Is Bitcoin mining sustainable?

[GODLIKE]

First of all, if this topic has been touched already, please don't bash me, I tried a search for something similar but couldn't find anything. If so, anyway, just let it drown down or even lock it, I won't be offended.

Second, it may seem that this topic is about mining, but it's not. This thread is about a possible huge point of failure that has never been considered about Bitcoin.



Now, considerations: at the actual price (let’s round it to 600$), each block found gives the miner around 600*12.5=7500$.
If we assume that miners are taking just a small share of overhead on that number, we can assume they pay something like 6000$ to mine those Bitcoins.
Some of those 6000$ are for hardware, a lot is for electricity.
Now, to make it easy, I’ll just assume that 5000 of those dollars are for electricity. This is, approximately, the cost in electricity to uphold the Bitcoin network for 10 minutes.

Assuming an average electricity bill for a house of 20$ per month, this means that 250 houses could be powered for one month with the power used to uphold the Bitcoin network for 10 minutes.
This means that 36000 houses could be powered for one month with the electricity used to uphold the Bitcoin network for one day.

Now, we are all here waiting that Bitcoin's value will rise where it deserves to, replacing or pairing up with fiat currency, and this means we are waiting for it to take off for good reaching 10000$ of value, or even 100000$ or even 1 million $, who knows.
Now you easily see where this is going.
If one Bitcoin is going to be as valuable as 100000$, upholding the network for one day will require the same electricity to feed 5976000 houses for one month.
Are we going to build nuke plants to uphold the Bitcoin network?

The problem is that, with the actual protocol and software, the energy necessary to uphold the Bitcoin network is directly proportional to its value, and nothing else.
In fact, the chain of events to consider is this:

1. Bitcoin price increases because there's more demand
2. People/companies add computational power to the network because there's more money to take from the market
3. The Bitcoin protocol adjusts the difficulty to suck even more electricity to uphold the network.
4. Goto 1.

This, if we look at raw data like I have done.
But there may be variables that I didn't consider, and I don't want to seem that one who knows everything.
I'll be glad to be proven wrong, because I see this as a hurdle coming fast onto the face of Bitcoin, and we may find that countries will ban Bitcoin mining because of this.
So if anyone can bring me additional variables to throw into the equation that could considerably lower the Bitcoin's network uphold cost, I'm all ears!
If there's some different Proof-of-XXX (Proof of Porn?) incoming on the software, that I'm not aware of, I'd be glad to read about it, as much as if there's any other software solution or protocol change that could clear the problem, and the same with anything else that can prove my reasoning wrong.

Again, if this matter has been treated already, just let the thread drown, but please give me the link to that thread!
Thank you!

[altcoinhosting]

I haven't done the exact calculations, but i would be supprised if 36.000 households could be powered with the combined power used by asics in 24 hours...

This would mean you could power 3000 households for a year with the power used in one day, or 1 million households for a full year with the power used for a year... Seems a bit high for me...

EDIT:

I'll try to look at this the other way around:
https://bitcoinwisdom.com/bitcoin/difficulty

The current hashrate is 1,840,166,130 GH/s.

Let's assume best case scenario first: everybody is mining with an antminer S9:
https://www.bitmaintech.com/productDetail.htm?pid=0002016052907243375530DcJIoK0654

14.000 GH/s for 1471 Watt at the wall...

1,840,166,130 / 14,000 =~ you'd need 132.000 antminer S9's to deliver the current network's hashrate... Seems a lot doesn't it?
132.000 * 1471 Watt   =~ 194 Megawatt... This power draw would power the full bitcoin network
Megawatt-hour = Megawatt × hour
194 * (24 hours*365 days) =~ 1.700.000 Megawat-hour =~ 1.700.000.000 Kilowatt-hour

An average household seems to use around 10 KWu a year (https://www.eia.gov/tools/faqs/faq.cfm?id=97&t=3), altough my household (me, my wife and my kid) use around 5 Kwu/year... Maybe our power usage is extremely low, maybe it's less for the EU as the US, maybe it has dropped significantly over the last 2 years???

So, by these calculations, 170.000.000 households could have been powered...

I must have missed something here... This can't be right   Probably a comma somewhere... 170 million households in the best case scenario probably means i did a convertion wrong somewhere...

[GODLIKE]

I haven't done the exact calculations, but i would be supprised if 36.000 households could be powered with the combined power used by asics in 24 hours...

This would mean you could power 3000 households for a year with the power used in one day, or 1 million households for a full year with the power used for a year... Seems a bit high for me...

EDIT:

I'll try to look at this the other way around:
https://bitcoinwisdom.com/bitcoin/difficulty

The current hashrate is 1,840,166,130 GH/s.

Let's assume best case scenario first: everybody is mining with an antminer S9:
https://www.bitmaintech.com/productDetail.htm?pid=0002016052907243375530DcJIoK0654

14.000 GH/s for 1471 Watt at the wall

That's exactly the point: we are not seeing how much electricity the Bitcoin network is sucking up at the moment, and that's why I made my calcs.

Imagine tomorrow 1 BTC would suddenly rise in value to 10000$, would YOU buy an ASIC and start mining?
I think I would.
You would recover the cost in the next 2 weeks and make quite some bucks over it.
Then, the price of Bitcoin could fall and you could just turn off your hardware.
But you don't need such sudden increase in value: miners are constantly calculating if it's convenient to add hardware and mine more.
And when is it convenient? When Bitcoin's price is higher.
In other words, Bitcoin's value is directly proportional to the electricity spent to uphold its network, AND NOTHING ELSE (<- the bad part).

Is there anything wrong in this reasoning?

[GODLIKE]

(1,840,166,130 / 14,000) * 1471 Watt = 193348884 Watt =~ 193 Megawatt... This power draw would power the full bitcoin network
Megawatt-hour = Megawatt × hour
193 * (24*365) = 1.700.000 Megawatt-hour

1 MILLION 700 THOUSANDS MILLIONS OF WATTS PER HOUR ?

Read it yourself, because if that's correct, I think I'm far off... and the network is sucking up much more than what I have written.

[GODLIKE]

So, by these calculations, 170.000.000 households could have been powered...

I must have missed something here... This can't be righ  Probably a comma somewhere...

Or maybe I am right and we are already fucking up a huge amount of electricity.

What I am wondiring is: is it possible that nobody hasn't yet noticed it???

[altcoinhosting]

(1,840,166,130 / 14,000) * 1471 Watt = 193348884 Watt =~ 193 Megawatt... This power draw would power the full bitcoin network
Megawatt-hour = Megawatt × hour
193 * (24*365) = 1.700.000 Megawatt-hour

1 MILLION 700 THOUSANDS MILLIONS OF WATTS PER HOUR ?

Read it yourself, because if that's correct, I think I'm far off... and the network is sucking up much more than what I have written.

Well... That's what i've written at the end of the calculation... I must be off somewhere... These numbers are way to high... I probably did a miscalculation or something, because this can't be right...
I'll re-read my calculations this evening (when i have some more time, currently at work)..

[NeuroticFish]

I don't know the numbers, but I also have the feeling that Bitcoin mining is quite power hungry on the big picture, mainly because of a ton of "hobby miners" who don't care of the power costs.
But when you figure out something close enough to the real numbers, please compare it with the power consumption of Vegas. I think that we are still much lower than that, so yeah.. we're fine.

[GODLIKE]

(1,840,166,130 / 14,000) * 1471 Watt = 193348884 Watt =~ 193 Megawatt... This power draw would power the full bitcoin network
Megawatt-hour = Megawatt × hour
193 * (24*365) = 1.700.000 Megawatt-hour

1 MILLION 700 THOUSANDS MILLIONS OF WATTS PER HOUR ?

Read it yourself, because if that's correct, I think I'm far off... and the network is sucking up much more than what I have written.

Well... That's what i've written at the end of the calculation... I must be off somewhere... These numbers are way to high... I probably did a miscalculation or something, because this can't be right...
I'll re-read my calculations this evening (when i have some more time, currently at work)..

OK I'm sorry I answered you real time.

[altcoinhosting]

(1,840,166,130 / 14,000) * 1471 Watt = 193348884 Watt =~ 193 Megawatt... This power draw would power the full bitcoin network
Megawatt-hour = Megawatt × hour
193 * (24*365) = 1.700.000 Megawatt-hour

1 MILLION 700 THOUSANDS MILLIONS OF WATTS PER HOUR ?

Read it yourself, because if that's correct, I think I'm far off... and the network is sucking up much more than what I have written.

Well... That's what i've written at the end of the calculation... I must be off somewhere... These numbers are way to high... I probably did a miscalculation or something, because this can't be right...
I'll re-read my calculations this evening (when i have some more time, currently at work)..

OK I'm sorry I answered you real time.

no problem... I was editing my post along the way
If i find an error in my calculation (or if somebody else does), i'll try to make a fresh post this evening, without editing my initial post, to keep everything clear

[Soros Shorts]

What I am wondiring is: is it possible that nobody hasn't yet noticed it???

People have already noticed it for years. There has been some arguments to replace proof-of-work mining with proof-of-stake mining, and power consumption has always been listed as one of the key advantages of PoS.

These discussions come up periodically, but lately they always gets buried by shit topics such as "What would you do if I gave you 765.43 bitcoins?"

[GODLIKE]

I don't know the numbers, but I also have the feeling that Bitcoin mining is quite power hungry on the big picture, mainly because of a ton of "hobby miners" who don't care of the power costs.
But when you figure out something close enough to the real numbers, please compare it with the power consumption of Vegas. I think that we are still much lower than that, so yeah.. we're fine.

The problem is: we are fine... NOW.
What when Bitcoin will value 100000$? (and that's what we are all waiting to happen and expect to happen...)

We will spend the same that is spent to feed 6 millions of houses for ONE MONTH to uphold the Bitcoin network for ONE DAY?
You see this can't be going anywhere.

[GODLIKE]

What I am wondiring is: is it possible that nobody hasn't yet noticed it???

People have already noticed it for years. There has been some arguments to replace proof-of-work mining with proof-of-stake mining, and power consumption has always been listed as one of the key advantages of PoS.

These discussions come up periodically, but lately they always gets buried by shit topics such as "What would you do if I gave you 765.43 bitcoins?"

Ahah and what would you do if I gave you 123.45 BTC?
Yeah Bitcoin is slowly entering mainstream and this is what happens...

I guess I'll have to dig better on the PoS paradigm then, because the last time I read about it, it seemed to me too much rewarding for people that own the currency, or to put it better, it become a bit similar to a Ponzi: the more you own, the more interests you have. And there's also inflation problems if I recall correctly.
However, once the electricity consumption will be solved, also the centralization problem will be solved.

[NorrisK]

You forget one important thing:

You assume linearity of power consumption and bitcoin price. This is not the case. The hashpower will become more efficient, so the power use will go down.

I agree with you that there is a problem with the amount of power that is used to mine bitcoin though. It just feels like a giant waste to only burn this much power for a chance at finding a bitcoin block.

[NeuroticFish]

I don't know the numbers, but I also have the feeling that Bitcoin mining is quite power hungry on the big picture, mainly because of a ton of "hobby miners" who don't care of the power costs.
But when you figure out something close enough to the real numbers, please compare it with the power consumption of Vegas. I think that we are still much lower than that, so yeah.. we're fine.

The problem is: we are fine... NOW.
What when Bitcoin will value 100000$? (and that's what we are all waiting to happen and expect to happen...)

We will spend the same that is spent to feed 6 millions of houses for ONE MONTH to uphold the Bitcoin network for ONE DAY?
You see this can't be going anywhere.

You missed the main point of my post. There are places / businesses much more power hungry than Bitcoin and with much less use than Bitcoin.
The power is not for free. If somebody is willing to pay for it, he will pay. Some will earn from it, some not.
Some will prefer to mine Bitcoin and pay for electricity, some will prefer to donate the money to the poor. If you want to donate to the poor instead of mining, it's your call.

Most of industries are not sustainable, and I'd start with the production of electricity.

[GODLIKE]

You forget one important thing:

You assume linearity of power consumption and bitcoin price. This is not the case. The hashpower will become more efficient, so the power use will go down.

I agree with you that there is a problem with the amount of power that is used to mine bitcoin though. It just feels like a giant waste to only burn this much power for a chance at finding a bitcoin block.

Somebody said (Antonopoulos I think) that we are already very near to the hardware wall.
The problem is not only Bitcoin's cost, but also the fact that this brings to centralization into countries with cheapest electricity.
I don't think this was forecasted by Satoshi Nakamoto when he released the software.
All this China centralization in mining is due to this effect.
We are just lucky that countries in Africa where electricity is even cheaper didn't start mining yet, because in those countries internet is even worse than in China...

[GODLIKE]

I don't know the numbers, but I also have the feeling that Bitcoin mining is quite power hungry on the big picture, mainly because of a ton of "hobby miners" who don't care of the power costs.
But when you figure out something close enough to the real numbers, please compare it with the power consumption of Vegas. I think that we are still much lower than that, so yeah.. we're fine.

The problem is: we are fine... NOW.
What when Bitcoin will value 100000$? (and that's what we are all waiting to happen and expect to happen...)

We will spend the same that is spent to feed 6 millions of houses for ONE MONTH to uphold the Bitcoin network for ONE DAY?
You see this can't be going anywhere.

You missed the main point of my post. There are places / businesses much more power hungry than Bitcoin and with much less use than Bitcoin.
The power is not for free. If somebody is willing to pay for it, he will pay. Some will earn from it, some not.
Some will prefer to mine Bitcoin and pay for electricity, some will prefer to donate the money to the poor. If you want to donate to the poor instead of mining, it's your call.

Most of industries are not sustainable, and I'd start with the production of electricity.

No no I got what you said, I only don't think Las Vegas is using in electricity in one hour anything near to what could feed 6 millions of houses for one month.
But that's exactly what is going to happen if Bitcoin will hit 100000$ value with the actual software.
And that's the whole point: Bitcoin is sustainable now, but it won't be sustainable ANYWHERE if it reaches, let's say, 50000$ of value per BTC.

[NeuroticFish]

No no I got what you said, I only don't think Las Vegas is using in electricity in one hour anything near to what could feed 6 millions of houses for one month.
But that's exactly what is going to happen if Bitcoin will hit 100000$ value with the actual software.
And that's the whole point: Bitcoin is sustainable now, but it won't be sustainable ANYWHERE if it reaches, let's say, 50000$ of value per BTC.

The mining hardware becomes better and better.
The mining is shifting more and more to the places with cheap electricity (and I know that this is extremely bad).
If BTC will ever reach such prices, the difficulty will rise so much the average Joe with an old miner will finally give up (which will be great for the power consumption and very bad for the not-so-decentralized-anymore network).
The power will become cleaner over time. If its price will rise, some businesses will stop mining too.
All in all, I expect the system to adjust itself and not get to the alarming numbers you expect.

[altcoinhosting]

No no I got what you said, I only don't think Las Vegas is using in electricity in one hour anything near to what could feed 6 millions of houses for one month.
But that's exactly what is going to happen if Bitcoin will hit 100000$ value with the actual software.
And that's the whole point: Bitcoin is sustainable now, but it won't be sustainable ANYWHERE if it reaches, let's say, 50000$ of value per BTC.

The mining hardware becomes better and better.
The mining is shifting more and more to the places with cheap electricity (and I know that this is extremely bad).
If BTC will ever reach such prices, the difficulty will rise so much the average Joe with an old miner will finally give up (which will be great for the power consumption and very bad for the not-so-decentralized-anymore network).
The power will become cleaner over time. If its price will rise, some businesses will stop mining too.
All in all, I expect the system to adjust itself and not get to the alarming numbers you expect.

I also think the system will adjust itself eventually...
That being said, https://coinmarketcap.com/ tells us that about 80 million USD worth of BTC is bought/sold per 24 hours... I don't know how much an average bank transacts in the same timeframe. Offcourse, this market cap has little to do with the ammount of BTC used to create outputs on a daily basis... But it's a start.
I do know an average bank has a big server room, a big network, ATM's, physical local bank offices filled with pc's, lights,...
Physical money is made out of paper, which has to be made from raw material, printed, transported,...

Instead of wondering how much power sustaining the BTC network consumes, it might be more interesting to compare FIAT to BTC regarding power consumption.

For example, how much power is used to circulate 80.000.000 * 365 USD (including printing, transporting, bank offices, ATM's,...). compared to sustaining the BTC network for a year...

I have no idear where to start such a calculation tough...

[DannyHamilton]

So if anyone can bring me additional variables to throw into the equation that could considerably lower the Bitcoin's network uphold cost, I'm all ears!

There are probably a few variables that you haven't included in your analysis.

The first one that comes to mind is the "halving" every four years.

You calculated:
"600*12.5=7500"

But in four years that 12.5 is going to be replaced with 6.25
Four years after that it will be replaced with 3.125
This will continue to reduce the mining revenue unless the bitcoin exchange rate doubles every four years.

I don't know that bitcoin will ever get to $100,000 per BTC, but if it takes 32 years for the exchange rate to reach that target, then the math would be:
100000*0.04882812 = 4882.81"

That's less miner revenue than today.

Another thing to consider is the law of supply and demand. There are costs associated with building additional power generation plants to support bitcoin mining, these costs would be reflected in higher electricity costs which would increase the cost of mining.  So the same amount of mining revenue in the future might not buy as much electricity as it does today.

I haven't looked too closely at the maths behind it recently, but I suspect that you've attributed too large a percentage of the revenue to electricity and not enough to equipment and other mining related expenses.

If we do altcoinhosting's maths again, we see that he's made a mistake:

Current network hashrate = 1,819,984,088,000,000,000 hash/s
Antminer S9 = 14,000,000,000,000 hash/s for 1471 Watt at the wall
1,819,984,088,000,000,000 / 14,000,000,000,000 = 130,000 antminer S9 units
130,000 * 1471 = 191,230,000 Watts
24 hours in a day
191,230,000 * 24 = 4,589,520,000 Watt*hours = 4,589,520 kWh per day
The average US electricity use is 911 kWh per month
4,589,520 / 911 = 5,037 households for 1 month

So, if I haven't made any errors in my maths, 1 day of Bitcoin mining currently uses as much electricity as about 5,000 average U.S. households do in 1 month.
(much less than the 170 million household calculated by altcoinhosting)

Your estimate of 36,000 households seems a bit high.  Clearly not everyone is running an antminer S9, but anyone mining with equipment half as efficient would have to have costs that are half as much in order to remain profitable.  As such, it seems unlikely that the daily electricity use is more than 10,000 households would use in a month.

I'm not sure where you got your $20 per month being the average cost of electricity for a household.  Here in the U.S. the average cost is somewhere around $0.11 per kWh. If we multiply that by the average usage of 911 kWh, it's about $100 per month.

So if mining electricity costs as much in a day as 5,037 average U.S. households spend on electricity in 1 month, then daily mining electricity costs should be about:

100 * 5037 = $503,700 per day
With 144 blocks per day, that's about $3500 per block spent on electricity (not the $600 that you estimated).

[franky1]

the biggest thing the OP is missing

4 years ago. someone spends $1600 on equipment. that has a power draw of ~1.4kwh = same costs, same electric...
but..... it only hashed at ~300 M/hash(4gpu used on 2mainboards with 2 700wPSU)
this was the olden days of GPU mining

today. someone spends $1600 on equipment. that has a power draw of ~1.4kwh = same costs, same electric...
but..... hashes at ~13 T/hash (1 s9 bitmain)
this was the current days of ASIC mining

in 4 years someone spends $1600 on equipment. that has a power draw of ~1.4kwh = same costs, same electric...
but..... probably hashes at XX P/hash (some crazy assed asic)