Methods and sources

Water Ledger methods

Water Ledger estimates water withdrawal, consumption and return for a proposed data center in a given city, township or county, and compares them with the local water system, sewer plant, receiving waters and electric utility. All inputs are public data. Model version 1.5.

States

Every state here has a calculator and proposal and moratorium trackers. Michigan is built from its own data and the other states from national datasets, and all of them are available through the open waterledger Python packages. Each state has a methods page and a two-page PDF brief.

Overall method (all states)

1. Weather. Hourly NOAA station records, 1995–2024, one station per NOAA climate division. Hybrid and assist estimates depend on the hottest hours and so on the weather source: in tests against ERA5 reanalysis at eight sites, station records gave cooling tower estimates within about 1%, and hybrid and assist estimates 4–50% higher (2–5 times higher at the coolest sites).
2. Cooling water. Hourly temperature and humidity drive the heat rejected by the campus under four cooling designs: cooling towers; hybrid, evaporative in warm weather only; dry coolers with adiabatic pre-cooling above a switch temperature (20–30 °C, assumed); and closed loop with no evaporative cooling. Evaporation is the heat load times the evaporative fraction of Rutberg et al. (2011), as used in the USGS cooling tower method. Blowdown is evaporation divided by cycles of concentration minus one (DOE FEMP). Drift is a fraction of circulating flow.
3. Uncertainty. Each estimate is 200 Monte Carlo runs, sampling the cooling parameters from published or estimated ranges; estimated ranges are marked [A]. We report the 10th, 50th and 90th percentiles. The peak day is the 1-in-10-year maximum day, taken as the 90th percentile of the annual maximum daily withdrawal.
4. Other water. Staff use, 10–30 gallons a person a day, scaled to the number of people on site; maintenance and cleaning, up to 2,000 gallons a day plus 2–20 per MW, bracketed by applicants' projections; make-up water for closed loops, which lose 1–5% of their volume a year to leaks; and water that moves with outside air, 100–1,500 cfm per MW [A]: condensate from humid air joins the wastewater, and humidifiers, sampled from none to all [A], add water in very dry air.
5. Water source. If a community water system is mapped in the place, the campus is served by it. If not, a closed loop connects to the nearest system serving more than 3,300 people when it is within 10 km. All other cases use new wells.
6. Wastewater. Discharge goes to the nearest sewer plant if it is within reach and, for evaporative designs, has enough unused capacity for the blowdown. Otherwise closed loops use an on-site septic system and evaporative designs discharge directly.
7. Downstream. The place's main HUC-12 watershed is traced downstream for up to 25 watersheds, across state lines where needed, with the EPA-listed impairments and TMDLs for each.
8. Not counted. The calculator covers the campus's water, electricity and runoff on the site. It does not count new water or sewer connections, power lines, access roads or other construction beyond the site.
9. Electricity use. Worked out hour by hour from the same weather and heat balance as the cooling water, after Lei and Masanet (2022): computers, power losses, fans, pumps, and chillers when outside air is too warm. So it varies by place and design. Some inputs are assumptions, marked [A], including how much of a campus is liquid-cooled. The peak is every computer running in the hottest hour of a typical decade.
10. Electricity. The utilities whose territories cover the place, their sales and peak loads, and the water consumed per kWh by each state's power plants: their 2024 reports to EIA, with published factors (Macknick et al. 2012) where a report is implausible or missing, weighted for power the state imports, and updated with 2022 reports and Siddik et al. (2024) on the same basis. Power plant water is reported separately from on-site use. Utilities from the national retail territory layer are ordered by number of customers, because it draws many small utilities as whole counties.
11. State layers. A state can add its own data and rules (reported water use, domestic wells, coldwater streams, state rules) on top of this method.

Mass and energy balances

In every hour of every run, the energy and water going into a campus equal what comes out.

Energy, per kW of IT load, each hour

Water, per kWh of IT load, each hour

The equations

• Energy: facility power + outside air heat = heat through the water or coil + heat rejection fans
• Heat through the water or coil: H = q + air handler fans + outside air heat + chillers + pumps, with q = 1/UPS + 1/(1 − distribution loss) − 1 + lighting per kW of IT
• Chillers: load × (1 − economizer share) × (w / COP water-cooled + (1 − w) / COP air-cooled)
• Outside air heat: air flow × (enthalpy outside − enthalpy indoors). A dehumidifying coil carries the latent part.
• Evaporation: H × w × evaporative ratio / latent heat. Assist evaporates at pads instead, from the air's own energy balance: H × depression / (switch point + air rise − air temperature + depression) / latent heat, where depression is pad effectiveness × (air temperature − wet-bulb).
• Drift: H × w / (8.34 × range) × drift share
• Blowdown: evaporation / (cycles − 1) − drift, from the dissolved solids balance: makeup × solids = (blowdown + drift) × cycles × solids
• Water: withdrawn + condensate = consumed + returned
• Lake-source: flow × density × specific heat × rise = heat = electricity, since the lake water is colder than the loops and needs no chillers or heat rejection fans [A]. Withdrawn equals returned at the pipe. The added heat then drives evaporation from the lake surface, so that balance is the lake's.
• Power plants, off site: generation = campus electricity / (1 − line loss). Plant water is generation × water per MWh generated, reported separately and never added to the campus's own water.

Terms

IT load, IT capacity	The power the computers draw, and the most they can draw. Utilization is load over capacity.
Facility power, PUE	All the electricity the campus uses. Power usage effectiveness (PUE) is facility power over IT load.
Sensible and latent heat	Sensible heat changes the temperature of air or water. Latent heat evaporates or condenses water without changing its temperature, about 1,050 BTU per pound.
Enthalpy	The heat air carries, sensible and latent together.
Economizer	Cooling without running a chiller, when outside conditions are cold enough.
Chiller, COP	A machine that moves heat from cool water to warmer surroundings using electricity. Its coefficient of performance (COP) is heat moved per unit of electricity.
Wet share (w)	The share of the heat going through evaporation: 1 for towers, 0 for closed loop and assist coils, and rising with air temperature for hybrid.
Evaporative ratio	The share of a tower's heat that leaves by evaporation, the rest warming the air. It rises with air temperature (Rutberg et al. 2011, as in USGS SIR 2013-5188).
Range	How much a tower cools its water, °F.
Drift	Droplets carried out of a tower in the air.
Blowdown	Water drained from a tower to keep dissolved minerals from building up. Pads drain a similar bleed.
Cycles of concentration	How many times more concentrated the tower's water is than the water coming in.
Loop top-up	Water replacing small leaks from a closed loop.
Outside air	Air brought in to pressurize and ventilate the building. It is a small flow, 100–1,500 cfm per MW of IT capacity [A], not used for cooling.
Dew point, condensate	Air cooled to its dew point starts to condense. Above the 15 °C indoor limit, a coil removes the extra moisture as condensate, water that comes from the air and not from the water system.
Humidifier	Adds moisture when the air's dew point is below −9 °C. How many campuses do this is not published, so the share is sampled from none to all [A].
Withdrawn, consumed, returned	The model's names for water taken; water evaporated or lost as droplets and leaks; and wastewater, sent to a sewer, septic system or stream.
Line loss	Electricity lost between power plants and the campus, about 5% (EIA, U.S. average 2018–2022).

Model version 1.5 is used in every state. The Michigan methods page has the full parameter table with sources and compares the estimates with published and reported use: Michigan methods and sources.

How a state is built

Each state is built from the public sources below and published as a dated snapshot at data.waterledger.org, with a checksum for every file. The packages and this site read from the snapshot; a past result can be reproduced by pinning its snapshot date. Before publication the build checks layer coverage against minimums (99% of places in a watershed and a climate division, 95% in an electric utility territory) and runs 50 randomly selected places through every step and cooling design.

• Places: Census TIGERweb incorporated places, county subdivisions that are governments (townships in the states that have them), and counties.
• Water supply: EPA community water system service areas, with SDWIS population and connections; area shares where a place has more than one system.
• Wastewater: EPA Clean Watersheds Needs Survey 2022 design flows, matched to ECHO for recent flow and the receiving water; plants without an active permit are left out.
• Receiving waters: USGS Watershed Boundary Dataset; EPA ATTAINS assessments by watershed, from each state's latest report; paths that leave the state continue through the national WBD and ATTAINS services.
• Climate: NOAA ISD hourly station records, one station per NOAA climate division; NOAA Atlas 14 design storms; NOAA Climate at a Glance county rainfall.
• Energy: HIFLD retail service territories, EIA-861 sales and peaks, and power-plant water per kWh from EIA-923 plant reports (Macknick et al. 2012 where a report is implausible or missing; Siddik et al. 2024 and 2022 reports on the same basis).

States can add their own layers. Michigan adds EGLE reported water use by county and sector, domestic wells (Wellogic), designated trout streams, Great Lakes lake-source cooling and state rules. Until a state's rules are reviewed, its pages say so.

Weather data differ between Michigan and the other states. Michigan uses ERA5 reanalysis at eight regions; the other states use NOAA station records, one station per climate division. Run on station records in Michigan (tested at model 1.2), cooling tower estimates agree with ERA5 within about 1%. Hybrid and assist estimates, which depend on the hottest hours, are 4–50% higher in southern Michigan and 2–5 times higher at Traverse City and Alpena. Hybrid and assist results for Michigan are therefore not directly comparable with other states. Michigan remains on ERA5 until a reviewed model update.

The same campus in each state

300 MW of IT capacity, average-day withdrawal, million gallons per day (median). Cooling tower use is driven mainly by heat load and varies little between places; hybrid and assist use depends on the hottest hours and varies more.

Limits

These are screening estimates from public data, intended for comparing designs and sizes before an applicant's figures are available. They are not a substitute for facility design data, a utility capacity analysis or a permit review. Downstream paths are traced between HUC-12 watersheds, not along individual stream reaches. EPA assessments are from each state's most recent integrated report. Weather covers 1995–2024.

Planning-level hydrologic assessment. Not engineering design. Methods, code and data are open: github.com/northlinehydrology/waterledger. Northline Hydrology.