The Winters in the Ground is a new Hearth & Horizon story, about 4,900 words. It's set at an invented Benedictine priory in the aspen parkland of central Saskatchewan, from January 2049 to January 2050.
The priory heats its guest house from thirty boreholes, each 150 m deep, drilled six metres apart under the car park. On the coldest night of January 2049, the fluid coming back from the ground reaches minus six and the retreat wing's heat pumps shut down. The ground at that depth started at five and a half degrees. The plant assistant, Wren, finds the reason in its own logs.
Why the first winters lie
The field was designed for the guest house and church, about 95 MWh of heat a year. In 2045 the cellarer joined the old school wing to it as well and took out the wing's propane boiler, raising the draw to about 150 MWh. He wasn't being careless. He had three winters of readings, and the coldest fluid had only ever reached minus 1.2 against a limit of minus six.
Those first readings only ever describe the rock next to the pipes. In the story's rock, the cooling from one borehole spreads about ten metres in a year, so within a year or two the holes, six metres apart, are drawing on each other's ground. After that the field behaves as one large cold block, and a large block recovers slowly. Each summer puts back a little less than the winter took. The story's model is a finite-line-source g-function summed over all 900 borehole pairs. It gives coldest-snap fluid temperatures of 0.0, -0.7 and -1.2 °C for the first three winters, then -4.0, -4.8, -5.4 and -6.0 once the wing is on the loop. With nothing changed, it reaches -8.1 by 2054.
Wren owns its part. Each month it reported how far that winter's minimum sat above the cut-out. It never put the winters side by side.
The fix that costs the price of the fans
Fifteen more boreholes would cost about 191,000 dollars, and the cellarer says no, for reasons the story treats as sound. The retired driller suggests the other remedy: since nothing puts heat into the ground in summer, pump the loop through a dry cooler on warm days. That's a large outdoor radiator with fans. The priory buys a secondhand one from a curling rink that is closing down.
The first try runs it only at night, to spare the retreatants in the garden. Nights are cool, so the season would return about 47 MWh and leave the coldest January at about minus 6.1, still past the cut-out. The novice points out that his lawnmower is louder than the cooler. After that it runs in the monks' working hours, with the fans slowed in mid-afternoon, and returns about 76 MWh. The projection for the next January moves to about minus 5.9.
The story keeps one limit in view. Recharging stops the slide, but it doesn't make the field any bigger. With afternoons every summer, about 97 MWh a year, the coldest fluid levels off near minus five and a half. On the hardest nights the wing will still need its baseboards.
What's real and what isn't
The physics is standard. Borehole fields lose temperature over years when the heat taken out each year exceeds the heat put back, and summer recharge with dry coolers or solar collectors is a real practice. Every figure in the story comes from one model we wrote for it. The ground properties, the loads, the climate cycle and the dry cooler's 5 kW per kelvin are stated assumptions, not data from a real site. The minus-six cut-out, the decibel readings and the drilling rate are invented. The story's notes list all of them. The priory, its people and Wren are invented too. The story's notes paraphrase the Rule of Saint Benedict rather than quote any one translation. It's fiction, not engineering advice.