The one idea: the fluid temperature is the ground temperature minus two drops
BoreTherm, like the classic monthly design method, checks one temperature: the mean fluid temperature Tf, the mean of the inlet and outlet temperature of the heat-carrier fluid in the boreholes. It starts from the undisturbed ground temperature Tm and subtracts two drops. The ground around the field cools as heat is taken out over the years: the g-function, the dimensionless temperature response of the field to a constant heat rate, carries that history. Between the borehole wall and the fluid there is the effective borehole thermal resistance Rb* (pipes, grout and the heat exchange between the pipe legs). On top of the monthly base load comes a short peak at the end of each month.
Tb = Tm + Σi≤n (qi − qi−1) · g(tn − ti−1) / (2πλ)Tf = Tb + qn · Rb*Tf,peak = Tf + (qp − qn) · [g(tp)/(2πλ) + Rb*]smallest H with Tf,min ≤ Tf(t) ≤ Tf,maxq is the heat rate per metre of borehole, q = Q / (N·H) for N boreholes of length H, positive when heat goes into the ground; λ is the thermal conductivity of the ground, and Tf,min, Tf,max are the design limits of the mean fluid temperature. Months are 730 h. A peak counts only when it is larger than the base load in its direction, and its energy is not added to the month. The loads are entered as building loads; the seasonal performance factor SPF of the heat pump (heat delivered per unit of electricity over the season) turns them into ground loads.
The workflow
Two questions are asked of the same model. Calculate: does this design keep the fluid inside the limits? Size borehole length: how long must the boreholes be, between the shortest and longest length allowed (Hmin, Hmax), so that it just does?
How the calculation works
Calculate: one simulation at a fixed length
simulate() for monthly loads. Nu: Nusselt number, Rp: pipe resistance. Hourly loads replace step 5 by an hourly convolution (FFT or load aggregation) and take the peaks from the hours themselves.Size borehole length: the search for the required length
Size borehole length does not optimise the layout: it finds the root of f(H) = max(Tf,min,limit − Tf,min(H), Tf,max(H) − Tf,max,limit), the violation of the worse limit. All years count unless Check the limits in the last year only (the rule of the classic monthly method) is set.
The theory behind the numbers
The g-function: the memory of the ground
The g-function is computed from finite line sources, segment by segment, with the same wall temperature in all boreholes (UBWT, as Eskilson’s g-functions). With λ = 2.6 W/(m K), 1 W/m held for 25 years gives 1.34 K at the wall; for 6 hours only 0.08 K.
Rb*: the borehole between wall and fluid
The multipole method, an analytical solution for the pipes in the grout, gives the borehole resistance Rb (all legs at one fluid temperature) and the internal resistance Ra (between the down- and the up-going leg); ṁcp is the capacity flow of one borehole. The fluid warms on its way down and back, so the legs exchange heat with each other. The uniform-heat-flux form of the classic monthly method adds that short circuit: Rb* = 0.1151 + (150/1988)² / (3 · 0.4397) = 0.1194 m K/W. It grows with H², so a longer borehole is a little less effective per metre, and Size borehole length recomputes it at every length.
Base load, peaks and years
How each result is derived
| Result | How BoreTherm derives it | Example at 150 m |
|---|---|---|
| Undisturbed ground Tm | Ts + qgeo · (H/2) / λ (mid-depth of the active length) | 9.0 + 0.06 · 75 / 2.6 = 10.73 °C; at 143.7 m: 10.66 °C |
| Time scale ts | H² / (9α), α = λ / C | 67 years |
| Borehole resistances Rb, Ra | multipole method (order 3), pipe film by Gnielinski | Rb 0.1151, Ra 0.4397 m K/W; Reynolds number 4958 |
| Effective resistance Rb* | Rb + (H/ṁcp)² / (3Ra), ṁcp of one borehole | ṁcp = 1988 W/K → 0.1194 m K/W; at 143.7 m: 0.1190 |
| Ground load, January | heat · (SPF−1)/SPF + DHW · (SPF−1)/SPF, per 730 h; peak · (SPF−1)/SPF | base −31.9 kW = −13.28 W/m; peak −73.7 kW = −30.70 W/m |
| Base temperature, January of year 25 | Tm + Σ Δq g / (2πλ) + q Rb* | 10.73 − 8.14 − 1.59 = 1.00 °C |
| Peak temperature | + (qp − q) [g(tp)/(2πλ) + Rb*], field g at tp | g(6 h) = 1.253: −3.42 K → −2.41 °C |
| Lowest / highest Tf | min / max of base and peaks over all months and years | −2.41 °C (year 25, January) / 16.03 °C (year 1, July) |
| Lowest inlet temperature | Tf + Q / (2ṁcp), field flow; freezing check | −3.57 °C, margin 6.2 K to −9.8 °C |
| Required length (Size borehole length) | smallest H with every Tf inside the limits | 143.72 m, total 2299.5 m; binding: Tf,min |
SPF on Heat pump › Setup: heating 3.8, DHW 2.8, cooling 5.0: ground factors 0.737, 0.643 and 1.20. The table repeats the code’s arithmetic with the result file’s g-function; the sum reproduces the reported temperature.
The example in numbers
| Step | What happens | Result |
|---|---|---|
| Open the example project | 4 × 4 field, 6 m spacing, 150 m, single-U PE 32 × 2.9, 25 % MPG, 25 years | limits −3 / 20 °C |
| Calculate | design check at 150 m | Tf −2.41 … 16.03 °C, limits met |
| Size borehole length | 12 simulations, 2 with the exact g-function; binding limit: minimum fluid temperature | 143.7 m per borehole, 2299 m in total; Tf −3.00 … 16.18 °C, inlet down to −4.16 °C |
| Apply length to project | all boreholes get the sized length; Edit → Undo restores 150 m | one undo step |
Shortening the boreholes from 150 m to 143.7 m lowers Tf,min from −2.41 to −3.00 °C: about 0.09 K per metre. The binding month is still the heating peak of January in the last year.
Limits to keep in mind
- Monthly loads assume a constant load within each month plus one peak at its end. Several cold days in a row are not in the model; hourly loads resolve them.
- The limits apply to the mean fluid temperature. In heating the inlet is colder (the example: −3.57 °C at 150 m). Keep the freezing margin, or switch on the inlet limit.
- One g-function for the whole field (UBWT): every borehole has the same wall and fluid temperature. Sizing keeps the positions and gives all boreholes the same length (unequal lengths keep their proportions).
- Tm depends on H through the geothermal flux. With cooling-dominated loads Tf,max can then fail at both short and very long lengths; BoreTherm scans the range and warns when the violation does not fall with H.
- The steady borehole model gives the fluid no heat capacity: Tf follows a load step at once. For short peaks this is conservative; the short-time and dynamic borehole models remove it (hourly loads).
- The compatibility option for the classic monthly method (Design limits & simulation) switches convection, peak response and last-year sizing to that method’s choices; the defaults check every year and use the field g-function for the peaks.
Numbers calculated with BoreTherm. Theory: BoreTherm theory manual.