Hacker News Reader

Comment by arsome | Hacker News Reader

Yeah, during the winter I mine on my video card because I'm basically just swapping some of my natural gas heat out for a few hundred watts of electric heat and my electricity is primarily provided by nuclear and hydro.

Kind of weird to think of mining as a way to reduce your carbon emissions but in some cases it certainly can be.

I'll definitely look more into heatpumps when it comes time to replace my furnace or air conditioner, but my understanding is they perform quite poorly in Canadian winters.

View on HN

> my electricity is primarily provided by nuclear and hydro. Kind of weird to think of mining as a way to reduce your carbon emissions but in some cases it certainly can be.

I think you are likely mistaken, although it is a surprisingly common misconception. Most places in the world, if you are using an extra kWh of power, the system needs to generate up to an extra marginal kWh from coal or gas.

Example 1: France generates power 70% from nuclear. However if a French person uses an extra kWh during base load, that could easily lead to one less kWh exported, which would most likely mean one extra kWh is generated from say coal somewhere else in Europe.

Example 2: New Zealand generates 60% of power from hydro. If I use an extra kWh of power at the moment (even at cheap rates), then if we have a drought in 6 months time, then an extra kWh needs to be generated from coal in 6 months time. I am presuming the coming critical storm[1] this weekend won’t fill our hydro lakes[2] and that the predicted long term dry weather pattern[0] will occur. I am a New Zealander.

You could be right if you are only using your power when it is cheapest (which usually implies the usage of base load nuclear), although it is surprisingly difficult to actually figure out whether you have generated zero extra carbon.

[0] https://www.rnz.co.nz/news/national/441393/hydro-power-more-...

[2] https://www.transpower.co.nz/system-operator/security-supply...

[1] https://www.newshub.co.nz/home/new-zealand/2021/05/rare-red-...

Edit: on second thoughts in NZ I am sure our lakes can’t fill that quickly from one good storm - the inflow rate is limited by the size of the river and the inflow rate is small compared to the storage size?

This seems like a silly way to look at production though - you could say every kWh in the system is a "marginal kWh" if you think about it like this. Turning on your oven, charging your car, etc would be similar in effect. If you really wanted to get involved you'd have to look at power imports and exports as well.

While it's true that fossil fuels do provide an important role in variable management of the grid still, this is also accomplished with hydroelectric quite heavily in my area and over time the rates of each production method will adjust accordingly so I think it's still a reasonable assessment to look at overall production.

I think you’re ignoring the concept of it being marginal, it’s marginal because they’re on the margin of choosing to mine or to turn up the heating. Those other activities aren’t marginal you need to cook, you need to drive, etc.

Cooking, though, is an activity that puts demand on the grid for a short time. Mining BTC to heat your house is a constant and predictable load. There are plenty of houses in canada with electric furnaces - Last I checked Hydro Quebec would help subsidize the cost if you switched from gas to electric heat.

Even a constant and predictable load will cause some extra carbon dioxide output. How much depends on a variety of factors.

For example, during any period that the nuclear and hydro are running at capacity. If you reduced your constant power usage by 1kWh then during those peak periods, 1kWh less gas generation would be used by the network.

If your constant 1kWh of electricity would otherwise be exported to the US if you didn't use it, then it could be that 1kWh less coal generation would be used at some times.

Most people find it difficult to think about how their marginal increase in usage very often will cause extra CO2. Even writing this comment down has made me think about how CO2 production also depends on the future predictions of generators. It is complicated, but worth spending time to think about.

I know - it is complicated. The commercial rate here in Ontario for large consumers is based on the delta between expected load and actual load. If the expected load is much higher than actual, it can actually go into negative pricing (if they have a big nuclear reactor they can’t or don’t want to spin down, they will literally pay you to use the electricity.) I had a friend at a big company here that had some large refrigeration units for medical related stuff, but they had quite a wide range of temperatures they were allowed to be. So my friend wrote a script that scrapes Ontario Hydro’s website to fetch the commercial rate, and when it was high he’d let the temperature climb a little, and when it was low he’d cool the units a little more. Ended up saving seven figures a year.

I don't know about Canadian winters, but heat pumps certainly perform well in Swedish winters, which I don't imagine are much milder than Canadian ones.

If you can get a water-water heat pump you'll be fine for sure (requires drilling a rather deep hole in your backyard tho, so the initial investment is a bit bigger)

While geothermal systems do use heat pumps, the term "heat pump" in general conversation almost always refers to air-source heat pumps (air to air), not geothermal. In the US/Canada, drilling additional wells is prohibitively expensive for most people, which is why geothermal systems haven't taken off.

Heat pumps don't perform very well, or at all, when it's -30° or -40°, so a secondary heat source is needed. Usually, that will be natural gas, in which case why bother with a heat pump when you can just burn much cheaper natural gas?

Oh, in Swedish they're both called heat pumps, i didn't realize there was such a distinction in the English language between the two systems.

In my mind they're just different kinds of heat pumps.

And yeah, drilling is pretty expensive, but once the system is installed, heating is ridiculously cheap. When my parents installed geothermal, ROI was estimated to about 5 years for their house iirc.

But then it may well be that drilling is cheaper here by virtue of being more common, idk.

There both is and isn't. If someone says heat pump in a generic heating context they don't care. However most of the time the difference is significant, so geothermo is added as a qualifier where that is what is meant, and heat pump alone means air to air. You have to guess from context which it is though - and you won't always be right.

Heat pumps are cheaper than propane at least, though I'm not sure if they are enough cheaper as to be worth the cost.

The context maters here. If a boat is being discussed, it won't be air to air.

What's your point?

I've heard before that the climate in Sweden is relatively mild considering its latitude, mostly due to the marine currents. A quick comparison of climate seems to suggest that this is the case: https://www.weather2travel.com/weather-comparison/?placename...

Air source heat pumps are fine for the places where most Canadians live including southern Ontario & Quebec and coastal British Columbia. If you live in Winnipeg or Saskatoon it might be dicey, but unless it routinely drops below -20C for extended periods you're fine. And you can have backup resistive heating elements for the occasional cold snap anyways.

> I'm basically just swapping some of my natural gas heat out for a few hundred watts of electric heat

Except now you have to compare efficiency of your natural gas system vs efficiency of the power plant + electric heater. Chances are that energy efficiency of the power plant + an electric space heater (which is what you're turning your PC into) is smaller than the efficiency of your furnace. Quick googling comes up with energy efficiency of 60% for natural gas powered plants, while I don't think you'll be able to buy a gas furnace that has efficiency smaller than 80% these days...

IIUC, the only way you can mine without worrying about energy usage is if you have electric space heaters in your home - then it doesn't matter if you turn on the PC or the heater. Or if you own solar panels and you are producing more than you are not able to use (or sell).

To summarize the part about efficiencies, by heating up your place with electricity instead of gas, the power plant needs to burn ~50% more gas (& so 50% more emissions) than you'd have to if you were to use your furnace. Intuitively, other fossil-fuel power plants will also cause larger emissions than you would locally (as they all start with burning stuff for heat & only then convert that to electricity with some losses, while you could stop there), but I haven't run the numbers. Thermodynamics is a b*tch.

Probably, though I live in an area where 50% of our electricity comes from the wind, so if the wind is blowing electric heat is far less CO2, while if the wind isn't blowing electric heat is far more.

In that case the problem becomes what other poster mentioned: it's a zero-sum game so you increasing your energy consumption takes green energy from someone else. Maybe if you mined only when you knew that the energy would be curtailed otherwise? Not sure how to know that though...

> Kind of weird to think of mining as a way to reduce your carbon emissions but in some cases it certainly can be.

Have you actually done that calculation out? You say "primarily provided by", but I wonder what that means, given just how much electricity you need to actually heat a home via GPUs.

(You may of course be saving money given the profit you generate, but that's not the same thing as reducing emissions.)

Well, just to give you an idea, my province typically ranges from 1-10% provided by fossil fuels and of the fossil fuels - it's just natural gas. I'm sure there's some line losses there, but I don't think they're enough to overcome the 90%+ of energy coming from other sources.

Of course, I'm not heating my entire home with a single GPU, merely offsetting a small portion of my overall heating cost.

I'm very curious how that can be better then an electrical heater.

It has to be more inefficient as you are not spending energy on calculation but if you substitute it with selling your bitcoins you would need to compensate your co2

It's pretty brilliant actually. Case 1 electric space heaters blindly pass electricity through dumb wires. Case 2 instead of dumb wires you use ASICs. Case 2 is a win win that helps secure a decentralized monetary asset, get paid for it, and heat your home which you would have done anyway, all with the exact same carbon footprint (assuming you're expending equivalent amount of electricity to just heat your home as originally planned). You've basically turned PoW into proof of heating where the miner gets paid and gets the benefit of heating his home on top. And incremental environmental impact is exactly 0.

edit: quick google -> and here it is :)

https://bitcoinminingheater.com

The only minor issue being that the materials used in fabricating the ASIC are likely more expensive (money and energy-wise) than those for a conventional heater!

Wow, this is actually amazing. In my country electric showers are really popular. They're literally just big resistances immersed in water. I wonder how long it'll take for them to put cryptocurrency miners in there.

My shower pays for itself :)

Now long showers aren't just relaxing but profitable too. Awesome.

But how much energy is used percentage wise for mining?

Technically a "perfect" miner would use zero energy (beyond microamps for driving the signal wires in the network cables), the heat comes from imperfect components that "waste" energy by turning it into heat.

So the mining itself doesn't use any energy, it's just that instead of dumping power into a dumb coil of wire, you dump it into a smarter coil of wires and various components that produce some output (that happens to be valuable to the Bitcoin network and it rewards you in exchange) as a side-effect.

For any given watt of electricity, the heat output is the same, just that with a miner you also get useful calculation results out of it.

The Landauer principle states that any computation involving information erasure (e.g., taking the hash of blocks) must correspond to some nonzero increase in entropy, which would be expressed here as heat.

That's not possible. Physically speaking

It definitely is.

Heat is the most retrograde form of energy - the form with the most entropy. You can put 100W into a resistor, and you’ll get out 100W of heat. You can put 100W into a lightbulb, and it will generate less than 100W of light and some heat (because it’s not perfectly efficient) but in a room with no windows that light will bounce around until it’s absorbed by the walls and turned into heat so the room will heat up the same as if you had a 100W heater. If you put 100W into a miner, it will mine bitcoins and shed 100W of heat.

Water always runs downhill, energy always follows entropy down into heat.

Edit: Maybe this example will make more sense - consider an old fashioned record player. If the turntable is magnetically levitated, and is in a vacuum chamber, you can spend a tiny amount of energy to spin the table up to 45 rpm, and then (with no needle on the record) it will spin at that speed basically forever, because it's in a frictionless environment (there's no real frictionless environment, of course - you'll need a tiny amount of energy to keep it going). If you have an old turntable from the 70s where all the cheap grease has hardened, then to keep it spinning, the motor will have to do some work. Let's say you're putting 5W in to keep it turning. That 5W is overcoming the friction from the air, and from the old grease, and from the belt drive that runs the turn table. That friction is generating 5W of heat. It has to be, because if it was generating less than 5W of heat then either we'd be using less than 5W to maintain a steady state, or the else the turntable would be accelerating until we were using 5W to maintain a steady state.

My understanding is it's essentially just as efficient as a resistive electrical heater as all the energy used in your video card is converted to heat at some point in the process. There may be some very minor loss here, but I'm not enough of a physicist to say what it'd be.

That's what I'm wondering. It costs energy to move electrons

That energy is all turned into heat. Waste from the point of view of a bitcoin mining attempt. However heat is a useful byproduct in this case when the local temperature is lower than the inhabitant would desire.

Because it also generates revenue via bitcoin. So it's "cheaper" insofar as the bitcoin is subsidizing the operation of the rig and, as a byproduct, heating the space.

Mitsubishi has heat pumps that maintain 100% heating capacity down to 23℉ (-5℃), and 76% down to -13℉ (-25℃). Fujitsu has similar technology.

Watch this: https://www.youtube.com/watch?v=7J52mDjZzto

Reply on news.ycombinator.com