This review requires multi-domain expertise — and thinking in reverse
1 · No single-domain reviewer can validate this assessment. The ACM spans microwave/RF physics, plasma chemistry, gas separation, PEM electrochemistry, hydrometallurgy, glass engineering, water treatment, DC power systems, and industrial automation. A reviewer expert in any one of these will correctly assess roughly one-sixth of the content and — without noticing — misread the rest through their home domain's assumptions. We recommend either a multi-domain panel or sequenced single-domain reviews, each scoped to its section.
2 · Every forward-industry heuristic runs backwards here. ACM is a vertically integrated manufacturing system operating in reverse (see the framing section below). Diligence instincts calibrated on waste-management, WTE, or conventional process plants can produce category errors unless system boundaries and operating doctrine are understood first. The expandable checklist below lists the specific inversions and the most common false-finding traps.
The inversion checklist · system context for standard diligence questions
| Forward-industry instinct | The ACM inversion | Where it is answered |
|---|---|---|
| Feedstock is waste to be disposed of at lowest cost | Feedstock is the raw-material input to a manufacturing system; the gate protocol is inbound QC, not disposal intake | Section 2 · Feedstock Acceptance Protocol |
| Judge the plant by residue minimization and disposal compliance | Judge it by conversion maximization: the system is designed to convert incoming mass into certified products and internal recycle streams, minimizing material requiring external disposition — and all non-product streams are mass-accounted and reported (see the verification pathway) | Section 1 Q2 · Section 4 Q13 |
| Ask for the stack test, flare records, dioxin formation data | No combustion, no flare, no atmospheric vent path — the oxygen-starved regime suppresses dioxin formation (DRE third-party verified); APS routes captured vapor back into the process | Section 4 Q11 · fate matrix · failure modes |
| Ask for the wastewater discharge permit compliance record | No routine liquid-discharge path in stable qualified operation; finished Water MAX product targets potable-or-higher specifications after treatment and analytical release (intermediate streams are not represented as potable); the national environmental regulator's limits function as internal design ceilings | Section 4 Q12 |
| Ask what the hydrogen sells for | H₂ is never sold — it is internal Island-Mode fuel for the Zero-E PowerBlock; the plant sells materials and water, not energy | Section 1 Q3 · Section 3 |
| Ask for catalyst consumption cost and replacement schedule | No bespoke catalyst — catalytic activity is inherent to the recirculated inert susceptor matrix; CRSCU is catalyst-free plasma | Section 2 Q3 · Section 1 Q4 |
| Assess construction risk with EPC heuristics (site engineering, commissioning float) | DFM manufactured modules ship to site and install into pre-provisioned frames; assess manufacturing quality + SIT instead | Landing · DFM-vs-EPC callout · Module Manifest |
| Ask for the TRL and pilot-plant data | DFM / ERL doctrine: composed from equipment classes with commercial operating history, validated at ASI System Integration Testing — not a research-progression pathway | Landing · design-basis callout |
| Scale-up risk: bigger plant = new engineering | Capacity replication: more identical 25 TPD modules, never a larger reactor | Landing · Modular Flexibility |
| Energy balance: how much grid power does it draw? | Island-Mode: the plant powers itself from internal H₂; surplus routes to CRSCU carbon recovery, MED water, and district heat — not to the grid | Section 1 Q3 |
| Ask for the single rigid design case (“what is THE flowsheet basis?”) | Envelope engineering, not point design: a broad, qualification-bounded input envelope; recipe-specific operating plans instead of one fixed flowsheet — engineered flexibility ranges across feedstock, energy, process, and outputs. Variation is the operating premise, not an off-design excursion; published balances are named recipe cases within the envelope | Reverse VI · Module Manifest · Agile Recipes |
| Ask “does it work?” as a binary pass/fail | There is no facility-wide or technology-wide binary threshold to test that against. The right question is how well it performs, across what range, and with what confidence — see the downside-stack case: even at 50% of the already-derated design-point throughput and a further 50% discount to design-basis product pricing, applied simultaneously, gross margin stays above 50% across three independent revenue streams | Section 3 · Reference module & downside case |
Suggested review-panel composition — chemical / process engineer (mass-energy balances, Sections 1–2) · RF / microwave engineer (Recyclotron™, Section 1 Q4) · plasma / high-temperature chemist (CRSCU, fate matrix) · PEM electrochemist (Section 3) · extractive metallurgist (Metal MAX) · water-treatment engineer (Section 4 Q12) · the host jurisdiction environmental-regulatory counsel (Sections 4–5).
AI-assisted review — recommended. Modern diligence teams run document review with AI tools, and we encourage it: the balances are internally consistent and machine-checkable. To make this practical we publish an AI Review Brief — a single plain-text consolidation of this entire response (mental model, inversion checklist, doctrine constants, all sixteen answers, verification framework) formatted for direct loading into an AI review tool, including a suggested review prompt. The prompt directs the tool toward the open technical questions and away from the false-finding traps above — Carbotura prefers a rigorous review over a lenient one. A separate companion note, Technology Validation Methodology, expands on the DFM vs EPC doctrine below with the specific fleet-wide operating-hours evidence and a risk-retirement comparison diagram.
Suggested reading order — before section deep-dives, read the four framing callouts on this page: Reverse Vertical Integration, The Periodic-Table Prime Directive, DFM vs EPC doctrine, and What MCR is NOT. Twenty minutes there prevents most of the traps above.
The questions a prudent technical reviewer will ask — answered before they are asked
The most important question — a shared question, held equally by Carbotura, the counterparty, and financing parties in a sound diligence process — is whether the plant operates within the ranges specified. If it does, bankability is supported. The model, component heritage, FOAK test plan, independent verification, contractual guarantees, and staged capital commitments all exist to answer that one question. Serious diligence asks hard questions. Rather than leave them for a review cycle to surface, this section answers the ten hardest directly — including the hydrogen disposition ledger, the water boundary table, and the bankability verification pathway that a prudent investor should require before treating any design basis as fundable. For AI-assisted review, the entire response is consolidated in the AI Review Brief (single plain-text file, suggested review prompt included).
The governing verification frame. The complete plant is modeled in a single elementally balanced multiphysics simulation at atom-count granularity, with a stated ~2.5% manufacturing tolerance — now concretely anchored: the design-basis feedstock elemental inventory (certified characterization) closes at ±2.4% (RSS) on total mass, with the dominant elements (C, H, O — 88% of mass) at ±2.1–2.6% each (see the certified inventory table in the fold below) — pending independent review and FOAK validation. The first fully integrated ACM plant (FOAK) is under way in Pennsylvania, USA — 400 TPD growing to 2,000 TPD, first 100 TPD production block (four 25 TPD Recyclotron™ modules) COD targeted Q4 2027–Q1 2028, expansion planned to begin approximately six months after successful COD and satisfaction of defined acceptance milestones. Every design-basis figure in this response — elemental balances, energy, water, destruction efficiencies, uptime — is subject to FOAK verification: independent model review beforehand, integrated demonstration and sustained-run acceptance at FOAK, third-party analytical thereafter. Constituent equipment classes carry commercial operating history; the composed system is what FOAK proves. Ranges, not absolutes: performance is engineered and stated in reasonable ranges. The design targets sustained Island-Mode operation across the qualified nominal feedstock envelope; startup, off-spec, low-LHV, maintenance, and transition cases may use approved supplemental energy — optional secondary generation from excess thermal energy (~2 MW module), temporary recipe steering toward increased internal H₂ production, or purchased hydrogen.
FOAK as an opportunity — not a risk to wait out. The FOAK timeline lands inside the counterparty's diligence window, which converts it from a risk item into an access position: the counterparty's reviewers are invited to observe the FOAK acceptance campaign; to receive FOAK operating data under NDA as it is generated (first-module commissioning, sustained-run, recipe transitions, third-party analytical); to sequence the host jurisdiction's investment decisions against FOAK milestones (independent model review → first-module COD → sustained-run acceptance → expansion start); and to position the host jurisdiction as the first international deployment of the FOAK-validated configuration — a regional first-mover position rather than a wait-and-see posture.
Access is secured by commitment. The FOAK access positions above — acceptance-campaign observation, NDA data rights, milestone-sequenced entry — attach to contract or investment, not to expressions of interest. More than 50 government entities are currently in negotiations for fleet-deployment queue positions; queue positions are held by contract or investment. Commercial mechanics belong to the investment discussion with your Carbotura officer contact, not to this technical response.
The ten hard questions · H₂ ledger · water boundaries · MCR differentiation · verification pathway
The ten hard questions — and the direct answers
| The question a prudent reviewer asks | The answer |
|---|---|
| 1 · Does the mass balance close elementally, not just arithmetically? | The representative feedstock elemental inventory is published below (design characterization basis pending third-party methodology review: ASTM D5231, 200 samples, 90% confidence; JRC/EPA table-row traceability; RSS uncertainty bands; total closes at ±2.4%). The full per-element process balances reside in the engineering model (model review package, under NDA), verified at FOAK. The pre-gate reflow is 240.6 kg/h — a physical stream (recirculated dried ash / bottom-solids susceptor make-up plus remix particulates) with composition, source, and destination in the model review package; a previously quoted “~76 kg/h” figure is retired as a netting artifact (317.0 internal water make-up − 240.6 reflow = 76.4 kg/h). |
| 2 · How do 657 kg/h H₂ and 17,500 kW LHV reconcile? (657 × 33.3 ≈ 21.9 MW) | They sit at different points of one ledger: 17,500 kW is the ~525 kg/h PowerBlock design-case consumption; 657 kg/h is the H₂ MAX-leaning illustrative recipe's boundary flow, not the steady-state production rate. Design doctrine is net balance: the balanced recipe steers production to ~8–11 t/d (333–458 kg/h), matching PowerBlock demand with a small surplus. The full Hydrogen Disposition Ledger is below. |
| 3 · Are the water figures on one consistent boundary? | Each figure has a named boundary — see the Water Boundary Table below. The Recombined-Water range is exactly the H₂-consumption cases × ~8.94 kg H₂O per kg H₂. External-boundary mass sources are atmospheric O₂ and startup H₂ only; circulating steam and internally generated PEM water are inside the boundary. |
| 4 · Is “no external water” absolute? | It is scoped: no routine external process-water demand after stable Island-Mode operation is established, subject to FOAK verification. Commissioning fill, firewater, and potable / sanitary supplies are separate utility categories. |
| 5 · How does zero-vent coexist with emissions monitoring? | Within the process envelope there is no routine uncontrolled atmospheric discharge: zero stacks; anything requiring pressure relief or venting routes to the APS (relief capacity, containment volume, and ultimate disposition defined in the HAZOP/LOPA safety case); CEMS monitors the permit-required compliance point; infiltration air is handled as captured atmospheric constituents routed through the process-gas handling system, with final recovery or disposition subject to FOAK validation. The PEM cathode air is separate and outside this scope: atmospheric air passes through the standalone Zero-E PowerBlock (outside the APS envelope), oxygen recombines to product water, and the nitrogen-rich depleted air returns directly to atmosphere — at the 525 kg/h H₂ case, roughly ~17,900 kg/h air in, ~4,170 kg/h O₂ consumed, ~13,700+ kg/h oxygen-depleted cathode air out (primarily nitrogen, residual oxygen, and water vapor); it never contacts process streams. The 15 kg/h N₂ in the vapor-lane balance is process purge nitrogen, unrelated to cathode air. |
| 6 · Is temperature margin sufficient to claim POP destruction? | No — and the claim is not temperature-only. The design basis is specified destruction and removal efficiencies: temperature margin + residence + mixing + rapid quench controlling de novo reformation during cooling, verified by third-party analytical at FOAK and post-COD across representative feedstocks. |
| 7 · Is MCR just pyrolysis by another name? | A reviewer may reasonably locate MCR within the thermochemical-decomposition family; the differentiation table below shows the engineering-decisive differences (volumetric microwave coupling, 17 coupled catalytic pathways, sub-atmospheric steam environment, recipe-steered product slate, downstream plasma tar cracking, module replication). The incineration / combustion / WtE disassociations remain categorical — no oxygen, no combustion. |
| 8 · Module replication removes reactor scale-up risk — what about system-level risk? | Correct distinction. Shared balance-of-plant — vapor headers, common CRSCU loading, turndown and transients, microgrid stability, common-mode failure, maintenance congestion — is validated at the aggregation level. That is precisely what FOAK demonstrates. |
| 9 · Component heritage is not integrated-system proof. | Agreed. Heritage de-risks the modules; FOAK proves the composition (Pennsylvania; 400 TPD growing to 2,000 TPD; first 100 TPD production block COD Q4 2027–Q1 2028; expansion planned ~+6 months after successful COD and acceptance milestones); the fleet inherits both. |
| 10 · What makes the design basis bankable? | The eight-element verification pathway below — independent model review, integrated demonstration, sustained run, contractual guarantees, feedstock-matrix qualification, third-party sampling, staged site-dependent process-safety studies (HAZID → HAZOP/LOPA → SIL → pre-startup review), and mass-accounted product disposition reporting. |
Hydrogen Disposition Ledger — one boundary, reconciled
Design doctrine: net balance with a small surplus. H₂ is an internal energy carrier, not an accumulating product. Production is recipe-steered to net against PowerBlock demand with a small designed surplus — generated through the production cascade: Recyclotron™ flash reformation (microwave energy + localized susceptor hot-spot micro-plasma, modeled explicitly, cascading through the 17 core / up-to-22 recipe-dependent reaction set — in-bed steam reformation, in-bed water-gas shift, hydrocarbon cracking) → CRSCU cracking to clean syngas → downstream WGS unit, which also reflows CO from PSA storage into additional hydrogen.
| Stream | Value | Note |
|---|---|---|
| Balanced design-recipe production | ~8–11 t/d (333–458 kg/h) | Recipe-steered to PowerBlock demand + small surplus; feedstock-dependent within the envelope |
| Full-internal-load design point (7,614 kWe) | ~457 kg/h (≈11 t/d) | ~50% LHV conversion; production nets demand with a small surplus to the 350-bar buffer |
| Gross-target case (8,756 kWe = 17,500 kW LHV) | ~525 kg/h | This is the figure the energy balance quotes — met by steering the recipe toward higher H₂ yield when the load case requires it |
| Recipe headroom (H₂ MAX-leaning case) | up to ~657 kg/h at boundary (~694 generated, ~37 to carbon-CVD) | The illustrative recipe used in the published Q2 mass balance (~21.9 MW LHV if fully consumed) — available on demand for surge, CVD, and high-power cases; not the steady-state production rate |
| Surplus disposition | small by design | 350-bar buffer → carbon-CVD reagent → CRSCU absorption, with surplus H₂ recycled as a controlled reducing / reforming gas into designated CRSCU reactions and surplus power driving RC5 nanomaterial-recovery cycles. Never sold; not flared and not vented under normal operation — emergency disposition per the site safety case. |
Steady-state disposition across recipes and load cases is resolved in the elementally balanced multiphysics simulation and is a FOAK acceptance-test parameter.
Where the hydrogen comes from — illustrative source ledger (ESTIMATED)
Hydrogen production starts in the Recyclotron™ MCR reactor: flash reformation driven by microwave energy and localized susceptor hot-spot micro-plasma events (modeled explicitly in the reaction platform; experimental characterization at FOAK), cascading through the 17 core / up-to-22 recipe-dependent reactions initiated in the reactor — in-bed steam reformation (steam injection into the fluidized susceptor bed), in-bed water-gas shift, and hydrocarbon cracking. Hot gases pass to the CRSCU for additional cracking into clean syngas (H₂, CO, CO₂, pristine carbon, H₂O vapor, plus uncracked aromatics that condense downstream); a downstream WGS unit then produces additional hydrogen, including from CO reflowed out of PSA storage. Illustratively at 100 TPD: FOAK feedstock carries 6–7% hydrogen on a mass basis (~6–7 t/d), with net steam consumed in reformation and WGS contributing the balance of the 8–11 t/d design range. Produced water is not a perpetual hydrogen source — water participates as a working fluid and hydrogen carrier, but net hydrogen ultimately originates from feedstock hydrogen plus net steam consumed, bounded by the energy balance. Hydrogen also exits in CH₄, aromatics, product water, and residual solid streams. The representative feedstock ultimate analysis and full per-element H/C/O/N/S/Cl/ash balance resides in the model review package (under NDA); until independently reviewed, treat the yield figures as ESTIMATED design outputs.
Design-basis feedstock elemental inventory — URVS-TBM-001 Rev.4 (design characterization basis; third-party methodology review pending)
Representative as-received elemental inventory per tonne of MSW feedstock, from the certified regional characterization (ASTM D5231, 200 physical samples, 90% confidence) resolved to elements with explicit JRC / EPA / USGS table-row traceability and RSS-propagated uncertainty bands. Document prepared for third-party methodology review by an internationally accredited inspection and certification firm. Proximate basis: 23.7% moisture · 55.1% volatile matter · 11.0% fixed carbon · 10.2% ash.
| Element | kg/MT (as-received) | Combined uncertainty (RSS) | Confidence |
|---|---|---|---|
| Carbon (C) | 376.6 | ±2.1% | HIGH |
| Hydrogen (H, incl. moisture H₂O) | 77.8 | ±2.3% | HIGH |
| Oxygen (O, incl. moisture H₂O) | 428.0 | ±2.6% | HIGH |
| Nitrogen (N) | 8.5 | ±5.3% | MEDIUM |
| Sulfur (S) | 1.77 | ±8.2% | MEDIUM |
| Chlorine (Cl) | 5.05 | ±12.1% | MEDIUM |
| Iron (Fe) | 25.0 | ±1.2% | HIGH |
| Aluminum (Al) | 13.2 | ±1.7% | HIGH |
| Silicon (Si) | 15.2 | ±1.6% | HIGH |
| Calcium (Ca) | 9.7 | ±2.0% | HIGH |
| Na / Mg / K / P | 3.85 / 1.11 / 1.54 / 0.88 | ±3–5% | HIGH–MED |
| Cu / Pb / Zn | 2.25 / 1.65 / 1.40 | ±20–35% | LOW (heterogeneous fractions) |
| Other trace (Ni, Cr, Sb, Cd, Hg, REE…) | ~3.5 | ±20–50% | LOW |
| TOTAL (as-received) | 1,000.0 | ±2.4% | HIGH |
Incoming-hydrogen accounting — total elemental hydrogen crossing the gate is 77.8 kg per tonne (±2.3%) = 7.78 t/d at 100 TPD: ~51.5 kg/t resides in the dry material (~6.7% of dry mass — the 6–7% feedstock figure) and ~26.3 kg/t arrives as moisture water (23.7% moisture × 2/18). With net injected steam (HRSG-generated) participating in reformation and WGS, the 8–11 t/d H₂ production range is bounded by this certified inventory plus net steam consumed — hydrogen also exits in CH₄, aromatics, condensate, and residual solids.
Uncertainty methodology: combined uncertainty = √(feedstock-composition MOE² + elemental-reference variance²), per element. The trace-metal bands (±20–50%) reflect fraction heterogeneity, represent <0.7% of feedstock mass, and do not materially affect the ±2.4% total. Full document (all material classes, JRC citations per row, trace-element matrix): independent-model review package.
Air & oxygen balance — PEM cathode passthrough (outside APS scope)
The PEM fuel cells breathe atmospheric air as a passthrough at the standalone Zero-E PowerBlock, which sits outside the APS envelope and the facility process scope. At the 525 kg/h H₂ design case: ~17,900 kg/h air enters the cathode side, ~4,170 kg/h O₂ is consumed into product water, and ~13,700+ kg/h oxygen-depleted cathode air (primarily nitrogen, residual oxygen, and water vapor) returns directly to atmosphere — never in contact with process streams. This stream is deliberately not counted in the process mass balance; the 15 kg/h N₂ there is process purge nitrogen.
Full-boundary energy balance — basis
The published energy Sankey starts at H₂ LHV; the complete balance starts at feedstock chemical energy (LHV) plus external utilities and closes through products' chemical energy, electricity, recovered heat, and losses. The itemized electrical parasitic loads (microwave generation, vacuum, plasma, WGS, PSA, 350-bar compression, CO₂ liquefaction, refining trains, negative-pressure systems, MED, materials handling) are already published in the 7,614 kWe consumer budget; the feedstock-LHV-basis balance with startup/restart events resides in the engineering model (model review package, under NDA). Island-Mode closure — the plant powering itself from feedstock-derived hydrogen — is exactly the claim FOAK exists to prove.
Water Boundary Table — every figure on its stated boundary
| Stream / figure | Value | Boundary and meaning |
|---|---|---|
| Vapor-lane OWS water + salt | ~154 kg/h (of 238 kg/h OWS inlet) | Recyclotron-gate DAG basis (Q2 Sankey) — internal steam/quench loops cancel and are omitted by convention |
| Process condensate to Water MAX | ~1,519 kg/h | Water-MAX intake boundary: primary-conversion + CRSCU-quench + syngas-cleanup condensates combined (a superset of the DAG-basis stream) |
| Recombined Water (PEM) | ~4,085 / ~4,697 / ~5,871 kg/h | PowerBlock boundary at H₂ consumption 457 / 525 / 657 kg/h × ~8.94 kg H₂O per kg H₂ |
| Process water make-up | ~317 kg/h | Drawn internally from Recombined Water into the steam / process loop (model v1.0 carried this as an external freshwater line before the reuse loop was wired — FOAK metering item). No routine external water. |
| Startup water inventory | ~17.9 t | One-time: ~2 t imported H₂ × ~8.94 — startup H₂ is an external material and energy input, stated as such |
| External-boundary mass sources | — | Atmospheric O₂ (~7.94 kg per kg H₂ consumed) and startup H₂ only. Internally generated PEM water and circulating process steam are inside the boundary and are not new mass sources. |
MCR vs conventional pyrolysis — precise differentiation
A reviewer may reasonably locate MCR within the broad thermochemical-decomposition family. The distinctions below are engineering-decisive, not cosmetic:
| Dimension | Conventional pyrolysis | MCR |
|---|---|---|
| Heat delivery | External / surface heat transfer through the vessel wall | Volumetric microwave coupling via the susceptor mineral matrix — heating from within the bed |
| Chemistry control | Uncontrolled thermal cracking; broad, feedstock-hostage product spread | 17 coupled catalytic pathways, atom-balanced; product slate steered by Agile Recipe |
| Environment | Inert / starved at ~1 atm | Sub-atmospheric ~0.5 atm with steam over susceptor catalysis |
| Products | Char + pyrolysis oil + gas of variable quality | OmniCrude™ two-phase state with recipe-steered composition |
| Tar handling | Chronic downstream fouling problem | CRSCU plasma cracks residual tars / CH₄ to H₂ + CO + pristine carbon |
| Scale path | Larger retorts (new engineering per size) | Replicated identical 25 TPD modules |
The bankability verification pathway — element by element
| Verification element | Status |
|---|---|
| Independent model review (reproduce elemental / energy / water / contaminant balances) | Standing offer under NDA — qualified process-engineering firm receives model assumptions and outputs pre-FID |
| Representative integrated demonstration | FOAK — Pennsylvania, USA · 400 TPD growing to 2,000 TPD · first 100 TPD production block COD target Q4 2027–Q1 2028 · expansion planned ~+6 months after successful COD and acceptance milestones; representative and worst-case recipes in scope |
| Sustained run (transitions, maintenance, trip recovery) | FOAK acceptance campaign — sustained-operation demonstration, recipe transitions, planned maintenance, trip recovery |
| Guaranteed performance envelope | Contractual guarantees (throughput, availability, parasitic load, H₂ purity, yields, emissions, water quality, consumables) delivered in the definitive agreement, intended to be supported by applicable equipment warranties and project-specific performance-risk instruments, subject to underwriting and final documentation |
| Feedstock matrix testing | Per-category qualified operating envelopes (moisture, Cl, S, metals, ash, calorific limits) — FOAK campaign + per-site gate QC |
| Third-party sampling (PFAS, dioxins/furans, acid gases, Hg, volatile metals, water, salt-cake, product contamination) | Accredited third-party plan at FOAK and post-COD at every site |
| Staged process-safety studies (site-dependent) | Sequenced per host-jurisdiction requirements: pre-FID HAZID / preliminary HAZOP → detailed-design HAZOP and LOPA → SIL verification → pre-startup safety review; FOAK safety case completes before commissioning |
| Mass-accounted product disposition (measured saleable / recycle / off-spec / maintenance residue / contingency) | The post-COD reporting standard — measured disposition reporting stands behind the “every atom becomes a product” design shorthand |
Your questions — click any row to jump to the answer
Every question below is quoted verbatim from the counterparty's Technical Assessment Questionnaire. Each row links directly to the corresponding response. Hover any row to see the full original question text.
The Exogenesis™ Protocol for legacy-landfill remediation (LLRP delivery, Urban Mining) is a future protocol and is out of scope for this Technical Assessment Response. This response covers the currently-operational Pregenesis™, Regenesis™, and Regenesis MAX™ protocols only. Questions specific to Exogenesis operating parameters, deployment timeline, or LLRP economics will be addressed in a separate future document to the counterparty.
Reverse Vertical Integration — a vertically integrated manufacturing system running in reverse
Traditional vertical integration takes raw materials, moves them through intermediates, and finishes them into a single product family — all in-house, with the discipline of an integrated production line.
ACM inverts this direction: a mixed feedstock stream enters, and the same integrated-manufacturing discipline runs it through six industrial refining lines, each producing five levels of valorization. One waste stream in → 30 refined product classes out. This is not a waste-management facility. It is a vertically integrated manufacturing system running in reverse.
All feedstock is a periodic-table inventory
ACM does not process thousands of unrelated discarded products. It processes a recurring elemental inventory whose physical forms, molecular arrangements, concentrations, and moisture levels vary. Carbotura analyzed 350 government-issued or government-commissioned municipal-material characterization studies (commonly published by their issuing agencies as waste-characterization studies) covering diverse geographic markets. After normalizing reporting categories and elemental composition, approximately 89% of the characterized municipal-material mass fell within a common design envelope; within that common-envelope mass, approximately 99% of as-received mass (including moisture) is carbon, oxygen, hydrogen, and nitrogen plus thirteen recurring minor elements — sulfur, chlorine, silicon, aluminum, iron, calcium, sodium, potassium, magnesium, phosphorus, and marginal members at the ~0.1–0.25% level (copper, zinc, titanium) whose ranking varies by regional profile — seventeen elements in total. The >99% holds on both as-received and dry bases (moisture contributes only H and O, already in the set); lead and other traces are accounted separately, with element membership and values governed by the characterization package. Regional variation is itself quantified in the characterization analysis — six world regions against a global baseline (table below), the largest single-element deviation being East Asia's construction-debris-driven carbon deficit (~−7.7 percentage points). The URVS-TBM-001 baseline is a North American regional characterization; each deployment region is qualified against the same envelope from its own gate characterization. The percentages shift with geography, season, moisture, consumption patterns, and collection practice — the underlying periodic-table inventory remains substantially the same.
The source studies are MSW fraction-composition data sets held in government repositories worldwide; URVS methodology steps 1 and 2 convert each study's reported fractions into a normalized elemental-mass vector, and common-envelope membership is assessed on the normalized elemental composition and proximate ranges against the qualified design thresholds. Studies are combined by weighted average — weights set by characterized fraction tonnage and each study's reported margin of error — with total characterized mass across the 350 studies running to millions of tonnes, giving the envelope statistics a large-sample basis. The detailed study list, weighting methodology, normalization rules, and elemental mapping are maintained in the supporting characterization package, which is the audit source of record.
Regional deviation from the global elemental baseline
Six world regions against the all-region baseline — percentage-point deviation by element (characterization analysis; top-ten elements shown):
| Region | C | O | H | N | Si | Ca | K | Al | Fe | P |
|---|---|---|---|---|---|---|---|---|---|---|
| Global baseline (%) | 43.7 | 31.5 | 6.1 | 1.9 | 3.0 | 3.1 | 1.0 | 1.1 | 1.3 | 0.8 |
| North America | +4.3 | −6.5 | +0.4 | −0.4 | 0.0 | −0.1 | −0.6 | +0.4 | −0.3 | −0.3 |
| Europe | +0.3 | −4.5 | +0.4 | −0.7 | +0.5 | +0.9 | −0.5 | −0.1 | +0.2 | −0.3 |
| East Asia | −7.7 | +1.5 | −1.1 | −0.4 | +3.0 | +2.4 | −0.4 | +0.4 | +1.2 | −0.3 |
| South Asia | +1.3 | +3.5 | −0.1 | +0.6 | −1.0 | −1.1 | 0.0 | −0.3 | −0.3 | 0.0 |
| Sub-Saharan Africa | −0.7 | +6.5 | −0.1 | +0.6 | −1.5 | −1.6 | +1.0 | −0.4 | −0.3 | +0.7 |
| Latin America | +2.3 | −0.5 | +0.4 | +0.1 | −1.0 | −0.6 | +0.5 | −0.1 | −0.3 | +0.2 |
The largest single-element deviation is East Asia's carbon deficit (−7.7 pp), driven by construction debris diluting the organic stream with silicon (+3.0 pp) and calcium (+2.4 pp). Every regional profile remains inside the qualified operating envelope; each deployment region is qualified from its own gate characterization.
The conventional question is “what kind of discarded product is this?” ACM asks instead: “which elements are present, in what quantities and chemical forms, and which combination of reaction pathways and refining trains will return them to specification?”
A bottle, food package, textile, tire, or composite object is a temporary physical arrangement, not a fundamental material identity. Dynamic range is therefore an elemental-ratio problem, not a product-recognition problem: the system measures elemental composition, moisture and bound water, heating value, ash and inorganic content, halogen / sulfur / nitrogen / metal loading, physical form, and the target product slate — and adjusts reaction emphasis, microwave-energy distribution, steam balance, residence conditions, internal reflow, separation settings, product-grade targets, and throughput or supplemental energy. The plant is never reconfigured for a new object class; recipe and refining conditions change within the qualified envelope while the facility architecture stays fixed.
The qualified envelope — five linked dimensions
| Envelope | What varies | How ACM responds |
|---|---|---|
| Feedstock | Elemental ratios, moisture, LHV, ash, halogens, metals, particle size | Accept, blend, pretreat, select recipe, restrict inclusion, or reject |
| Process | Temperature, pressure, steam-to-carbon ratio, residence time, susceptor circulation, CRSCU duty | Adjust operating set points within qualified ranges |
| Energy | Internal H₂ production, plant load, recovered heat, buffer status | Island-Mode, recipe steering, supplemental energy, derating, or transition mode |
| Product | Yield, purity, contaminant level, RC grade, market destination | Up-value, reflow, release at lower qualified grade, or quarantine |
| Equipment & safety | Treatment capacity, gas generation, storage, APS, salt cake, guard beds, cooling | Hard interlocks, derating, isolation, shutdown, or rejection |
Operating states — resolved across the physical taxonomy
Every operating condition is classified into one of five states — and the classification is resolved at each level of the physical taxonomy: Zone → Line → Frame → Module → Equipment, not plant-wide. Each level carries its own state; the facility state is the aggregate. This is where normal ~50% design loading becomes visible in operations: a single module can sit AMBER (constrained) while its line remains GREEN, because each parallel line is normally loaded at approximately 50% of its design flow and the partner line absorbs the pair's flow within its existing rating. CAFI continuously recommends or executes state classification and approved operational adjustments within defined authority limits — safety-critical actions are performed by independent control and safety systems; hard safety interlocks, safety-instrumented functions, and operator emergency authority remain independent of the optimization layer (see the control chain below).
| State | Meaning | Required response |
|---|---|---|
| GREEN | Normal qualified operation | Automatic recipe optimization |
| BLUE | Qualified but non-nominal | Adjust steam, power, residence time, throughput, reflow, and product slate |
| AMBER | Temporary constrained operation | Derate throughput, increase reflow, blend feed, or use supplemental energy |
| RED | Outside stable continuous range | Quarantine, pretreat, blend, separately qualify, or reject |
| BLACK | Prohibited material or unsafe condition | Immediate isolation and shutdown response |
Dynamic range includes rate of change, not only composition. A feed composition can be inside the qualified envelope and still be unacceptable as an instantaneous transition. Each critical variable carries a minimum sustained value, nominal range, maximum sustained value, short-duration excursion limit, maximum allowable rate of change, and required stabilization time after a recipe transition — applied in particular to moisture, LHV, chlorine, sulfur, metals, gas generation, hydrogen production, throughput, guard-bed loading, and product-train availability.
Control chain & system invariants
Before receipt: contractual feed specifications, supplier history, manifests, and prequalification define the expected elemental and contaminant envelope. At the gate: rapid characterization determines whether material is accepted, blended, restricted to a recipe, quarantined for laboratory analysis, pretreated, or rejected. During processing: online measurements continuously update the recipe — moisture, composition proxies, calorific value, reactor temperature, microwave reflected power, gas composition, acid gases, bed pressure drop, guard-bed condition, and product-quality indicators. This loop is managed by the CAFI core operating in the Digital Triplet — the physical module, its live-data multiphysics simulation (400-plus measured and derived variables used by the module's supervisory model and control domain), and CAFI as the optimization and decision layer, acting within approved authority limits. The control philosophy is model-predictive optimization bounded by hard safety, environmental, equipment, and product-quality interlocks; deviation triggers pre-emptive intervention before damage occurs. The control architecture is layered, and the layers are independent: (1) CAFI optimization and decision layer → (2) basic process control system → (3) safety instrumented system → (4) emergency shutdown system → (5) operator command authority. Safety-instrumented functions and operator emergency authority are never subordinate to the optimization layer. CAFI is distributed — it executes at each controller level, over a common data lake that is not required to be continuously available; loss of the data lake or of any CAFI instance transitions the affected control domain to an approved deterministic fallback mode or a controlled shutdown.
System invariants — these rules hold in every state:
- No operation outside approved temperature, pressure, and containment limits
- No feed introduction without sufficient APS, downstream, storage, and treatment capacity
- No hydrogen production above safe storage, process-use, and PowerBlock-consumption capacity
- No contaminant loading above guard-bed, salt-cake, quench, corrosion, or product-quality limits
- No product release without analytical qualification to the applicable RevCon™ grade and destination specification
- No recipe transition that violates thermal, mechanical, material-handling, or safety ramp limits
- No throughput above the capacity of the slowest active downstream unit
- No supplemental material or energy input outside the stated mass and energy balance
The advantage is not that ACM can process anything. The advantage is that the visible diversity of municipal solid material is largely a superficial arrangement of a recurring periodic-table inventory. The physical forms vary. The elemental foundation repeats. The periodic table defines the process architecture.
Envelope engineering, not point design. A conventional process facility is designed forward from a tightly specified input to a single product family — a point design, where feedstock variation is an off-design excursion to be survived. ACM faces the opposite requirement: the feedstock is a broad but contractually and analytically bounded mixture of elements and molecules (gate exclusions, moisture / ash / halogen / heavy-metal / heating-value limits, recipe qualification), and the plant applies recipe-specific operating plans within that qualified envelope rather than one fixed flowsheet condition. The engineering object is an envelope — wide, deliberately engineered flexibility ranges across feedstock (certified distribution, sorted or unsorted, 40+ qualified classes), energy (recipe-steered H₂ / power net balance), process (17 core and up to 22 recipe-dependent reaction pathways, multizonal control, hot-swap module manifest), and outputs (six refining lines × five grades, steered per Agile Recipe). Variation is the operating premise the plant is engineered around — not a deviation it must survive. Operating decisions are not binary: recipe, energy, and disposition choices adjust continuously to targeted outputs, giving the plant materially greater agility and resiliency than a dedicated single-product plant. The system disintegrates and refines raw materials to various levels to a specification, and the RevCon™ grade and reflow architecture is designed to materially reduce stranded-product risk by providing up-value (further refinement to a higher RC grade), reflow (return to the process for yield), and down-value (release at a lower qualified grade) pathways, subject to specification, market demand, processing capacity, and regulatory approval.
The 6 refining lines · multi-domain comparison · industrial-diversification alignment
The 6 industrial refining lines — what each one produces
- Water MAX — potable → industrial-DI → softened/distilled → semiconductor 18 MΩ → pharma-WFI
- Gas MAX — industrial N2 → CO/CH4 syngas → liquefied CO2 → high-purity gases → ultra-pure specialty
- Carbon MAX — activated carbon → thermal black → synthetic graphite → CNT / graphene → fullerenes / diamond
- Metal MAX — EAF billet steel → hydromet non-ferrous → alloy stock → REE concentrates → precious metals
- Glass MAX — container-grade → fiberglass → borosilicate → semiconductor → ultra-pure optical
- Aromatics MAX — BTX → heavy aromatics → refined solvents → specialty chemicals → pharma intermediates
Why this scales — the China multi-domain manufacturing model, applied to circular industry
China's industrial ascent over the last two decades did not come from single-product mega-plants. It came from multi-domain manufacturing integration: modular production zones, standardized units repeated at scale, concurrent multi-product output from shared infrastructure, and rapid site-by-site replication. ACM applies the same organizing principles to circular manufacturing:
| Dimension | Traditional Western industrial pattern | Multi-domain / ACM pattern |
|---|---|---|
| Plant scope | Single-product mega-plant; each product family gets its own plant | Multi-product concurrency from one facility; 6 refining lines × 5 grades share one utility footprint, one workforce, one permit envelope |
| Engineering approach | Custom-engineered per site; unique drawings; year-scale design | Standardized module catalog; DFM manufactured on production line; ERL-qualified; ship-and-install |
| Deployment rate | Decade to build one plant class in one region | Month-scale module installs into pre-provisioned frames; year-scale for full facility rollout; concurrent multi-site deployment from the same catalog |
| Product breadth per facility | 1–3 product classes (often just one) | 30 product classes concurrent (6 refining lines × 5 valorization grades) |
| Feedstock envelope | Locked at design; changing it = major refit | Manifest-updatable; new feedstock class = new modules into reserved frames (see Module Manifest) |
| Capital efficiency | Sunk CAPEX per product family; asset stranded when product exits market | Shared CAPEX across 30 product classes; retire a module → redirect its RC slot; no stranded asset |
| Regulatory pattern | Per-plant per-product permitting; slow scale | the national waste-management regulation's Cat-1 baseline + Cat-2/3 expansion by operating record; single permit covers 30 product classes; scope grows as record supports |
| Multi-site rollout | Custom-engineered site-by-site; sequential; capital and schedule diverge from copy 1 to copy N | Same catalog replicated across sites; DFM manufacturing line; 100s–1000s of facilities deployable from one catalog |
What this means for the host jurisdiction
Every ACM site deployed in the host jurisdiction brings the multi-domain manufacturing pattern — the specific pattern that scaled Chinese industry — to its circular-manufacturing base. For its industrial-diversification mandate, this maps directly to industrial competitiveness objectives:
- Locally-manufactured product breadth from waste streams — 30 refined product classes are produced locally, from local feedstock at every site. Water, industrial gases, carbon products, metals + REEs, technical glass, and aromatic chemicals all come out of the same facility — a full slice of the industrial material economy.
- Import substitution across six domains — the host jurisdiction currently imports significant volumes across each of the six refining lines. Each ACM site substitutes imports concurrently across all six — not sequentially, not per product family, but concurrently from one facility.
- Rapid national scaling — standardized DFM module catalog + pre-provisioned frames means new sites deploy at the pace of module manufacturing + civil shell construction, not at the pace of bespoke engineering. This is why the host jurisdiction can plan 29 modules across 7 regions in a single Phase 1 rollout rather than sequential per-region custom builds.
- Scope expansion by operating record — a site's licensed feedstock and product scope grows as its operating record supports the national waste-management regulation's Cat-2/3 expansion. Its industrial capacity in circular manufacturing grows on the same time-axis as the operating fleet, not on a separate permitting axis.
- Community augmentation delivered per site — potable water and clean electricity to the surrounding community are catalog features, not custom retrofits. Every site can carry them; sites that need them switch them on at manifest.
- Positioned as the region's most sophisticated circular-manufacturing platform — no other regional neighbor has a multi-domain circular pattern at manufacturing scale. Deploying this under its industrial and economic-zone authority structure positions it as the region's reference platform for circular industrial capacity.
This document responds directly to the Technical Assessment Questionnaire received from the counterparty. Every question is quoted verbatim, with a Response at architectural / summary level and an expandable Engineering Detail layer for reviewers who require deeper substance. Beyond these two layers — exact process temperatures, pressures, residence times, catalyst chemistry, and reagent formulations — are trade secrets of Carbotura, Inc. and are not disclosed at any stage of the counterparty investment process. Verification proceeds through independent third-party review under NDA and through post-COD operating data (commissioning results, continuous emissions monitoring, CAFI Digital Triplet performance).
Disclosure Framework
Why our answers look different from an EPC vendor's answers
Most engineering questionnaires implicitly assume an EPC (Engineer-Procure-Construct) delivery model — where each facility is a bespoke construction project with site-specific engineering, one-off procurement chains, and long-tail commissioning risk. Carbotura operates under a fundamentally different doctrine: DFM (Design for Manufacturability) at Engineering Readiness Level (ERL) — factory-fabricated modular units, standardized across the deployed fleet, delivered per site.
Full DFM-vs-EPC doctrine
Strategic takeaway: DFM is the only viable delivery model for rapidly deploying hundreds — if not thousands — of facilities. EPC's per-site engineering and one-off procurement chains cannot scale on that curve; DFM's factory-fabricated, standardized modules can.
This is why our responses reference design envelopes and ranges rather than site-specific set points: the process architecture is fixed at manufacturing. Per-site variance lives in feedstock characterization, recipe variations, and AI tracking algorithms — the ACM commercial model is economically and energy driven, and agile from 0–100% throughput, letting each module dynamically optimize throughputs, product yields, and product routing (output, feedback, or upgrade). Each module carries a dynamic Nameplate Capacity and Function, not a fixed spec. Facilities operate in true Island Mode: no utility interconnection. Design duty is 7×24×365 continuous at 50% of potential maximum capacity, with redundant lines and failovers maintaining 100% nameplate output and surge headroom to 200% of nameplate for seasonal peaks, emergencies, and growth (frame-level process capability, confirmed per qualified recipe; certain esoteric feedstock recipes do not carry the full 200%). It is also why we report cumulative operating hours across the deployed fleet rather than facility-by-facility commissioning risk: the module is the facility.
- Design envelopes are LOCKED at architecture level; ESTIMATED at per-site level pending Intake Study.
- Certification sequencing: framework certifications (ISO management systems, national/local-content certification) pursued pre-COD; first-of-kind facility certifications require operating history (Section 5).
- Risk retirement is sequential (module fabrication → deployment → site COD), not the traditional EPC stacked design-procure-construct-commission chain.
If you are reading this from a waste-to-energy or combustion background, you will be looking for the wrong picture.
The Recyclotron running the Microwave Catalytic Reforming process (MCR) process is emphatically not:
- ✗NOT incineration. No open flame, no combustion air, no oxidation-driven destruction. The Recyclotron runs oxygen-starved. Feedstock carbon does not exit as flue-gas CO2.
- ✗NOT waste-to-energy (WTE) as the industry knows it. WTE burns waste to raise steam → turbine → power, with mass CO2 emissions and giant flue-gas cleanup trains. The Recyclotron converts waste into material products (RC3–RC5 grades) and hydrogen. Power is a by-product of the on-site PEM PowerBlock, not the primary output.
- ✗NOT conventional pyrolysis or gasification. No hours-long batch cycles, no bespoke catalyst bed to poison, no single-pathway thermal cracking. The MCR runs 17 coupled atom-balanced reaction pathways concurrently under oxygen-starvation at 6–9 min residence, driven by targeted microwave energy on a mineral susceptor — chaotic disintegration, not selective molecular cracking.
- ✗NOT plasma gasification. Plasma gasification feeds solid waste directly under a plasma arc. The Recyclotron is microwave-driven, not plasma, and it does not gasify solids under an arc. The ACM does use plasma downstream — the CRSCU is a plasma cracker of already-gaseous syngas (not solids), one per vapor line at 1,300–1,800 °C. Plasma cracking of gas and plasma gasification of solids are fundamentally different unit operations.
- ✗NOT refuse-derived fuel (RDF) production. No pelletizing for downstream combustion. The Recyclotron converts feedstock directly into hydrogen, CO2, syngas, pristine carbon, glass/mineral, metals, aromatics, and water — full-value products, not intermediate fuel.
- ✗NOT anaerobic digestion or composting. No microbial pathway, no wet-feed constraint, no methane-limited yield. The Recyclotron accepts >500 tested waste streams (MSW, plastics, tires, ash, sludge, biomass; see Section 2 Q5) and converts them at industrial rates.
- ✗NOT a scrubber facility of any kind. The closed-loop capture subsystem is a materials capture system, not an emissions scrubber. Solids, liquids, and gases it captures are reflowed into the process as recovered materials — not treated and vented as waste. Because the process runs oxygen-starved (no combustion, no gas expansion), the capture manifold is also a fraction the size of any combustion facility's scrubber train — but the fundamental difference is functional: capture-and-reflow, not scrub-and-vent.
Why each of these does not apply
An oxygen-starved microwave-catalytic materials converter that runs 17 coupled reaction pathways concurrently on a susceptor mineral matrix, breaking waste feedstock into atom-balanced streams (H2 at 4–5 nines — consumed internally in the Zero-E PowerBlock, never sold — plus CO2, syngas, pristine carbon, glass, metals, aromatics, water products) under Recyclotron gate-flow control — the ACM (Advanced Circular Manufacturing) architecture.
Carbotura generates 4–5 nines hydrogen — used internally, never sold
On-site hydrogen purity is 4–5 nines (99.99% – 99.999%) — comfortably above ISO 14687 PEM fuel-cell grade (99.97%), with no external polishing. All hydrogen is consumed internally in the Zero-E PowerBlock (Island-Mode power + Recombined Water); H2 is not a product and not a revenue line. The purity spec matters because it protects the PEM catalyst and maximizes conversion efficiency — it is an internal engineering standard, not a sales specification.
PowerBlock augmentation for community benefit
Where a specific site would benefit the local community, the ACM PowerBlock can be augmented with:
- Solar arrays — additional daytime power generation
- Electrolysis units — additional H2 and clean-water production (water from PEM H2+O2 recombination)
Result: the site can provide surplus clean water and power to the local community beyond baseline ACM operation.
Per-location decision — determined during site engineering, driven by local infrastructure needs and community-benefit priorities.
PFAS, mercury, lead, dioxins, POPs — where they go
Every atom entering the ACM must exit through one of four routes. The full contaminant-fate matrix is in Section 4; the summary below shows the four categories and where they land.
Category table · capture routes
| Category | Examples | Fate |
|---|---|---|
| A. Persistent halogenated organics | PFAS, PCBs, PBDEs, dioxins/furans, chlorinated solvents, legacy POPs pesticides, ozone-depleting substances | Destroyed at CRSCU plasma (1,300–1,800 °C); halogens → salt-cake product; C/H → syngas. Oxygen-starved regime suppresses de novo dioxin formation; rapid quench controls downstream reformation — DRE verified by third-party analytical. |
| B. Volatile heavy metals | Mercury, arsenic, selenium, thallium, cadmium (partial), antimony (partial) | Captured on sulfur-impregnated activated-carbon guard beds (self-produced) or in condensate; recovered via Metal MAX Trace Metals train. |
| C. Non-volatile heavy metals + REEs + precious metals | Pb, Cr (as Cr(III)), Ni, Cu, Zn, Sn, Co, V, Mn, Li, Be; REEs (Nd, Dy, La, Ce, Y…); Au, Ag, Pt, Pd, Rh | OmniCrude™ solid → Solids Separation → Metal MAX streams (EAF steel + hydromet + REE + precious + battery metals + trace). |
| D. Excluded from feedstock envelope | Radionuclides, cyanide chemistry; asbestos only under dedicated HAZMAT protocol | Not accepted — contractual boundary. Asbestos (when qualified): vitrified in Glass MAX at 1,200–1,600 °C for permanent immobilization. |
No default disposal path. Under routine operation, every atom exits as product — via one of the six MAX Processing Trains. N2, CO2-biogenic, and H2O are all captured as product streams (industrial N2, liquefied CO2, RC1–RC5 water). There is no ash landfill, no wastewater discharge, no incinerator stack, and no atmospheric vent path under routine operation. Verification is by accredited third-party analytical + post-COD CEMS + the national waste-management regulation's residue classification.
Full contaminant fate matrix → Section 4 · 30+ compound classes, destruction thresholds vs. CRSCU condition, architectural framings, regulatory framing caveats.
Modular Flexibility — the ACM design advantage
The module is the design unit, not the plant. An ACM site is a composition of standard catalog modules with defined I/O interfaces and RC-grade slots. New feedstock arrives → add the appropriate front-end + destination MAX modules. New product grade requested → add a refining module upstream of the target RC slot. Requirement retires → retire the module, redirect the RC slot. No plant redesign. No custom engineering. No shutdown to retrofit.
Interface contracts · fixed-plant comparison · design library
Three interface contracts every module honors
- Mass I/O — kg/h in, kg/h out at defined composition envelope
- Utility I/O — kW electrical, steam, N2, cooling water, H2 at defined tolerance
- RC-grade slot — which RevCon level the module feeds (RC1 potable minimum → RC5 pharma-WFI / ultra-pure)
The physical enabler is a DC microgrid at 800 / 1,500 / 30,000 VDC (30 kVDC deployed as the site power backbone at second module) plus an overhead modular Nexus Utility Delivery system with standardized Nexus Connection Points at every module. Any module can be shut down, disconnected at its NCP, and swapped without affecting any other module — no facility outage. Full detail → Section 2 Physical Interconnect.
If a module honors its three contracts, it fits into any frame slot that accepts that class. Base frames stay; modules slide. Site manifest = a subset of the catalog. Day-one selection is against the site's feedstock envelope + product demand. As those shift over the site life, the manifest updates — often at planned maintenance windows, without commissioning delay.
Why this beats the fixed-plant playbook
| Scenario | Fixed-plant WTE / gasification / thermal | ACM (modular) |
|---|---|---|
| New feedstock class enters the region | Custom engineering study → permit amendment → capital retrofit → plant shutdown for install. 12–36 months; often infeasible. | Select existing module from catalog → ship → install into pre-provisioned frame at planned maintenance. Weeks to months, no commissioning re-do. |
| New product grade requested (e.g. RC4 semiconductor water) | Not possible without redesign. Fixed products by construction. | Add refining module upstream of RC4 slot; base Water MAX untouched. Incremental. |
| Feedstock composition drifts over 10 years | Plant progressively off-design; economics degrade; eventual major refit. | Manifest updates as composition drifts. Plant stays on-design across decades. |
| Product no longer commercially viable | Sunk equipment; write-down. | Retire module; redirect its RC slot to a still-viable product. No stranded asset. |
| Multi-site rollout at scale | Each site custom-engineered. Sequential. | Same catalog at every site. DFM manufacturing line. 100s–1000s of facilities from one catalog. |
Feedstock Design Library — ready-to-deploy classes
Beyond the standard envelope (MSW · C&I dry waste · C&D · post-consumer plastic · tires · WEEE · agricultural biomass · mineral tailings site-qualified feedstock), Carbotura holds engineered module designs for:
- Petrochemical residues (heavy fractions, catalyst fines)
- Refinery sludges + oil-water emulsions
- Automotive shredder residue (ASR)
- Battery recycling — Li-ion, NiCad, lead-acid, alkaline
- Medical waste (autoclave + CRSCU destruction chain)
- Fly ash / bottom ash from legacy WTE
- Spent catalyst / refractory
- Contaminated soil (industrial spill / remediation feedstock)
- Chemical process residues
- Textile / apparel waste (incl. PFAS-contaminated)
- Food processing / rendering residues
- Municipal + industrial wastewater sludge
- Dredge material / contaminated sediment
- Spent solvent / industrial-solvent recovery
- Fiberglass / composite waste
- Rubber / elastomer waste (beyond tires)
Full class-by-class fate mapping + a worked example manifest + a Li-ion-add diff in Section 2 → Feedstock Design Library · Module Manifest.
Same modules · different product mixes
The ACM plant is not a fixed-output factory. Via AI tracking algorithms adjusting recipe variables per shift, the same 4 × MCR modules can be tuned to prioritize different products. The primary levers are: feedstock blend (augment ratio), steam-to-carbon ratio, WGS aggressiveness, CRSCU residence time, reforming severity, and condenser temperature profile.
Recipe table — H₂ MAX / Carbon MAX / Syngas MAX / Aromatics MAX / Balanced
| Recipe mode | H2 (kg/t) |
Carbon (kg/t) |
Syngas CO/CH4 (kg/t) |
Aromatics (kg/t) |
Primary lever |
|---|---|---|---|---|---|
| Balanced (baseline) | ~157 | ~53 | ~157 | ~20 | Standard operating point |
| H2 MAX | 200–260 | 30–40 | 100–130 | 15–20 | High steam / plastics augment · WGS aggressive |
| Carbon MAX (pristine C / nanomat) | 130–140 | 80–100 | 130–150 | 15–20 | Extended CRSCU residence · carbon-heavy feed |
| Syngas MAX (CO/CH4 priority) | 120–140 | 40–50 | 200–250 | 20–30 | Reduced reforming severity · WGS bypass |
| Aromatics MAX (BTX / petrochemical) | 130–150 | 45–55 | 140–160 | 30–45 | Condenser profile shifted · lower CRSCU severity |
| H2 + Carbon (dual-priority) | 175–210 | 65–80 | 130–150 | 15–20 | Steam-rich + CRSCU extended |
Ranges are design-basis illustrative at the 100 TPD Recyclotron gate; actual yields depend on incoming feedstock chemistry and OEM operating cases. Recipe switching is a shift-level decision, not a facility rebuild — the modules are physically identical; only the operating envelope changes. Higher RevCon™ grades (RC4/RC5) further shift the value-mix without changing throughput.
The design basis is not simulation from scratch
Every ACM module is modeled on the existing operating history of the equipment it comprises — actual throughputs, mass balances, and performance envelopes from equipment currently in commercial service. The engineering model composes these validated building blocks into the ACM-1 configuration; the composed system is then verified at the Authorized System Integrator facility during SIT.
Model accuracy · warranty hierarchy
Model accuracy: the ACM engineering model uses multiphysics atom-count-level accounting across mass, energy, and species. Design-basis vs. measured deviation is calibrated to ~2.5% at the atom-count boundary — a substantially tighter closure than conventional process simulation typically achieves. This is why we quote ranges (envelope) rather than single set points: the ranges reflect real feedstock variability, not model uncertainty.
The physical architecture is hierarchical:
Each level — Facility, Building, Zone, Line, Frame, Module-in-Frame, and Equipment-in-Module — carries:
- Defined scope (functional and physical boundary)
- Warranty (bound at that level)
- Testing regime (at that level, executed at ASI facility during SIT)
Warranties and tests cascade upward from equipment-in-module through modules, frames, lines, zones, buildings, and into the composed facility. Every scope is bound; every level is warranted; every warranty is tested.
Before reading the answers
All figures are design-basis values from a validated engineering model — not certified operating results from the composed ACM-1 configuration. The full limitations list is required reading.
Read the full limitations list
- This Assessment is on the Modular Capacity unit rated at 100 TPD, itself rated at 50% of potential maximum module capacity, designed for 7×24×365 continuous operation with redundant lines and failovers that maintain 100% nameplate output, and with surge headroom to 200% of nameplate for seasonal peaks, emergencies, and growth (frame-level process capability, confirmed per qualified recipe; certain esoteric feedstock recipes do not carry the full 200%). Figures are design-basis values from a self-consistent engineering model that draws on the existing operating history of the equipment comprising each module — actual throughputs, mass balances, and performance envelopes from equipment currently in commercial service. The composed ACM-1 configuration is validated at the Authorized System Integrator facility during System Integration Testing (SIT), then at commissioning and post-COD.
- The ACM engineering model is built on a 100% circular ("near-zero waste, zero vent, near-zero discharge") design principle. That is a design intent enforced inside the model, not a demonstrated regulatory-compliance result. Routine operations target zero vent by APS design and near-zero across waste and discharge; accidental events (spills, upsets, maintenance operations) are not zero and are managed under the national environmental regulator's and national waste-management regulation's standard operator-permit conditions. A model that routes every stream to a product boundary is not evidence that a physical plant discharges nothing.
- No claim is made that ACM-1 is currently certified, permitted, or operating in the host jurisdiction. Permitting and conformity must be confirmed by Carbotura's local permitting team and the competent authorities (the national waste-management regulation, the national environmental regulator, the national standards body) for the specific site, throughput, and product slate. That said, the systems have been integrated to meet and beat all known current Environmental Protection standards for emissions, waste, discharge, and noise, odor, and community impact.
- Mass balance is presented on three explicit boundaries: MCR reactor core closes to <0.01%; under the owner-confirmed Recyclotron-gate capacity definition, the ~241 kg/h number is not an unreconciled residual — it is the algebraically required pre-gate reflow that closes the balance. Reconciled basis: gate 4,166.7 kg/h = fresh feed 3,926.1 + susceptor reflow 240.6 (physical stream); water make-up 317.0 kg/h drawn internally from Recombined Water (model v1.0 carried this as an external freshwater line before the reuse loop was wired — FOAK metering item); products 4,242.6 kg/h. A previously quoted “~76 kg/h reflow” is retired — it was the net artifact 317.0 − 240.6 = 76.4, not a physical flow. 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 the PSA tail gas. The whole-plant balance is algebraically closed at the design intent; the connected kilogram-level ledger resides in the project engineering model. Note also: the 240.6 kg/h reflow determination encompasses the ~2.5% multiphysics atom-count accuracy and real-world feedstock materials variability (actual MSW composition varies from the design EPA_MSW reference). At 4,166.7 kg/h gate flow this reflow is ~5.8% — well within the combined tolerance envelope of the 17 concurrent atom-balanced reaction pathways plus real feedstock variance; full site including PEM PowerBlock shows the plant is a net water producer because atmospheric O2 is fixed into fuel-cell water, bringing total output to 130–160% of MSW feed mass. Reconciliation of the mineral-stream definitions and metering at ASI SIT is required before any balance is bankable — see Section 1 Q2.
- The 100 TPD basis is owner-confirmed as the total mass crossing the Recyclotron capacity gate — the reactor throughput including all pre-gate reflows (dried Trough-2 ash / sludge solids, blend-returned particulate, condensate-derived recycle). Fresh external feed = 100 TPD − pre-gate reflows; reflows raise the internal circulating load without adding external site mass. The chemistry / process-table sub-model in the repository still treats 100 TPD as MSW alone, while the equipment flowsheet also feeds Trough-2 WTE/coal ash + WWTP sludge into the blend without a quantified mass split — a basis reconciliation to be frozen in the site-specific counterparty engineering package.
- Microwave operating frequency is frozen per generation: Gen 1 Recyclotron™ (current generation, deployed in the ACM-1 reference plant) operates at 915 MHz (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.
The counterparty provides · Carbotura delivers, insured & warranted
Every technical item on the site path is resolved by contract, applicable equipment warranties, and project-specific performance-risk instruments, subject to underwriting and final documentation. What the counterparty provides for the specific host-jurisdiction site is listed below; everything else is Carbotura's responsibility, delivered under commercial commitments.
Counterparty-provides / Carbotura-delivers breakdown
- Site selection — designated deployment site within the host jurisdiction
- Feedstock analysis — actual waste-stream chemistry (moisture, ash, Cl, calorific value) from the specific stream
- Feedstock supply contract — contractual acceptance envelope tied to the actual stream
Note: this list is the minimum baseline. If the counterparty elects other Contracted options (e.g. alternate deployment tier, expanded LLRP scope), additional counterparty-side requirements may apply per the specific Contract terms.
- Site ownership + operation — Carbotura owns and operates every deployed ACM facility. The counterparty provides feedstock under the definitive agreement; Carbotura runs the plant.
- All host-jurisdiction permits — Carbotura is the licensed permittee for classification under the national waste-management regulation, the national environmental regulator's environmental permit, and the national standards body product conformity for every site, throughput, and product slate
- Module insurance — commercial insurance intended on every delivered module, subject to underwriting and final documentation
- Performance guarantees — nameplate throughput, product yields, emissions envelopes
- Warranties — Facility · Building · Zone · Line · Frame · Module · Equipment level (see hierarchy above)
- ASI SIT validation — composed system tested at the Authorized System Integrator facility
- Commissioning + post-COD — CEMS + CAFI Digital Triplet performance data
- Mass balance — validated stream-by-stream at ASI SIT
- Energy balance — validated at ASI SIT + OEM PEM PowerBlock guarantee
- Feedstock blend curve — solved from counterparty-provided analysis, delivered as operating recipe
- Reagent / catalyst schedule — from OEM datasheets, backed by operating warranty
- Guard-bed sizing + PEM protection — engineered to OEM contaminant limits, validated at SIT
- Water: zero discharge, net producer — the ACM is a net water producer (PEM fuel-cell water + purified process condensate) and operates on zero liquid discharge. Where a site needs water, Carbotura can supply clean water to the site or the surrounding community — there is no discharge scenario
- Emissions performance — meets and beats known Environmental Protection standards
- Product conformity — RC3 / RC4 / RC5 grades to the national standards body + international standards, accredited testing
- Third-party / OEM certifications — procurement + integration by Carbotura
Any item not explicitly listed above as counterparty-provided is Carbotura's responsibility — delivered under applicable equipment warranties and project-specific performance-risk instruments, subject to underwriting and final documentation.