Topic: [RENAMED] 800 amp, ~300kW, <$0.03/kwh, warehouse for $3,000 a month (Read 3812 times)

We host in Eastern Washington with our office in Seattle.  Our rate are .023 which are amazing.

Dalkore that's interesting but you're in Seattle, power rates there are closer to 6c that's more expensive than certain areas in Colorado with a terrible wetbulb.  100 per kw is but whats your cost per KW, 40 to 50?

Sorry meant to say Moses lake

This is exactly what has happened.   The PUD I am working with change the 2015 pricing to basically double the build-out cost for high-voltage commercial power projects because of all the tire-kickers.   We were lucky enough to secure our power in 2014 when it was cheaper.   

Currently we have room for some investment capital to build-out out multi-MW facility.  We have a 5,000 sqft. building with a very long-term lease.   If anyone is serious and want a serious plan and financials, please contact me directly via PM. 

-D

strangely enough, the pot growers, coupled with the bitcoin miners all trying to move into the area are the reason we have the power upgrade moratorium.

it *sounds* like they had a lot of looky-lews that were requesting power studies for buildings with insufficient power to them. after finding out the cost to get a MW of 480 is six figures and change, most of them backed out... thus the power company is tired of wasting its time doing power studies for people that have no realistic idea of how much the price tag was going to be. i know one pot grower that is looking for 144,000 watts of 120v. thats 1200 amps for lighting alone... he is gonna need another 4-800 for ACs, water pumps, etc.

Good thoughts on this one.... except there is currently a moratorium on upgrades in excess of 1MW here, so upgrades aren't as easy as one might think. Even if you can get them approved, upgrades still cost tens of thousands of dollars, unless you happen to have a big building right next to an existing power sub-station. That's not to say there aren't locations that already have over a MW. For instance, the ASICSPACE facility up in Cashmere, they lucked out as TreeTop already had that much draw to the location from decades ago. Morman Lake is in Arizona, wrong state. Growing pot, if mining were to go south is not exactly easy, the application process is closed, so you would have to buy an existing license, which are going for $40-350k depending on how much growing square footage was requested on the original application and your building would need to meet a list of about 20 state requirements.

I can confirm the above. You need at least evap (cross building) cooling or miners and surroundings that will tolerate that heat.

Having lived in Wenatchee most of my life(where this building in question is located), i can tell you that it gets hotter here in the summer then it does in LA. despite being fairly far north, we are a high desert region and frequently spend weeks in the summer with daily highs of 100+. There is no way you would be able to operate 24/7 from mid/late may to about October without some kind of cooling. I would even question the viability of running at night, as our nightly lows can stay 70+ for weeks as well.

good luck, main cost will be in wiring / cooling infrastructure.

I'm guessing you're around Morman Lake in middle Washington for the .03c per kw price point.

One thing to consider is can you add additional service to your address (additional power) - call the power company and check time, and pricing (easy way to take a 5k square foot warehouse) and make it handle 1mw+ of power

Second, consider your wet bulb, Washington state (west of cascades) is high, out east, relative humidity is low.  You can get away with evaporate cooling easy enough and cheaply.

Third, if mining goes south, whats another high yield / power hungry industry in your state (Washington) that is now legal? With all the circuits, power and the cooling, consider splitting the investment between bitcoin mining and a growing operation.  This might possibly be the best way to guarantee the FBI would kick down your door, but hey, you would be doing everything legally!

I decided not to go with this deal as they weren't very flexible about a termination clause in case things go sour. And, they were charging $10 a sq ft, and about $10/kw of capacity. I'm going to wait on another new construction in a different location that supposedly will go live this summer that they quoted me would be almost 10 times cheaper - $1.6/sq ft and $1.7/kW of capacity, could do 0.5 MW at 5k sq ft or 1.0 MW at 10k sq ft. Thanks for chiming in all.

Like some others that may bounce around I won't put too much on the table publicly as this is one of the aspects of my consulting business (mining design, deployments, etc.) but I will say you need to work out the labor for the electrical, and work the details on cooling although you may very well be onto something.. Even in CO I have some places that only have chillers running for a few months out of the year, if you are further north you can mimic some of the larger mining operations but you need to understand the thermals and design the right hot/cold setup..

No way, I don't think I would be investing in back up generators, at least not for a long time. They are more important in critical hosting environments, such as high frequency trading for example. For bitcoin mining, if it goes down, then it would be rare, <24 hours at a time, maybe every other month or something. Plus, I think the location brags high uptime and low power outages, but maybe that's what they all say. My friend has hosted with Toom.IM for example and I think they had a power outage one time in a 3-4 month period and they were only down for a couple hours. Not worth the investment in my opinion. I don't think any of the bitcoin hosting competitors who offer <$100 per kW pricing are investing in back up generators, but maybe ASICSPACE or toom.im or HashPlex can chime in.

Back-up internet connection may be more dooable, sure. It's true we didn't even cover internet connection, but these days I take Internet for granted so I'm going to guess that's not going to be a significant expense in relation to rent + electricity. Correct me if I'm wrong, I should probably look into this, too.

I think Jamphone is probably referring to Cooling. Nobody really talked about cooling much, but this warehouse is in an environment high north in latitude, so it's winters most of the time with a modest summer. Also, these aren't high-end servers/computers, but bitcoin miners. I think they can operate in less-than-typical "data-center" friendly environments. This isn't Texas or California, even though some have been able to host there as well.  Some have suggested that I may be able to do away with HVAC completely, or at a very bare minimum as long as I set up the racks properly and have smart ventilation. Are they bsing me? I don't think so, but maybe.

Ahh ok, don't forget about backup generators and fuel in case the power goes down, as well as backup and sometimes even backup of backup of the internet connection.

I highly doubt you've received any decent information because I see two critical errors in your plan that no one has pointed out. Really bold ones. So I'm surprised to hear you're getting so much help.

Anyway, I love finding new business partners. PM me your phone number if you are serious about this project. I can offer you cheaper hosting than you can provide yourself.

I haven't yet started looking for investors, and am only in the gathering information stage (hence speculation). Also, for a size this small, I may be able to fund all of this myself, incrementally. For example, I may not need to worry about installing HVAC until summer approaches and can shoulder off those costs until later. Keep in mind, this is just barely a .3 MW operation. ASICSPACE has ~5MW (I think), Toom.Im has ~1M, so I don't consider this a mega large operation. If and when I get to the point to commit and start up a legitimate business and/or seek investors/customers, I'll have a company profile, my information, my identity, my linkedIn profile, etc, and don't have any hang up in revealing this. It's too soon for that as currently I'm just seeking information to determine if this will be profitable and how long would it take to recoup my initial investment if I pursued this. I estimated that after rent, electricity, insurance costs, I should take in just about ~$4-5,000 barely, (tbh, much less than I would take in at my previous software engineering jobs, but I seek the thrill of entrepreneurship and learning experience). I'm only wondering how much I should estimate for initial capital to invest into racks, fans, electrical wiring, etc, to determine how long it would take to break even on my original investment. I'm estimating in the best case scenario I'd need $10-25k. Hopefully it doesn't go over $50k, as that would make things very difficult for me. Is this reasonable? Thanks!

Jamphone- you'd be surprised, not everyone wants to close off knowledge away. I'm following an "open" business model here where I reveal almost everything up front, the rent, the electricity cost, etc. If people want to participate with me on this, that's their choice, and in the end if this succeeds, they may have the benefit to access early-promotion hosting discounts. There are a lot of examples on this forum of people helping people, being open. Indeed, it seems some of the most respected members on here are those that are open to sharing knowledge. And, in general, Bitcoin itself is premised on an open-source model. Anyways, I digress. Of course there's always a threshold that people are not willing to cross in terms of revealing information, and I respect everyone's boundaries and appreciate whatever help I can get. Again, seeing that this is a <0.5 MW operation, I don't think this is rocket science, and expect to learn a lot along the way. Again, if I was shooting for a 1-5MW operation, I'd be a fool, but I think for 277 kW, it may be a neat experiment. I've already received a good amount of help via private messages, so not everyone is as closed off as you portray. Also, I tip for those that help, showing my appreciation and that I don't take it for granted.

I just meant that you're asking some very technical questions and not offering people anything in return.

"Please come design my data centre for free."

And I'm telling you that anyone who knows the answers to those questions isn't going to post them for free on bitcointalk. Might as well just give our businesses away.

If you don't know the answers to a lot of those questions, don't go dropping the kind of money you're talking about. I've been hosting for two years. Jumping into that kind of a set up will be tough, there is a lot to know.

And you budget is way, way too low. Wait until you start pricing out the electrical work you need and the infrastructure you'll need to have in place.

I am a little unclear in what you are actually trying to do...

I see you are debating about buying a warehouse, but are you only going to run your miners there or are you planning to host miners for other people?

You have made a post about potentially looking for investors, but to be honest, without a little more information you won't find anyone willing.  I am not saying post the address of the warehouse, but at least give a state that it is in, as people will not be interested in investing without of course going to the place and meeting you, and spending a couple hundred plus dollars per flight for a small operation probably isn't that enticing to most.

If you are planning to host for others, my recommendation would be to pair with a pretty trusted individual, that is probably the best security you can do for the site as more people will be willing to send their units to you without needing to know specifics about location and whatnot.

Just my general input, good luck with your endevour though.

That wasn't very helpful, but thanks.

I've spent two years learning all of this. Hostin', much like pimpin', ain't easy.

Good luck. I don't think anyone's going to divulge trade secrets.

It doesn't look like they want to negotiate the price lower at this time. They just recently put it on the market and they'd rather wait and see who bites on it at the old rent price. The other tenant paid $3k, so they think they'll be able to find someone. I either gamble and hope they'll accept my $2700 offer 2-3 weeks from now, or lose it to someone else who'll take it for $3,000. I don't want to risk losing this spot since it's been pretty dry with those power rates and electricity already built in without needing improvements. Thinking about going with the $3k asking price, but before I commit, I'd like to first try gauging how much capital I'd need for infrastructure equipment. Any thoughts? I'm trying for as economic as possible, and smracer has been kind to give some pointers regarding whips, for example. I'd appreciate any more advice & tips!

How much should I estimate for initial capital for equipment? Best case scenario, average, and worst case? Respectively, I'm thinking $10k, $25k, $50k.

Capacity: 800 amps, 208v, 3-phase, ~277 kW. 100-300 machines

Security deposit: $6k
Cheap/Ghetto Racks: ? (Home Depot? Where's a good spot and how much)
Whips: ~30 30Amp whips at $50 each, $1,500
Additional Cords/Wiring: ~$750
Misc. Wiring Electrical Costs: ~$2,000?
Used Industrial-size fans (2x): ~$600??
Vents: $600

What about HVAC>> How much would that run me.

So, I want to say $10,000 if everything goes smooth, $25,000, more likely scenario, but hopefully no more than $50,000, at worst case scenario? Anyone have any approximations? Thanks again!

The temps that refrigeration operate at usually are at a level where you see condensation or icing.  When you get below the dew point, you will see this effect.  Even with de-humidifcation, it will not be fast enough not to potentially hard your electronics.   Try this, get a humidity meter and set the frig to the operating temp you want to run it at and take some readings.  Next step would be to add some load in there and run it for 3-4 days and see what builds up on the unit before committing to building it out.   

What you could do if you find it doesn't work, is use the existing ducting and to that attached your cooling equipment and still use the room and any electrical outlets it may be wired for. 

Does that explain my thinking on this a little more?   We actually looked at a frig unit as a space way back when and this was a concern for us.

-D

I though refrigerator unit take away humidity like air-conditioner? What makes you think it raise the humidity?

The space could work but I would not run the refrigeration unit, this will raise the humidity in the air and your electronics will not like that.   Using it though as a quality insulated room and getting some other from of cooling piped in could work well.   It is surprising that no 480v volt is in the building with all those compressors and motors. 


-D

On HorseRider's point:

Old equipment won't ROI. Costs too much to ship it. I'm presuming 2.5 cents/kwh is somewhere extremely remote, so shipping will be even more expensive.

If you have very cheap electricity power rate, mining is still quite profitable right now. You can buy older generation of mining rig to save the investment and leverage the cheap electricity cost.

Yeah, I'm kinda leaning against option B. I think Option A is already wired, ready to go. Already has like 5 panels. It has that large freezer taking up space, so leaving only 1500-2000 sq ft, but that should be plenty. I contacted them to see if it's still available, hopefully it is. If it is, I'm going to offer them 6 months up front for $15,000 (i.e., $2,500 per month), with option to renew for 3 years at $3k a mo. If they accept, I'd be starting next week. But, let's not jump the gun here. I'll be able to answer more questions if I can sign the deal at least. My company is registered in the state of New York. Again, don't want to jump the gun, if I can sign the deal, I'll be opening up to more questions.

okay then you could rent/host at 8 cents a kwatt and turn a profit.

I think at 8 cents  you could find renters/hostees

So do you want some advice don't do the improvements go base setup.

Tell us the location in general.
 Ie USA, New York State.
 and how will you insure safety of our gear.

Lets say you are in New York State and  I am in New Jersey .  I say I will send my  8 sp20's to you  with power supplies. Since we are 1 state apart I will drive them to you.  Or ups them really does not matter.  But I know the location of the warehouse basically I would not sent 4k in miners and 1.5k in psu's to you unless I know the location.
the upside is I know the miners location.  the downside is 10 or 20 guys like me do this.  we all know the location.  so breaking entering and grabbing gear is a temptation to all of us that shipped the out gear to you. 

sometimes the details suck, but can you keep it safe from theft?  I am just tossing out stuff to consider on your end. Not trying to be an A-hole.  but 100 sp20s set to use 800 watts and 200 s-5's set to use 600 watts is a lot of temptation.  and that list is not quite you max  it is about 200k of power.

the other 30 k need some for cooling as 200k in power is almost 800k btu's.

It's 3-phase, so you have to multiply by the sqrt(3) or ~ 1.73, and at 208v, so we're looking at about 800A * 208 * 1.73 * .8 = 230kvA. So you run for 24 hours, ~5,520 kwh. It's ~2.5 cents approximately per hour. No joke - $0.025 per kwh. Per day, that's about $140 per day in electricity costs, or $4,200.  So roughly, $3000 for rent, let's assume I negotiated it down to $2700, so $2700 + $4200 = $6,900 monthly cost for 165k kwh. $6,900/165k = $0.0418 per kwh, not 6 cents a kwh.

here is your math 3000 rent  a month.  and lets say 800 x 208 x.75 =about 125kwatts safely   125 x 24 = 3000 x .029 = 87 dollar in power each day or 2610 this means 3000 + 2610 = 5610 a month for the real cost of ops so if you charge 6 cents a kwatt to me to mine you break out even.

So you need to charge me 8-9 cents to hopefully turn a profit.   lets say you do 9 cents rent it all to miner you get 8100 -5610 =  2500   

not counting internet .
 not counting cooling
 not count labor

so 9 cents is most likely not going to work for you try 10 cents  that comes to 9000 -5610 = 3500 a month   this may be okay on your end but on the miners end 10 cents is high. 

I am in my house I pay 13 cents but for the next 3-4 months it is reduced to 10 cents due to heat savings.  So I have no need to use a 10 cent  hoster until apr-may.

Also I am in USA, NJ I have to ship my 8 sp20's to you which will cost money to ship and money lost to shipping + setup time.   So it is a tough sell to me.

Now if it was May  and  I do not need the heat and my rates are turning to 16 cents for the summer on June 1. sending you  8 miners may make sense.

I think your timing is off a bit to start your business. A lot of people are like me and they will mine in their house for the next 100 days.

Electricty isn't billed by kilowatt. It's billed by killowatt hour

Just a little example:

Remembering from our high school science class that 1A = 1 coulomb/second and 1V = 1 Joule/ Coloumb and a 1W = 1 Joule/ Second

And of course, we remember that a coulomb is 6.241509×10^18 electrons

800A at 120V = 96000W = 96kw

Say you run for 24 hours.

That's:

2304kwh

At $0.029 (are you sure this is correct? My rate is $0.23 (23c per kwh) at home)

That's:

$66/day

Although it's more likely to be $660/day

 

Option B seems to be a better deal if you are going to mine for really long term, but I prefer option A as I dont want to spend so much money to upgrade and it seems risky as bitcoin halving is coming soon.

That's awesome. Thanks very much. I've looked into Washington State but it looks like all the good spots have been taken up already . I'm keeping an eye out though

I'm having trouble deciding which route to take.

Option A) 800 amps, 3-phase 208v/120, service already installed at $2-3k a month, ~3000-5000 sq ft
Option B) 200 amps, 3-phase 240v/120 (older "delta" "CT" installation) at $2k a month, ~5000-7000 sq ft. Spend anywhere between $30-$80k upgrading to 0.5MW-1MW, possibly more.
Option C) Neither, find a place you can buy, not lease. I should mention there large 7000-8000 sq ft warehouse building available made of concrete foundation, but that's around $250k for the property to buy (not sure about the electric but probably have to upgrade either way as in option B). That's the major leagues though.
*Presumably I can do 2-5 year lease terms.

Option B would require more capital because I'd have to upgrade with the utilities company but it could turn out to be a better long-term solution. But, I also think upgrading to 0.5MW-1MW, would be a bit over my head.

I think, given my experience, I should probably just go with Option A for now, and if I can build up some experience there, consider moving or starting another operation in something like Option B (where I'd have to make bigger expenditures, i.e., upgrading with electrical company). Thoughts? I also pmed "smracer." It seems like, he was doing something similar with the 1200 amp warehouse, but lucky for him, he bought the place, so he has a lot more leeway there.

If you're in WA, and need help with capital, build expertise, and operations, let me know.

Hey ASICSPACE, thanks for chiming in. I'm not currently disclosing the location of the warehouse that I'm considering to lease. But yes, I'm sure unfortunately, that it's currently at 208v. One can upgrade to 480v and get more power as well, but again, those costs are heavy $30-80k all inclusive I'd guess conservatively; still, could be worthwhile if I could raise the money. Got around $30k of my own to invest, but probably would need more to be safe.

Where are you located? Are you sure you power isn't coming in at 480V?

Thanks for the input guys. Pretty much scratching that idea.

This place is going for $3,000 a month and has 800 amps at 208/120v, could possibly negotiate down. Electricity on average would be about $.025-$.029 a killowat. Anyone think there's a small business/mining opportunity here? Anyone think the community would be interested in pitching in for some kind of a "group host"/"group buy" option? Been thinking. Or 800 amps too little since that gives ~290kW. Another option would be to spend $30-80k alone upgrading to 3000 amps (1 MW), but that would be a much bigger project of course, and more capital.

Freezers and refrigerators are designed to keep objects that do not generate heat cool. Mining hardware generates ridiculous amounts of heat and they will burn out motors of coolers not suited to them in no time, no matter how large they are. There are old tales of people sticking mining hardware into their refrigerators here only to find an expensive mess made of both in short time.

Don't think it can be payed off form mining.

This is an interesting idea. Though I don't think it'd work for the same reasons listed by others ^
Kudos for thinking outside of the box though.

More importantly, if it is a giant freezer, a/c that can make a big space cold. That means it is going to consume a lot of power. Mining profit margin is already so small and if you add that, you might not ROI forever...

Ventilation is cheaper and better..

It probably won't, like a fridge I expect that this freezer equipment is not made to run 24/7. They cool the stuff inside down and then the insulation takes care of it for a good while. That won't work if you have a heater (aka miner) in there producing heat all the time.

I've been looking for some space for a potential small-scale mining operation, and I came across a property that has a freaking large-scale commercial freezer, like 30 ft by 60 ft by 8 ft, around 13,000 cubic feet. I'm still waiting to get more information on the freezer - how much power draw, walk-in or, wide-open doors like in the super market - and will get back to you.

I read over this archived thread here, but not much information on a large scale freezer
https://bitcointalk.org/index.php?topic=114296.40

Any way to make it work for cooling? Even to cool the place, can I open that baby up and let the cold air through during the summer months? Or leave the miners inside? Or just forget about it, no use whatsoever, just takes up space. Even if it's less effective than A/C, if it can save costs on purchasing new A/C, then it may be worthwhile.

EDIT: Renamed title: from "13000 cubic feet freezer - any use to a potential mining rig?" to ^^^