Skip to content

Infrastructure Capacity Decision — Template

Version: 1.1.0 | Licence: CC BY 4.0

Tools for deciding whether new digital infrastructure capacity, expansion or a cloud move is justified, and for making energy, water, resilience and lifecycle assumptions visible. Capacity should follow a governed service need.

Related: Digital Sustainability and Data Centre Energy Governance. Definitions are in the KPI reference, Section 2.

1. Frame the service need

Question Entry
Which service or user need requires capacity?
Current and forecast workload, and how reliable is the forecast?
Evidence that new capacity is actually necessary
Latency and availability requirements
Security, sovereignty and recovery constraints
Can optimisation, consolidation, data reduction, reuse or workload redesign meet the need?
What will be measured before and after the change?

Keep three things distinct: reserved capacity, forecast demand and actual utilisation.

2. Options comparison

Options to consider where realistic: optimise existing facilities; consolidate or retire underused capacity; reuse a brownfield site; colocation or cloud; build or expand; distribute workloads; defer; redesign the service.

Decision dimension Question Option A Option B Option C
Service value and demand Linked to a defined outcome and evidenced demand?
Energy and grid Load profile, connection readiness, energy source, price exposure, flexibility
Climate and site Ambient conditions, land use, local constraints, community impacts
Water Direct and indirect needs, local conditions, seasonal constraints
Carbon and lifecycle Operational emissions and equipment impacts over the lifecycle
Resilience and security Continuity, redundancy, recovery, security needs met?
Economics Total lifecycle cost: energy, water, cooling, equipment, maintenance, exit
Delivery maturity Planning, connection, equipment, skills and operating plans credible?
Reversibility Can it be scaled, migrated or retired if assumptions change?

Do not collapse non-negotiable service, safety, environmental or legal constraints into a single average score. Record evidence, assumptions, confidence and the accountable decision-maker. Where an infrastructure choice is hard to reverse, make portability, workload placement, phased growth and service continuity explicit, and register material resource and infrastructure dependencies in the Strategic dependency register.

3. Measurement boundary sheet

Field Entry
Measure
Boundary (what is in and out)
Period and baseline
Source system and owner
Method: measured or estimated; allocation method for shared infrastructure
Emissions factors and accounting method (location or market), if relevant
Denominator (for intensity measures)
Reporting frequency and verification approach

Report absolute consumption alongside intensity measures. Do not treat heat reuse or water reduction as established benefits until measured.

4. Decision gates

Gate Focus Decision, conditions, owner, review trigger
1. Need and alternatives Validate demand and service outcome; assess optimisation and reuse first
2. Feasibility Site, grid, energy, water, resilience, delivery maturity
3. Investment Lifecycle cost and resource scenarios; trade-offs and owners
4. Design assurance Facility, workload, security, environmental and operational requirements
5. Operational readiness Monitoring, energy procurement, water plans, incident response, reporting
6. Benefits review Actual service and environmental outcomes against baseline; adapt or retire

A decision record should include outcome, options, assumptions, evidence, environmental boundary, capacity forecast, decision owner, conditions, review trigger and exit path. Use the Decision record example.

5. Procurement requirements

Requirements should be specific, measurable, proportionate and verifiable. Prefer evidence and reporting duties that support operational decisions over unsupported environmental claims.

  • [ ] Service outcome, demand evidence and capacity forecast
  • [ ] Site, grid, energy, water, carbon, equipment and resilience assumptions
  • [ ] Total lifecycle cost and sensitivity to energy-price or demand changes
  • [ ] Supplier evidence and measurement rights for relevant resource and service metrics
  • [ ] Data boundaries, reporting frequency, source systems and verification
  • [ ] Architecture and exit considerations, including portability and continuity
  • [ ] Benefit owner, baseline, target, review date and corrective-action triggers

6. Fictional decision record

Illustrative only. "the fictional services organisation" and its decision are invented; no saving is assumed.

Field Entry
Decision Approve a time-bounded optimisation and feasibility phase; defer the expansion decision
Rationale Demand evidence and capacity forecast are not yet credible enough to justify new capacity
Conditions Baseline for energy, utilisation, water and cost agreed; named infrastructure and service owners; defined measurement boundary
Revisit when Demand evidence changes, measured energy per successful transaction is available, or an option cannot meet service requirements
Evidence status All results remain to be measured

Template for adaptation. Independent professional guidance; not an engineering design, environmental assessment or regulatory opinion.