Design Basis for Data Center

Electrical
1. Document Purpose and Scope

This Design Basis Report (DBR) section defines the Electrical Design Philosophy for a 100 MW
(IT Load) Data Center including the incoming utility interface, HV/MV/LV distribution, UPS
systems, standby generation, grounding, lightning protection, protection coordination
philosophy, metering, power quality, and monitoring/control integration. It establishes the
electrical standards, redundancy targets, operating modes, sizing philosophy, and functional
requirements to be used for detailed design, procurement, construction, commissioning,
and operations.

This DBR is intended to be consistent with hyperscale data center practices, EPC tender
documentation, and applicable international standards.

2. Basis of Design: Key Definitions

• IT Load (Critical Load): Power delivered to IT equipment (servers, storage, network)

at rack/PDU output. Here it is 100 MW.

• Non-IT Load (Support Load): Cooling plant, pumps, CRAH/CRAC fans, lighting,
security, fire systems, office loads, etc.

• Total Facility Load: IT Load + Non-IT Load + losses.

• PUE (Power Usage Effectiveness): Total Facility Power / IT Power. Target range
depends on climate and cooling technology.

• Critical Power Path: Utility/HV intake → MV distribution → LV → UPS → PDU/RPP →

Rack.

• Redundancy Terms:

o N: exactly required capacity

o N+1: one additional module/unit

o 2N: two fully independent power trains each capable of carrying full load

o 2N+1: two independent trains plus extra backup

3. Design Objectives

The electrical system shall be designed to meet the following objectives:

1. No Single Point of Failure (SPOF) in the critical power path (for the selected
Tier/redundancy).
 2. Concurrent Maintainability: planned maintenance without IT outage (minimum).

3. High Availability: fast and automatic fault isolation and load transfer.

4. Safety: arc-flash mitigation, safe isolation, robust earthing, compliance with electrical
codes.

5. Scalability: modular growth in blocks without rework.

6. Power Quality: stable voltage/frequency, low harmonics, surge protection, ride-

through events.

7. Operational Clarity: clear operating modes, interlocks, SOP-friendly switching.

8. Energy Efficiency: optimize losses and support PUE targets.

9. Monitoring & Control: end-to-end metering and event recording integrated to

DCIM/BMS/SCADA.

4. Capacity and Expansion Philosophy

4.1 Capacity Block Approach

The 100 MW IT load shall be implemented in modular blocks (typical hyperscale practice),
for example:

• 4 × 25 MW IT blocks, or

• 5 × 20 MW IT blocks, or

• 10 × 10 MW IT blocks

Each block shall be electrically repeatable, enabling standardized procurement,

commissioning, spares strategy, and fast expansion.

4.2 Oversizing and Margins

Electrical equipment shall be sized with defined margins for:

• Future IT growth within the block

• UPS and switchgear thermal margin

• Transformer temperature rise margin

• Short circuit duty margin

• Degradation and harmonics margin

Typical philosophy:

• Transformers: sized to support full block load + losses with thermal headroom.
 • UPS: sized as modular N+1 at the block level.

• Generators: sized with step loading and motor-start allowances.

5. Reliability Target and Redundancy Philosophy

5.1 Tier Target (Project-Specific)

The electrical philosophy supports either of the following, depending on Employer’s

Requirements:

• Tier III style: concurrently maintainable, typically N+1 or dual utility with robust
backup.

• Tier IV style: fault tolerant, typically 2N end-to-end with physically separated paths.

For a 100 MW hyperscale facility, the recommended baseline philosophy is:

• 2N MV/LV power trains to IT load (Train A and Train B)

• UPS 2N at critical distribution (each train has its own UPS plant)

• Generator redundancy: each train supported by generator capacity with N+1 at train
level or 2N generator plants depending on fuel strategy and reliability requirement.

5.2 Dual Power Path to IT

Each IT rack shall receive power via:

• Dual-corded IT loads (preferred), fed from A and B sources.

• Where single-cord equipment exists, it shall be fed through rack ATS or equivalent
with defined transfer philosophy.

6. Utility Interface and High Voltage Intake Philosophy

6.1 Utility Supply Configuration

The facility shall be fed by utility in one of the following typical configurations:

• Option 1: Dual independent utility feeders at 220 kV (preferred for 100 MW scale)

• Option 2: Dual utility feeders at 132 kV

• Option 3: One utility feeder + one dedicated onsite/IPP feeder (site-specific)

Philosophy requirements:

• Utility supplies shall be diverse (separate substations / separate routes) where

possible.

• Utility protection, metering, and operational switching shall be coordinated with the
utility operator.
6.2 HV Substation

An on-site HV substation shall be provided (AIS or GIS based on land, pollution, and
reliability needs). The HV yard shall include:

• Incoming line bays (minimum 2)

• Busbar scheme (typically double bus or breaker-and-a-half for high availability)

• Transformer bays feeding MV

• Protection relays, disturbance recorders, revenue metering as required

The HV system shall be designed for:

• High fault levels and appropriate fault clearing times

• Safe switching and interlocking

• Maintainability with minimal disruption

7. MV Distribution Philosophy (33 kV / 11 kV)

7.1 MV Voltage Selection

For 100 MW, MV is typically 33 kV for primary site distribution (preferred for lower current
and fewer feeders), with local step-down to 11 kV or LV near load centers.

Recommended philosophy:

• HV → 33 kV primary distribution

• 33 kV rings/sections feeding multiple block substations

• Step-down transformers near blocks to 11 kV or 415 V, depending on chosen

topology

7.2 MV Switchgear

MV switchgear shall be metal-clad, withdrawable (where applicable), with:

• Internal arc classification per applicable standard

• Protection relays with communication and event logging

• Busbar sectionalizing to allow maintenance and fault isolation

• Space for expansion (future feeders)

7.3 MV Topology

The MV topology shall be either:

• Two independent MV systems (Train A and Train B) with physical separation, or

 • Ring main arrangement with sectionalizers and dual feeds (if approved by reliability
target)

Each IT block shall be fed from both trains, with clear demarcation.

8. Transformer Philosophy

8.1 Transformer Stages

Typical stages:

• HV/MV transformers (220/33 kV or 132/33 kV)

• MV/LV transformers near UPS/LV switchrooms (33/0.415 kV) or (33/11 kV and

11/0.415 kV)

8.2 Transformer Type Selection

• Oil-filled outdoor transformers are typical for large capacity and efficiency.

• Dry-type transformers may be used indoors for specific areas (subject to thermal and
fire strategy).

8.3 Redundancy

Transformers shall follow the selected redundancy model:

• For 2N trains: each train has transformer capacity for full train load.

• For N+1: spare transformer capacity supports failure/maintenance.

8.4 Performance Requirements

Transformers shall be specified with:

• Low losses (to support PUE)

• Appropriate impedance to manage fault levels

• Temperature rise suitable for local ambient conditions

• On-load tap changer (OLTC) philosophy where voltage regulation is required

• Online monitoring (oil temperature, winding temperature, DGA optional for critical
units)

9. LV Distribution Philosophy (415 V)

9.1 LV Main Distribution

Each power train shall include LV main switchboards supplied from MV/LV transformers,
feeding:
 • UPS inputs

• Mechanical loads (chillers, pumps, CRAH, cooling towers)

• General services

LV switchboards shall be designed with:

• Busbar rated with thermal margin

• Fault withstand ratings aligned with calculated short-circuit levels

• Selective coordination philosophy

• Segregation and safe maintenance access

9.2 LV Topology near IT

Preferred approach:

• UPS output → PDU/RPP → Busway → Rack PDUs

• Minimize distribution losses and reduce cable congestion using bus ducts in white
space.

10. UPS System Philosophy

10.1 UPS Role

UPS shall provide:

• Ride-through for utility disturbances and transfer events

• Clean and regulated power for IT

• Power quality improvement (voltage regulation, harmonic control depending on UPS

type)

10.2 UPS Topology

For hyperscale, common choices:

• Centralized modular UPS plants per block (preferred for 10–25 MW blocks)

• Distributed UPS for smaller deployments (less typical at 100 MW scale)

10.3 Redundancy

Typical philosophy:

• N+1 modular UPS per train per block, or

• 2N UPS if Tier IV style is mandated

10.4 UPS Technology Selection

Options include:

• Double-conversion (VFI) UPS (high power quality)

• Eco-mode / high-efficiency modes (subject to risk acceptance)

• Static bypass and maintenance bypass with full interlocking

10.5 Battery Energy Storage (UPS Batteries)

Battery system options:

• VRLA (legacy)

• Lithium-ion (common for modern hyperscale due to footprint, lifecycle, monitoring)

Battery autonomy philosophy:

• Commonly 5–10 minutes to bridge generator start and stabilization.

• Autonomy time to be confirmed based on generator start reliability, site conditions,

and operational philosophy.

Battery monitoring:

• Per string/cabinet monitoring

• Temperature management

• BMS alarms integrated into DCIM/BMS

11. Standby Generator Philosophy

11.1 Role and Start Sequence

Generators provide long-duration backup during utility outages.

Philosophy:

• Detect utility failure

• Start generators automatically

• Transfer critical loads to generator via ATS/STS/switching sequence

• Synchronize and load-share across multiple generator sets

• Maintain stable voltage/frequency under step-load conditions

11.2 Generator Plant Configuration

For 100 MW IT plus support loads, generator capacity is significant and typically arranged as:

• Multiple medium-speed or high-speed diesel generator sets in parallel

 • Dedicated generator sets per train (A and B)

• N+1 or N+2 at train level

11.3 Fuel Autonomy

Fuel storage autonomy shall be defined (common philosophy: 24 to 72 hours on-site, with
refueling contracts for extended outages). Fuel treatment, polishing, and quality assurance
philosophy shall be included to avoid failures due to contamination.

11.4 Compliance and Emissions

Generator design shall comply with local environmental regulations, noise limits, stack
design, and emission requirements.

12. Transfer, Switching, and Operating Modes Philosophy

12.1 Normal Mode

• Utility supplies both trains (or each train from its dedicated utility source)

• UPS supplies IT loads in normal online mode

• Generators in standby, warmed and ready

12.2 Utility Disturbance / Outage Mode

• UPS carries load immediately without interruption

• Generators start and reach stable conditions

• Load transfers to generator supply as per sequence

• Non-critical loads may be shed automatically (load shedding philosophy)

12.3 Maintenance Mode

• Equipment can be isolated and maintained without taking down IT load (for Tier III+
targets)

• Maintenance bypass systems and interlocks ensure safe switching without backfeed
risk

12.4 Emergency Mode

• Fire alarm integration may trigger selective shutdown of non-essential systems

• EPO philosophy: EPO shall be carefully zoned to avoid accidental total shutdown; IT
hall EPO shall comply with applicable codes and client requirements.

13. Protection Philosophy and Coordination

13.1 Protection Objectives

 • Protect personnel and equipment

• Isolate faults quickly

• Maintain continuity of supply for unaffected sections

• Ensure selectivity / discrimination

13.2 Coordination Philosophy

Protection coordination shall be performed from HV to LV including:

• Relay settings coordination

• Time-current grading

• Differential protection for transformers and busbars where applicable

• Generator protection (reverse power, over/under frequency, over/under voltage, loss

of excitation as applicable)

• UPS and LV selective coordination

13.3 Short Circuit Levels and Ratings

Short-circuit calculations shall define:

• HV fault levels (utility contribution)

• MV bus fault levels

• LV fault levels (transformer impedance and contribution)

Switchgear interrupting ratings shall exceed calculated fault levels with margin. Arc-flash
studies shall determine PPE levels and mitigation requirements.

14. Power Quality and Harmonics Philosophy

The electrical system shall be designed to minimize:

• Voltage dips and swells

• Harmonic distortion

• Transient overvoltages

• Flicker due to large motor starts

Philosophy measures:

• UPS with input rectifier harmonic mitigation

• Harmonic filters where required (active/passive)

• Surge protection devices (SPDs) at strategic locations

 • Proper transformer vector group selection to mitigate harmonics and circulating
currents

• Power factor correction philosophy (if required and compatible with UPS/generator
operation)

15. Earthing and Bonding Philosophy

The site shall have a robust earthing system providing:

• Safety against touch and step voltages

• Effective fault current return path

• Stable reference for sensitive electronics

• Lightning discharge path

Philosophy includes:

• Site-wide earth grid with target resistance based on soil conditions

• Equipotential bonding of all metallic structures, cable trays, and enclosures

• Dedicated bonding for IT areas, raised floors, and busways

• Separation philosophy between lightning protection down conductors and sensitive

systems bonding (while maintaining overall equipotential concept)

16. Lightning Protection Philosophy

Lightning protection shall be designed to:

• Protect buildings, HV yard, generator yards, and cooling plants

• Prevent direct strikes to critical equipment

• Control lightning surge propagation

System shall include:

• Air terminals / masts as per risk assessment

• Down conductors and bonding

• SPDs at incoming services and critical panels

• Coordination with earthing grid

17. Metering and Energy Management Philosophy

A comprehensive metering system shall support:

• Utility revenue metering (as required)

 • Sub-metering for energy accountability

• PUE measurement (Total facility vs IT load)

• Capacity management and forecasting

Metering points philosophy:

• HV incomers

• MV feeders

• Transformer secondaries

• UPS input/output

• PDU/RPP outputs

• Mechanical plant major loads

• Tenant/customer billing metering where applicable

18. Monitoring, Control, and Integration Philosophy (SCADA/BMS/DCIM)

The electrical infrastructure shall be integrated into:

• BMS for facility monitoring

• DCIM for capacity/asset/energy analytics

• SCADA/EMS for HV/MV operations if required

Monitoring shall include:

• Breaker status, alarms, events

• Power parameters (V, I, kW, kVA, PF, THD)

• UPS status and battery health

• Generator status, load share, fuel levels

• Transformer temps and alarms

• Protection relay events and disturbance records

Communications philosophy:

• Redundant networks for critical monitoring

• Time synchronization (NTP/PTP) for event correlation

• Cybersecurity segmentation between OT and IT networks

19. Safety Philosophy (Arc Flash, LOTO, Access)

The facility shall implement:

• Arc flash risk assessment and labeling

• Arc-resistant switchgear where applicable

• Remote racking / switching where needed

• Clear isolation points and interlocks

• LOTO procedures supported by design (lockable disconnects, key interlocks)

• Safe access routes and maintainability clearances

20. Commissioning and Testing Philosophy

Commissioning shall be performed in structured stages:

1. Factory acceptance tests (FAT) for major equipment

2. Site acceptance tests (SAT)

3. Integrated Systems Testing (IST) including:

o Utility outage simulation

o Generator start and load transfer tests

o UPS battery autonomy tests

o Protection trip tests and selective coordination validation

o Black start and recovery procedures (as applicable)

4. Performance verification:

o Load bank testing for generators

o Thermal imaging and power quality verification

o End-to-end power path validation

21. Typical Electrical Architecture Philosophy (Reference Topology)

A common reference architecture for a 100 MW data center is:

• 220/132 kV utility intake (dual feed)

• HV GIS/AIS with high-availability bus scheme

• HV/MV transformers feeding 33 kV systems

• Two independent 33 kV trains (A & B)

• Block substations with MV/LV transformers

 • UPS plants per block per train (modular N+1)

• UPS output to PDUs / busways

• Dual feeds to racks (A and B)

• Generator plant per train with load-shed logic

22. Assumptions and Design Inputs (To Be Confirmed During Detail Design)

The following shall be confirmed in project execution:

• Utility supply voltage, capacity, reliability, and fault level

• Site ambient temperature and altitude

• Cooling technology and resultant support load

• Target Tier level / redundancy model

• Fuel autonomy requirement and refueling logistics

• Rack density assumptions and distribution strategy

• Harmonic limits and utility interconnection requirements

• Space constraints and equipment placement constraints

• Local code requirements (fire, electrical, environmental)

Deliverable Outputs from this Philosophy (What EPC Will Produce Next)

Based on this DBR philosophy, the next stage design shall produce:

• Electrical SLDs from HV to rack

• Load lists and demand factors

• Equipment sizing schedules (transformers, switchgear, UPS, generators)

• Short circuit study, protection coordination, arc flash study

• Earthing and lightning protection design

• Cable schedules and routing philosophy

• Metering architecture and PUE measurement design

• Control/SCADA/BMS/DCIM integration architecture

• Commissioning and IST procedures