OT(ish) - bitcoin mining rig as heating appliance?

Feb 26, 2022 Last reply: 4 years ago 7 Replies
formatting link

The new proposed Intel device allegedly consumes 3.6kW. This is a bit more than a fan heater.



A combi boiler for a larger house can be rated up to around 30kW, although probably a system boiler would be a more reasonable application.



Hmmm...our smart meter display suggests that we have used about 39 kWh so far today but I've never bothered to check when it resets the counter. Assume midnight?



Anyway, that would be around 10 -11 hours output from a mining rig. Then again that includes all electric devices including TV and kettle, various computing stuff, lighting.



Quick fag packet suggests that feeding 3.6 kW into a massive heat store for 24/7/365(6) would produce more heat than we are likely to use.



The hypothetical device would have to switch between heating (to bring the heat store up to temperature) and cooling (to cool the rig once all required heat had been stored).



A larger house with potentially more heat loss could possibly use more of the output.



Here my ignorance really comes to the fore. How many {whatever} coins could you expect to mine with this rig over the year, and what would be an approximate value (in pieces of string)?



My rough calculation would be (cost of mining - (cost required for heating



  • value of coins mined))

I assume that people make some income from high end mining rigs. Although to be really successful you would need to reduce the unit cost of electricity. Or co-locate with a cannabis farm that is pinching power from the street lighting. The larger the building the more scope for reusing the heat, I assume.



So. team, is there any future in a Worcester Bosch bitcoin mining system boiler?


Cheers


Dave R



OK...

That's much bigger than needed to heat the water. A typical house probably needs about 10kW.

Yes.

Our baseload is about 10-12kWh, the rest is heating.

Yes, but...

What you're missing here is the cost of the rig. Bitcoin is a deflating asset, which means the earlier you produce the more your output is worth. Your income is the value of the coins, minus $x,000 for the rig, minus the cost of your electricity, minus the return in coins. It is better to earn those coins earlier rather than later.

That means it's imperative to earn asap, and hence you have to run the rig

24/7. This is not commensurate with being a heating system, where half the year you don't need any heat and the rest you often don't need much of the heat.

If you wait too long the value of the coins mined will decline (it becomes harder to mine over time), the payback is less good, and the resale value of your rig drops away.

There are calculators for your return, eg:

formatting link
Set that against your other options:

- gas, which is ~25% of the price of electricity per kWh

- invest in insulation, which will payback against every bill for years to come

- heat pump, where you get 2x-4x return on electricity input

- oil, LPG, etc if mains gas isn't available

It would require doing the numbers which are very dependent on the price of crypto, but I doubt you can make the 'bitcoin boiler' to work economically, even if you get free heat out of it.

In November our oil boiler broke down, so I set up an RTX3060 to mine Ethereum as a source of background heat. NVIDIA cripples that GPU so it only gets 27Mhashes/sec, but that's still middle of the road. Took about

160W for the GPU, plus the rest of the PC.

For about 3 weeks I ran it, turning it off at night. I made about $12 (excluding electricity). Unfortunately the transaction cost of the Ethereum blockchain means you can't do transactions worth less than about $300, unless you sign up for a third party service and jump through hoops. So I couldn't claim my $12. It was an interesting lesson in how futile is all is.

For example, using it to mine bitcoin at 20p/kWh:

formatting link
45p/day. At that doesn't count for the upfront installation cost, which is amortised over only a couple of years.

They either steal electricity, or run in places like Kazakhstan where power is cheap.

You need something that is a constant heat demand, for which gas, oil or another fuel is not an option, and heat pump doesn't make sense. I suspect any kind of industrial process heat is going to be using gas if at all possible. The heat you get from servers is also low grade ~50C, which is no good for things that need to be hot rather than warm.

No. It shares some of the same problems as CHP - you can't use much of the heat it generates, and modulating it down defeats the point.

Theo

Why can you not use much of the heat? To use rather vague finger in air figures, ave 1kW electricity consumption might get you ave 4kW heat, useful for over half the year.

Because, as I posted, not every day is a freezing winter one where you can use 24h worth of heat output. Running it less than 24/7/365 trashes your economics.

Theo

Oh, if you mean about CHP (4kW of heat for 1kW *generation*, no idea if those numbers are right), the issue with domestic CHP is you can't modulate it down enough. On a spring day when you only need a little heat, you can't turn down the CHP system to produce only a handful of kW. You could I suppose run it into a big storage tank and then empty that gradually, but cycling the CHP isn't great either. If you run it for the generation but throw the heat away it's more efficient to run that in a power station CCGT than your domestic CHP unit.

It's less of an issue on a commercial CHP where there's always hundreds-of-kW load for heating somewhere like a factory.

One place where the mining/CHP 'free heat' could work is somewhere like a laundrette: 50C is about the right temperature, you don't want it much hotter, and they run all day. These days they mostly run on gas but maybe somebody could make a 'mining' tumble dryer... (it seems like commercial heat pump dryers are a thing, so that's the competition)

Theo

I'm not convinced that bitcoin systems are viable, since it all revolves around very complex encryption. If that costs more than the value of the coin to generate in energy, then what is the point? Besides, if Quantum computers are as good as they say they are then what is to stop the users of these nicking your bitcoins or making them worthless. We seem to have forgotten what money was supposed to be for, It represents values and values sooner or later are related to the amount you can pay people to produce what others need but cannot produce themselves. Brian

Fair nuff. I've been looking at a asystem where there's always use for some of the heat, and always available electrical loads for all the output.

Fanned dehumidified cold air is very effective at drying, putting CHP heat out of a job. I don't know if launderettes have caught up yet, last one I saw in the 90s was still using electric heated dryers.

According to this:

formatting link
laundrettes use gas power. I suppose heat pump drying (ie the dehumidified cold air you mention) is more efficient, but the equipment in that article looks ancient. Maybe it's because the machines cost a lot upfront and last for ever, so that there's no incentive to invest in new machines.

Theo

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required