Help sizingsystem

usao · March 2017

I need power at a remote site, off-grid.
What I need is 12v @ 1KW 24/365

This is approx 90A continuous.
I don't need an inverter, 12v DC only.
From what I understand about solar, I need to pull 90A * 24 hours = 2.1 KAh per day from batteries, and that number needs to be 50% of total battery capacity, so total battery capacity should be around 4.5 KAh
In addition, I need to push 2.1 KAh into the batteries over a 5-8 hour period each day.
If I assume 5 hours, the the solar panels need to generate about 450 A.
I havent found a charge controller which does 450A yet. Looked at Morning Star and others, but they all seem to be in the 60-80A range.
Are my numbers wrong somewhere?

westbranch · March 2017

Sketchy but in the ball park.
You need multiple Charge Controllers. Most will be in the range you stated. It is not good practice to push electronics steadily above ~80% of their rating max, strange smoke cab be released and other bad things happen... fire?...
50% is the max you want to draw so iif this is every day you need a safety cushion as well.
You also need a huge array(s)

mcgivor · March 2017

usao said:

I need power at a remote site, off-grid.
What I need is 12v @ 1KW 24/365

This is approx 90A continuous.
I don't need an inverter, 12v DC only.
From what I understand about solar, I need to pull 90A * 24 hours = 2.1 KAh per day from batteries, and that number needs to be 50% of total battery capacity, so total battery capacity should be around 4.5 KAh
In addition, I need to push 2.1 KAh into the batteries over a 5-8 hour period each day.
If I assume 5 hours, the the solar panels need to generate about 450 A.
I havent found a charge controller which does 450A yet. Looked at Morning Star and others, but they all seem to be in the 60-80A range.
Are my numbers wrong somewhere?

Regarding charge controllers, see link.
http://www.morningstarcorp.com/parallel-charging-using-multiple-controllers-separate-pv-arrays/

Horsefly · March 2017

usao said:

I need power at a remote site, off-grid.
What I need is 12v @ 1KW 24/365

This is approx 90A continuous.
I don't need an inverter, 12v DC only.
From what I understand about solar, I need to pull 90A * 24 hours = 2.1 KAh per day from batteries, and that number needs to be 50% of total battery capacity, so total battery capacity should be around 4.5 KAh
In addition, I need to push 2.1 KAh into the batteries over a 5-8 hour period each day.
If I assume 5 hours, the the solar panels need to generate about 450 A.
I havent found a charge controller which does 450A yet. Looked at Morning Star and others, but they all seem to be in the 60-80A range.
Are my numbers wrong somewhere?

I don't think your math is wrong. However, back in January you approached a similar but smaller problem and decided maybe it was too much (http://forum.solar-electric.com/discussion/comment/366513#Comment_366513 ). I'm guessing that if that one was too big to justify, this one is even more so.

Anyway. 4,500AH battery bank at 12V, will have to get recharged per a 12V bank rate, like at 14.4V or so. Assuming typical losses along the way (wiring derating, panel derating, charge controller derating), you probably need to assume about 70% efficiency getting to those batteries. So:

4500AH * 14.4V / 0.70 * 0.10 (10% charging rate per hour) = ~9.2KW of panels. That's lots of panels.

You are right, that is too much to push through a single charge controller. You would need to split that array up into something like 10 separate charge controllers.

This big of a problem doesn't seem to be something anyone tries to solve with a 12V system. More likely a 48V system. I see you can't do that, but it makes the problem (e.g., number of charge controllers) at least 4x as big.

This is all assuming you can cobble together a 4500AH 12V battery (yikes!). Needless to say, your continuous 90A draw would mean you were intentionally drawing down your batteries to 50% DoD every single day assuming there was nothing but bright sun outside. You'd need a pretty good backup generator for all those days that didn't have plenty of sun.

I'm not a pro at this and I am generally keen to try and solve a problem, but... Your problem seems like one to walk away from and find a different approach.

Steve

mike95490 · March 2017

12V @ 90A 24/7 is a wrong number. You need to consider a 48V @23A system, and then downconvert to 12V at the final usage point.
You won't save any solar panels, but you will save on charge controller costs and copper costs.

usao · March 2017

Im still trying to figure a way to do this, perhaps on a smaller scale. I may shoot for a 500W, 250W or other and just run on a smaller scale.
I figured that if I had the numbers for 1KW, then I could easily scale up/down and reach a number I could work with.
I did find a 2V battery (http://www.rollsbattery.com/wp-content/uploads/batteries/2YS62P.pdf) which might work.
Im a vit confused about how to get a 12V draw out of a 48V source though. Would I need to use an inverter to go to 220 then a PSU to get back down to 12V? Seems like a lot of loss inbetween.

Estragon · March 2017

To get 12v for loads from a 48v system you would use a DC-DC converter, likely one that outputs constant 12v out from variable (eg) 40-65v input to allow for varying battery charging voltages.

mcgivor · March 2017

usao said:

Im still trying to figure a way to do this, perhaps on a smaller scale. I may shoot for a 500W, 250W or other and just run on a smaller scale.
I figured that if I had the numbers for 1KW, then I could easily scale up/down and reach a number I could work with.
I did find a 2V battery (http://www.rollsbattery.com/wp-content/uploads/batteries/2YS62P.pdf) which might work.
Im a vit confused about how to get a 12V draw out of a 48V source though. Would I need to use an inverter to go to 220 then a PSU to get back down to 12V? Seems like a lot of loss inbetween.

Without defining the loads, it's near impossible for anyone to suggest a solution, it's all speculation, there are many ways things can be done, but without details ,suggestions are just shots in the dark, no offense implied.

oil pan 4 · April 2017

The extreme aversion to saying what the loads are makes me think it's something illegal.

usao · April 2017

oil pan 4 said:

The extreme aversion to saying what the loads are makes me think it's something illegal.

Not sure what you mean by "extreme aversion to saying what the loads are".
Im looking for a way to keep bitcoin mining equipment running as the profitability goes down over time.
Mining rigs use a lot of electricity and their revenue drops over time. Electric costs are a significant component which often renders rigs break-even or worse even after a few months, meaning I can no longer mine and am forced to turn off.
If I can get something going which will allow me to keep them running for a much longer time their revenue generation lifetime can be greatly extended. Most mining rigs run on 12V. My smallest rig uses 40W and my largest rig uses 1KW. I have a total of 10KW currently running my garage. I have a 50A @ 208 circuit dedicated to my rigs.
So, that is why im flexible on the loads as I have a variety of rigs I could use in this way.
I already know the cost of the solar setup is way more than the revenue, thats not the point for me. Rathe having a solar setup of some sort is (in my mind) always a useful thing to have for other purposes, for example, im considering a chevy spark and have looked at the tesla powerwall also.

Photowhit · April 2017

usao said:
Im looking for a way to keep bitcoin mining equipment running as the profitability goes down over time.

So if we can say with a certainty that the cost of solar will be higher per Kwh than the cost of grid then your hunt would be over?
This will pretty much always be the case.

Might hunt for lowest cost of electric in the US, some places charge less for higher usage... If you were to try to do cheapest solar and you didn't have to store energy and just use grid tied energy to offset higher prime daylight energy times...

Not sure how you are doing bitcoin mining without internet, I thought the loading and bundling of transaction would require pretty good internet service. ...but it's not my end of the pool.

mcgivor · April 2017

@usao said Electric costs are a significant component which often renders rigs break-even or worse even after a few months, meaning I can no longer mine and am forced to turn off.

I already know the cost of the solar setup is way more than the revenue, thats not the point for me.

Sorry I'm a little confused by the two statements made, they seem contradict one another.
As @Photowhit suggested, grid tied would be the only economic solution.

Estragon · April 2017

Assuming grid tied would run about $2-3/watt and grid power at 0.25/kwh (pretty high), breakeven is 2000/.25=8000hrs. At 4hrs/day that's about 6 years ignoring cost of funds and depreciation. To get a real number you would want to find out how the local utility handles grid-tie buy/sell rates and billing. IMHO, there's also a risk that utilities may strand investment in grid-tie if it gets big enough to start causing significant network issues.

If you really are remote and off-grid like you said in your first post, initial cost, depreciation, and end to end system power losses will all be higher. If the numbers don't work well with grid power, there's no way they work off-grid.

If I was going to run a server farm commercially, I'd locate it somewhere with cheaper power and climate for lower cooling costs (eg Manitoba in Canada).

BB. · April 2017

An interesting article about bitcoin mining and cost of power:

https://www.washingtonpost.com/news/worldviews/wp/2017/03/10/bitcoin-mining-is-big-business-in-venezuela-but-the-government-wants-to-shut-it-down/

CARACAS, Venezuela — Venezuela has become widely known as an economic basket case in recent years. But with its cheap electricity and volatile national currency, the country has at least one competitive advantage: It’s a good place to make the digital cash known as bitcoin.

Bitcoins are increasingly accepted online for buying real-world goods and services. And, unlike the Venezuelan bolivar, the virtual currency has been going up in value.

Making bitcoins is known as “mining,” but it requires a powerful computer instead of a pick and shovel. Those computers produce bitcoins by creating elaborate algorithms, but they also suck up a lot of electricity. In many countries, the cost of running a “mining terminal” can run higher than the value of the actual bitcoins.

That’s not the case in cash-poor, oil-rich Venezuela, where state-subsidized electricity is so cheap it’s virtually free. But Venezuela’s government isn’t pleased. It’s cracking down on bitcoin mining, even though the country has no laws on the books outlawing the currency or its manufacture.

In November, Venezuela’s secret police raided the house of two brothers in Caracas and found more than 90 mining terminals. The agents demanded $1,000 in bribes for each machine, according to the brothers, who spoke on the condition of anonymity because they fear arrest. The brothers said they paid the bribes to stay in business...

And a history about a major Bitcoin exchange... Basically the rise and fall of MtGox:

https://en.wikipedia.org/wiki/Mt._Gox

-Bill

usao · April 2017

mcgivor said:

@usao said Electric costs are a significant component which often renders rigs break-even or worse even after a few months, meaning I can no longer mine and am forced to turn off.

I already know the cost of the solar setup is way more than the revenue, thats not the point for me.

Sorry I'm a little confused by the two statements made, they seem contradict one another.
As @Photowhit suggested, grid tied would be the only economic solution.

There's no right answer here. I have a lot of fun with bitcoin and at the same time I really like the idea of off-grid solar.
I know the 2 ideas are somewhat contradictory in that bitcoin mining is not environmentally conscious yet at the same time I am into environmentally conscious activities such as setting up recycling for businesses, composting and growing my own veggies and peppers and other activities.
The other reason I said "remote" was just to re-enforce the idea that I wanted to use 12V only without the inverter in between the batteries and the mining rigs. In my mind it was simply short hand for saying I wanted stand-alone setup, but in the end I had to give the full blown explanation anyhow, so im sorry for that. I really didn't expect it to become a big issue, and I guess that's my bad.
After all that said, ive decided to start small. Ive ordered a morning star SS-20L-12V, and will get batteries this week from a local supplier. Im going to start by verifying that I can run one of my older Gridseed mining rigs using the batteries through the load ports on the Morningstar for 24 hours. Thats a total of 2.4KWh/day, and I will recharge using a battery charger during off-peak hours. If that works, ill start shopping for some panels to generate 2.4KW in 5-8 hours (I live in AZ so get lotsof sun).
Again, sorry for all the confusion, I was trying to keep it short and simple Q&A just to validate my ideas, but that didnt work.
Also, I know the investment outpaces the revenue, but I also figured at some point I could start using the solar system for other purposes over the coming years.

Estragon · April 2017

Good plan... start small and get some real world experience with solar and batteries. I'm glad I did - avoided a lot of mistakes that way.

BB. · April 2017

No problem... We try to look at the needs in full. Just designing a "cost effective" solar power system powering a black box does not always work out well. Load calculations and conservation are a very important part of a successful solar design.

Plus we are always curious about what other people are using solar power for.

-Bill

Help sizingsystem

Comments

Categories