Common service scope
Set workloads, usable IT capacity, resilience, security, latency and the required opening date before comparing locations.
EBG Studios · Development Atlas · DRAFT — For review & feedback
The investment decision · from shortlist to commitment
Build the case before you build the campus.
A guided pro forma walkthrough: connect business need, site evidence, utilities, cost and risk. Start in discovery, develop the case in feasibility, and revisit it as the project changes.
Decision 1 of 6
What investment are we actually evaluating?
A pro forma is a forward-looking model of assumptions, costs, timing and outcomes. It starts with the business need and changes as evidence improves.
Primary responsibilityMaya · Development lead
With: Jo (operations lead); Noor (power & cooling engineer).
Owner versus developer economics
For this enterprise owner, compare whole-life cost and business outcomes with equivalent alternatives. A rental developer also needs lease-up, rent, operating income, financing and exit assumptions.
Decision criteria
Agree which requirements are non-negotiable and how cost, speed, resilience, water use and community effects will be weighed.
1 / 6
Noor + Eli · cooling and source capacity
Water is a site decision, too.
Choose a cooling concept and investigate the source. Annual consumption and peak deliverable flow answer different questions. Blank fields mean unknown, not zero.
Direct-to-chip liquid cooling describes how heat leaves the servers. The final heat-rejection method determines much of the water demand. Selecting a concept does not assign a consumption rate.
Obtain a site-specific provider response covering capacity, pressure, upgrades, service timing and restrictions.
Estimated annual cooling makeupAwaiting an assumption
IT MW × 1,000 × 8,760 hours × liters/IT kWh. Cooling makeup only; this is not a full-site WUE audit, source-withdrawal permit or peak-flow calculation. It excludes domestic/fire water and upstream electricity-related water.
Peak capacity is unverified. A positive margin alone does not establish pressure, quality, reliability, discharge permission or authorization to withdraw.
What Noor and Eli must bring to the decision
- Peak-day and seasonal flow/pressure assessment, including domestic demand and fire-flow coordination.
- Quality and treatment compatibility, blowdown/discharge route, and storage or alternate operating strategy.
- Supply rights, applicable permits, drought restrictions, sustainable yield and the cost/date of infrastructure upgrades.
- Water, sewer, treatment and pumping costs in the operating pro forma, with consistent scope and no double counting.
A national or cross-border shortlist
Compare evidence before choosing a winner.
Use up to three candidate locations across all 50 states, Washington, D.C. and Alberta. Every candidate begins unverified. These are your screening notes, not live utility quotes or government determinations.
Power capacity & service date
Noor with the electric provider
Water supply & cooling
Noor and Eli with the water provider
Land use & building approvals
Morgan and Eli with Maya and the authorities
Fiber routes & latency
Jo with the carriers
Security & data requirements
Jo with IT, security and counsel
Buildability & delivery schedule
Sam with Morgan and the utility providers
Add a preliminary cost comparison
Use the same capacity, workload, service level and cost scope for all candidates. Enter an all-in initial budget and a flat annual operating allowance. This deliberately simple screen uses ten year-end operating payments, an 8% annual discount rate and initial capital at time zero; it excludes later refresh, escalation, financing, tax effects and residual value. Use the full enterprise lab for a more detailed USD case.
| Candidate | Evidence readiness | 10-year screening present cost · USD | Cost basis |
|---|---|---|---|
| Norfolk candidate | 6 evidence checks unresolved | Incomplete inputs | Estimate basis not recorded |
| U.S. alternative | 6 evidence checks unresolved | Incomplete inputs | Estimate basis not recorded |
| Alberta candidate | 6 evidence checks unresolved | Incomplete inputs | Estimate basis not recorded |
Read the full decision walkthrough
Define success: What investment are we actually evaluating?
A pro forma is a forward-looking model of assumptions, costs, timing and outcomes. It starts with the business need and changes as evidence improves.
Common service scope. Set workloads, usable IT capacity, resilience, security, latency and the required opening date before comparing locations.
Owner versus developer economics. For this enterprise owner, compare whole-life cost and business outcomes with equivalent alternatives. A rental developer also needs lease-up, rent, operating income, financing and exit assumptions.
Decision criteria. Agree which requirements are non-negotiable and how cost, speed, resilience, water use and community effects will be weighed.
Compare locations: Which sites deserve detailed investigation?
Screen a national or cross-border longlist before spending heavily on individual parcels. A cheap site that cannot meet a critical requirement is not the low-cost solution.
Hard constraints before weighted preferences. First test power, water, land use, connectivity, data requirements and delivery. Then compare preferences among feasible candidates; a good cost score cannot cancel a blocking issue.
Same scope, different place. Compare the same facility and service assumptions. Document differences in land, labor, construction, climate, insurance, taxes, incentives, tariffs and infrastructure upgrades.
Cross-border basis. Keep Alberta estimates in their quoted currency; use a stated exchange-rate convention, date and source to compare in USD. Review data residency and cross-border transfer requirements with counsel and application owners.
Prove the utilities: Can the site sustain the campus when demand is highest?
A nearby line, substation, river or well is an investigation lead. Available capacity, service conditions, rights, timing and upgrade costs require their own evidence.
Power and fiber. Confirm the actual provider, load study, connection route, firm versus interruptible service, upgrades, service date and separate fiber paths. Average IT load does not size peak electrical service.
Cooling and water. Choose how heat is rejected. Evaluate annual makeup volume, peak-day flow and pressure, quality, treatment, discharge, drought conditions and source permissions. Closed-loop liquid cooling alone does not establish zero water use.
A dependable operating envelope. Reconcile cooling water, domestic demand and fire protection with the provider. Test drought, hot weather, maintenance and outage conditions; do not assume storage alone resolves a continuing shortfall.
Build the pro forma: What is the full cost, and when is cash required?
Translate the coordinated concept and utility commitments into a cost and delivery model. Keep quantities, unit rates, exclusions, dates and evidence visible.
Capital and timing. Include acquisition, diligence, design, permits, site work, building, power and cooling, utility upgrades, IT, commissioning, migration, contingency and the spending schedule.
Operating costs and funding. Include energy plus demand/network charges, water and sewer, treatment, staffing, maintenance, insurance, property taxes and refresh. Identify equity, corporate capital or financing needs and when approvals are required.
Compare equivalent outcomes. For the enterprise, use discounted ownership cost with separately supported business benefits and alternatives. For income property, test NOI, debt service, returns and exit assumptions in the property model.
Challenge the plan: What could make the preferred site stop working?
A base-case result is only one possible outcome. Test the assumptions that could change the decision, especially correlated events such as hot weather, power demand and water restrictions.
Downside scenarios. Test energization delay, construction escalation, workload ramp, tariff changes, water constraints, incentive eligibility and exchange rates. Model costs in the period when they occur.
Mitigation versus avoidance. Price alternative cooling, a different connection, phased capacity or another site. Assign mitigation owners and deadlines; do not represent an unresolved dependency as approved.
Decision headroom. Ask how much cost or delay the business can tolerate and which failure would require revising or abandoning the investment.
Make the recommendation: What should the enterprise authorize now?
The pro forma supports a decision; it does not issue permits, secure utility service or prove operational readiness. An approval can release only the next bounded phase.
Explain the choice. Present the shortlist, common assumptions, financial comparison, unresolved issues and reasons for the preferred alternative. Keep unsupported incentives out of the committed case.
State the conditions. Identify the next authorization: continue diligence, secure conditional site control, revise, pause or commit a defined work package. List conditions, responsible people and dates.
Keep the model alive. Refresh it at site control, concept design, approvals, procurement, construction changes and operations. Compare actual performance with the original decision basis.