There Is Nothing Virtual About Bitcoin’s Energy Appetite

In the virtual currency world this creation process is called “mining.” There is no physical digging, since bitcoins are purely digital. But the computer power needed to create each digital token consumes at least as much electricity as the average American household burns through in two years, according to figures from Morgan Stanley and Alex de Vries, an economist who tracks energy use in the industry.

The total network of computers plugged into the bitcoin network consumes as much energy each day as some medium-size countries — which country depends on whose estimates you believe. And the network supporting Ethereum, the second-most valuable virtual currency, gobbles up another country’s worth of electricity each day.

The energy consumption of these systems has risen as the prices of virtual currencies have skyrocketed, leading to a vigorous debate among Bitcoin and Ethereum enthusiasts about burning so much electricity.

The creator of Ethereum, Vitalik Buterin, is leading an experiment with a more energy-efficient way to create tokens, in part because of his concern about the effect that the network’s electricity use could have on global warming.

“I would personally feel very unhappy if my main contribution to the world was adding Cyprus’ worth of electricity consumption to global warming,” Buterin said in an interview.

But many virtual currency aficionados argue that the energy consumption is worth it for the grander cause of securing the Bitcoin and Ethereum networks and making a new kind of financial infrastructure, free from the meddling of banks or governments.

“The electricity usage is really essential,” said Peter Van Valkenburgh, the director of research at Coin Center, a group that advocates for virtual technology. “Because of the costs, we know the only people participating are serious, that they are economically invested. That creates the incentives for cooperation.”

This dispute has its foundations in the complex systems that produce tokens like bitcoin; ether, the currency on the Ethereum network; and many other new virtual currencies.

All of the computers trying to mine tokens are in a computational race, trying to find a particular, somewhat random answer to a math algorithm. But the algorithm is so complicated that the only way to find the desired answer is to make lots of different guesses. The more guesses a computer makes, the better its chances of winning. But each time the computers try new guesses, they use computational power and electricity.

The process essentially encourages people to use lots of fast computers, and lots of electricity, to find the right answer and unlock the new bitcoins that are distributed every 10 minutes or so.

This process was defined by the original bitcoin software, released in 2009. The goal was to distribute new coins to people on the Bitcoin network without a central institution handing out the money.

Early on, it was possible to win the contest with just a laptop computer. But the rules of the network dictate that as more computers join in the race, the algorithm automatically adjusts to get harder, requiring anyone who wants to compete to use more computers and more electricity.

These days, the 12.5 bitcoins that are handed out every 10 minutes or so are worth about $145,000, so people have been willing to invest astronomical sums to participate in this race, which has in turn made the race harder. This explains why there are now enormous server farms around the world dedicated to mining bitcoins.

This process is central to bitcoin’s existence because in the process of mining, all the computers are also serving as accountants for the Bitcoin network. The algorithm the computers solve requires them to also keep track of all the new transactions coming onto the network.

The mining race is meant to be hard so that no one can dominate the accounting and fudge the records. In the 2008 paper that first described bitcoin, the mysterious creator of the virtual currency, Satoshi Nakamoto, wrote that the system was designed to thwart a “greedy attacker” who might want to alter the records and “defraud people by stealing back his payments.” Because of the mining and accounting rules, the attacker “ought to find it more profitable to play by the rules.”

The rules have kept attackers at bay in the nine years since the network got going. Without this process, most computer scientists agree, bitcoin would not work.

But there is disagreement over the real value of bitcoin and the network that supports it.

For people who consider bitcoin nothing more than a speculative bubble — or a speculative bubble that has enabled online drug sales and ransom payments — any new contribution toward global warming is probably not worth it.

But bitcoin aficionados counter that it has allowed for the creation of the first financial network with no government or company in charge. In countries like Zimbabwe and Argentina, bitcoin has sometimes provided a more stable place to park money than the local currency. And in countries with more stable economies, bitcoin has led to a flurry of new investments, jobs and startup companies.

“Labeling bitcoin mining as a ‘waste’ is a failure to look at the big picture,” Marc Bevand, a miner and analyst, wrote on his blog. The jobs alone, he added, “are a direct, measurable and positive impact that bitcoin already made on the economy.”

But even some people who are interested in all that innovation have worried about the enormous electrical use.

De Vries, who keeps track of the use on the site Digiconomist, estimated that each bitcoin transaction currently required 80,000 times more electricity to process than each Visa credit card transaction, for example.

“Visa is more centralized,” de Vries said. “If you really distrust the financial system, maybe that is unattractive. But is that difference really worth the additional energy cost? I think for most people that is probably not worth the case.”

The figures published by de Vries have been criticized by Bevand and other bitcoin fans, who say they overstate the energy costs by a factor of about three. Many critics add that producing and securing physical money and gold also require lots of energy, in some cases as much as or more than bitcoin uses.

Van Valkenburgh, of the Coin Center, has argued that bitcoin miners, who can do the work anywhere, have an incentive to situate themselves near cheap, often green energy sources, especially now that coal-guzzling China appears to be exiting the mining business. Several mining companies have opened server farms near geothermal energy in Iceland and hydroelectric power in Washington state.

But the concerns about electricity use have still hit home with many in the industry. The virtual currencies known as Ripple and Stellar, which were created after bitcoin, were designed not to require electrically demanding mining.

Perhaps the biggest change could come from the new mining process proposed by Buterin for Ethereum, a process that some smaller currencies are already using. Known as “proof of stake,” it distributes new coins to people who are able to prove their ownership of existing coins — their stake in the system.

The current method, which relies so heavily on computational power, is called “proof of work.” Under that method, the accounts and people who get new coins do not need existing tokens. They just need lots of computers to take part in the computational race.

Energy concerns are not the only factor encouraging the move. Buterin also believes that the new method, which is likely to be rolled out over the next year, will allow for a less centralized network of computers overseeing the system.

But it is far from clear that the method will be as secure as the one used by Bitcoin. Buterin has been fiercely attacked by bitcoin advocates, who say his proposal will lose the qualities that make virtual currencies valuable.

Van Valkenburgh said that for now, throwing lots of computing power into the mix — and the electricity that it burns — was the only proven solution to the problems bitcoin solves.

“At the moment, if you want robust security, you need proof of work,” he said.