SENSE
Level, dual flow, rain
Pressure level sensor. Inline flow meters on both the roof inlet and the outlet. Tipping-bucket gauge for on-site rainfall. Dual flow is what lets the conservation check close.
A metered, verifiable rainwater network across the Hill Country. The hardware on the tank, the proof layer above it, the credit it issues, and a permanent trust that returns the value to the people who catch the water.
Eight layers between a roof and a check. Each one is a separate failure mode, so each one is specified separately.
Every quantity carries a confidence marker. A number without one is incomplete copy, not a claim.
One enclosure, mounted at the cistern. The tank and the install are the real cost, not the electronics. Each node is minted to the registry as a deed to a catchment; roof area is verified once at install and caps issuance forever.
Pressure level sensor. Inline flow meters on both the roof inlet and the outlet. Tipping-bucket gauge for on-site rainfall. Dual flow is what lets the conservation check close.
ATECC608-class secure element. Every reading signed on-device with a monotonic counter and a device-unique key burned at manufacture. This is the entire trust root.
LoRaWAN over Helium's existing IoT network, already covering Central Texas, for pennies in Data Credits. Do not build a radio network. That lesson cost someone else hundreds of millions.
Enclosure switch, accelerometer, sealed sensor run. Any tamper event marks readings unattested until re-inspection. The meter will be attacked, because it will be worth attacking.
The catchment train the meter sits on is standard practice, not invention: screened gutters sized for local peak intensity, a first-flush diverter, a calming inlet that does not stir the sediment layer, a floating extraction filter on the outlet, and an overflow sized at least as large as the combined inlet.
The attack that matters is a hose in a tank. If a node can claim gallons that never fell, the credit is worthless. Four independent checks, all of which must pass — and credits issue only on beneficial use at the outlet, which kills tank-farming by construction.
Live model, mirrored from sim/rainsim.py. Honest nodes issue; the hose fails WEATHER and NEIGHBORS at once; the tank-farmer passes the physics and still earns nothing, because credits require beneficial use at the outlet.
Not the homeowner. The installer pulls the permits, signs the workmanship, and carries repeat-business risk in a small market. Verified fraud burns the bond.
2–3% of nodes per quarter, drawn at random, inspected in person against the signed record.
Unsigned or unattested firmware marks every downstream reading unattested until re-inspection.
Four checks are a floor, not a moat. Two upgrades change what the company is, not just what the device does: re-found the credit on a standard corporations already audit against, and fuse signals from physics domains a fraudster would have to fool all at once.
Re-found the credit on WRI Volumetric Water Benefit Accounting 2.0 — the standard corporate water-stewardship teams already trust. It quantifies beneficial hydrology change in cubic meters with a documented six-step baseline-and-additionality method, and it explicitly prices a cubic meter higher in a stressed basin. The Trinity–Edwards region is exactly such a basin, so a Dripping Springs credit is worth more under VWBA than a generic one.
The effect: buyers become corporations with published water-positive targets instead of individuals, and the meter becomes the automated, tamper-evident data layer under a recognized accounting standard. That is a far bigger moat than a proprietary credit.
A fraudster now has to simultaneously fool four different physics domains, a bonded installer, a permit record, and a registry reconciliation.
Ingest the forecast, compute expected inflow, release exactly enough now to capture it, and log the released volume as flood-peak reduction. The same water and the same meter earn under two distinct benefit types — supply and detention. Proven in existing stormwater practice, so it is the fastest of the three upgrades to ship. Guard: a hard firmware floor reserve, so a drought cannot drain a supply tank.
The firmware, the registry integration, and the trust instrument are specified and not yet fielded. There are no signed credit buyers at volume, no adopted ordinance, and no third-party validation of the measurement method. This phase exists to convert those from thesis into evidence.
One credit equals 1,000 verified gallons, beneficially used. The mental model is a Renewable Energy Certificate: issued on measured production, sold once, permanently retired on claim. Retirement is the difference between an instrument and a casino.
| Object | On chain | Why |
|---|---|---|
| Node registry | YES · NFT PER NODE | A deed to a catchment. Holds the roof-area cap that limits issuance forever. |
| Epoch Merkle root | YES | One hash anchors every signed reading in the epoch. |
| Issuance + retirement | YES | Double-counting prohibited by construction. Vintage year stamped at mint. |
| Raw telemetry | NO | Object storage, hash-anchored. Sensor data on a blockchain is the classic expensive mistake. |
| Settlement | DOLLARS | Nothing here requires a volatile token, so nothing gets one. |
In the Firma Elemint architecture this is e(H2O): an on-chain fingerprint of verified water reality, sibling to e(photon), which mints per verified kWh. Same doctrine — audited physical production, minted as proof, never marketed as an investment.
Chain: firmamint or Solana. Do not build a chain. Helium itself migrated rather than keep maintaining one.
A credit paid for by its own issuance is a subsidy with extra steps. Real buyers, ranked by how close the money already is. The first row is the anchor tenant: a legal obligation with a price attached, 25 miles from the catchment.
| Buyer | Why they buy | What has to change |
|---|---|---|
| Large Austin developments | 250,000+ sq ft projects must install onsite reuse since April 2024 VERIFIED. Austin Water pays up to $1.5M per project in incentives VERIFIED. | An offsite-offset pathway in Code Ch. 15-13. Cheaper per gallon than a basement cistern plant. |
| Groundwater districts | Hays Trinity GCD has held roughly 40% curtailment since 2022–23 with only blunt tools, and is in litigation over pumping above allocation VERIFIED. | A district rule recognizing credits as offsets. A board vote, not legislation. |
| Utilities and WSCs | "Negagallons" at $2/kgal beat new supply at any price PROPOSED. | A demand-reduction procurement line. |
| Corporate water-positive | Data centers hold public replenishment commitments VERIFIED. Metered beats modeled. | Nothing. This demand already buys lower-quality attributes elsewhere. |
Annual value per 10,000 sq ft catchment — roughly 174,000 gal/yr PROPOSED — and the payback on a $25K install as each layer stacks. The honest conclusion: residential retrofits do not pencil today. Large roofs do.
All four rows are PROPOSED — modelled from the standard yield formula and current water rates, not measured on a fielded node.
A single 50,000 sq ft roof yields about 2.7 acre-feet a year from one install PROPOSED. Bee Cave Builders' roughly 20 active projects are the pilot fleet, already standing.
0.623 gallons per square foot per inch of rain is the volumetric constant VERIFIED. C is the runoff coefficient: 0.75–0.95, and 0.90–0.95 for smooth metal.
A $25K residential retrofit does not clear its own payback at today's credit prices. The trust's tanks-not-checks dividend is the mechanism that fixes that, not a better spreadsheet.
Do 2,000 homes catch more rain than 20 warehouses? Yes — about 87M gallons against 17M PROPOSED. Gross rainfall favors the homes. What the warehouses win is cost per verified gallon.
| Axis | 2,000 homes | 20 warehouses |
|---|---|---|
| Sales conversations | 2,000 | 20 |
| Installs and inspections | 2,000 | 20 |
| Audit surface | 2,000 NODES | 20 NODES |
| Install cost per sq ft of catchment | High — retrofit | A fraction — tank before slab |
| Time to a provable number | Years | One quarter |
The strategy is not "warehouses instead of homes." It is warehouses first, so the trust gets funded, so the Commons dividend can buy homes their tanks later. The fund before the fleet.
Satellite plus permit records. The number is knowable before anyone commits to anything.
Node and tank under a lease model — the trust or a lessor owns the hardware, the church hosts it.
Paid on verified beneficial use. Real money into the general fund, from a roof that was already there.
An Alaska Permanent Fund for rain. 25% of every credit sale goes into a corpus that can never be spent, only invested — written into the trust instrument, not left to an annual vote. Alaska's fund holds $80B+ and has paid a dividend every year since 1982 VERIFIED. Copy the structure, not the vibe.
Paid to node holders proportional to verified gallons beneficially used. This is what makes people want a tank.
Early years: the residents' share funds fully subsidized installs for households that could never afford a $25,000 system. A check spends once. A tank yields for thirty years.
A board seating the county, the groundwater district, landowners, and node holders — with per-holder caps so no large-roof operator captures the commons.
One signed LOI from a non-crypto buyer, priced in dollars per thousand gallons, plus three instrumented roofs, inside 90 days. No LOI, no network.
A rainwater credit is only worth what its measurement is worth, and the weakest point in any environmental credit is the party who benefits from overstating the number. The installer is the one actor in the chain who cannot afford to lie.
Matt Spannaus / Bee Cave Builders posts the bond, pulls the permits, and signs the workmanship. If a tank underperforms or a meter reads high, the callback lands on his crew and his warranty — in a small market where reputation is the whole business. We are not asking him to be honest as a favor. His economics already punish an inflated number harder than any auditor could.
He is not a vendor recruited to a pitch. He reached the water problem on his own and called metered rainwater capture one of the top solutions to the Texas water crisis before we wrote a spec.
The Hill Country is building homes faster than the Trinity Aquifer can legally supply them.
Austin's onsite reuse ordinance is enacted and being permitted 25 miles away.
Data-center water demand is landing in Texas with public replenishment commitments attached.
RECs proved the instrument. Voluntary carbon proved what bad measurement costs.
Why it is inevitable. Texas will keep building where the water is already curtailed, because the growth is not optional and the aquifer will not stretch. Somebody is going to require capture, and Austin has already shown the region how to write that rule. Once capture is required it will be metered, because unmetered mandates are unenforceable. Once it is metered, the captured volume is a countable, tradeable asset, and someone will build the ledger that counts it. The only open questions are who builds it, whether it is honest, and whether the value returns to the watershed or leaks out to middlemen.
If credits are bought mainly by token holders, this is a Ponzi with a rain gauge. Test: one non-crypto buyer signs an LOI in dollars per thousand gallons before a single node ships.
One viral hose video ends the program politically. Over-invest in the attestation stack and the audit program from day one. It is cheaper than the recovery.
SEC v. Nova Labs (April 2025) dismissed with prejudice the claim that Helium's hardware and token distribution were unregistered securities VERIFIED. Favorable, fact-specific, not a license. Market the device as equipment, never as an investment.
Diffused surface water belongs to the landowner while it stays on the land. The moment it reaches a watercourse it becomes state water. Sell the credit, not the water.
Non-potable only for the first two years: irrigation, livestock, cooling, flushing, construction.
Pitched as a crypto project in Hays County, this dies. Pitched as private property rights, local control, metered accountability, and no new taxes, it is one of the more conservative ideas in Texas water. Say the meter line in every room.