Miners, what's a good rate for electricity these days?

[ColoMiner.com]

I'm curious about electricity prices.

What do you pay for electricity? What's the most you've ever paid per kwh? Are you satisfied with the current rates? What is your outlook on future prices? Doing some homework, here. Trying to get a better understanding. Thanks.

[philipma1957]

I'm curious about electricity prices.

What do you pay for electricity? What's the most you've ever paid per kwh? Are you satisfied with the current rates? What is your outlook on future prices? Doing some homework, here. Trying to get a better understanding. Thanks.

Once upon a time 12 cents could work.

Now 6 cents is no good for mining ⛏ btc directly with asics.

a s19pro does 110th that is about $ 7.70

it burns 3.3kwatts an hour or 80kwatts a day at six cents power would be $4.80

7.70-4.80= $2.90 profit a day. since the s19pro costs over 3000 usd it would take

1000 days to earn 2900 usd. and just about 1040 to pay the unit off.

1040 payback is simply no good for a miner.

so I would argue 5 cents is the beginning of good for a miner.

and 6 cents is the beginning of bad for a miner.

for me I want better then 5 cents.  I would want 4 cents or lower.

[ColoMiner.com]

Thank you, I appreciate your insight.

What options are there for 4 cents or lower electricity? I've read about Greenland, China's wet/dry seasons, and a few other cheap power options but they seem out of reach for all but the largest mining operations.

[mikeywith]

IMO if you can't compete with miners in China such as those in Xinjiang and Inner Mongolia then making a profit, in the long run, will be quite difficult, miners in Sichuan have about 6 months (April to October) of about 1 cent per KhW, but I'd say the average year would be 3-4 cents for the majority of miners in China.

You might think that since the Chinese don't makeup 100% of the total hashrate and thus the average electricity cost is higher when considering other countries, well, the bad news is, there are many other countries with much cheaper power rate, such as Iran and most of the Middle-East countries, East-Europe has some very cheap rates, countries like Venezuela and Malaysia have even better rates.

Not being able to compete with those people in terms of electricity price puts you in great danger of losing money, people who paid 2.5-3k for S19 and mine at 5-6 cents are unlikely to ROI, let alone make a profit unless of course bitcoin price was to moon unrealistically, but in that case, you are better off buying the coin directly.

[ColoMiner.com]

... 3-4 cents for the majority of miners in China.

Do miners in the United States near hydroelectricity pay similar rates, or are there no US competitors in that price range?

... people who paid 2.5-3k for S19 and mine at 5-6 cents...

Would those miners be satisfied at 4.5 cents? Thank you for your insight as well.

[philipma1957]

As mining is getting more industrial with units designed to use 3000-3400 watts margins of profit get smaller.

I have been doing this since 2012. Yeah we had a nice spike for 6 months Sept 2017 to Feb 2018. But other then that nice profit spike not a lot after that.

We did have the 2013 spike which was shorter.

If I were to build a super mine I would want a way to keep my costs down for power. I am not tiny but with todays numbers I am small to low middles sized user under 120kwatts an hour.  We were higher before the ½ ing but we cut out all the s9's and the L3+ gear. which had us up to 160kwatts an hour.

I know how to make money at this ,but it is hard to find places with proper power. I may have found a new spot that has as much as 1000 amps at 240 volts for me which means 800 amps. We may go check the site out in 2 weeks.

He is further from me then I would like. About 2 hr 45 minutes. each way.

With him we would have all the same gear.

Maybe used s17pros. With aftermarket firmware on a down clock. I can do 26 watts a th so it is more efficient then a s19pro but does 38th.

BTW that is like taking 4 cent power and making it 2.8 cent power.  as 26/37 = .702 and   2.8/4 = .70

It is another way to compete.  Which is why bitmain makes it harder and harder to adjust settings.

In the USA under 4 cents in bulk is really hard to find.  It can be done.  What does happen a lot is you find a guy that purchased power on a 2-3 year deal. It is paid for and the guy lost a customer for that power.

So if he is buying 4000 x 240 = 960,000 kwatt say a mega watt an hour. for 3 years at 4 cents and he loses a sub contract for ¼ of that he would be willing to sell cheaper then 4 cents maybe as low as 3 cents. since he is selling the other ¾ at 5 cents. and 3 cents looks good as it is better then zero.

We got in on our last warehouse like this. the guy has 7 buildings some freezers and a five year contract. He lost clients. so we cut a deal with him.

[mikeywith]

Phill you should also consider cost per terahash, downclocking an S17 pro to 26w/th means you essentially paid a higher price, you pay $1000 for a used S17 pro which is supposed to do 54TH, you downclock it to 38th and then cost per terahash becomes  $26, so it's like paying $1421.

Compare that to S19 pro which is almost as efficient (29w/th) if you pay $3000 for it, then the cost per TH = 3000/110 = $27, so the price is almost identical, pros here: S19 is brand new, S17pro is used, cons: slightly less efficient.

Of course, this is based on $1000 vs $3000, which is just a rough estimate, if S17 pro goes for $800 and S19 pro $3800 then the maths will favor the S17 pro.

[philipma1957]

Phill you should also consider cost per terahash, downclocking an S17 pro to 26w/th means you essentially paid a higher price, you pay $1000 for a used S17 pro which is supposed to do 54TH, you downclock it to 38th and then cost per terahash becomes  $26, so it's like paying $1421.

Compare that to S19 pro which is almost as efficient (29w/th) if you pay $3000 for it, then the cost per TH = 3000/110 = $27, so the price is almost identical, pros here: S19 is brand new, S17pro is used, cons: slightly less efficient.

Of course, this is based on $1000 vs $3000, which is just a rough estimate, if S17 pro goes for $800 and S19 pro $3800 then the maths will favor the S17 pro.

yeah and if you have run the s17pro for a year already. downclocking it from 57th turbo to 38th

drops power from about 2600 to 980 watts the cost is 19th.

watts saved is about 1600.

so I have 17 of them.

buy 23 s17pro used for 20k

I am running 40 units at 980 watts lets pretend 40kwatts

And 40x 38 th = 1520th at 40kwatts

right now i do 17 x 2600 watts = 44.2 kwatts and get 17 x 57 = 969th

hell of an upgrade for 20k

vs buying 8 s19 pros for 24k. that is 880th using 26.4kwatts
keeping the 17 s17pros clock to 38th does about 650th
 using 17kwatts

so you are at 1530th and use 43.4 kwatts.

so getting the used s17pros about the same hash costs less money uses less power.

and if btc moons  your 1520 can jump back to 2280th.

if you have the s17pro s19 pro and btc moons you can clock to about 1900th.

if i am at 4.5 cents as the op asks it could work.
if i can have a reserved amount of power it would be an attaction.

knowing my underclocked setup can boost power at same 4.5 cent rate makes deal nice.

if the data center sells me 100kwatt for 4.5 cents and says you can add 50kwatt at same price for the next six months it could be an incentive.

[ColoMiner.com]

if the data center sells me 100kwatt for 4.5 cents and says you can add 50kwatt at same price for the next six months

100kw @ 4.5 cents is an absolute certainty. Ideally, additional power would always be available to you. There may be limited delays in availability as we expand overall power capacity.

I wonder if miners would appreciate one price for everyone? Even small miners, those with 1 single unit. Is 4.5 cents the magic number? Seems like a sweet spot for mostly everyone on this side of the world.

Can anyone beat 4.5 cents in real life without sailing the seven seas and crossing oceans to do so? I'd imagine you would have to be a fairly large operation if you could.

[philipma1957]

It is a good price. Especially for guys with a few miners.

I am sending a pm.

[mikeywith]

Would those miners be satisfied at 4.5 cents? Thank you for your insight as well.

I am not in the U.S so I can't give you an accurate answer, but according to my knowledge 4.5 cents in the U.S is below the average, so I think the should be satisfied, although, I am not sure if YOU will be satisfied as well, I am not aware of your mining background but hosting mining gears isn't easy and there are other costs involved, adding that to your initial power cost might put you at loss, so I would run the math a hundred times before getting involved.

And by the way, phill is great resource for these kind of information and he is in the U.S so he would help you better.

[DaveF]

I have posted about it before, but there are some places near hydro power in the NW US that you can get  $.05 power but you are paying for committed rate.

So it's a use it or loose it thing. Very popular with the server hosting, less so with miners. Mostly because of the lack of reliability. I can fairly certainly know that I am going to have an HP server pulling in X watts running in the data center for at least a year at a time. Miners, who knows. BTC drops I might want to shut some down. One or 2 die, I might be pulling 5000+ watts less for a month.

But if I asked for 100 amps I'm paying for 100 amps. Even if I pull all the equipment. Want another 100 amps fine here is the price. Never use it, not their problem you are paying for it for a year no matter what.

And on the flip side those kind of data centers suck for interconnectivity between providers.

-Dave

[ColoMiner.com]

... I am not sure if YOU will be satisfied as well...

We're still running fleets of S9s at a profit but there is serious consideration taking place about switching to a colocation service focus. Instead of replacing S9s with newer hardware, it may be wiser to invest in more infrastructure and the opportunity to provide US customers with an option that will allow them to be competitive again at 4.5 cents. We would service anyone, from people with one asic to companies running hundreds of miners. Larger scale customers could nearly be plug and play using their own shipping containers. We have multiple sites in place. Each site is fenced, gated, secured, insured, easily accessed and ready for you to pull up, drop your unit, and plug in to our system.

There are lots of things that need further consideration such as insurance for other people's equipment and other minor hills to cross like offering fan / board replacement or additional options, but the mountain of being able to offer 4.5 cent service in the first place has already been conquered.

... you are paying for it for a year no matter what.

We've discussed monthly, quarterly and annual contracts. With a goal of 4.5 cents for everyone, it may end up being monthly for everyone. Ultimately, we would be shooting for flexibility in our contracts and satisfied customers.

[philipma1957]

May I ask a few questions.

1) The general location the areas are in?
2) How does a miner send you his gear?

If data centers are all in one country say china shipping gear from USA to China = too costly

So for a USA guy to want to put gear in a china based data location he will need to source gear in china.

For instance  some large resellers are in our marketplace

this gear is in North America
https://bitcointalk.org/index.php?topic=5138318.0

this gear is in Asia
https://bitcointalk.org/index.php?topic=5187550.0
this gear is in Asia
https://bitcointalk.org/index.php?topic=5272509.msg55098419#msg55098419

A miner may want to source used gear from above and send to you.

[ColoMiner.com]

Absolutely. I welcome all questions as we speculate about moving our company in this direction.

1) All of our sites are in the United States, east of the Mississippi River. I'd rather not pinpoint any locations further at this time, but as a customer your site location wouldn't be a secret to you.

2) Dealing with such valuable equipment belonging to other people, I wouldn't have a problem with miners personally dropping off their equipment and viewing the site where their equipment will be running. We haven't solved the issue of insurance for customers and that may cause friction with people other than employees being on the site. Perhaps customers would be able to find insurance on their own as a different route. We're not exactly sure how to insure customers yet. That is one of the hills that have to be crossed in order to accomplish this business transition. Large miners could simply have their containers shipped to us directly or personally deliver to one of our sites themselves. Smaller miners could alternatively just send their miners though the mail. I really like the idea of a face to face meeting and a more personal approach to doing business. I think it would give everyone more peace of mind and cultivate better business relationships. That may not be what everyone wants and we would try to maintain flexibility so that a customer, if they choose to, can just send us their equipment and pay us, keeping the personal interaction to a bare minimum for those who prefer that way of doing things.

[philipma1957]

Well I am in New Jersey So when you are ready I would consider locating a few pieces with you to start.

Maybe 2 s17 with special firmware clocked to 1000 watts each  so 2 kwatts an hour or about 1500 kwatts a month. I know there are some people here that 4.5 cents could work for them.

[ColoMiner.com]

2 s17 with special firmware clocked to 1000 watts each  so 2 kwatts an hour or about 1500 kwatts a month.

I appreciate your input here. 4.5 cents is not a problem but how would we account for intricacies such as updated firmware on a few miners out of hundreds owned by other customers, causing those few miners to consume less electricity than the exact same model sitting next to it? There are obvious solutions and we could simply measure the actual usage to confirm the power consumption and then bill accordingly. But with that being said, we are only interested in becoming a mining colocation service provider. As such, ideally we would have monthly billing for the space customers lease at our facility and not for the actual electricity usage. It will always break down to 4.5 cents, but perhaps if it were to be billed in such a way with exact usage specifics... we might possibly be viewed as an electric company, which is not the type of service provider we are intending to become. Overall, it seems your situation may present a few extra steps in the process but it can be done.

I suspect these types of situations and others are why most providers often have long contracts and minimum requirements. We think we can disrupt that. What is going to blindside this operation and change the dynamics enough to convince us that we can't? Further speculation is welcome and would be appreciated. Thank you.

[DaveF]

There are obvious solutions and we could simply measure the actual usage to confirm the power consumption and then bill accordingly. But with that being said, we are only interested in becoming a mining colocation service provider. As such, ideally we would have monthly billing for the space customers lease at our facility and not for the actual electricity usage.

There are high end PDUs that will give you power usage by outlet.

There are bandwidth considerations too, who provides it? Remote access to miners? Remote hands for when something needs to be physically handled?

Switch hardware? Network segregation?

Warehousing and shipping and install?

-Dave

[mikeywith]

I appreciate your input here. 4.5 cents is not a problem but how would we account for intricacies such as updated firmware on a few miners out of hundreds owned by other customers, causing those few miners to consume less electricity than the exact same model sitting next to it?

You would have to sell your services in packages, the 4.5 cents Kwh is just a pretty rough figure for your clients to know how much they are paying, but then you would have to price every miner based on its power consumption, for example:

S9 > 14 TH > 1250w = (1.25kw*0.045$*24H) = 1.35$ daily
S9 > 12 TH > 1050w = (1.05kw*0.045$*24H) = 1.134$ daily

Obviously, you would want to limit the options you offer to perhaps 3 profiles per miners, so if you have a total of 5 different models you would have a total of 15 packages to sell from, this will work only if you don't allow the owners to remotely access the miner and modify the mining profile, if you want to give them access then you will have to use a meter, some PDUs like DaveF explained, but this will be treated differently.

You would have to force your clients to buy a fixed number of spots, for example, 1 PDU is good for 10 miners, then they have to buy your host service in batches of 10s, 15 won't work, 9 won't work either, if you don't do that, you will encounter an extra set-up cost as well as extra operational cost.

[philipma1957]

Here is a for instance I have a wire with a built in meter .  It has  hourly kwatt readout.

l30-p to>>>>> small accurate meter>>>>> to l30-r>>>> to 4 plug pdf >>>>>> to 2 s17pros.

l can provide that  and set my s 17pros to run at 2000 watts an hour = 1440 kwatts a month = $64.80 a month at 4.5 cents

I am in on that big time. I am east coast usa based he said he has east coast gear.

this is the pdu

https://www.ebay.com/itm/Lot-of-8x-HP-411273-002-HSTNR-PS03-Modular-PDU-Extension-Bar/164316086633?

it attaches to the cable  I have.   https://i.imgur.com/SUmBDs5.jpg. it reads kwatts so if sent at 1400 after a month should read close to 2840 or  1440 kwatts

so for a small miner. it is pretty bullet proof. setup. fairly easy to do 50 people with the exact same setup. or 1000 s17pros.

That power bar limits to 2800 watts.
The cable limits to 4000 watts.

2 miners set to do 38th at 1000 watts is easy peasy to get on a shelf and at 65 in power are earning about 165 or 100 a month profit.
I would be willing to have no direct access to gear but contact to 'help' would need action.
The setup I describe could need under 5 checkups in a year. As I have had s17pros run with zero issues for more then 100 days in a row.

If you want to use more power you can use this pdu
https://www.ebay.com/itm/HP-228481-002-EO4501-Modular-PDU-Power-Distribution-Control-Unit/233528878767?

and wire this meter to it

https://www.ebay.com/itm/EKM-Metering-Inc-Single-Phase-3-wire-kWh-Meter-EKM-25IDS/264878819515?

that can do a full 30amp circuit.

which means the s17pro could be set as high as 2400 watts each or 2 units at 4800 watts.

well within safety limits.

I know of no cheaper way to build a safe pdu then using the 2 above.  not counting labor you can have it for under 100 bucks. If you diy it.