Topic: Question for experienced "semi-large" scale bitcoin miners - page 5. (Read 10287 times)

What are the humidity levels like there? I'm sure you're not running your miners out in the open without regulating humidity, but they're inside where the humidity is controlled to some degree like I want to do with a dehumidifier. 

it is humid here in southeast asia...none of my 280x's since 2013 died because of humidity..all are still fine.

looking at that warehouse reminds me of something....ahhhh a brick oven hahaha

btw OP i think you are underestimating the power of ventilation, the sheer amount/volume of air passing thru will make things good enough for miners just take a look at a typical cpu heatsink with fan, it simply relies on air going thru the heat sink fins.

I'm no HVAC expert, been mining for some time now so if I'm at your shoes this is what comes to my mind as a solution...lets get creative  



an example of roof turbine..


use this filters for air..they are cheap .... dust problem eliminated...there are different type of materials used and rating (there are different standards in filtration measurement like MERV rating)...typically filter rolls are in pre-filter category, but for dust they are very good specially the highest rating and thickest.



i think assembling your own evap cooler is cheaper, good for customization (special case to case basis), parts are bought separately (you can go high quality per parts), you can go bigger...just look at this pump https://www.amazon.com/Zoeller-Mighty-mate-Submersible-Sump-Pump/dp/B000H5PYR2/ref=sr_1_1?s=hi&ie=UTF8&qid=1494031508&sr=1-1&keywords=zoeller+sump+pump   , this is the one i bought..

i haven't assembled my evap cooling so my creative setup isn't finished and i can't show a running setup and pitfalls in making one...I'm also in the process of acquiring a filter roll...

also buy some humidity measurement devices.

I don't know about 'nobody' but sure, it's definitely uncommon. I'm a special case though since I'm in Florida, no? I just feel like exhausting the heat out won't be enough.

Nobody does it so why do you want to? Ventilation is king

There have to be some HVAC experts here that can give their input to my dehumidifier/evaporative cooler idea posted above!

FYI if you buy any new S/T9s they are autotune. There is no overclocking.

I've done my calculations based upon 1250w for the 11.5TH/s batch, plus some head room for overclocking or if they happen to consume more for whatever reason, and it's still plenty profitable at $0.10/kWh, even before this spike in price (and difficulty). Like I said I'm not too concerned with that right now, but with heating/cooling. If I can't come up with a solution for that, there is no point in contemplating any further.

Depending on the batch they use more than that. I have some older S9s here running nearly 1500w at the wall socket.

Also there is no S9 that only pulls 1125....you are forgetting to add another 10% to those figures.

From bitmains site:

 Power Consumption 1127W (11.5TH/s batch), 1225W (12.5TH/s batch), 1274W (13TH/s batch), 1323W (13.5TH/s batch), 1372W (14TH/s batch) (at the wall, with Bitmain’s APW3 PSU,93% efficiency, 25°C ambient temp). Expected discrepancy of +10%

so thats a minimum of ~1239 for the slowest up to ~1509 for the fastest

Thanks! Yeah I do keep mixing up kW and kWh, my bad on that. Baikal miners have been around for a little while now, they make multi-algorithm ASIC miners, which are very profitable as of recently with the rise of a few altcoins. They used to sell out in the first day then be resold on ebay for twice as much because of how profitable they were, but for some reason they still have some in stock and it's been days. We'll see how bitcoin does in the coming months, right now I'm only considering doing this. If it continues to rise in price along with the difficulty as it's done for years now, I should be able to remain profitable (I was considering doing this months ago before the huge price increase, and even then with 10 cents kWh it was still very very profitable). Once a new miner comes out you sell your old equipment to recoup a lot of those costs, and upgrade. If the price of bitcoin tanks, so does the difficulty, thus allowing the "larger" miners to remain profitable. If it's ever not profitable for whatever reason, so long as I can run for a few months before whatever catastrophe would cause that, I sell everything off, eat the cost of the lease, and still make profit, just not as much. $0.05 / kW-h would net insane profits, I can't even imagine making that much money from mining. If I had $0.05/kWh there would be no hesitation to start mining lol. I really want to know though if that dehumidifier/cooler combo will work as well as I think it will, that's the only thing I'm unsure of at this point.

The wattage / antminer S9 varies per batch (total TH/s) and ranges from 1125 - 1325 Watts. Assuming 1225 Watt 20 x S9's will pull 24,500 watts. It's not that you're wrong I guess but you keep incorrectly labeling kw-h and kw.. I'll just run through it. So yes, at 24,500 watts the total energy usage in a day is 24,500 * 24 / 1000 = 588 kW-h consumed / day (note, not kW consumed / day as you said). A kW-h is a unit of energy (can be converted to Joules). A kW is a unit of power or rate of energy consumption. If you're running at 220 V then your expected current draw is 24,500 / 220 = 112 Amps. Plenty of room on 220 Amp service. You're on the right track, I have no doubt you can figure out how to set up a nice little mine, just confusing on your units.

Just keep this in mind - right now BTC price is awesome and economics look good for a mine at a wide range of power costs. However, time has proven that the network will self-regulate; difficulty will increase to match price increases so that only the low cost miners can mine more profitably than simply buying bitcoin. I personally would not build out a sizable mine unless my operating cost was $0.05 / kW-h or less. Hosting is a pretty cheap solution as the going hosting rate equates to running miners at a power cost of around $0.06-$0.09 depending where you host with, and you do NOT have to pay off the other capital costs as you would with your own mine.

PS - I've never even heard of Baikal miners... Right now it is widely accepted that Bitmain provides the most efficient / profitable miners. Be VERY careful about buying other brands, there are tons of scams on the go. Personally I only trust Bitmain and Avalon right now with my money, but other names will undoubtedly turn up soon with the price going crazy as it is.

Good luck brotha

Thanks for the advice. I'm just not sure how it would be cheaper to rent space at a reputable hosting location than spend a few more cents on power. Even with rates as high as 10 cents per kWh, I will still make a very nice profit (for me at least), including costs such as rent for the warehouse, the heating/cooling equipment, installation of additional power, etc. I've ran the numbers, and those profits don't even include reselling the equipment at a later point in time to recoup some of the costs. Even if it was cheaper by a small margin to go that route, what fun would that be?

I'm also not sure how my power calculations are wrong, I've run through them so many times. 20 Antminers requires 23000 watts of power (20 miners x 1150w per), also known as 23 kWh. 24 hours a day makes 552 kW consumed per day, which makes 16560 kW consumed per month, or as I said, 16.5 MW, resulting in a very expensive power bill, yet still cheap enough so I can generate large profits. Having a 220 amp 220v line run to my unit would provide me with 44000 watts of power, plenty enough for my size operation. If that's not how you properly calculate power usage then I'm totally lost and I have my head on backwards. I've ran the numbers, it comes out to over $2000 a month in power costs, but even after paying that huge bill I'll still be pulling in even more than that in profits. Also I may throw in a few Baikal miners, they're much more profitable than Antminers as they consume far less power, but I'll only have a few of those for diversification. Like I said, I'm not too concerned with startup costs like the dehumidifier or swamp cooler (if those would even be viable) because I know how to resell them and recoup a lot of their costs, so don't factor those into the equation. I've done plenty of research on bitcoin mining over the years, and have mined multiple times in the past, just not on this scale. I'm fairly confident that I already know the basics

I just want to hear from some guys that know HVAC really well to see if the scenario I posted above is viable, or if the dehumidifier and evaporative cooler would end up working against each other and produce no results. Theoretically it seems like it should work to both reduce humidity and drop temperatures, and they should play off each other. Dehumidifier to keep the humidity stable and low enough for the evaporative cooler to work, which would then keep the temperature in check.

GMPoison - If you expect your power cost to be $0.10 / kW-h I recommend that if you want to pursue mining to use your money to buy just the Antminers and send them to a reputable hosting operation. Not only do you not have to pay any extra capital cost for power / ventilation / property renovations etc, but you also don't have the headache of running the operation. You also need to consider that renting a commercial property has it's own operating costs, which add to the overall $ / kW-h.

Or, you can also buy bitcoin with your money and earn with a price increase. This is often a better option than mining when your power costs are > $0.10 (picked that # arbitrarily).

Without getting into detail on your recent replies, your calculations are incorrect. There are plenty of resources online how to calculate current draw (Amps) based on rate of power consumption (Watts) and voltage. I like your optimism, but I recommend you spend a lot more time getting down the basics (spend time researching the forum and internet, there's plenty of details on building small bitcoin mines and what to consider). If you can't spend time on this now and learn what you need to, you will end up making a bad, rushed investment. Good luck

So now what I'm trying to figure out is heating/cooling. Here in Florida it gets hot. Not only does it get hot but it gets humid too. I have exhaust worked out which would be three of these (http://www.homedepot.com/p/MaxxAir-Pro-24-in-Industrial-Heavy-Duty-2-Speed-Multi-Purpose-PRO-Tilt-Drum-Fan-BF24TFORGPRO/204631701) lined up like this (http://images.wisegeek.com/garage-exhaust-fans.jpg), so now I'm figuring out how to deal with humidity and temperatures, because I don't think those exhaust fans will be enough. (650 sq foot warehouse/garage btw)

I'm considering some sort of dehumidifier or portable a/c. I know it's very bad practice to try and cool miners with an a/c rather than spend the money on venting the hot air out, but this is a very small and confined space, so I think some sort of combination between a dehumidifier and an evaporative cooler could work. Dehumidifier to deal with the humidity, and the cooler to deal with the temperatures.

A combination of something like this (http://www.homedepot.com/p/PORTACOOL-16-in-Vertical-Tank-3900-CFM-3-Speed-Portable-Evaporative-Cooler-for-900-sq-ft-PAC163SV1A1/206757987) and this (http://www.homedepot.com/p/MaxxAir-145-Pint-Bucketless-Portable-Dehumidifier-DH-075-ORG/207155737) I'm thinking could work. I don't know if there is any way to test that however before dropping a bunch of money. There are so many variables involved that I don't see how anyone could tell me if something like this could work or not unless they've actually done it before, but it seems like theoretically it would work well enough to keep the temps down and humidity down so the miners would be comfortable, even in the summer heat. I'm going to find people that are pro's in hvac to speak to but I figured I would ask here. I have no clue if the dehumidifier would make the room dry enough for the evaporative cooler to work well enough, and I don't know if the evaporative cooler would work well enough to give me a 15-25 degree drop in temps that it claims it would. I don't care so much about how much those two things cost as long as they work, but if anyone has a more cost efficient way I'd love to hear it.

Any input from guys that know heating/cooling?  

edit: For that cooler specifically the manual says the following: "For optimum performance, the temperature should be 85˚ F or higher and the relative humidity should be below 75%. However,
Portacool evaporative coolers will reduce the temperature in almost any environment, making it more comfortable." That sounds like it would work nicely as I should be able to keep humidity below 75% with the dehumidifier, and the temps will definitely be at least 85 degrees. It also says that it consumes it's entire 42 gallon reservoir every 2 to 10 hours depending on humidity, so I have 0 clue about how I would regulate the flow of water so it would get enough and not too much every day since the humidity levels fluctuate outside... then again with the dehumidifier, I guess the goal is to get a somewhat stable environment humidity wise?

Thanks, but I just posted above how I got my megawatts calculation which was per month by the way not per hour or anything like that. 22.6 kWh is 16.2MW per month, at least as far as I understand.

As far as profits go, you have to believe that the price of bitcoin will rise as the difficulty rises. That's how it has pretty much always worked, and how I believe it will continue to work. I will probably hold onto the miners until Bitmain releases the next most efficient miner, who knows when that will be, but I don't believe that I will ever *not* make a profit by mining. If I do, I sell the miners off and make a lot of that money back.

Thanks for your reply. I got 10-20 megawatts from the following calculation: 20 antminers x 1130W = ~22600W or 22.6 kWh. 22.6 kWh is 16272 kW/month or 16.2MW per month. I may be doing the calculations wrong but that's how I got 10-20 megawatts per month. A 220 amp 220v service is 44000 watts which is twice as much as I would need, so that would be plenty, even if I wanted to purchase more miners one day. Any rough estimate that you could give me on how much something like that *might* cost? I know I'll have to get quotes for actual numbers.

Thank you too for your suggestion about exhausting the heat. I'm sure I could come up with something like that, somehow securing the door and the fans to the ground. I was thinking more along the lines of 3-4 feet though not just 3-4 inches, so I could put a couple industrial high cfm fans like this one next to each other for really good airflow: goo.gl/dRQJ8E I could see something like that working where I would just close the door to the top of the fans, and secure the door to the chain of fans with some heavy duty metal chains so there would be no way to get in or out without completely busting the fans as they would all be linked together to the door and to the ground. Even then there would be security tapes and whatnot.

This kind of cooling is cost prohibitive and unnecessary for what you are doing. If it is a dusty environment just blow them out with a compressor every couple months or so. You wont find a profitable large scale mine out there that gives two shits about dust and spends a ton of money on air filtration. At best they use swamp pads and consider those their filtration system.

20 T9s is only 32kva. Any commercial building can handle that no problem. The biggest problem you will face is actually making a profit. 10 cent power is pushing the limits of viability.

You will also have to purchase a small amount of electrical infrastructure, network gear, power supplies, and find a way to cool it all. You also need to find a warehouse that is not a concrete bunker as it will just absorb the heat and make it even harder to cool. Also most commercial establishments are not going to enjoy the sound of a small airplane buzzing around their property 24/7/365 either so you will have to factor noise into the location you pick.

Making $2200/month means it will take close to a year to recoup the initial investment. In another year the difficulty and network hashrate will be so much higher the chances are you wont be making close to $2200/month at the end of the first 12 month period it will be more like $1500. Im not saying you wont eventually make a small profit but dont expect to make a bunch of money on this investment. To make real money mining bitcoin you have to go large in a dedicated facility with ridiculously cheap power. At that point you are looking at millions of dollars to make it through year 1.


P.S. You will be using like 35-40kilowatts, not even scratching the surface of the megawatts. 1 megawatt of power is like 650+ T9s on its own.

Make sure you can vent a LOT of air through the space.
The recommendation of "more CFM than the total of your miners flow" is a good one - but IMO it's better to use filtered positive-pressure INTAKE fans over exhaust fans and ghods only know what kind of dust pollen and whatnot come into your space from wherever.

Tampa is to damp for evaporative cooling to work, unfortunately.

20 Antminers should be handled comfortably by 200 amp (220) service, which is pretty commonly available but NOT universal anywhere I've ever been on spaces in the 1000-2000 square foot range.
 That's only 26 KW or so, well within the 43KW+ rating of a 200 amp service (even with derating for continuous draw service you should have over 33 KW available COMFORTABLY).

 Be sure to CHECK the space for service availability and venting capability before you actually lease it though.

 I dunno where the MEGAWATTS figure came from.


 In theory, it should be able to keep the garage-type door open 3-4 inches without being a security issue, if you can figure out how to lock the door at that point without having to make modifications the landlord would find unacceptable. Perhaps reuse an existing bolt in the side of the door and a bracket that bolts to the door "rails", so you can remove it and put it back how it was when the time comes to move out.