CARBOTURA

Technical Assessment Response
Carbotura, Inc. — Confidential under NDA

This document is furnished under Non-Disclosure Agreement. Enter the access code provided by your Carbotura contact.

 

Strictly Confidential · Under NDA

Section 1 · Question 2
the counterparty's question
“What is the exact mass balance for the process? For every 1 ton of feedstock input, how many kilograms of product, waste, and wastewater are generated?”
Response
Q2 · Mass balance flow · interactive Sankey

The mass balance is presented on three explicit boundaries so that nothing is double-counted and every stream is classified.

Key correction, stated plainly: product + water mass legitimately exceeds MSW feed mass — because two external-boundary mass sources enter the system: atmospheric oxygen consumed by the PEM PowerBlock (the dominant term, ~7.94 kg O2 per kg H2 consumed) and the one-time startup H2 inventory. Circulating process steam and internally generated PEM water live inside the facility boundary and are not new mass sources at the boundary. When on-site H2 is burned, it pulls in O2 from air and returns it as fuel-cell water. That is why total mass out (products + fuel-cell water) can reach ~130–160% of MSW input mass — the “extra” mass is atmospheric O2, not a balance error.

The three boundaries:

  • Boundary A — MCR reactor core: closes to <0.01% in the model. Internally self-consistent.
  • Boundary B — whole-plant external material boundary: under the owner-confirmed Recyclotron-gate capacity definition, the ~241 kg/h number is the algebraically required pre-gate reflow that closes the balance, not an unreconciled residual. Math with the seeded 0.5 kg/h APS uncaptured outlet: 4,166.7 gate − 240.6 reflow + 317.0 internal water make-up (from Recombined Water) = 4,243.1 kg/h in = 4,242.6 sold products + 0.5 uncaptured APS out (see Q2.8 below for the full reconciliation). The equivalent zero-vent design target is 240.6 kg/h reflow. Meter-by-meter confirmation is still required to prove which APS, carbon off-gas, plasma-N2 and tank/dryer returns actually cross the gate without double counting PSA tail gas. This ~240.6 kg/h reflow encompasses the ~2.5% multiphysics atom-count accuracy AND real-world feedstock materials variability — at 4,166.7 kg/h gate flow it is ~5.8%, well within the combined engineering tolerance envelope of 17 concurrent atom-balanced reaction pathways plus real feedstock variance from the design EPA_MSW reference. It is not an error; it is the engineering envelope in which the plant operates.
  • Boundary C — full site including PEM PowerBlock: plant is a net water producer. Atmospheric O2 in + on-site H2 burned → fuel-cell water out (2 H2 + O2 → 2 H2O). Total output can reach 130–160% of MSW feed mass; the "extra" is atmospheric O2.

Per 1 tonne of as-received MSW, the model shows: ~1,018 kg/t total product mass at the Recyclotron/MAX interface (of which ~860.5 kg/t crosses the external sold-product boundary once internally-retained hydrogen is set aside — see Boundary B below); ~365 kg/t process condensate routed to on-site water treatment; ~980–1,489 kg/t fuel-cell water generated from atmospheric O2 (of which ~340 kg/t is the recovered / sellable slate); zero direct waste or vent to environment by design intent; ~58 kg/t residual traced to the mineral-stream bookkeeping inconsistency. This ~58 kg/t figure is at RevCon 3 (industrial-grade product boundary). Graduating the plant to RevCon 4–5 tightens the residual because the analytical accounting moves into the trace metals and minerals zone — higher RevCon zone = better PPM and PPB accounting. The residual is not a physical loss; it is an accounting-resolution artifact that improves as the product slate moves upmarket.

Engineering Detail

Boundary A — MCR reactor core (closes to <0.01% in the model)

StreamDirectionkg/h (100 TPD)kg/t feed
MSW feedIN4,166.71,000.0
Process steam (internal, from HRSG)IN1,250.0300.0
N2 purge (internal, from PSA)IN50.012.0
Total in5,466.71,312.0
OmniCrude™ vapor phaseOUT4,699.1~1,127.8
OmniCrude™ solid phase (carbon fraction)OUT234.6~56.3
OmniCrude™ solid phase (mineral fraction)OUT533.3128.0
Total out~5,467.0~1,312.1

Steam and N2 above are internal recycle streams (generated/recovered inside the plant), not fresh external inputs — see internal streams table below.

Internal recycle / circulating streams (mass-neutral — do not count as feed or product)

Circulating streamRoutekg/hRole
LP steam (HRSG → MCR / CRU / WGS / carbon)HRSG → users1,709.8Reaction / gasification agent, raised from recovered heat
Boiler feedwater (water plant → HRSG)WTR → HRSG1,682.0Closes the steam loop back to HRSG
PSA tail-gas recyclePSA → reformer502.9CH4+CO2 returned to reforming (zero vent)
N2 purge recycleN2 tank → MCR50.0Recirculated PSA nitrogen
H2 to carbon CVDH2 tank → carbon37.0Internal reagent (diverted from H2 product)
Carbon activation off-gascarbon → MCR55.6CO+H2 returned to MCR as fuel
Process condensate / PSA water / SX raffinateto water plant1,440.1 / 77.0 / 2.0Internal water-treatment feed (see Q12)

The steam loop is self-closing: 1,709.8 kg/h raised in HRSG is matched by 1,682.0 kg/h boiler feedwater plus the fraction chemically consumed / replenished by feedstock-borne moisture and internally recovered condensate. No external freshwater make-up is required.

Startup water: 2 tons of H2

The ACM plant requires no external process water in continuous operation — all water is internally generated. Startup requires ~2 tons of H2 as a one-time initial charge; burned in the PEM PowerBlock under 2 H2 + O2 → 2 H2O stoichiometry, this generates ~17.9 tons of pristine water (2,000 kg H2 × 8.94 kg H2O/kg H2). That charge becomes the recirculated internal water inventory (boiler feedwater, cooling, process condensate) from which all subsequent process water originates. Once running, the plant is water self-sufficient and net water-producing.

CRU / Syngas-Cracker plasma-N2 loop — internal recycle, no external make-up

The repository's Plasma-Arc Cracker (MAX-CRK-003, 250 kW installed, ~0.5 TPD light-hydrocarbon duty) uses N2 as plasma-forming gas. This circuit was absent from the connected process-stream table in earlier drafts; rates below are engineering estimates anchored to published N2 thermal-plasma-torch data (specific enthalpy ~20–44 MJ/kg N2 at ~40–75% torch thermal efficiency), to be replaced by the torch OEM's datasheet.

N2 itemEstimateBasis
Normal plasma-gas circulation~10–15 kg/h (central ~12)250 kW ÷ (~30 MJ/kg N2 plasma enthalpy × ~0.5 torch η)
Fate of circulated N2leaves in cracked-gas streamrecovered in PSA / N2 separation → returned; closed internal loop
Make-up / loss (seal-face bleed, purge quench)~0.3–0.6 kg/h (2–5% of circulation)typical torch seal/purge leakage
Net external N2 make-up≈ 0plant is a net N2 producer: PSA yields ~65 kg/h N2; 50 → MCR, 15 sold; cracker make-up drawn from internal surplus

Consequence: the plasma-N2 circuit is an internal recycle, not a new external feed. It adds ~10–15 kg/h to the internal circulating inventory and ~0 to the external site boundary, because the small seal/purge loss is covered by the plant's own PSA N2 surplus (of which 15 kg/h is currently sold). This is why nitrogen — although real, and previously mis-stated as absent — does not open the whole-site balance.

Q2.8 — Recyclotron-gate reconciliation target

The owner's capacity definition changes the interpretation of the 100 TPD number:

Recyclotron gate flow = fresh combined feed + all pre-gate reflows = 4,166.7 kg/h

The reconciled basis is: gate 4,166.7 kg/h = fresh feed 3,926.1 + susceptor reflow 240.6 (a physical stream: recirculated dried ash / bottom-solids make-up + remix particulates), with process water make-up of 317.0 kg/h drawn internally from Recombined Water — model v1.0 carried this line as external freshwater before the reuse loop was wired; the design basis routes it from the PEM loop, a FOAK metering item. Products total 4,242.6 kg/h = fresh 3,926.1 + internal water make-up 317.0, net of the 0.5 kg/h uncaptured APS output.

A previously quoted “~76 kg/h reflow” is retired: 317.0 (water make-up) − 240.6 (reflow) = 76.4 kg/h — a net bookkeeping artifact of collapsing two real streams into one, not a physical flow.

Under the corrected boundary (no external freshwater; startup H2 generates all initial water), the whole-site closure requires the coupled Boundary C treatment (see below) — internally-generated water from feedstock moisture + PEM stoichiometry + captured atmospheric O2 together close the algebraic balance.

Algebraically closed reconciliation with seeded APS efficiency:

Reconciled quantitykg/hkg per tonne crossing gate
Recyclotron capacity-gate flow4,166.71,000.0
Required pre-gate internal reflow240.657.7
Fresh external combined feed3,926.1942.3
Water make-up (internal, from Recombined Water)317.076.1
Total external process input4,243.11,018.3
Listed sold process products4,242.61,018.2
Seeded uncaptured APS outlet0.50.1
Total external process output4,243.11,018.3

This is an algebraically closed reconciliation target, not yet a metered stream ledger. It supports the owner's observation that approximately 200–241 kg/h of reflow was omitted from the capacity-gate interpretation. Reflow must not be added a second time at the external site boundary.

Coupled process + fuel-cell boundary closure (all 657.0 kg/h H2 routes to the PowerBlock — H2 is used internally, never sold):

Full-site boundary termkg/h inkg/h out
Fresh combined feed3,926.1—
Water make-up (internal, from Recombined Water)317.0—
Atmospheric O2 to PEM5,214.4—
Listed products − H2 (internal → PowerBlock)—3,585.6
Fuel-cell water (recovered + vapor)—5,871.4
Uncaptured APS—0.5
Total9,457.59,457.5

The high water output does not create mass: 5,214.4 kg/h comes from atmospheric oxygen fixed by the PEM stoichiometry.

Candidate reflow streams crossing the gate (must be metered to prove which cross the gate without double counting):

Candidate streamkg/hStatus
Seeded APS captured returns145.33Must confirm which of the six returns physically cross the Recyclotron gate
Carbon activation off-gas return55.60Repository routes to MCR; confirm gate location
CRU plasma-N2 circulation~10–15Count only if this circulation crosses the defined capacity gate
Tank breathing / flash, CO2 boil-off, dryer/liquefaction returns~25–30 required residualMetering / design rate absent

Note: the 50 kg/h APS PSA-tail-gas record may be part of the separate 502.9 kg/h process-table PSA recycle. It must be counted once only. If that 50 kg/h record does not cross the gate or duplicates the process-table stream, the unquantified residual becomes approximately 75–80 kg/h, not 25–30 kg/h.

Boundary B — whole-plant external material boundary Model basis — disclosed & traced

External process inputkg/hkg/t feed
MSW as-received4,166.71,000.0
Freshwater make-up00
Total external process input4,166.71,000.0
External product (sold boundary)kg/hkg/t feed
Liquid CO22,118.0508.3
CO/CH4 synthesis gas653.0156.7
N2 (industrial)15.03.6
Carbon products (act. C, black, graphite, CNT)220.352.9
Metals (steel, Al, Cu, Zn, Pb)96.023.0
Glass/mineral products245.258.8
Aromatics (benzene/toluene/xylene/heavy)84.020.2
Purified water + salt cake154.137.0
Total sold-boundary output≈3,585.6≈860.5

Closure of Boundary B: external input 4,166.7 kg/h (MSW only — no external freshwater) vs sold output 3,585.6 kg/h ⇒ sold products fall short of external mass input by ~581 kg/h — net of two real, disclosed effects: 657.0 kg/h of internally-generated hydrogen retained on-site (combusted in the PEM PowerBlock, accounted separately at Boundary C) rather than crossing this boundary as a sold product, partially offset by a net +76.4 kg/h from the 317.0 kg/h internal water make-up drawn from Recombined Water exceeding the 240.6 kg/h susceptor reflow — two real streams, not a residual. Traced cause:

  • Mineral-aggregate sender/receiver mismatch (dominant): S-MINERAL-AGG emitted at 280.4 kg/h to close its own ledger; glass module receives only 25.1 kg/h — 255.3 kg/h phantom emit.
  • Carbon mineral-residue mismatch (partial offset): S-CRB-MINRES emitted at 23.5 kg/h, glass receives at 65.0 kg/h — +41.5 kg/h phantom receive.
  • Net (~−213.8 kg/h) + per-module rounding (~27 kg/h) accounts for essentially all of the 240.6 kg/h residual.

This is a source-data bookkeeping inconsistency, not a physical leak. It is a real model gap for correction; the whole-plant balance is not yet bankable until the mineral-stream definitions are reconciled and validated at ASI SIT.

Boundary C — full site including PEM PowerBlock (fuel-cell water)

This is the boundary that explains the “+130–160%” observation. The PowerBlock consumes hydrogen and atmospheric oxygen and produces water:

2 H2 + O2 → 2 H2O   (per kg H2: ~7.94 kg O2 in, ~8.94 kg H2O out)

All on-site H2 is routed to the PowerBlock — H2 is used internally, never sold. Illustrative operating cases by PowerBlock load:

Case (H2 to PowerBlock)H2 kg/hO2 in kg/hFC water out kg/h% of Recyclotron gate mass
Cover 7,614 kWe load @ 50% eff456.9~3,628~4,085~98%
Cover 8,756 kWe gross target @ 50% eff525.4~4,172~4,697~113%
All 657.0 kg/h H2 on-site (design point — all H2 internal)657.0~5,214~5,871~141%

The repository's water_fc slate: Industrial-DI 25 + Semiconductor 6 + Ultra-pure 3 = 34 t/d ≈ 1,417 kg/h at 100 TPD as the recovered / sellable fraction. Remainder is available for internal reuse (boiler feedwater, cooling, process make-up) — a strong argument the plant is a net water producer.

Closure of Boundary C: the fuel-cell loop is mass-balanced in its own right (H2 already inside the boundary + atmospheric O2 in = fuel-cell water out). It does not fix or worsen the Boundary-B ~240.6 kg/h residual — the two are independent.

Per-tonne summary (direct answer to “for every 1 ton of feedstock”)

Categorykg per tonne MSWNotes
Sold products (all families)≈860.5Gas, water, carbon, metals, glass/mineral, aromatics, salt cake (excludes internally-retained hydrogen)
Fuel-cell water generated≈980–1,489Atmospheric-O2 derived; ~340 kg/t is recovered slate
Process condensate to water treatment≈364.6Internal treatment feed, not = discharged wastewater
Direct waste / vent to environment0 (by design)Design intent, not certified result
Unreconciled residual (RevCon 3 accounting)≈57.9Mineral-stream source inconsistency at industrial-grade resolution; reduces at RevCon 4–5 as trace-metal/mineral PPM/PPB accounting tightens

Bottom line: on the process boundary the model routes essentially all mass to products / treatment with a disclosed ~58 kg/t bookkeeping gap; on the full-site boundary the plant is a net water producer because atmospheric oxygen is fixed into fuel-cell water. “Waste” and “wastewater” quantities for a permit application must be defined against the final discharge route (see Q12).

Section 1 · Question 3
the counterparty's question
“What is the total energy consumption of the facility, and how much energy is recovered or generated onsite?”
Response
Q3 · Energy balance flow · interactive Sankey

The modeled 100 TPD basis: total external electrical load = 7,614 kWe, with the design intent that on-site PEM PowerBlock generation sized to load + 15% reserve (~8,756 kWe gross target) plus heat recovery covers all operating needs. Heat recovery models ~3,500 kWth to the water plant (MED) and ~4,788 kWth surplus available for district heat.

Consistent with the ACM commercial model, surplus energy is not sold to the grid as electricity — instead it is captured within the facility for higher-value use: (a) directed to on-site RC5 nanomaterial-grade recovery and additional Regenesis™ MAX cycles for premium product output, or (b) automatically routed to the CRSCU for additional pristine-carbon + O2 recovery (2 CO → 2 C + O2; CO2 → C + O2). Hydrogen is never a sold product — all H2 is consumed internally in the Zero-E PowerBlock.

Automatic excess-power routing to CRSCU. Any electrical surplus beyond consumers + reserve is automatically directed to the Carbon Recovery / Syngas Cracker (CRSCU) subsystem. With minimal recipe change, the CRSCU begins cracking additional CO and CO2 into pristine carbon and pure oxygen — 2 CO → 2 C + O2 and CO2 → C + O2. Surplus electrical capacity converts directly into higher-value solid carbon products (RC4/RC5 grade) plus pure O2 as a saleable byproduct. The plant self-optimizes to the recipe envelope without operator intervention — it is a design feature of the ACM control loop, not a separate operating mode.
Optional secondary generation module. The plant has design provision for additional generation modules that recover approximately 50% of the surplus heat and convert it to an additional ~2 MW electrical. This is a per-site option installed where local energy economics justify further electrical output beyond the primary PEM PowerBlock. Total on-site generation with the option installed: ~8.8 MW (PEM PowerBlock) + ~2 MW (heat-recovery module) = ~10.8 MW electrical, with the remaining surplus heat still available for MED and district heat.

Integrated fuel-cell dispatch and internal hydrogen allocation require OEM guarantees and an operating campaign to confirm. All H2 is consumed internally — there is no merchant-sale or export path.

Engineering Detail

Electrical consumption (modeled 100 TPD)

AreakWe
MCR microwave power2,400
Glass processing1,540
Carbon processing1,473
Aromatics processing531
Mini-EAF steel495
Non-ferrous recovery335
Water treatment266
Pregenesis preparation233
CO2 liquefaction85
PSA45
Balance (reforming, HRSG, condenser, WGS, OWS, cyclones, compression)~256
Total7,614

On-site generation & recovery (design intent)

  • Electrical generation: design-intent PEM PowerBlock sized to total load × 1.15 (15% reserve) ⇒ gross design target ~8,756 kWe. This is a model closure rule, not an OEM guarantee.
  • Hydrogen fuel available: process models 657 kg/h H2 at the PowerBlock boundary in the H2 MAX-leaning illustrative recipe — all of it consumed internally; H2 is never sold. Design doctrine is net balance with a small surplus: the balanced recipe steers production to ~8–11 t/d (333–458 kg/h) against PowerBlock demand, with the H2-rich recipe available as headroom for surge and high-power cases. At illustrative 50% conversion efficiency: ~457 kg/h H2 covers the 7,614 kWe load; ~525 kg/h covers the 8,756 kWe gross target. The operating point between load cases is an internal allocation decision requiring OEM efficiency data and a reconciled operating case.
  • Fuel-cell water (mass link to Q2 Boundary C): whatever H2 is burned in the PowerBlock returns as water — ~8.94 kg H2O per kg H2, drawing ~7.94 kg atmospheric O2 per kg H2. At illustrative cases: ~4,085 kg/h (7,614 kWe), ~4,697 kg/h (8,756 kWe), up to ~6,204 kg/h if all 694 kg/h on-site H2 is consumed. The repository's water_fc slate models 34 t/d (~1,417 kg/h) recovered / sellable fuel-cell water; the balance is available for internal reuse, making the site a net water producer.
  • Heat recovery: heat-exchanger network models ~3,500 kWth to water plant (MED) and ~4,788 kWth surplus available for district heat.

Whole-facility energy-balance qualification Model basis — disclosed gap; resolved at SIT

Individual kWth annotations in the process table include feed / product lower-heating values and recovered heat. They must not be summed and described as independently generated net energy. A signed heat-and-material balance must reconcile feed LHV, electrical work, reaction duty, heat losses, internal recycles, recovered heat, and product LHV before the energy balance is treated as bankable.

To close this gap: commissioning + operating campaign with power metering + OEM PEM fuel-cell performance guarantee + integrated electrical/heat study reconciling generation, parasitic load, and internal H2 allocation.

Section 1 · Question 4
the counterparty's question
“What are the actual operating temperatures, pressures, and residence times for the Pregenesis™, Regenesis™, Regenesis MAX™, and Exogenesis™ protocols?”
Response

Modeled operating conditions are provided below at Level 2. Three important qualifications apply upfront:

  • “Actual” vs modeled. These are design-basis conditions, not measured operating results. MCR reactor-simulation hotspot temperatures are simulation outputs and must not be quoted as validated operating temperatures. Confirmation via ASI facility System Integration Testing (SIT) and post-COD commissioning data.
  • Microwave frequency is frozen per generation. Gen 1 Recyclotron™ (current, deployed in ACM-1 reference plant) operates at 915 MHz, a standard industrial ISM band. Gen 2 Recyclotron™ is in design at 433 MHz, target release late-2027 to early-2028. Restricted-internal at both generations: measured material-permittivity data, dielectric-property library, and tuning parameters. Equipment certification and EMC / electrical documentation follow the frozen frequency at each generation. Verified in the engineering repository; vendor-dependent for Generation 2.
  • Exogenesis™ — out of scope for this Technical Assessment. Exogenesis is a future protocol for legacy-landfill remediation (LLRP delivery); it sits outside the standard operational protocol suite (Pregenesis™ → Regenesis™ → Regenesis MAX™). Its operating parameters, deployment timeline, and economics will be addressed in a separate future document to the counterparty. See the scope note on the landing page.
Engineering Detail

Design-basis operating conditions across the ACM-1 pyrolysis flow

Protocol / unitTemperaturePressureResidence / notes
Pregenesis™ (prep P1–P6)~25 °C ambientatmosphericcontinuous mechanical train; product <50 mm
Regenesis™ MCR~550 °C~0.5 atm (sub-atmospheric)6–9 min simulation design reference; 4 × 25 TPD; susceptor: WTE ash / coal ash / recirculated OmniCrude™ mineral phase (not a bespoke catalyst — catalytic activity is inherent to metals + oxides in the ash matrix, see Q7); steam + N2 purge
CRSCU (Carbon Recovery / Syngas Cracker Unit)1,300–1,800 °C~1 atmplasma-driven; no catalyst; quench boiler + dedicated cyclone for pristine carbon recovery. One CRSCU per vapor line; each line sized for the vapor balance from 200 TPD of Recyclotron output (2× nameplate) for failover — 2 vapor lines / 2 CRSCUs at 100 TPD
Water-gas shift~250 °C~1 atmFe/Cr; ~75% CO conversion (modeled)
Condenser200 → 40 °C~1 atm—
Oil-water separation~40 °C~1 atmresidence ≥ 15 min
PSA~40 °C8 barg ads / 0.1 barg desorbzeolite 5A + activated carbon
Regenesis MAX™ — carbon200–3,000 °C—under N2/H2 atmosphere
Regenesis MAX™ — EAF steel~1,600 °C—mini-EAF melt
Regenesis MAX™ — non-ferrous60 °C leach / 25 °C SX / 3–5V EW—H2SO4 leach, solvent extraction, electrowinning
Regenesis MAX™ — glass1,200–1,600 °C—glass / mineral melt
Regenesis MAX™ — aromatics60–350 °C10 mbar (heavy ends)vacuum distillation; closed-loop solvent recovery
H2 storageambient350 barType-IV composite vessels
CO/CH4 storageambient200 bar—
N2 storageambient10 bar—
CO2 product−20 °C20 bargliquefied; modeled >99.5% purity
Beyond this response

Beyond the design-basis engineering disclosed in this Assessment, additional proprietary internals of the ACM engineering model — measured material dielectric properties, dielectric-property library, tuning parameters, specific catalyst formulations beyond generic type names, precise reagent dosing schedules, and other restricted design internals — remain trade secrets of Carbotura, Inc. and are not disclosed at any stage of the counterparty's diligence, term-sheet execution, Cornerstone Round commitment, or subsequent commercial operation. Verification of the disclosed design basis proceeds through independent third-party review under NDA, OEM guarantees, ASI System Integration Testing (SIT), commissioning, continuous emissions monitoring (CEMS), and CAFI Digital Triplet post-COD performance data. Disclosure level here is partner-safe; proprietary internals restricted in the source engineering system are excluded.