flow
  • CRAC/CRAH unit selection and placement for HAC/CAC containment
  • In-row cooling and overhead cooling for high-density racks
  • Cooling tower and chiller plant — free cooling integration
  • CFD thermal analysis — hotspot identification before build
  • HAC (Hot Aisle Containment) and CAC system design
  • Airside economiser design and free cooling hours calculation
  • Chilled water hydraulic analysis — pipe sizing, pump selection (AFT Fathom / PIPENET)
Deliverables: Cooling system design basis, hydraulic calculation report (AFT Fathom / PIPENET), chilled water pipe sizing, pump data sheets, PUE model, CFD analysis report.
3
🔥
Fire Suppression System Design

Data centre fire suppression must protect equipment without causing water damage — every system hydraulically calculated to NFPA standards. NFPA 2001 clean agent design for IT halls, NFPA 13 pre-action for adjacent spaces.

  • FM-200 (HFC-227ea) clean agent system design (NFPA 2001)
  • Novec 1230 (FK-5-1-12) — low GWP clean agent alternative
  • Inert gas — INERGEN, Argon, IG-541 suppression design
  • CO₂ system for UPS rooms and high-value equipment (NFPA 12)
  • VESDA aspirating smoke detection layout and design
  • Water mist system — alternative to gaseous suppression (NFPA 750)
  • Pre-action dry pipe sprinkler for IT halls and raised floors
  • Fire pump sizing and room design (NFPA 20)
Deliverables: Fire protection design basis, hydraulic calculation reports, suppression agent quantity calculations, piping layout drawings, equipment schedules, fire authority submission package.
4
📐
MEP BIM Design & Coordination

Fully coordinated REVIT BIM models eliminating design clashes before construction begins. Multi-discipline coordination across power, cooling, fire protection, cable management, and containment — in a single federated model.

  • REVIT MEP model — power, cooling, fire, containment in single coordinated model
  • Multi-discipline clash detection and interference check reports
  • Cable management design — overhead trays, raised floor pathways, conduit routing
  • Containment system coordination — HAC/CAC within REVIT model
  • Coordinated drawings — ceiling plans, riser diagrams, section views
  • Material take-off (MTO) for procurement and tendering
  • IFC export for contractor use and quantity surveying
  • As-built BIM model update post-construction
Deliverables: REVIT MEP model (coordinated), clash detection report, issue register, coordinated drawings, MTO schedule, IFC export package.
5
📊
Energy & PUE Optimisation

PUE reduction directly reduces operating cost and improves sustainability metrics — every percentage point improvement translates to significant annual energy savings at scale. A 5MW facility moving from PUE 1.5 to PUE 1.35 saves approximately ₹1.5 crore annually at Indian electricity rates.

  • PUE baseline calculation and target PUE modelling
  • Cooling efficiency — partial load performance, economiser hours analysis
  • Power system efficiency — UPS efficiency curves, transformer losses
  • Airside vs waterside economiser comparison and selection
  • Free cooling hours calculation for Indian climate zones
  • DCIM integration strategy for real-time monitoring
  • ISO/IEC 30134 energy efficiency metric reporting (PUE, WUE, CUE)
  • Annual energy cost model — full-year simulation with seasonal variation
Deliverables: PUE analysis report, annual energy model, economiser feasibility study, ISO/IEC 30134 reporting framework, DCIM integration specification.
6
📋
Data Centre PMC — Owner’s Representative

Independent owner’s representative services for data centre construction — protecting the client’s interests from design coordination through commissioning and acceptance testing, with no conflict of interest.

  • Pre-construction — design coordination, contractor procurement, value engineering
  • Construction stage — site supervision, QA/QC, programme management
  • SMSTS (Site Mechanical Systems Test & Survey)
  • CMSTS (Component Mechanical Systems Test & Survey)
  • IMTS (Integrated Mechanical Test & Survey) execution and management
  • IT load bank testing coordination and acceptance
  • UPS walkthrough, generator test witnessing
  • Variation order management and contractor claim assessment
Deliverables: Construction stage reports, commissioning test protocols, SMSTS/CMSTS/IMTS documentation, acceptance testing records, handover pack.
Next-Generation Cooling

Immersion Cooling for AI GPU Data Centres

Traditional air cooling cannot handle the rack densities required for AI and GPU workloads. KVRM designs immersion cooling systems for next-generation hyperscale and edge AI data centres.

New Capability — AI & GPU Infrastructure
Single-Phase & Two-Phase Immersion Cooling Design for AI GPU Racks at 50–100 kW Rack Density

AI GPU servers (NVIDIA H100, H200, GB200 NVL72) generate 5–10× the heat of standard compute — exceeding the thermal capacity of conventional CRAC-based air cooling. KVRM engineers single-phase and two-phase immersion cooling systems sized for 50–100 kW per rack, integrated with facility-level heat rejection systems (cooling towers, dry coolers, heat exchangers).

Suitable for AI training clusters, HPC installations, hyperscale GPU pods, and edge AI inference facilities. We design the complete thermal infrastructure — from tank selection and dielectric fluid specification through to facility-level heat rejection and secondary coolant loop hydraulic analysis using AFT Fathom.

100kW
Rack density supported
<1.1
PUE achievable with immersion cooling
95%
Heat captured vs 40% for air cooling
Types designed: single-phase & two-phase
Selected Project

Data Centre Engineering Delivered

An example of KVRM data centre engineering scope — the same rigour applied to every facility from enterprise data rooms to hyperscale campuses.

Case Study — Hyperscale Tier-3 Data Centre
5MW Tier-3 Data Centre — Full MEP & TIA-942 Engineering
5MW
IT Load
TIA-942 Rated-3
Tier Compliance
1.35
PUE Achieved (target 1.4)
NFPA 2001
Clean Agent Suppression

Scope: Critical power MEP design (2N UPS, N+1 generator), chilled water precision cooling (variable primary flow, airside economiser), NFPA 2001 FM-200 suppression in all IT halls, REVIT BIM multi-discipline coordination (clash detection, MTO, IFC delivery), and PUE optimisation study. Final PUE of 1.35 exceeded the 1.4 design target — achieved through airside economiser integration and variable speed pump controls. Full SMSTS, CMSTS, and IMTS commissioning documentation provided.

Serving data centre projects across India:

📍 Delhi NCR 📍 Mumbai 📍 Hyderabad 📍 Bangalore 📍 Pune 📍 Chennai 📍 International
Standards & Codes

Applicable Standards — Data Centres

Every data centre design delivered to the applicable international standard. TIA-942 Tier compliance is established at design concept — not verified at the end.

Infrastructure Standards
  • TIA-942-B — Data Centre Infrastructure (Tier I–IV)
  • ASHRAE TC 9.9 — Thermal Guidelines for Data Processing
  • BICSI 002 — Data Centre Design & Implementation Best Practices
  • ISO/IEC 30134 — Data Centre Energy Efficiency Metrics (PUE, WUE, CUE)
  • EN 50600 — European Data Centre Standards
  • Uptime Institute Tier Standard (Topology, Operational Sustainability)
Engineering Standards
  • NFPA 2001 — Clean Agent Fire Suppression Systems
  • NFPA 12 — CO₂ Extinguishing Systems
  • NFPA 750 — Water Mist Fire Protection Systems
  • NFPA 20 — Fire Pump Installations
  • NFPA 70 (NEC) / IS Codes — Electrical Systems
  • IEC 62040-3 — UPS Performance (VFI, VI, VFD classification)
  • NBC 2016 — National Building Code India
Frequently Asked Questions

Data Centre Engineering — Common Questions

Questions we regularly receive from data centre developers, operators, and EPC contractors.

What does TIA-942 Tier III actually guarantee — and how is it different from Tier IV?+
TIA-942 Tier III (Rated-3) means the data centre is “concurrently maintainable” — any single component of the power or cooling infrastructure can be taken offline for maintenance without interrupting IT operations. The facility has N+1 redundancy in both power and cooling paths, with a single path active. Tier IV (Rated-4) goes further — “fault tolerant,” meaning any single failure in any component does not cause an outage. Tier IV requires 2N redundancy throughout, dual power paths, and the capability to sustain a second failure while the first is being repaired. The key practical difference: Tier III has an expected annual downtime of 1.6 hours, Tier IV 0.4 hours. The additional capital cost of Tier IV vs Tier III is typically 25–40% — so most enterprise operators choose Tier III unless they have regulatory or contractual requirements for Tier IV.
How do you reduce data centre PUE — what are the most effective measures for Indian climate conditions?+
In Indian climatic conditions, the most impactful PUE reduction measures are: (1) Airside or waterside economisers — using ambient air or cooling tower water to supplement or replace mechanical chilling during cooler months. In Delhi NCR, a waterside economiser can reduce chiller operating hours by 20–30% annually. (2) Variable speed drives (VSDs) on all chilled water pumps, cooling tower fans, and AHU fans — typically the fastest-payback PUE measure. (3) Hot aisle/cold aisle containment — reduces CRAC unit air mixing losses and can improve PUE by 0.05–0.15 alone. (4) Raising supply air temperature to ASHRAE TC 9.9 A1 class limits (27°C supply) — reduces chiller energy consumption. (5) UPS right-sizing and high-efficiency UPS selection — modern double-conversion UPS can reach 96%+ efficiency at load. A well-designed 5MW data centre in India targeting PUE 1.35 (vs a typical 1.6–1.8 in older facilities) can save ₹1.5–2 crore annually in electricity costs.
When is immersion cooling required — and does it work with standard servers?+
Immersion cooling becomes practically necessary (rather than just advantageous) when rack densities exceed approximately 30–40 kW per rack — a threshold regularly exceeded by modern AI GPU servers. An NVIDIA H100 SXM GPU server draws approximately 10–12 kW alone, and a standard 8-GPU rack reaches 80–100 kW. Standard CRAC/CRAH air cooling is designed for 5–15 kW racks, making it fundamentally inadequate for dense AI workloads. Single-phase immersion cooling (e.g. with 3M Novec or mineral oil-based fluids) is compatible with most standard servers if they are rated for immersion by the manufacturer — major OEMs including Intel, AMD, and NVIDIA now offer immersion-ready hardware. Two-phase cooling (using fluorinated fluids that boil at low temperature) is more efficient but requires purpose-built hardware. For new AI data centre projects, we recommend designing for immersion cooling from the outset — retrofitting an air-cooled facility for immersion is significantly more expensive than designing for it initially.
What is the difference between FM-200 and Novec 1230 — which should I specify for my data centre?+
Both FM-200 (HFC-227ea) and Novec 1230 (FK-5-1-12) are NFPA 2001 clean agent suppression systems suitable for data centres — both extinguish fire without water and leave no residue that would damage IT equipment. The key differences are: (1) Environmental impact — FM-200 has a Global Warming Potential (GWP) of 3,220; Novec 1230 has GWP of 1. Many operators and sustainability-focused organisations prefer Novec 1230 for ESG reasons. (2) Cost — FM-200 is typically 20–30% cheaper in initial system cost; Novec 1230 agent is more expensive. (3) Atmospheric lifetime — FM-200 has a 36-year atmospheric lifetime; Novec 1230 degrades in 5 days. (4) Both systems require the same concentration (approximately 6–7% by volume) and are equally effective. Our recommendation for new data centres: Novec 1230 if the project has sustainability/ESG reporting requirements or green building certification aspirations; FM-200 if capital cost is the primary constraint. Inert gas systems (INERGEN, Argon) are a third option with zero GWP but require larger cylinder storage and higher pipe sizing.
Related Services

Explore Our Engineering Services

🔧
Engineering Services
MEP design, fire protection, piping stress (Caesar II), hydraulic analysis (AFT Fathom, PIPENET), process simulation, static equipment, and 3D plant modelling.
Engineering Services
📋
PMC — Owner’s Engineer
End-to-end project management as independent owner’s representative — planning, procurement, QA/QC, commissioning (SMSTS/CMSTS/IMTS) and variation order management.
PMC Services
🔋
Battery Manufacturing / Gigafactory
Dry room HVAC (dew point to −40°C), NFPA 855 lithium-ion fire protection, process utility MEP, and Gigafactory infrastructure engineering.
Gigafactory Engineering
Ready to Discuss Your Data Centre Project?

Our data centre engineering team in New Delhi, Navi Mumbai, and Faridabad responds within 24 hours — hyperscale, colocation, or enterprise data room.

Email
info@kvrm.in
Phone / WhatsApp
+91 8447784536
Offices
New Delhi · Navi Mumbai · Faridabad
Scroll to Top