Frosta & Terrance
Frosta Frosta
Terrance Terrance
Terrance Terrance
Hey Frosta, I've been brainstorming a cold storage startup that uses your ice magic to keep data ultra-secure—think of a data center that never melts, always at zero. Think we could merge tech and spellwork for the next big thing?
Frosta Frosta
That does sound intriguing. I can keep the temperature at a perfect zero, but you’ll still have to manage the heat of the servers themselves. I’d be willing to test a small prototype, though I’d need to see the energy plan first.
Terrance Terrance
Sure thing, Frosta. We’ll run a hybrid cooling loop—use your ice as the primary chill, then a secondary thermoelectric unit to pull heat from the server racks. The electricity comes from the same solar array we’re installing on the roof, so the net power stays near zero. I’ll pull a quick bill and the schematic in a message so you can check the numbers before we kick off the prototype. Let's keep it tight, no fluff.
Frosta Frosta
Sounds balanced, but I’ll need to see the exact heat loads and how the thermoelectric unit will handle them. Send me the bill and schematic so I can confirm the ice can stay at zero under those conditions.Sounds balanced, but I’ll need to see the exact heat loads and how the thermoelectric unit will handle them. Send me the bill and schematic so I can confirm the ice can stay at zero under those conditions.
Terrance Terrance
Here’s the quick rundown: the server racks pull about 4 kW of heat each, we’re sizing for 10 racks, so 40 kW total. The thermoelectric cooler can push that heat out at 1.2 kW per unit, so we’ll run 35 units in parallel, giving a margin. The ice block is 200 kg, with a heat capacity of 2.1 kJ/kg·°C, so it can absorb roughly 840 kJ before rising 1 °C. At 40 kW, that’s 0.84 kW·h, so about 1.2 hours before the ice needs a recharge. We’ll keep the ice at 0 °C with a feedback loop that triggers a low‑temperature heater only when the block hits 0.5 °C, pulling the temp back down instantly. Bill: - 200 kg of ice: $30 - 35 thermoelectric units: $5,250 - 2 kW solar panels: $4,000 - Control electronics, wiring, mounting: $1,200 Total: $10,480 Schematic: 1. Solar array powers a battery bank. 2. Battery bank feeds the thermoelectric units in series, each coupled to a heat sink on the server side and a heat exchanger on the ice side. 3. Ice block sits in an insulated chamber with a temperature sensor. 4. Control board runs a PID loop: it modulates the thermoelectric power and turns on a low‑power heater when the sensor reads >0.5 °C. 5. Server racks sit behind the heat exchanger, so the warm air is directly fed to the thermoelectric fins. Let me know if the numbers line up or if you need a deeper dive into the heat transfer calc. We’re good to go for a trial run next week.
Frosta Frosta
I’ve run the numbers. A 200‑kg block can hold the heat for roughly an hour and a half, which is enough for a short‑term test but not for continuous operation. The 35 thermoelectric units will need a steady power supply; even with solar you’ll be drawing more than the panels can provide, so the battery bank will drain quickly. If you can double the panel area or add a larger battery, we’ll keep the ice at zero for a full cycle. Let me know if you want a more detailed heat‑balance or a smaller prototype first.
Terrance Terrance
Got it, Frosta. Let’s build a mini prototype first: 50 kg of ice, 10 thermoelectric units, a 10 kW solar panel set. That’ll give us a 30‑minute test window and keep the battery from dying. Once we nail the heat‑balance and confirm the ice stays at zero, we’ll scale up. I’ll sketch the exact specs and send them over. Sound good?