For the first part of this post go here.
As we discussed earlier, Bitcoin is not everyone’s golden panacea for anonymous, decentralized online payment.
In other words, it consumes a lot of energy.
That did come out rather unexpected.
For some, Bitcoin eats up a bit too much energy.
But how much exactly?
By some generous estimates, about three thousands times more than a payment made through the traditional credit card system.
We discussed, in the previous post, about the data points we would need to truly gauge how much energy does Bitcoin consume.
The first of those data points was the theoretical maximum number of transactions that Bitcoin, as a system, could manage.
Right now, that theoretical number is about seven online transactions every second.
Of course, this number could change with newer implementations of Bitcoin.
But for our purposes here, we’ll continue to go with the number seven.
You might ask, what is that number seven again?
Seven is basically the number of transactions Bitcoin can handle every second.
Moreover, the average number of Bitcoin transactions that take place every day is 302,150.
A fact that we mentioned in the previous post as well.
Do take note that this 302,150 comes directory from the authority site blockchain.info.
Now we’ll move on to some of the other data points that we need to measure Bitcoin’s consumption power.
The second data point is hashrate.
More specifically, bitcoin hashrate.
This number indicated, or rather measures, the real computation power of the Bitcoin network.
There are many ways to go about calculating this number.
But the most convenient one is the go forward with the number that is based on weekly averages.
This makes sure that the daily noise factor is smoothened out without causing many problems.
Right now, the latest statistics (true as of March 1) put this weekly average to about 3.3 million terahashes every second.
That is, 3.3 million terahashes/s.
Simple enough right?
Moving forward then.
We almost know everything there is to know if we want to calculate Bitcoin’s consumption potential.
Now all that is left is to measure the efficiency of the miners that play a big role in the overall Bitcoin network.
Luckily, we now know the power consumption of Bitcoin miners as well.
But to make the comparison fair, let’s take the power consumption of some of the most efficient Bitcoin miners.
That is, the most efficient Bitcoin miners on the market.
If we look at the numbers, it seems like Antminer and BitFury are the top dogs with a huge lead.
Antminer S9 eats about .098 W/GH/s while BitFury is about two times more efficient than Antminer.
That’s because BitFury uses the more efficient data center that is liquid cooled.
Of course, these numbers don’t represent the general mining community.
So in order to calculate a fairer lower bound for the total amount of network energy consumed, we should do the following.
First, we should go ahead and multiply the Bitcoin miner power consumption per hash per second by another special number.
This special number should be the global hashrate per second.
Again, we’ll go by the numbers that are available as of March 1, 2017.
Do the calculations and you should come up with a number that closely resembles 332 megawatts of constant energy consumption.
To be fair to the Bitcoin community, this number doesn’t seem like an awfully large number.
It is actually pretty reasonable considering other reports about Bitcoin’s energy consumption numbers.
Moreover, you also have to take into account the newly announced plans from a particular Bitcoin entrepreneur.
We won’t go into the details of those plans here.
But just for information’s sake, the plans consist of setting up a 130 megawatt Bitcoin mining facility in China of all places.
You can read more about that here.
Since most people won’t be able to relate to a number like 332 megawatts of constant energy consumption, let’s express that in another way.
A way that will hopefully be more relatable to everyone reading this post.
More Easily Understood Bitcoin Energy Consumption Number
Let’s assume that all Bitcoin miners were great and ethical people or businesses.
In other words, let’s assume that all of them ran computer hardware that was extremely efficient.
Not just efficient.
But extremely efficient.
If we consider that case to be true, then the overall Bitcoin power consumption would come down.
By a bit.
In other words, not a lot.
Regardless, if we take the above-mentioned assumption then the total Bitcoin power consumption number would come at slightly lower than 332 megawatts.
To put it another way, Bitcoin would then consume power that could provide for the daily needs of around 268,990 average American citizen homes.
As you can probably see, even when we have considered Bitcoin miners to have a very efficient hardware, the consumption number is still huge.
It is massive.
But we’ll get to that part a bit later.
Now keep that number 302,150 in mind.
This is the average weekly number of Bitcoin transactions.
What this basically means is that each Bitcoin transaction represents about 26 kWh of electricity.
Most of this electricity is spent during Bitcoin mining.
As we have mentioned before as well, the average electricity required to power an American household for a period of 24 hours is slightly above 26 kWh.
Okay. The exact number is about 29.2 kWh.
More Bad News Regarding Bitcoin Energy Consumption
So if each American household requires an average of 29 kWh.
And each Bitcoin transaction eats up 26 kWh then that means that each Bitcoin transaction represents about 89 percent of an average American household energy needs.
That is bad right?
But apparently, the situation was even worse before.
Back in the year 2015, Bitcoin consumed about 157 percent of an average American household’s electricity requirements.
What’s the main take here?
The main takeaway here is that it is true that Bitcoin miners have improved their hardware.
In terms of efficiency and consumption of power, Bitcoin miners have become better miners.
Right now, Bitcoin miners can perform much more computation processes for each unit of power than before.
Moreover, compared to 2015, the number of Bitcoin transactions per day have also increased.
So Bitcoin should be more sustainable now, right?
Or at least the energy situation with Bitcoin is getting better by the year, right?
Nothing to worry about right?
More like not necessarily.
Just because Bitcoin has become more efficient per transaction doesn’t mean it is a sustainable payment system.
There are many reasons.
First Reason Why Bitcoin May Not Be A Sustainable Payment Method Just Yet
As far as our analysis goes, the first reason is simply our assumption.
Or rather our huge assumption.
Our assumption that all Bitcoin miners are ethical enough to be running the most efficient of computer machines.
That simply cannot be true.
Take a look at this post by Marc Bevand who works as an engineer.
The post says Bitcoin transactions are not wasteful or unsustainable.
Let’s leave that aside for a minute or two.
The more interesting part is that Marc Bevand uses or rather assumes a different figure for global Bitcoin energy consumption.
He uses a figure of .15 J/GH.
Without going into too many details, this figure would add an instant fifty percent to the amount of electricity consumed for every Bitcoin transaction.
Taking his figures, the Bitcoin energy consumption would go up to 134 percent of an average American household’s daily energy needs.
Nobody Knows The Exact Number On How Much Energy Bitcoin Eats Up.
Which brings us back to our original point.
Right now, no one really knows how much energy each Bitcoin transaction consumes.
But one thing should be clear by now.
That consumption number cannot be low regardless of the fact that no one knows the exact number.
That also holds true for the real efficiency number.
No one truly knows that either.
But the 0.15 J/GH quoted by Marc might not be useful because of it having gone out of date.
So we’ll go ahead and stick with the newer 0.098 J/GH for our analysis.
But here is another problem.
Some Bitcoin Experts Have A Different Idea About Bitcoin Energy Consumption
Mostly, they think that the overall Bitcoin energy consumption is not majorly affected by Bitcoin miner efficiency.
In other words, Bitcoin energy consumption is not affected by Bitcoin miner efficiency in a significant way.
Here is what Ittay Eyal, who works as an assistant director of the Initiative for Cryptocurrencies and Contracts (and is also a Cornell University computer scientist) told a reporter at Motherboard.
He said via email that Bitcoin’s energy usage was essentially a function of the currency’s exchange rate in the market and of course, the block reward.
If that didn’t make any sense to you then hang on for a second.
What that really means is that it is inevitable that the price of Bitcoin will go up sometime in the future.
To counter for that increase in price, Bitcoin miners will not stop from adding more computer power to their systems.
In fact, they will add any amount that is necessary in order to restore the balance in favor of profitability.
That is the only way they can afford to keep their companies alive in the Bitcoin mining industry.
Of course, they earn rewards for their services but that doesn’t mean they want more profits as well.
That’s Not The Whole Equation Though
See, experts in the Bitcoin community know this fact.
Bitcoin miners will continue to add more computing power as the prices go up.
And hence we have the Bitcoin Electricity Consumption Index (BECI).
Right off the bat, let’s ask the questions, what is it and what does it do?
The Bitcoin Electricity Consumption Index is actually an online product from Alex De Vries who works for Digiconomist.
More specifically, though, it is a real-time modeling tool.
Now for the ‘what does it do” part.
The main function of Bitcoin Electricity Consumption Index is to gather figures related to Bitcoin on a daily basis.
To accomplish this along with other tasks it makes use of a plausible set of assumptions.
These assumptions are then used to estimate the amount of electricity that could be used by the whole network.
So what is the whole network in this situation?
Well, the whole network consists of many players.
But the major ones are the Bitcoin miners.
The Bitcoin Electricity Consumption tries to estimate the amount of electricity that would be sufficient and viable (economically speaking) so that the whole Bitcoin network could consume and still be able to cover Bitcoin miners’ costs related to electricity consumption.
Of course, the actual process is a bit more complex than that but you get the idea.
Perhaps you should also know that most of the data that is collected for Digiconomist’s index is collected through automated means.
Additionally, the figures related to the output are also recalculated on a daily basis.
Why Should These Values Be Calculated On A Daily Basis
The logic behind the daily recalculation is as sound as it gets.
What we mean by that is the, there should be a system in place that can account for price rise as well as price drop.
Needless to say, the estimated cost of power consumption on part of the Bitcoin network should also be recalculated as the price fluctuates.
How Is De Vries Calculation Different From The One We Have Mentioned Here?
It should be stated that all of our calculations are simple in nature.
And while our analysis is useful to find the lower bound for the total cost, it isn’t a comprehensive future.
In other words, economic incentives drive the market to a much larger extent.
Bitcoin miners will always try to approach the break-even point.
The break-even point is the point where the Bitcoin price is exactly equal to the marginal cost it takes to produce a new Bitcoin.
Makes sense right?
To make more sense of the whole equation take a look at this 2015 Hayes’s paper.
The paper is titled ‘A cost of production model for Bitcoin’.
You don’t need to read the whole paper to figure out that Haye’s model does have a shortcoming.
In other words, it is not always accurate in terms of predicting the current energy consumption.
With that said, there is no doubt that the paper has its merits.
It can provide reasonably accurate averages of energy consumption in the long run.
And that is in itself enough to justify his work.
We’ll talk more about how more accurate estimates can be made and what effect with Bitcoin have in the future of energy consumption on a global scale.
Stay tuned to securitygladiators.com
3 thoughts on “Bitcoin Energy Usage: Still A Lot”
I’ve seen estimates of kWh/transaction (averages) with ranges from 26 to over 90. Something does not add up. since the cost of electricity for a kWh is usually more than 5 US cents and often much more. I don’t know where most miners are located, but in places where they are how does this translate into actual dollar (or whatever currency) costs?
You linked to my blog, and I wanted you to refer you to a more recent research & estimation I worked on: https://blog.zorinaq.com/bitcoin-electricity-consumption/
Also, FYI the index by Digiconomist (BECI) is flawed and overestimates the consumption by most likely 2× to 2.5×: https://blog.zorinaq.com/serious-faults-in-beci/
As of November 24th it is over 10 GW of power consumption and fast approaching 20. That’s because the equilibrium energy consumption is:
mining reward/block time * (the inverse of the cheapest cost of electricity)
On Nov. 24th, 2017 that would be $350,000 reward, 10 minute block time or 600 seconds, 8 cents a kwh = 8 cents/3600000J.
So the equilibrium would be ($350000/600s)*3600000J/$0.08 = 26.25 GW of power. It will take 2 months for mining equipment manufacturers to catch up to the latest price rise so if the mining reward stays the same expect mining to consume 26GW of power by February 2018.