METHODS, SOURCES & LIMITS
Illustrative scenarios.
Visible assumptions.
Last reviewed September 6, 2026
These tools help residents and project teams ask better questions. They are not engineering forecasts, permit findings, utility commitments, acoustic studies, economic-impact studies, or promises of project performance. Replace every planning assumption with cited project records before making a decision.
Site-selection checklist
The tool intentionally does not calculate a score. Distance to a residence and a native-planting target are shown independently because no universal threshold or weight proves compatibility.
- Residential distance means the shortest project-reported horizontal distance from the defined facility or equipment boundary to the nearest residential parcel or occupied dwelling. The applicant must state which boundary it used.
- Native replacement target means the applicant's proposed percentage of replacement trees, shrubs, and groundcover that are locally native. The applicant must state whether the denominator is plant count, planted area, canopy projection, or cost.
- Validation requires the local zoning ordinance, comprehensive or future-land-use plan, surveyed plan, equipment layout, viewshed, traffic analysis, landscape plan, and maintenance commitment.
Power formulas
PUE, IT load, grid share, on-site share, and annual load factor are user inputs. The model does not predict coincident peak, ramping, phased energization, backup-generator hours, carbon intensity, interconnection cost, or rate treatment. The 8,760 coefficient is the number of hours in a 365-day year; the selected load factor prevents the annual scenario from silently assuming continuous full-load operation.
Lawrence Berkeley National Laboratory: 2024 United States Data Center Energy Usage Report ↗
Water consumption ranges
The planning profiles estimate on-site consumption, meaning water not returned to the immediate water system. They do not estimate utility withdrawal, discharge, domestic use, peak utility demand, drought operations, water quality, or off-site water used to generate electricity. LBNL reports substantial variation by cooling system, climate, workload, and facility characteristics; project WUE or documented peak-day gallons is therefore preferred.
| Planning profile | Illustrative range | Coefficient status and source |
|---|---|---|
| Room-air / evaporative rejection | 2,000–5,000 gal/MW/day | Site-owned screening range informed by the cooling-system categories in the LBNL report; not a published standard. |
| Open evaporative or once-through | 3,000–6,000 gal/MW/day | Site-owned screening range informed by the LBNL report; verify system design and make-up/blowdown balance. |
| Closed loop with dry chillers | 25–150 gal/MW/day | Site-owned allowance informed by the LBNL cooling context; confirm whether heat rejection is truly non-evaporative. |
| Liquid-to-chip with dry rejection | 10–100 gal/MW/day | Site-owned allowance informed by the LBNL cooling context; not a vendor guarantee. |
| Dry / near-zero-water design | 0–25 gal/MW/day | Site-owned operational allowance informed by the LBNL cooling context; domestic use is excluded. |
| Household comparison | 300 gal/household/day | Rounded contextual benchmark from EPA WaterSense; not a resource-allocation equivalence. |
Sound comparison
Ambient and facility-source levels are combined logarithmically and dBA and dBC remain separate. The model applies only the user-entered solid-barrier value. Equipment controls can be tracked, but receive no generic numerical credit and are never stacked.
FHWA says an effective noise barrier can typically achieve about 5–10 dB of reduction when it is solid, sufficiently tall and long, and has no openings. The slider allows 0–15 dB so a user can enter a value from a project study; values outside the cited typical band require project-specific support. FHWA barrier guidance ↗
Community-outcome scenario
This section is an editable, site-owned scenario—not a statement of typical DMV outcomes. Its starting coefficients are deliberately visible: 12 construction job-years per MW, 0.45 operating jobs per MW, 0.35 maintenance jobs per MW, $110,000 construction payroll per job-year, $120,000 annual operating payroll per job, $60,000 maintenance spend per MW, 5% local-vendor share, and a 50% / 25% / 25% split of user-entered commitments among schools, small business, and workforce development.
Those coefficients are not empirical forecasts. Replace them with the project budget, county assessment, adopted tax rates, incentive agreement, workforce plan, vendor commitments, and a jurisdiction-specific economic study. For national employment context—without treating it as a local coefficient—see the U.S. Census Bureau data-center employment analysis ↗.
Editorial and verification standards
- Use 7×24 Exchange in display copy, Washington, D.C. in prose, and sentence case for Community guide.
- Describe the core region as three DMV jurisdictions—Washington, D.C., Maryland, and Virginia—with West Virginia noted only as additional AFCOM Potomac Chapter coverage.
- Date time-sensitive claims and event verification. Upcoming events are sorted by start date; expired events remain available only through the “Show past events” control.
- Prefer original government, research, organizer, and association sources. Link health is checked regularly, but readers should still verify decisions against the current official record.