Methodology

How each tool works: the equations, the material data, the simplifications, and where the inputs come from.

Will My Pipes Freeze Tonight?

The tool estimates, hour by hour for the next 48 hours, the temperature of the air around a pipe and of the water inside it, including the freezing process. The approach follows the heat-transfer fundamentals in theASHRAE Handbook — Fundamentals (heat transfer and thermal property chapters): a steady-state thermal resistance network around the pipe combined with a transient lumped-capacitance energy balance for the water.

1. Inputs and data sources

• Location: ZIP code → latitude/longitude from theGeoNames US postal code dataset (CC BY 4.0). The lookup runs in your browser from static files split by the first three digits of the ZIP.
• Weather: hourly temperature and wind speed from theNational Weather Service API(/points → forecastHourly). Where NWS gives a wind range (e.g. "5 to 15 mph"), the upper value is used.
• Safety margin: every forecast temperature is lowered by 2 °F (1.1 °C) before use.

2. Temperature of the space around the pipe

Unheated spaces sit between indoor and outdoor temperature. At steady state the model keeps a fraction f of the indoor–outdoor difference:

T_space,ss = T_out + f · w · (T_in − T_out)

where w = max(0.6, 1 − 0.015 · wind_mph) accounts for wind-driven air exchange (not applied to heated rooms). The space responds to changes with a first-order lag that represents the thermal mass of the structure and ground:

dT_space/dt = (T_space,ss − T_space) / τ

These coefficients are engineering estimates chosen at the cold end of typical field behavior. The attic value is very low because ventilated attics track outdoor air closely and can drop below it on clear nights through sky radiation. The heated-room case represents a pipe in a closed sink cabinet on an exterior wall.

3. Heat loss from the pipe

Per metre of pipe, heat flows from the water to the surrounding air through a series of cylindrical resistances:

R′ = ln(r_o/r_i) / (2π k_pipe) + ln(r_ins/r_o) / (2π k_ins) + 1 / (h_o · 2π r_ins)   [m·K/W]

q′ = (T_water − T_space) / R′   [W/m]

The outer coefficient h_o combines natural convection and long-wave radiation. The water-side film resistance is neglected, which slightly overstates heat loss (conservative). No credit is taken for the critical insulation radius effect beyond what the equation itself produces.

Dimensions follow ASTM B88 (copper Type L), ASTM F876/F877 SDR 9 (PEX), ASTM D2846 SDR 11 (CPVC) and Schedule 40 steel pipe. Foam insulation:Foam sleeve, ½ in wall, Foam sleeve, ¾ in wall, k = 0.038 W/mK (polyethylene / elastomeric foam).

4. Cooling and freezing of the water

The water and pipe wall are treated as one lumped mass per metre of pipe:

C′ = ρc_water · π r_i² + ρc_pipe · π (r_o² − r_i²)

Above 0 °C:   C′ · dT_water/dt = −q′

At 0 °C the temperature is held while latent heat is removed. The energy needed to freeze all water in one metre is

L′ = ρ_water · h_sf · π r_i² ,   h_sf = 334 kJ/kg

and the ice fraction is the latent heat removed so far divided by L′. Water properties: ρ = 1000 kg/m³, c = 4186 J/kgK. The equations are integrated with a 1-minute time step, linearly interpolating the hourly forecast. The water starts at the current space temperature (but not below 0 °C, with no ice). The model assumes no flow, no supercooling below 0 °C, and no heat from adjacent ducts or the ground. Lumped treatment is reasonable because the Biot number of a small water-filled pipe in still air is well below 0.1 for the outer resistances used here.

5. Risk levels

"Risk starts" is the first time the modeled water reaches 35 °F. Research on freeze-induced pipe bursts by Gordon (Building Research Council, University of Illinois) found that rupture is driven by pressure building between an ice blockage and a closed fixture, and identified about 20 °F as the temperature at which unprotected pipes in unheated spaces become vulnerable; the 20 °F criterion comes from that work. The 25% ice fraction is our own conservative stand-in for "an ice plug has likely formed somewhere along the run".

6. Limitations

• Real spaces vary with air leakage, sun exposure, nearby ducts and ground contact; the coefficients are typical, not measured for your home.
• Forecast error grows with lead time; the second night is less certain than the first.
• Water use during the night (a running dishwasher, a dripping faucet) lowers the risk compared with the model.

References

• ASHRAE. ASHRAE Handbook — Fundamentals, chapters on heat transfer and thermal properties of materials.
• Incropera, DeWitt, Bergman, Lavine. Fundamentals of Heat and Mass Transfer. Wiley.
• Gordon, J. R. An Investigation into Freezing and Bursting Water Pipes in Residential Construction. Building Research Council, University of Illinois at Urbana-Champaign, 1996.
• National Weather Service API documentation — weather.gov.
• GeoNames postal code data — geonames.org, licensed CC BY 4.0.