Will crypto mining kill polar bears?

Bitcoin mining uses as much electricity as a small country. Many people hate it for this reason, its one of the more popular arguments against crypto currencies. Will crypto mining kill polar bears? I think not. I think it will help save polar bears. “Bear” with me.

Germany produces a significant part of its electricity from renewable energy: wind and solar. As we all know, these sources are intermittent and seasonal, as is demand. When the share of renewable energy in the overall energy mix becomes large enough, the result is inevitable: temporary and seasonal overcapacity. This isnt just theoretical, energy prices in germany and the UK where effectively negative last Christmas: http://www.businessinsider.com/renewable-power-germany-negative-electricity-cost-2017-12//?r=AU&IR=T

As explained in the above article, this isnt a rare freak occurrence, its expected and this will have to be become much more common if as a society, we want to transition away from fossil fuels. Because to do that we need (much) more renewable energy sources. A study I saw for Germany calculated they needed at least 89% more capacity, just to handle peaks loads. But that also implies an incredible amount of overcapacity when demand isnt anywhere near peak, or when supply is above average due to favorable weather. Storing excess renewable electricity in most places is very expensive and inefficient. So much so that its rarely even done. This is a major problem. Wind turbines are therefore feathered, solar panels turned off, excess electricity dumped in giant electrical heaters, offered for free or even offered at negative prices. Renewable energy may have become cheaper than other forms per KWH, but thats only if when you can use all of the production. Building capacity that can only be used 50% or even 10% of the time, or building infrastructure to store surplus electricity is still very expensive.

I know what you’re thinking. Mining wont help here, because mining intermittently is something that seems crazy today; miners keep their expensive machines on 24/7. But thats only because today, the overall cost structure of a (bitcoin) miner is heavily titled towards writing off hardware. Particularly for anyone paying retail prices for mining asics. This will change completely, because of two related reasons:

1) mining efficiency improvements will taper off.

Mining asics have been progressing extremely rapidly, from FPGA’s to using 20 year old obsolete 130nm process technology in the first asics to modern 16nm chips now. This has resulted in at least a million fold improvement in efficiency in just a few years, which in turn lead to hardware investments that needed to be recovered in a few months or even weeks (!) before they were obsolete. Opportunity cost has been so high, that miners have literally chartered 747s to transport new mining equipment from the manufacturer in China to their mines in the US.

This cant and wont last. 12nm and 7nm asics are about to be produced, or are being produced now. It doesnt get better than that today, and it wont for many years to come. Moore’s law is often cited to show efficiency will keep going up. That may be true, but until now the giant leaps we have seen had nothing to do with moore’s law, which “only” predicts a doubling every 18 months. Moore’s law is also hitting a brick wall (you cant scale transistors smaller than atoms), and only states that transistor density increases. Not that chips become more efficient or faster. The end of transistor scaling is already becoming evident. Compared to a few years ago, CPUs have gained more cores (ie, more transistors), but a 4 year old highend cpu like a Haswell core i7 runs at comparable speeds, and comparable power efficiency with today’s chips. If it were useful for mining, it would still be a competitive chip, and probably remain so far a few more years at least.

What all this means is that these upcoming state of the art mining asics will remain competitive for many years, at least 3, possibly more than 5 years, and thus can be used and written off over that many years. But they will still consume electricity during all those years, shifting the overall costs from hardware to electricity.

2) Mining is still too profitable (for anyone making their own asics) and mining hardware is therefore still too expensive (for everyone else)

Miner hardware production rate simply hasnt yet been able to keep up with demand and soaring bitcoin prices. This leads to artificially low mining difficulty, making mining operationally profitable even with expensive electricity, and this also leads to exuberant hardware profit margins. You can see this easily, just look at the difficulty of bitcoin. When the price dropped by 70%, did you see a corresponding drop in difficulty? No, no drop at all, it just keeps growing exponentially. That only makes sense because we are not yet near saturation, or near marginal electricity costs for bitmain & Co. Its not worth it yet for them to turn off their miners. Its not even worth it yet for residential miners. Another piece of evidence for this, is bitmains estimated $4 billion profit. But mining is a zero sum game, over time, market forces will drive hardware prices and the mining itself to become only marginally profitable. We’re clearly not close to that -yet. You might think so as a private miner, but thats only because you overpaid for your hardware.

Lets look at todays situation to get an idea. An Antminer S9 retails for $2300 and uses ~1300W at the wall. If you write off the hardware over a year, electricity and hardware costs balance out at an electricity price of $0.2/KWH. Anything below that, and hardware becomes the major cost. But how will that evolve?

Developing an asic, even one as relatively simple as an SHA engine, costs a lot of money, particularly on an advanced process, but silicon wafers are cheaper than you might think. Based on some experience in the industry, my guess is the BOM for a miner like the S9 without PSU is below $200; depending on yields and binning, and the wafer prices they can negotiate with TSMC, it could be as low as $100. The fan might literally be the most expensive part. As difficulty keeps going up, demand for asics will eventually taper off, and market prices will head towards marginal costs. Let say a 1300W S9 equivalent at that point gets sold at $400 leaving bitmain a healthy profit margin; that would mean each year a miner would spend 5x more on electricity than hardware. Hardware will remain competitive for more than a single year though. Say you write it off over 3 years, now you’re spending 15x more on electricity than the hardware. Intermittent mining with free or virtually free electricity 50% of the time will become feasible long before that, hardware costs will become almost a moot point and mining even a few hours per day of a few days per week might actually make sense.

The result is that crypto mining will give green energy producers a way to efficiently monetise local, seasonal or intermittent excess energy production. Instead of paying people to use their excess electricity, they will be paid for it, and they will earn pretty much the global average electricity rates for it, as mining difficulty will adjust to around that level. That means investing in renewable energy is now much more lucrative, because you dont have to worry about utilisation rates or what to do with your excess production.

By now, I will hopefully have convinced you of the viability of mining with intermittent excess renewable energy; but its not only viable, it will become the only way to do it profitably. Renewable energy *at the source* is already cheaper than any carbon burning source. Even in Quatar, they install solar plants because its cheaper than burning their own gas. Its transporting and storing the electricity that usually is the problem. Gas can easily be transported and stored. Wind and solar energy can not. And thats a massive problem for the industry. But mining doesnt need either. You can mine pretty much anywhere and anytime. All you need besides electricity, is a few containers and an internet connection for a solar plant or wind farm to monetize excess energy.

Moreover, mining is a zero sum game, a race to the bottom. As long as its profitable for green energy providers to deploy more hardware (which will be true as long as they can at least recover their hardware investment), difficulty will go up. Until it becomes unprofitable for anyone who has to pay for his electricity. No one gives oil, coal or gas away for free, so anyone depending on those sources of electricity, can not remain competitive. If bitcoin price were to go up so much, that there isnt enough renewable electricity production in the world to accommodate the hashrate, bitcoin miners will simply install more solar and wind farms. Not because of their ecological awareness, but because it makes the most financial sense. And during peak demand periods, why wouldnt they turn off the miners and sell their electricity to the grid for a premium?

Basically crypto mining would fund renewable energy development, and solve the exact problem laid out in the article linked above: provide **over**capacity of renewable energy to handle grid peak loads, without needing any government funding or taxation on carbon based sources, without needing expensive and very inefficient energy storage. The blockchain becomes a (financial) battery that is ~100% efficient at converting energy and that finances itself.

TL:DR, deploying more renewable electricity overcapacity is both very expensive and very necessary if we want to save polar bears. Financing for these large scale green energy projects will either have to come from tax payer money to store or subsidise the largely unused excess electricity, or it will come from crypto mining. Market forces will drive crypto mining to use the cheapest energy. Renewable energy already is cheaper per KWH than carbon based power, and nothing is cheaper than excess and thus free (or negative value) renewable energy. Bitcoin mining’s carbon foot print will therefore become ~zero. If you take in to account the effect of financing and subsidizing large scale renewable energy development that can also be used to supply the grid during peak demand periods, its carbon footprint will be hugely negative.

BTW, if you wonder what Blockchains LLC is going to do with 61K acres near Tesla’s factory; my guess is solar plants and crypto mining. Expect to see renewable energy development and crypto mining to merge in to one single industry. Check out [envion](https://www.envion.org/en/) to get a glimpse of this future. Im not endorsing their token as an investment, I havent researched it at all, but the market they are going after is a very real one and its about to explode.

View Source

View Reddit by Vertigo722