← All decksPrepared for CarbonBridge · bioreactor digital twin · July 2026

Your reactor, made legible.

A live twin of CarbonBridge's stacking modular reactor, wired to what MESSAI can actually do for you. Move the stack; the data, models and economics follow.

Digital twin CH₄ → CH₃OH Scale by stacking Illustrative

01 · The twinModular direct-gas bioreactor

Drag to rotate. Each plate is a self-contained chamber with its own gas inlet; the stack scales by adding plates.

drag to rotate Feedstock layer · gas inlet ports
Reactor geometry modeled on CarbonBridge's publicly described stacking modular direct-gas bioreactor. Highlight follows the selected layer in section 03.
12.0kg / day
Modeled methanol output
95% CI 10.1 – 13.9 · conformal-calibrated
26gCO₂e/MJ
Carbon intensity
fossil methanol reference 95
−73%
vs fossil methanol
carbon-intensity reduction

Illustrative Demo digital-twin estimates, not validated CarbonBridge data. The method is real; the numbers are placeholders until your bench data is ingested.

02 · What MESSAI can actually doThree honest tiers

Not a four-equal-features pitch. One thing is ready now, one is a build worth doing, and one is squarely yours. We won't pretend otherwise.

Available now

Turn your bench data into ground truth

Every reactor run captured as a provenance-tracked, calibrated record: conditions structured, uncertainty attached, comparable across the team. "Which conditions gave our best runs, and why" becomes a query. Your ARPA-E and investor claims get defensible intervals instead of hopeful point numbers.

Doesn't need our corpus. It's our method on your data.
Buildable

Rank your feedstock sites

You deploy where the waste gas is. We model the biogas landscape and can fuse it with facility data (EPA CWNS, DOE AgSTAR) into a ranked, characterized site map: flow, methane fraction, contaminants, so units land where the feedstock supports them.

A build, not off-the-shelf. We'd scope it to one region first.
Stays yours

Your reactor, strains & science

Strain performance, gas-liquid mass transfer, the hardware breakthrough: that's your edge, and we have no data on it. We won't claim to predict your reactor better than you can. We organize and calibrate around your science; we don't replace it.

Naming this is the point. Trust beats overreach.
03 · See it on the reactorFour layers, one stack

Each layer touches a real part of the reactor. Select one to highlight it in the twin above and see what MESSAI brings. Everything recomputes from the current stack size and feedstock.

A

Feedstock intelligence

MESSAI · wastewater / AD surface

CarbonBridge sources the biogas that's flared today. Pick an archetype and the twin re-solves for its gas composition. MESSAI hands back a ranked, characterized site map so units land where the feedstock actually supports them.

62%
Methane fraction, CH₄
420 L/min
Typical gas flow
16,000+
Candidate sites (US)
04 · The closed loopFeedstock in, methanol out

MESSAI wraps the cycle you already run.

StageNodeRole
SourceWaste biogasWe map and rank the sites.
ConvertYour reactorYour modular direct-gas stack.
ProductLow-carbon methanolMarine-fuel grade output.
PredictModeling + TEACalibrated output, cost, carbon.
ValidateData & provenanceMeasured vs predicted, traced.
What this is not
  • All output, cost and carbon-intensity figures are illustrative digital-twin estimates from a transparent demo model, not validated CarbonBridge performance data.
  • The demo model is simple by design: methanol output scales as 2.4 kg/day per plate × (CH₄ fraction / 0.60); CI is ±16 %; carbon intensity falls 0.9 gCO₂e/MJ per plate from 31 to a floor of 18; levelized cost starts at $1.92/kg and drops $0.055/kg per plate to a cap of $0.62; module CAPEX is $4.8k per plate.
  • Numbers become real only once your bench data is ingested through MESSAI's provenance-tracked pipeline.
  • Strain performance, gas-liquid mass transfer and the hardware are CarbonBridge's science. We have no data on them and do not model them.