386 m Well Potential

Well log

KFN-L, drawn to depth scale

Technology comparison

Peak heat output from the well, at your delivery temperature. Select a row for detail.

Technology	Well potential	Covers your building	Seasonal COP	Status

Detail

Fluid temperature over the well's life

Heat pump at the well’s potential. Coldest cold-snap fluid temperature each year, which reaches the minimum at the end of the design life.

Output by delivery temperature

The ground caps how much heat can be drawn. Hotter delivery adds more compressor electricity on top, so output rises a little while COP falls.

What moves the answer

Well potential when each input moves across a realistic range, everything else held.

Your building: cost and emissions

Your building through the year

Loop fluid comparison

How it's calculated, and what would firm it up

What's measured, and what's still missing

Done (2023): temperature profile and thermal response test. The fibre-optic cable gave temperature every 0.25 m and conductivity of 1.9 ± 0.1 W/m·K in clay from 203 to 373 m.
Pump test on the base screen: the biggest open question. Water enters only through 3.4 m of screen at 384–387 m, and its yield sets open-loop output.
Static water level and winter freezing: the casing water froze in the permafrost zone and was steam-thawed in 2022. Any design that moves water must keep it from freezing there.
Role of the well: grouting a closed loop into the casing ends its use as a monitoring well but roughly triples output. A removable loop in the casing water has to run above freezing to avoid ice, which limits it to about 5–6 kW.
Water chemistry and heat load: iron, manganese and hardness for open systems, and the actual heating demand of the buildings to serve.

Model

Ground temperature: T(z) = T_surface + gradient · z, held at ≥ 0 °C above the permafrost base. The mean over the heat-exchanging length sets the undisturbed temperature.
Ground response: finite line source (Claesson & Javed), buried below any insulated section. Monthly ground loads are superposed over the whole design life, with a cold-snap pulse at full heat pump output on top of every month. Checked against pygfunction: within 0.1%.
Borehole resistance: line-source multipole for the real pipe positions inside the casing, with water (0.57), bentonite (0.8) or enhanced grout (1.6 W/m·K). Pipe convection uses Gnielinski (turbulent) and laminar Nusselt numbers for pipes and annuli. Heat leaking between the down and up legs is solved along the full 386 m. Effective R_b checked against pygfunction's full multipole: U-tubes within 11% (conservative), coaxial within 1.5%.
Climate: ECCC 1991–2020 normals for Burwash A. Each month's hourly temperatures are a daily sine (from the mean max and min) plus a normal day-to-day spread (σ ≈ 11 K in January), fitted to the normals' counts of days below −30, −20, −10 and −2 °C and scaled to the published heating degree-days (base 15 °C). January 2.5% design temperature: −44 °C.
Building: annual fuel use × appliance efficiency gives heat. The hot water share is a constant load; the rest scales with degree-hours below 15 °C. The heat pump serves each hour up to its capacity and the backup covers the rest.
Well potential: the largest heat pump, heating a building of the same shape sized to it, that keeps the coldest cold-snap fluid temperature at the minimum in the final design year. Your building's heat pump is the smaller of its design load and what the well supports for its load pattern.
Uncertainty: 40 Monte Carlo runs over conductivity, ground temperatures, heat capacity, heat pump quality, borehole resistance and cold-snap length, shown as a 10th–90th percentile range.
Emissions: heating oil 2.75 kg CO₂/L, propane 1.51 kg/L, diesel generation 2.68 kg/L at the plant output you set.
Loop fluids: CoolProp data at the minimum loop temperature. Design flow gives 4 K across the heat pump, raised to reach turbulence where that takes less than double the flow.
Pumping: pipe friction plus 40 kPa for the heat pump exchanger and fittings, at 30% circulator efficiency. Well pumps run at 45%.
Heat pump: Carnot COP × an efficiency that falls with temperature lift (−0.0033 per K), using the brine leaving the heat pump minus a 3 K approach (2 K for water). The default of 64% gives COP 4.6 at B0/W35, 3.7 at B0/W45 and 3.0 at B0/W55, typical of current rated brine-to-water units.
Ice check (water-filled casing): coldest pipe wall = loop temperature − ΔT/4 + (heat per metre ÷ pipes) × pipe resistance. Below 0 °C, ice forms in the casing water.
Standing column well: closed-loop model with low borehole resistance, plus bleed heat ṁ_bleed · c_p · (T_ground − T_min) at peak. This is an approximation.
Open loop: Q = flow · c_p · ΔT (≤ 5 K, and at least 3 °C leaving). Pumped water cools as it rises past colder ground (U ≈ 2.5 W/m·K bare, in series with foam where insulated).
Permafrost insulation: an insulated closed loop exchanges heat only below it; the insulated top leaks through ln((r+t)/r) / (2π·0.04) of foam.
Direct use needs well water hotter than delivery + 3 °C. ORC power needs about 80 °C or more. Electrical efficiency is 40% of Carnot, less 20% parasitic load.

KFN-L data (Chapman et al. 2025)

Construction: drilled 2012 beside the Water Treatment Plant (61.35355° N, 138.97817° W). 6-inch steel casing (155.6 mm ID) to 384 m, 10-slot screen from 384 to 387.4 m. Outer casing and cement seal to 84.1 m.
Geology: entirely in sediments on the north side of the Denali fault, mostly clay and silt with sand and gravel lenses. Permafrost roughly 2–55 m; the profile sits near 0 °C from 24 to 50 m. A schist boulder was hit at 379 m.
Groundwater: horizontal flow zones at 48–65, 80–85, 112–132 and 150–203 m, all behind the casing. Heat moves mainly by conduction from 203 to 373 m.
Thermal: gradient 45.2 °C/km measured, 47.0 °C/km after topography and paleoclimate corrections. Conductivity 1.9 ± 0.1 W/m·K (203–373 m); 1.6–2.7 W/m·K apparent in the flow zones above. Heat flux 89 [84–94] mW/m².
Used here: the 45.2 °C/km measured gradient, extrapolated to about −2 °C at surface and 15.6 °C at 386 m (the 2012 drilling record reported ~16 °C water). Heat capacity 2.8 MJ/m³·K is an estimate for saturated clay; the paper doesn't report it.

Limitations

The building is inferred from annual fuel use with a 15 °C balance point, using an average climate year. Cold years, solar gains and heat pump capacity loss at low loop temperatures are not modelled.
Uncertainty ranges cover ground and equipment inputs, not the building, yield or prices.
Groundwater flow in the 48–203 m zones and latent heat from freezing clay are ignored. Both would raise real output.
The line source ignores the heat stored in the ~19 L/m of casing water, so 6-hour peaks are slightly pessimistic.
Contact resistance between the steel casing and the clay is ignored. Freeze–thaw cracking of bentonite grout at sub-zero loop temperatures is not modelled.
Ground temperature is a straight line; the top ~50 m is approximated to within 1–2 °C.
Insulation is modelled for steady running only. Freezing of standing water during shutdowns is not modelled.
Screening-grade only. Use it to decide what to test and which option to price, not to finalize equipment sizing.