Uptime Institute Tier IV - Constructed Facility

Every IT device and infrastructure component must be connected to two independent power distribution systems sourced from separate utility feeds or on-site generation capacity, with automatic transfer switches capable of seamless failover without load shedding.

Each PDU circuit must be independently breaker-protected and routed through physically separated cable trays and conduits to prevent common-mode failures from single cable conduit damage.

Battery backup systems must provide 15+ minutes of runtime at 100% facility load, sufficient for generator synchronization and load transfer, with redundant UPS units sized for N+1 configurations and automatic switchover logic preventing bus transfers that would interrupt critical loads.

Electrical distribution must eliminate single points of failure at every tier through redundant automatic transfer switches (ATS) at multiple hierarchical levels, with no centralized distribution point that could disable the facility if compromised.

Generator capacity must equal or exceed peak facility load requirements with N+1 redundancy, meaning any single generator failure maintains operational capacity; fuel systems must support minimum 72-hour continuous operation at design load.

Switchgear, transformer banks, and distribution panels must be physically separated and independently operational, with no shared busses, common neutrals, or grounding points that could cascade failures across redundant systems.

Facility cooling infrastructure must consist of at least two completely independent systems capable of maintaining design temperature specifications (typically 18-27°C) while simultaneously handling full facility thermal load, even during maintenance on the primary cooling system.

Each cooling loop must possess dedicated chillers, condenser pumps, distribution pumps, and piping with no shared components; hot aisle/cold aisle containment must be physically separated by independent systems to prevent thermal crosstalk.

Cooling delivery must be sub-metered at each zone level enabling immediate identification of thermal failures, with redundant sensors feeding to independent Building Management Systems (BMS) that can operate autonomously if either BMS experiences failure.

All network connectivity must utilize multiple independent carrier circuits from geographically diverse providers, with network device redundancy ensuring automatic failover without manual intervention or packet loss exceeding single Round-Trip Time (RTT).

Each network path must enter the facility through separate conduit systems, cross separate distribution frames, and terminate on redundant switching infrastructure with no shared router, firewall, or load balancer that could create a single point of failure.

Border Gateway Protocol (BGP) or equivalent dynamic routing must enable transparent failover between carriers within subsecond timeframes, with traffic engineering ensuring equal or near-equal load distribution across redundant paths during normal operations.

The defining characteristic of Tier IV requires that any infrastructure component—power distribution, cooling, network, or security systems—may be taken offline for maintenance, upgrades, or repair without operational impact on IT systems or facility uptime.

This demands N+1 or greater redundancy across every critical infrastructure domain, with automatic failover mechanisms requiring no human intervention and no load shedding during transitions.

Maintenance procedures must be documented for every infrastructure element, specifying safe shutdown sequences, failover validation checkpoints, and restoration procedures that prevent damage to redundant systems during maintenance activities.

Critical infrastructure components must be physically separated to prevent common-mode failures from single physical events such as flooding, fire, or cable conduit damage; power and cooling distribution systems must follow completely independent physical pathways through the facility.

Tier IV facilities should incorporate geographic redundancy where feasible, with critical infrastructure distributed across multiple buildings or campuses to protect against localized environmental disasters.

Cable routing specifications must prohibit bundling of redundant systems, require minimum physical separation distances (typically 6+ feet), and mandate installation in separate conduit systems with documented separation verification through post-construction inspection.

All automatic transfer mechanisms must operate without manual intervention, administrative approval, or any timing delay that would result in service interruption; Automatic Transfer Switches must transfer loads in less than 4 milliseconds to prevent IT equipment brownout conditions.

System switchover logic must incorporate health monitoring that validates target infrastructure health before completing transfer, preventing cascade failures where equipment transfers to a failed redundant system.

Comprehensive testing protocols must validate switchover mechanisms quarterly, documenting transfer times, load distribution, and any anomalies; test procedures must include simulated failures that intentionally isolate redundant systems to verify failover without manual bypass.

Tier IV certification requires comprehensive as-built documentation validated against design specifications through post-construction audit, including electrical schematics, network diagrams, cooling system specifications, and maintenance procedure manuals.

Annual third-party verification assessments must validate that all redundant systems remain functional, automatic failover mechanisms operate within specification, and facility operations teams demonstrate competency in emergency procedures and system restoration.

Change management protocols must capture all infrastructure modifications, with reassessment required when changes could impact redundancy, failover mechanisms, or maintenance procedures; facilities must maintain detailed logs of all maintenance activities, system failures, and manual interventions for historical trending and predictive maintenance analysis.