Basically they heat up the soil and rock beneath their park during the summer using hot water/glycol from their solar arrays, and then during the winter they pump cold water through it which soaks that heat back up.
What's amazing is that it apparently took 4 years to fully saturate the area with heat energy.
TL;DR ground heat is remarkably persisting and takes years to dissipate, if deep enough. Coolth, the absence of heat, is really the same in this regard, because the heat transport through the ground is not rapid.
This is also why deep hot rocks need fracking to get thermal energy working unless its happening naturally. Rocks all around the place "down there" are hot. Very hot. A lot of ground water comes out hot. Bath (uk) for instance. its hot. So how come the rest of bath is cold and wet? Because ground heat doesn't move fast.
According to the same page though, the tunnels only add 4% to the heat and humans add 7% — remaining 89% comes from the trains; unless I misunderstood something.
(1) 10 year report on the project, which has detailed charts on the performance of the system:
https://www.dlsc.ca/reports/swc2017-0033-Mesquita.pdf
(2) More mathematical explanation of the dynamics of the system, specifically: “Because of their construction principle, BTES are not thermally insulated to the bottom and the side; only a top insulation layer reduces the losses to the environment. As the thermal conductivity of underground material is rather moderate, in a range of 1–5 W/m·K, heat losses can be kept low if the total volume is large enough to achieve a good surface-to-volume ratio. Size is important because heat losses are proportional to the storage surface while the storage capacity is proportional to the volume.“
https://www.sciencedirect.com/topics/engineering/borehole-th...
(3) What is the specification W/mK
https://forum.digikey.com/t/what-is-the-specification-w-mk/3...