Offgrid power options for crypto mining

ggse000 · November 2017

Hi all, I am software professional and recently has been doing a lot of mining. There is one thing about world's largest cryptomining is that they waste lot of energy and so I am sort of pulled apart: I am for cryptocoin because of its decentralized currency nature but against it because it wastes some much energy. So I am kinda wondering about possibility of off-grid charging system for cryptomining rigs.
Basically the idea is setup large number of computer systems with Graphics cards can possible consume about 1-1.5kw and earn cryptocoins. The typical configuration of such systems are usually 1 personal computer with 4 or so graphics processing unit. However what I lack experience deeply is whether several of these rigs (or possible large number of it) can be ran off-grid on renewable energy.
What I visualize is build small scale solar or wind power system but also batteries, that is much larger than home-based but less than industrial one. Home based is too small for this sort of project and industrial systems with several MWs are way beyond my financial ability. There are lot of products out there, but what I see is those are mostly home and usually include 2-4 batteries and that number of solar panels. I am visualizing for this project is that the solar installation as well as its batteries can scale and pool its stored charge to power several mining rigs.

Another concern is that land usage and laws related to it. I dont think it is possible in metro area, so if it is possible to buy a land toward central valley with cell reception providing internet connectivity, then it should be possible to build a solar-powered mining rig.

I am even open to collaborate and start venture with the ones who are interested!! I supply the computer science expertise with someone who has expertise building something like I mentioned above. Of course, the big part will be the technical feasibility of such system.
Thanks!

mcgivor · November 2017

For any proposed system, an analysis of investment versus return must be conducted, how longl the solar system will pay itself off. Although it's often perceived solar energy is free, there has to be investment into equipment, the return on investment depends largely on the cost of grid energy, using any form of battery storage will cost more than grid power. Grid tied would be the only way to actually get ahead of the game, using the grid as a battery, feeding into it at a greater rate then consumption during the day to offset night time consumption. The hours of productive input are limited so the ratio of input would have to be calculated depending on location, let's assume 4 hours per day, the PV would have to produce 6 times the amount consumed just to break even. This is obviously a hypothetical calculation, not every day is sunny, but it demonstrates, to some degree what is required. The feed in tariff is another consideration, if the amount is, for arguments sakes, 50% of the purchase price per Kwh, then the ratio doubles to 12 times input over consumption. So unless the mining is very profitable the return on investment could be extensive, considering off grid is almost impossible, unless the cost of energy is extremely high, say in the neighborhood of $3 US per Kwh, then there is the maintenance and equipment replacement costs to consider.

Most off grid users do not have the luxury of the grid, so needs take preference over wants, because of the expense, operating an energy hungry business, has no business being off grid. Opinions and thoughts YMMV, but probably not.

Estragon · November 2017

+1 on @mcgivor comments.

IMHO, a better way to approach it would be to look for a location with relatively cheap renewable grid energy. For example, Manitoba and Quebec in Canada generate mainly from Hydro, and are reasonably priced relative to many other N.A. locations. As a bonus, much of the year provides free cooling, which can be a big part of electrical load in server farms.

I have to wonder about the sustainability of crypto-currencies as presently implemented though. Was really surprised when listening to an interview with a researcher recently who quoted some numbers for power consumption used in verifying transactions (over and above that for initial mining). Pretty mind boggling if crypto got anything like mainstream.

WaterWheel · November 2017

I had a friend ask me the same question a few days ago. At first he wanted wind power and I suggested solar but I didn't conceder it economically viable.

ggse000 · November 2017

Thank you for precious inputs. I am still looking around for an alternative. Here in the california, small home based electricity rate is over the roof once certain threshold pass it doubles the rate. Not only I am interested in cheaper energy but also renewable energy to power if not 100% then large share of the power used in mining. The places with cheap energy coupled with mostly from renewable makes it very ideal location, quick googling shows up around >7c / KW. Another attractiveness is that large effort and costs involving and designing your own system is relieved by small rate.

In my home country (the asia) the electricity rate is also cheap but the fact it is almost powered from close to 100% on coal makes it less ideal although economically. I will keep it posted!!

Now lets say if I decide to move to places like Manitoba then how about i go for renting the electricity. I dont think typical home will provide ample enough power, there must be some sort of site that can provide large amperage.

ggse000 · November 2017

mcgivor said:

For any proposed system, an analysis of investment versus return must be conducted, how longl the solar system will pay itself off. Although it's often perceived solar energy is free, there has to be investment into equipment, the return on investment depends largely on the cost of grid energy, using any form of battery storage will cost more than grid power. Grid tied would be the only way to actually get ahead of the game, using the grid as a battery, feeding into it at a greater rate then consumption during the day to offset night time consumption. The hours of productive input are limited so the ratio of input would have to be calculated depending on location, let's assume 4 hours per day, the PV would have to produce 6 times the amount consumed just to break even. This is obviously a hypothetical calculation, not every day is sunny, but it demonstrates, to some degree what is required. The feed in tariff is another consideration, if the amount is, for arguments sakes, 50% of the purchase price per Kwh, then the ratio doubles to 12 times input over consumption. So unless the mining is very profitable the return on investment could be extensive, considering off grid is almost impossible, unless the cost of energy is extremely high, say in the neighborhood of $3 US per Kwh, then there is the maintenance and equipment replacement costs to consider.

Most off grid users do not have the luxury of the grid, so needs take preference over wants, because of the expense, operating an energy hungry business, has no business being off grid. Opinions and thoughts YMMV, but probably not.

I think plugging into grid will make it complicated project. Perhaps better way is to start with just grid power and then gradually move to renewable energy to offset some of the energy costs. I think my fear is confirmed that going off grid will be technically close to infeasible due to not be able to collect and store enough energy.
As far as the mining revenue, it is also a moving target, lot of things has to come under one place to get going. But I think I can fund enough to start off with a small to small-medium project and extend if it goes well.

Estragon · November 2017

> @ggse000 said:
>
> Now lets say if I decide to move to places like Manitoba then how about i go for renting the electricity. I dont think typical home will provide ample enough power, there must be some sort of site that can provide large amperage.

Winnipeg has lots of commercial and industrial space with quite large power supplies.

If you try doing something like this in a residential space there's a decent chance you'd get flagged as an illegal grow operation.

cupcake · November 2017

crypto-currency=killowatthour-currency

.

ggse000 · November 2017

cupcake said:

crypto-currency=killowatthour-currency
Yes, the very reason wanna doing something about it.

.

myocardia · November 2017

ggse000 said:

Yes, the very reason wanna doing something about it.

.

Unfortunately for you, there is no way whatsoever, besides moving to somewhere where the electricity is cheap, to mine crypto-currency on multiple rigs for not too much moola in electric bills. Your computers/mining rigs each use 30-35,000 watt hours/30kWhr-35kWhr per day, assuming they have 4 fast GPUs each, and use an efficient Bronze or better ATX power supply. That is roughly 10x as much power as the average small to medium-sized offgrid home uses per day, and that is for each of your rigs. Each rig would require you to not only produce, but to also store somewhere in the neighborhood of 60,000 watt hours per day of solar or wind, even during the winter when there is comparatively little solar available. I'm sure it would cost you upwards of $100,000 per rig (installed price, not cost of equipment) to be able to run them offgrid reliably. And then, even after that initial outlay, you then have to start planning ahead/saving for the cost of battery, charge controller, and inverter replacements, since all of those things wear out.

mike95490 · November 2017

FPG arrays are better ways then general purpose GPU & CPU's (for mining)

pdh · November 2017

@ggse000 -- I think you may have a basic misconception, thinking that the cost of off-grid solar power is less per kWH than grid power. That's not true -- I think that's what most of the comments here are saying (in more detail).

ggse000 · November 2017

Ок, i think i have gone too optimistic amd slowed view on the off grid option. Thanks for clarifying, it was hard realization but enlightening.

ggse000 · November 2017

How much of a engineering and technical challenge to do some solar installation to plug into grid and yourself? Is it better than buy a OOB (out-of-box) unit or can it be done in DYI environment to save cost?
Actually, when I pursued my EE degree back in college ~15 years ago, we did have lab environment where mech-to-electrical motor was plugged into grid and vaguely remember all sort fo phase etc. etc which I barely remember. Now I am all software guy, the only thing I remember is V=IR, oh yes and P=VI too.

WaterWheel · December 2017

With what little I know of crypto mining most of the popular "coins" have risen their level of difficulty so high that some of them are barely worth mining unless you live in Iceland where power is cheap and cooling requirements are lower.    And if they make the algorithms much harder that will be the only choice for "mining".

I'd suggest just buying coins if that's what you want to do.    One more level of difficulty and even the S9 won't be profitable in most of the US with our power costs.

I'm dabbling with Genesis Mining with some Alt-Coins on a 2 year contract.    It looks like recouping original costs will take about 43 weeks. Profit will depend on coin value and difficulty levels.

ggse000 · December 2017

I have scoured around US States and averaging about 10-15c per kw. I am also considering home country there is a 4c and less. But my ultimate though is if i manage to get some sizable revenue, wanna spend big chunk of it in a renewable energy integration. It is a two order but granduous idea than simply just mining for profit. I also formulating an idea of smart mining software with full blown features something are like: sophisticated utility rate input, controlling and scheduling of mining rigs, and selection of multiple crypto coins based on difficulty and other factor andwith am these factors, make an automated decisions to maximize revenue or minimize wasted efforts etc. I already started laying groundwork for it.

ggse000 · December 2017

> @WaterWheel said:
> With what little I know of crypto mining most of the popular "coins" have risen their level of difficulty so high that some of them are barely worth mining unless you live in Iceland where power is cheap and cooling requirements are lower.    And if they make the algorithms much harder that will be the only choice for "mining".
>
> I'd suggest just buying coins if that's what you want to do.    One more level of difficulty and even the S9 won't be profitable in most of the US with our power costs.
>
> I'm dabbling with Genesis Mining with some Alt-Coins on a 2 year contract.    It looks like recouping original costs will take about 43 weeks. Profit will depend on coin value and difficulty levels.

Yes currently i am earning about 3$ coin per single day with most powerful gpu gtx1080 in a span of day. It is close to worthless.

ggse000 · December 2017

Now continuing this, I am putting less emphasis on how much worth $ of coin I earned today vs. how much I am spending on eletricity. There are 100s of coins on the market now and many of them are one way or better than bitcoin, each one of them has unique feature(s) that solves the shortcomings of original Bitcoin. This means that sooner or later Bitcoin will possibly lose its dominant position in the market and some other "better" coins will accent to take place of bitcoin. Not to say bitcoin sucks, what I think is even though if bitcoin lose its importance, I think it will still be remembered as the most iconic, representative and classic symbol of crypto. Now regarding othercoins there are mining possibilities exists and if your mines earned today might not worth much in few years time it can possibly multiple by order of hundreds.

ggse000 said:

> @WaterWheel said:
> With what little I know of crypto mining most of the popular "coins" have risen their level of difficulty so high that some of them are barely worth mining unless you live in Iceland where power is cheap and cooling requirements are lower.    And if they make the algorithms much harder that will be the only choice for "mining".
>
> I'd suggest just buying coins if that's what you want to do.    One more level of difficulty and even the S9 won't be profitable in most of the US with our power costs.
>
> I'm dabbling with Genesis Mining with some Alt-Coins on a 2 year contract.    It looks like recouping original costs will take about 43 weeks. Profit will depend on coin value and difficulty levels.

Yes currently i am earning about 3$ coin per single day with most powerful gpu gtx1080 in a span of day. It is close to worthless.

ggse000 · December 2017

pdh said:

@ggse000 -- I think you may have a basic misconception, thinking that the cost of off-grid solar power is less per kWH than grid power. That's not true -- I think that's what most of the comments here are saying (in more detail).

It is a relative isn it? From what I understand upfront cost is very high on solar equipment that is why companies trying to make a financial program to spread out the costs over the long term. Now on the long term since the energy from sun is being used, it should pay off but then I do realize there are recurring costs in regards to equipment i.e. inverter, solar panels and depreciating assets like that.

Estragon · December 2017

For cost modelling small scale solar pv reasonable replacement cycle time numbers to use would be; 5yrs for batteries, 10yrs for electronics, and 25yrs for pv, racking, wire, etc.

There are very few places where battery-based solar will come close to or below grid price, even in the long term. In order to make sense long term, you would have to make some really big assumptions about the path of future grid rates.

ggse000 · December 2017

Estragon said:

For cost modelling small scale solar pv reasonable replacement cycle time numbers to use would be; 5yrs for batteries, 10yrs for electronics, and 25yrs for pv, racking, wire, etc.

There are very few places where battery-based solar will come close to or below grid price, even in the long term. In order to make sense long term, you would have to make some really big assumptions about the path of future grid rates.

It looks like something like a tall order for now, but I will definitely be watching over it. Does battery in operation deplete and charge fully? From what I understand EVs only do up to certain amount of depletion and full (never fully charge and fully deplete) and that makes it last longer.

mcgivor · December 2017

There is no way any battery storage system will work in your application and save money, it simply can't be done. EV use lithium technology, they are more efficient, able to deep cycle, would last longer with shallower depth of discharge, as would any battery, best not to fully charge and also way more expensive. If grid tied feeding into the grid is not an option, then a zero injection grid tied arrangement would offset useage during the day, but return on investment would be extensive.

ggse000 · December 2017

Estragon said:

> @ggse000 said:
>
> Now lets say if I decide to move to places like Manitoba then how about i go for renting the electricity. I dont think typical home will provide ample enough power, there must be some sort of site that can provide large amperage.

Winnipeg has lots of commercial and industrial space with quite large power supplies.

If you try doing something like this in a residential space there's a decent chance you'd get flagged as an illegal grow operation.

Estragon, do you think you can give me a more lead on this? WHen i check the winnipeg hydro electric rates, it is extremely competitive. Please feel free to PM me, thanks!

Estragon · December 2017

Not sure what other info you're looking for.

ggse000 · December 2017

I am curious about what the unit price of powerpack battery? I looked up everywhere on the internet, absolutely no info.
Is there a computing product from lesser known manufacturer of similar nature?

Photowhit · December 2017

ggse000 said:

I am curious about what the unit price of powerpack battery? I looked up everywhere on the internet, absolutely no info.
Is there a computing product from lesser known manufacturer of similar nature?

I missed where a 'powerpack battery' was mentioned. But as a long time off grid solar person (15+ years now) I have posted that with self installed system and serious bargain shopping I was close to grid parity for my expensive electricity. Forecasting a cost of 26 cents a Kwh over 30 year life of my system, including replacing a battery at 15 years and electronics every 10 years...

Many question my cost figures and it's not uncommon for off grid electric to be quoted at roughly $1 a Kwh here.

Just to reach the figure of 26 cents a Kwh, I use no backup generator, and my heaviest loads fluctuate with the amount of sun available, since I use air conditioning. The per Kwh cost if you wanted to do a steady 16Kwhs a day through out the year would be staggering, likely well over $1 a Kwh, and require a backup generator. Perhaps you could figure something out but you will be looking for a very unique situation. a Place where you can get a T1 line, and put in a cave home, with geothermal pool water for cooling. Some where high and south facing exposure, with very sunny year round climate.

By the time you get everything setup, the market will have a 25% correction when the FED and other currency banks start calling in some of the 20 Trillion dollars they have printed flooding the economy with cheap money.... I don't like the long game in Crypto currency, but many have made some coin...

Roner · December 2017

@ggse000 I %100 believe in your vision as this is something I would like to do myself and am currently doing research on how to make this profitable without adding to our environmental problems. I have to admit though I am neither an engineer or a programmer, but a business grad that loves challenges and most of all, profit. I truly believe blockchain and digital currency will be used long term but I think the average person is quickly going to be priced out of any mining at all, much less on large scale (this might already be the case). I also believe that renewable and sustainable energy is the only responsible way to go forward and if we aren't there yet as far as our technology some of these amazing profits people are making need to be spurring the research.

I'd love to help any way I could, or at least collaborate to some degree. Either way, if you break down this barrier I'll be waiting for a how-to guide!

Photowhit · December 2017

First you have to make off grid energy less expensive than on grid energy!

Do that and we all win!

ggse000 · January 2018

sorry, have been away for few weeks, lot of updates and perhaps keep the discussion going.

ggse000 · January 2018

mcgivor said:

There is no way any battery storage system will work in your application and save money, it simply can't be done. EV use lithium technology, they are more efficient, able to deep cycle, would last longer with shallower depth of discharge, as would any battery, best not to fully charge and also way more expensive. If grid tied feeding into the grid is not an option, then a zero injection grid tied arrangement would offset useage during the day, but return on investment would be extensive.

i got so many no no-s on this one so I already abandoned this idea and instead either do a hybrid option (grid connected system) or just move the operation to an area where electricity is largely generated from hydro or other form of electrical source. IT appears N. American like Washington State's chelan county is getting flood of inquiries already due to its proximity to grand coulee dam and also overal Canadian provinces.

Dude above mentioned Manitoba, Winnipeg as one option but from where I am it is about 1900 miles of distance, I am looking to see any comparable hydro dam on the west coast area of Canada closer to perhaps Vancouver. I did see one show on Netflix where the const. crews were building some sort tunnels off the mountain and showing the dangers of building it and if I remember correctly it was sort of system that would use the large water current resulting from melting ice off the mountain to harness its power.
Lol.

manzanita · January 2018

With the cost of components, building a big PV system would cost tens of thousands. A former neighbor spent $60,000 for a decent sized system. I don't think that you'd ever get a positive cash flow for a good number of years, but it could depend on how much you could earn per kilowatt hour of energy used in calculations. Utility power would be cheaper.

If you were doing a large amount of calculations per minute, wouldn't you need a better Internet connection than a cell phone? Sometimes reception and bandwidth is lousy when you get further from town. I feel that you'd need a physical connection to the Internet. You might check on how much bandwidth you'd need to do what you want. It almost sounds like you'd need a T1 connection, but I hope that I'm wrong.

I used to toy with the idea of running a computer cluster with solar PV during the day and shutting it down at night. Ideally, the system would power up or shut down individual nodes to match the available power. If nodes changed their power demand rapidly, the power up / shut down system wouldn't be able to keep up.

But it would take lots of money to build the solar equipment and a smaller amount to build a small cluster. Not certain any of it made sense.

Offgrid power options for crypto mining

Comments

Categories