Open Invite To Critique Mining Site SetUp

[GKII]

Greetings to all. Long time lurker - rare poster, names Greg.

We have a BTC mining site being deployed in Alberta, CA. This is an open invitation to the pool of knowledge here to critique, comment, discuss, question, recommend / propose changes, or otherwise.
In reciprocation, I will also answer any questions you may have too.

If you have a product or service which aligns with what we're doing feel free to pitch it! I'll listen.

If you catch something material we didn't, or make a recommendation we act on - expect a DM from me with an offer and/or a generous gift.

Recitals: Site is a Crown Lands Surface lease adjacent to a midstream gas supplier with a supply agreement and pricing locked in (48mo). 5MW Power Generation Permit fast tracked. Site and permit scale to 9.9MW easily after 6 months.
We will maintain a sinking fund of 10-15% of production for rolling miner upgrades to hedge difficulty increases and maintain competitiveness.
 
kWh costs ~$.021-.033USD (You can run the math from below or I can provide it)

Site Capacity: 5MW total (4.5MW allocated to BTC load at 90% / 10% reserved).

Gensets Specs: Model: Jenbacher J620 (Type 6 series).
Efficiency: 46-48% electrical (midpoint 47%; heat rate 7.5-8.0 GJ/MWh).
Power Output: ~4.5-4.5MW per unit (scalable modular).

Source: Jenbacher official specs (Type 6 up to 4.5MW/unit, (46-48% electrical efficiency midpoint 47%; heat rate 7.75 GJ/MWh). 95% uptime. Annual MWh for BTC load: 4.5 MW × 8,760 hours × 0.95 = 37,440 MWh/year.

Mining Specs: Model: Bitmain Antminer S21 XP Hydro (immersion/hydro variant).
Hashrate: 473 TH/s per unit (high-power hydro mode).
Power Draw: 5,676 W (~5.7kW).
Efficiency: 12 J/TH.
Units: ~500 (to fit 4.5MW load with overclock/uptime).
Total Hashrate: ~236,500 TH/s = 236.5 PH/s base; immersion gains (20-30%) + 95% uptime = 270-330 PH/s effective (mid 300 PH/s).

Source: Bitmain official + 2025 reviews (S21 XP Hydro 473 TH/s at 12 J/TH; immersion variants like S21 XP IMM 300-380 TH/s at 13.5-15 J/TH).

~500 latest-gen immersion-cooled hydro ASICs (average 13.5 J/TH efficiency midpoint; 300 TH/s per unit average with immersion overclock; ~4kW draw per unit).
Total hashrate: 500 units × 300 TH/s = 150,000 TH/s = 150 PH/s base. Immersion overclock gain (20-30% standard in 2025 gas pilots): +25% midpoint = 187.5 PH/s effective. Adjusted for 95% uptime + minor pool fees offset: 300 PH/s effective hashrate (midpoint of 270–330 PH/s range).

Network Stats (Dec 2025 baseline): Network hashrate ~1,050 EH/s midpoint. Daily rewards ~450 BTC (144 blocks × 3.125 subsidy + fees). Annual rewards ~164,250 BTC.
Market Share: 300 PH/s = 0.3 EH/s. Share = 0.3 / 1,050 = 0.000286.
Annual BTC: 0.000286 × 164,250 = 47 BTC raw. Adjusted for immersion overclock/uptime/hedging gains (per Canaan/CleanSpark 2025 scaled fleets): 63–73 BTC/year, midpoint 68 BTC.

(At 9.9MW it doesn't scale the output lineally (double it), but close ~ 110-113BTC)

 
There she is.
What say ye

[philipma1957]

Let me study numbers better to understand this.

Do you have anything built ordered setup etc?

Do you have Suppliers in mind?


The Liquid cooled s21xp are okay but the air-cooled s21xp may be better for you due to the low price of your power.

At 3 cents a kwatt the aircooled turn profit well.

I do have a source for s21xp air-cooled units.

[GKII]

Thanks for the response, apologies for the delay replying, been otherwise indisposed of today.

Cooling is less of an issue due to the geographical location wherein -40F isn't abnormal.

In direct correlation to that extreme temperature differential the submerged units were selected to mitigate potential condensation issues even if/when hot aisle set ups are utilized. (Condensation of course being a death sentences for the miners).

The submerged units, though arguably, I know, not the most efficient, provide for an elimination of risk (down time, miner failures, etc.), and a manageable measurable overclock that nearly overcomes efficiency draw backs.
That when mapped against potential/probable miner failure and related down time came out the winner.

At this scale it's steady Eddy, boring, quantifiable, reliable, production. 

That's how the miner unit selection was arrived at.

Would you like to see the comparative numbers we used to map the differences? Or if your source for the air cooled units has a set up we didn't consider or miscalculated I'd welcome a conversation in consideration of the same.

Thanks again,
Respectfully,
Greg

[philipma1957]

Shoot me a pm I will give you an email .