Grid dependency
Every kilowatt is rented
A grid-connected mine pays retail or industrial rates that rise with demand from data centres, electrification and heat. The machines are efficient; the power contract is the liability.
Concept and business plan · October 2026
Falling water
into sovereign compute.
HydroBit is a factory-built shipping container holding a small hydro turbine, hydro-cooled Bitcoin miners, switchgear and a satellite uplink. Set it beside a penstock, connect the pipe, and it mines. No grid. No fuel.
Capex per kW · 1 MW
$2,073
Down from $3,984 at 50 kW. Controls and integration cost the same on every box, so they spread over more kilowatts.
Gross revenue · 250 kW
$17,445/mo
At BTC $80,761, 90% uptime and free water power. Civil works, water rights and permitting excluded.
Sats per dollar of capex · year
510
At 1 MW, against 260 at 50 kW. The metric to watch if you intend to hold rather than sell.
The engine inside
HydroBit is the engine powering Echo.
Echo is a social network built so that no company can switch it off — the last social you'll ever need. HydroBit's compute, and the Bitcoin reserve it mines, are what keep it running.
See EchoThe problem
Compute got expensive the moment it got serious.
Mining Bitcoin on grid power is a bet that your electricity stays cheaper than everyone else's. That bet has been losing for years, and the cost of power is only heading one way. HydroBit takes the load off the grid entirely and puts it where the water falls.
Water has no fuel cost. So a HydroBit container is
cash-positive whenever it runs — and when power is worth more than
mining, an export port lets the site sell it instead.
Sovereign compute
Someone else's switch
Grid power can be curtailed, repriced or refused. A container on a penstock with its own generator and its own satellite link does not ask permission to run.
The arbitrage
11.7¢ against the market
A 100 kW container's output is worth roughly 11.7¢/kWh-equivalent mined, against 14.47¢ US commercial retail and $0.25–0.50+/kWh displacing remote diesel. The gap is the business.
The unit
One box. Water in, Bitcoin out, nothing else required.
A standard ISO container, integrated and factory-tested, that arrives complete. The containerized hydro plant is not a new idea — what is unbuilt is putting the mining load, its cooling and the satellite link inside the same box.
Hydraulics
Penstock flange, inlet valve, Pelton or Turgo runner — crossflow or Francis on low-head sites — jet deflector to survive a load rejection, tailrace outlet.
Generator
400 V three-phase synchronous, so hydro miners can be fed without a step-down transformer. Electronic load controller with a resistive dump load.
Mining bay
Insulated partition, racked hydro ASICs, closed glycol loop, coolant distribution unit and dry cooler. Hydro ASICs take 20–50°C coolant.
Electrical
Breaker, generator protection, metering, PDUs, surge protection and earthing — plus an export port for selling power instead of mining it.
Safety
Clean-agent suppression, smoke and heat detection, flood and leak sensors, E-stop, UPS for black-start.
Controls
PLC/SCADA, cameras and a Starlink uplink. Mining traffic and telemetry use a few GB a month, so the smallest business plan should do.
Ratings
50 kW to 1 MW. A 1 MW unit exceeds what one box holds usefully, so that size pairs a 40 ft power container with a separate dry-cooler skid. Smaller sizes stay single-box.
Scale economics
Larger is better, and the reason is boring.
Controls, safety and integration cost roughly the same on every box. Spreading them over more kilowatts cuts capex per kW from $3,984 at 50 kW to $2,073 at 1 MW, and raises the Bitcoin captured per dollar of capex from 260 to 510 sats a year.
| Container | Mining load | Gross / month | Capex | Per kW | Paybackflat difficulty | Sats / $ / yrpre-halving | Payback2028 halving at mo. 18 |
|---|---|---|---|---|---|---|---|
| 50 kW · 20 ft | 45 kW | $3,489 | $199,180 | $3,984 | 69 mo | 260 | 120 mo |
| 100 kW · 20 ft | 91 kW | $6,978 | $312,110 | $3,121 | 52 mo | 332 | 86 mo |
| 250 kW · 40 ft | 227 kW | $17,445 | $608,350 | $2,433 | 39 mo | 426 | 60 mo |
| 500 kW · 40 ft | 460 kW | $35,326 | $1,121,420 | $2,243 | 35 mo | 468 | 52 mo |
| 1 MW · 40 ft + skid | 925 kW | $71,088 | $2,072,820 | $2,073 | 32 mo | 510 | 46 mo |
Landowner-owner case at BTC $80,761, flat price and difficulty, after a 2% pool fee, $55/mo Starlink and 3%/yr O&M and insurance. Free water power; 90% uptime. Civil works, water rights and permitting are excluded and site-specific. Payback stretches to the far column if the 2028 halving lands at month 18. Every figure is a HydroBit estimate, not a quote, and no supplier quotes have been obtained.
$3,984 → $2,073
Capex per kW, 50 kW to 1 MW
260 → 510
Sats per $ of capex per year
2,160 kWh/day
Output of a 100 kW container at 90% uptime
0.07%
Network share of a 100 × 100 kW fleet
Site requirements
This box needs elevation, not a connection.
Output is P = ρgQHη, with water-to-wire efficiency estimated at 0.80. Low-head sites need a different machine entirely.
| Net head | Flow | Flow (US) | Output | Turbine |
|---|---|---|---|---|
| 30 m | 0.21 m³/s | ~3,300 gpm | 49 kW | Crossflow / Turgo |
| 100 m | 0.064 m³/s | ~1,010 gpm | 50 kW | Pelton / Turgo |
| 150 m | 0.085 m³/s | ~1,350 gpm | 100 kW | Pelton |
| 50 m | 0.64 m³/s | ~10,100 gpm | 251 kW | Francis / Turgo |
| 200 m | 0.16 m³/s | ~2,540 gpm | 251 kW | Pelton |
Head net of penstock friction; design for under 10% loss (estimate).
Minimum
30 m net head
50–300 m preferred. Below 30 m the economics stop working with this machine.
Flow
Firm, after bypass
Seasonal streams are the norm on the coast, and dry-season flow governs annual output. A year of gauged data is strongly preferred.
Penstock
Under 10% friction loss
A route that delivers the head you paid for, not the head you measured at the intake.
Access
Truck and crane
For a 20 or 40 ft box. A flood-safe pad and a clear sky view for the uplink finish the list.
Legal
The right to the water
The one requirement that cannot be engineered around. Every site needs a water right, and permitting is the long pole.
Candidate regions
Thirteen places on earth where the water already falls.
These are candidate regions for a first deployment, chosen on head, flow and the presence of high-gradient terrain. They are not secured sites, and no land, water right or permit has been obtained anywhere on this globe.
Candidate region
Land
Orbital track
- 01Pacific Northwest
- 02British Columbia
- 03Alaska
- 04Sierra Nevada, California
- 05California Central Coast
- 06Colombia
- 07Peru
- 08Chile
- 09Iceland
- 10Norway
- 11Nepal
- 12New Zealand
- 13Eastern Africa
Read this before the map. These are candidate regions, not secured sites. HydroBit holds no land, no water rights and no permits, and has not approached a single landowner or authority in any of these places.
Two pots, not one
The reserve is collateral. It is not the wage account.
Projects that want the credibility of mining without the arithmetic tend to blur these two numbers together. Keeping them apart is the whole point of HydroBit.
Bitcoin reserve
HydroBit mines Bitcoin continuously and holds it. That reserve is the collateral behind Echo coin — the asset that gives the coin something to stand on. It is not paid out as wages, and it was never meant to be.
A hundred 100 kW containers mine about 103.7 BTC in year one, worth $8.4 M at today's price. Real money, and none of it is a wage.
User wages
Wages come from Echosystem revenue: ads, data sales, marketplace fees, subscriptions and outbound transfer friction. That revenue has to exist before wages can be paid at any daily rate, and the honest way to say it is that the wage scales with what the revenue actually is.
HydroBit does not pay wages. It pays for the infrastructure that makes the wage possible later, if the revenue arrives.
The arithmetic nobody wants to print
A million users at $17.50 a day needs $6.39 B a year of Echosystem revenue.
Nobody should pretend that arrives in month one. For scale: a 100 MW fleet — a hundred 1 MW containers — mines roughly $85 M a year, and every dollar of it goes to the reserve, not to wages. Mining is not a wage line, and drawing it as one is the mistake this design exists to avoid. Two constraints hold: the same Bitcoin cannot sit in the reserve and be spent as wages, and no amount of mining substitutes for revenue.
Echo coin is plumbing. Users see dollars.
Users never see a token balance, a ticker, a chart or a wallet address. The interface shows dollars, and the coin does its work behind that number. A user with $412 inside the Echosystem has $412 of spending power inside it — and leaving with value carries a disclosed outbound transfer friction band of 30–50%, stated in dollars before she moves anything.
Design proposal, not a plan of record. Echo coin has not been minted, is not audited and has not been reviewed by counsel. A coin backed by a reserve, supported by issuer buybacks and described as gaining value from scarcity invites analysis as an investment contract. No coin should be sold, promised or marketed until securities, commodities, money-transmission, tax and consumer-protection lawyers have reviewed the design in each target jurisdiction. Early participants would be buying something with no established value.
Where power is worth more
Mine when it pays. Sell when it pays more.
A 100 kW container makes about 2,160 kWh a day at 90% uptime. What that is worth depends entirely on whether anyone nearby will buy it — which is why there is an export port in the design.
- Run the local load first. If a co-located customer will pay more than mining, sell to them.
- Mine the surplus. Water has no fuel cost, so mining is cash-positive whenever it runs.
- Shut down below the instream minimum. The stream's minimum flow is not negotiable for a few sats.
Status, plainly
Nothing has been built. That is the honest starting point.
Verified as of October 2026
Units built
None
Not a prototype, not a shell, not a mock-up in a yard.
Sites operating
None
No site has been surveyed, leased or commissioned.
Permits
None
Hydro is the most legally intensive renewable to site. No application has been filed.
Partners, customers, letters of intent
None
No commercial agreement of any kind exists.
Supplier quotes
None obtained
Cost figures are estimates benchmarked against published listings.
Team
Founder only
No employees, advisors or contracted partners. The intended hires are listed below.
Legal entity
Not yet formed
HydroBit has no company behind it today. Independence is the intent; the structure is unsettled.
Echo coin
Not minted
Unaudited, unreviewed, and not for sale or marketing in any form.
What this is
A concept and business plan
Which is exactly what a seed investor is being asked to fund.
Plan
First units targeted Q3–Q4 2027. That can slip.
The near-term milestones are supplier quotes and a detailed design, then one 100 kW pilot built and tested with measured performance published. Sales come before the fleet, because the buyer funds hardware, civil works and permits while HydroBit earns product margin plus monitoring and service fees. A fleet keeps the mining upside but needs roughly $31 M of containers before a single civil work.
Intended hires as funding allows: a licensed hydro or civil engineer for site studies, an electrical engineer for protection and controls, a California water-rights and environmental attorney, a China-based sourcing and quality manager, and a mining operations lead. None has been retained.
Now · Q4 2026 – Q1 2027
Supplier quotes, detailed design, and a pilot site with existing water rights identified.
2027
Build and test one 100 kW pilot. Publish measured performance. Open pre-orders only after acceptance.
Q3–Q4 2027
First units targeted. This date is the plan, not a promise, and it can slip.
2027–2028
Small-batch production and the first fleet tranche, subject to financing and site agreements. Economics re-run before every purchase, especially ahead of the April 2028 halving.
The ask
A seed round to fund the first unit.
The money covers going to China to build the first 100 kW unit, securing a site, deploying it and commissioning it. Everything after that is a function of whether that unit works.
Seed round · 2026–2027
$300,000 – $500,000
An opening range, not a fixed term sheet. The figure will be refined once supplier quotes and a site survey are in hand — which is precisely what the round pays for.
- Travel and build in ChinaFirst unit
- Turbine, generator and container fit-outHardware
- Site agreement, access and padSite
- Freight, installation and commissioningDeployment
- Measurement and published performance dataProof
No unit sale price is published. Units are sold only to vetted operators, by application, under license. The design is kept confidential because some engineering solutions inside the box are trade secrets.
Ownership and secrecy
HydroBit owns its fleet; it builds units for itself first. Units may be sold only to vetted operators under license, by application. There is no public price list, and these units are deliberately expensive at first, or not sold at all.
Structure and independence
HydroBit is not yet a legal entity. It is intended to keep some independence while serving as the engine powering Echo — the last social you'll ever need — built with Grok. HydroBit mines and holds BTC as the reserve behind Echo coin. It does not pay wages.
Licensing the design
Other ventures that want to model Echo's structure can license the design. That is a separate conversation, by application, and it is not a product you can buy from a page.
What a seed investor is buying
A founder, a design, and a documented plan to prove it with one unit. Not revenue, not a customer base, and not a token. The downside case is that the unit does not work, and that case should be priced accordingly.
Introductions
If you invest in hard things before they exist, let's talk.
HydroBit is looking for a small number of seed investors, and for hydro operators and landowners with existing water rights who want to be the site of the first unit. Both start the same way.
Gabriel Marciel, Founder · capmarciel805@gmail.com
Every inquiry is read by the founder. There is no sales team, no deck gate and no funnel. If the timing is wrong, the honest answer is that the timing is wrong.