Operations & Maintenance for data centers in India

Where reputations are made

The reliability a Tier III design promises is delivered only in operations: every UPS, chiller, generator and switchboard maintained to its OEM schedule, every change executed to a method statement, every incident rehearsed before it happens. Operating cost over 15 years exceeds construction cost, and most of it is energy and maintenance, which makes the operations contract the longest-lived commercial decision an owner makes.

The Indian market offers three routes: integrated facility management from large FM companies, equipment AMCs from OEMs and their partners, and specialist critical-facility operators that take full responsibility for uptime under SLA. The choice is about accountability. Fragmented contracts produce finger-pointing during incidents; a single accountable party with the equipment AMCs beneath it produces one SLA and one phone call.

The asset register and the maintenance plan

Everything starts from a complete asset register: every maintainable item, its OEM maintenance schedule, criticality, spares list and warranty status. The maintenance calendar is built from it, and so is the AMC tender, because a contractor cannot price what it cannot see. Build the register during commissioning from the as-built documentation and keep it in the DCIM or CMMS, not in a spreadsheet that leaves with the facility manager.

Procedures and drills

A critical facility runs on its MOP, SOP and EOP library: method statements for every piece of maintenance on live systems, standard procedures for routine operations, and emergency procedures for every failure mode the design is supposed to survive. The library is validated during Level 5 commissioning and then drilled, quarterly at least, until the response is automatic. Uptime's Management and Operations Stamp of Approval is the recognised external check on this discipline and is increasingly asked for by customers.

Energy and predictive maintenance

Energy is the other half of operating cost. PUE improvement programmes (containment, raised set points, chiller plant optimisation, AI-driven controls) pay back in two to three years at Indian tariffs, and PUE should be reported monthly so drift is caught early. Predictive maintenance on rotating equipment, using vibration, thermal and electrical signatures, has moved from pilot to standard practice and reduces both failures and unnecessary interventions on critical plant.

Contract structure and benchmarks

Integrated FM for a 10 MW facility runs indicatively INR 3 to 5 crore a year excluding equipment AMCs and energy; comprehensive AMCs run 1.5 to 3 percent of equipment value a year. The contract should carry a single SLA with availability, response and repair-time commitments and penalties, a monthly KPI report with incident and near-miss reporting, critical spares held on site or within four hours, and a transition plan with a shadow period of six to ten weeks on either side. Renegotiate annually against the KPI record rather than rolling over.

How we procure operations

We define the operating model and the SLA, build or audit the asset register, tender integrated FM, AMCs or a critical-facility operator from our network against it, and run the transition. For operating sites we start with a condition and documentation audit, because the incoming contractor should accept the facility in a known state.