Comment by yodelshady | Hacker News Reader

Comment by yodelshady | Hacker News Reader

Parent Full thread

Oh boy, fermi estimation time! Big caveat: all of the below assumes ideal mixing, which is not a given at this scale at all. So I'm not rubbishing the residents. (Nor disputing the idiocy of burning 45 MW to "mine" numbers for that matter.)

From wiki, the lake has a volume of 15 km3, or 15 trillion litres. The daily discharge of 135 million gallons is ~ 600 million litres. So, per day, it's cycling about 40 parts per million. Alternatively, you'll cycle the volume of the lake once every 67 years at that rate.

That same discharge is ~ 7000 litres per second, which will require 30 MW to heat by one degree Celsius. It's listed at at most 45 MW, so one and a half degrees rise, unless any of the water is evaporated, which it quite possibly is (after all, more is licensed to go in the plant than come out. I wonder if they have some scheme where a small volume of water is heated to a point where evaporation matters, then it is mixed back in?).

Once again, the locals may be noticing, but only if very significant stratification is occurring. Water is good at dissipating heat.

Running 45 MW of miners != Producing 45 MW of energy != Producing producing 45 MW of waste heat. First you have to consider the thermodynamic efficiency of the power plant. With a ~40% efficiency, 1 W of produced heat wastes 1.5 W of heat. Then you have to consider the overhead for running the miners. Ballpark, about half of the energy probably goes to HVAC. Furthermore, that 45 MW of mining power is also eventually getting released as heat, some of which may be returning to the lake.

Overall, 45 MW of bitcoin produced could mean 135 - 180 MW of waste heat.

It's quite likely that the real situation isn't anywhere close to ideal mixing, especially since warm water is lighter and the deep part of the lake might not be mixing efficiently with the warm discharged water. Then the surface water might be warming much more than your calculation suggests, to the point where locals are noticing a major difference.

The Finger Lakes are famously deep as well. Seneca averages 291' with a max depth of 618'

If this is a critical issue it probably wouldn't be overly difficult to spend some energy to pump in from a colder / deeper part of the lake if that is the concern, then heat and dump on the top. May be actually COLDER still than existing surface waters.

Or if it is a critical issue, just shut it down. We don't need to let someone ruin a lake to mine Bitcoin.

So putting it all together, assuming no evaporation and ideal mixing, at its current max rate of 45MW it would take 67 years to raise the lake temperature 1.5 degrees Celsius?

I think cooling rates may increase as lake temperature increases. That said, over 67 years may be just in the noise.

Reply on news.ycombinator.com