feat: ACM0014 v8.0 Treatment of Wastewater dMRV -- Bounty Submission

[BikramBiswas786]

ACM0014 v8.0 Treatment of Wastewater dMRV -- Bounty Submission

Submitted by Bikram Biswas for the DLT Earth methodology bounty.

Contents

Path	Purpose
ACM0014_v8_0_Policy_C_BOUNTY_SUBMISSION.policy	Final policy file. Import directly into Guardian.

. Real project data basis

I sourced my core test values from the Chok Chai Starch Co. Wastewater Treatment Project (Uthai Thani, Thailand) -- a CDM Project Design Document under ACM0014, DOE-verified by JQA (Japan Quality Assurance Organisation). UASB digester, 4,650 m3 active volume, 1.87-day HRT, biogas to hot-oil burners and a 450 kW generator.

Reproduced against my engine:

• BE_EL: 1,525.9 vs. real PDD's 1,526 -- near-exact match
• BE_HG: 5,135.5 vs. real PDD's 5,135 -- near-exact match
• PE_y,total: 13,712 vs. real PDD's 13,712 -- exact match
• BE_CH4 (van't Hoff-Arrhenius term): 41,904 vs. the real project's registered (v2.1) figure of 65,048 -- partial match, disclosed below

Why BE_CH4 doesn't match exactly: the PDD's public tables don't disclose month-by-month COD/temperature data at the granularity my v8.0 engine needs (that data sits in an Appendix 1 spreadsheet the registry lists but doesn't make downloadable). My engine also implements a structurally different Arrhenius-weighted baseline model than the simpler v2.1 approach this specific project actually registered under. BE_EL, BE_HG, and PE_total all reproduce cleanly because those figures come directly from disclosed PDD table values, independent of that structural difference.

What this is not: a VCS-native ACM0014 project. Chok Chai is CDM-registered (DOE: JQA), not Verra VCS. I haven't confirmed a currently-active Verra VCS-registered ACM0014 project in the public registry.

Project 9045 (Sapthip, Thailand): the specific reference project I was asked about directly. Confirmed from UNFCCC's registry -- registered under ACM0014 v4, 101,083 tCO2e/yr, registered 28 Dec 2012, crediting period through 27 Dec 2022. Monitoring report covers 28 Dec 2012 - 31 Dec 2020, with issuance status still "Awaiting issuance request." The three appendices (TH_Sapthip_ER_Calculation_20Dec12.xls, TH_Sapthip_FA_Calculation_05Dec12.xls, Thai_Grid_EF_06-08_V4.xls) are confirmed to exist on the registry, but the file-storage endpoint blocks automated retrieval. I'll need direct browser access to pull those and reconcile them further.

Current, Verra-native comparanda in the same sector: Monsoon Carbon / Cenergi's "Monsoon Methane Avoidance from Industrial Wastewater" Grouped Project, with Project Activity Instances in Malaysia (Langkap Biogas Plant, Perak -- traded on Bursa Carbon Exchange June 2025) and Indonesia (Verra Project ID 5292). These are real, live, 2025-2026-active Verra VCS projects in industrial wastewater biogas recovery. I'm disclosing that these likely register under AMS-III.H / AMS-III.Y (small-scale, grouped-project methodologies), not ACM0014 itself, which is typically used for larger single-facility projects.

. Methodology currency

Verra received a proposal to revise ACM0014 (vinasse waste management) and rejected it in April 2025, citing overlap with the existing methodology -- confirming the current version is Verra's live, endorsed reference, not something that's fallen behind. ACM0014 is also directly cited as a required calculation reference inside Verra's own VM0052 methodology (2025), which operates under VCS Version 5 -- Verra's current standard as of December 2025.

One dated item I'm tracking, not a current blocker: VCS v5.0's transition rules give v4-era methodologies (which describes ACM0014's structure) a grace period through December 2026 before they need updating to v5.0's standardized-methods format.

Test evidence

I ran two independent, complete lifecycle tests -- submit, VVB verify, mint -- with zero credentialSubject.undefined errors, each cross-checked by hand against the formula chain and matching to the last available decimal:

Run	BE_y,total	PE_y,total	ER_y (mint field)	Minted amount
1	32,856.46	452.11	32,384.347205130754	32384.34
2	33,228.40	408.74	32,799.65905950902	32799.65

5. Compliance with methodology-digitization guidance

• I implemented calculations in mathBlock, not customLogicBlock, per reviewer guidance.
• CDM tools (TOOL01 additionality, TOOL05 electricity, TOOL14 digester PE/LE) are embedded inline and sourced from Project Description / Monitoring Report template fields, rather than wired as separate Tool policies -- this avoids schema-IRI resolution errors entirely.
• I collect 12 months of monitoring data within a single annual report and aggregate them in-block, rather than via cross-VC aggregation across monthly submissions.
• I used TOOL01 (not VT0008) for additionality, since ACM0014 has no VMR revision requiring the Verra replacement tool.

6. Remaining open items

• TOOL07 / TOOL09 remain simplified to disclosed manual inputs rather than fully re-derived sub-calculations.
• Flaring / land-application / solid-material-leakage branches remain validated only against hand-calculation and my pinned test dataset, not against a real project that uses them.
• I have not confirmed a live Verra VCS project specifically under ACM0014 for full numerical cross-validation.
• Sapthip's underlying Appendix 1/2/3 data needs manual (browser) retrieval before I can reconcile it against this engine.
• My Policy Integrity Test baseline is specified but not yet registered inside a live Guardian instance -- I still need to do that through the UI.
• VCS v5.0 transition deadline (December 2026) for v4-era methodology format.
  policy_1784319662206.xlsx
  6a5a8e88367d2b94db0ca920.json

[BikramBiswas786]

@Pyatakov No one assigned yet . Please look into this

[dubgeis]

@BikramBiswas786 This needs to be conformant with the Verra approach to the methodology not the CDM. If you have a list of questions on tools please let us know. Please leave tagging of our team to us, thank you.

[BikramBiswas786]

@BikramBiswas786 This needs to be conformant with the Verra approach to the methodology not the CDM. If you have a list of questions on tools please let us know. Please leave tagging of our team to us, thank you.

Thank you for the review. I've checked the Verra methodology registry to make sure I understand the comment correctly.

ACM0014 v8.0 appears on Verra's Active CDM Methodologies list — there is no separate Verra-native revision of ACM0014. The methodology is the CDM version, accepted by Verra in the VCS Program. The tools it references (TOOL01, TOOL07, TOOL09, TOOL14) are all CDM tools; VT0008 is not referenced anywhere in ACM0014 v8.0.

Based on guidance from the @swidnikk on Mural Dashboard ("if VT0008 is not mentioned and there's no Verra revision available, only use the tools mentioned by CDM"), I implemented TOOL14 as specified in the methodology — which I believe is the correct approach.

I'd like to make sure I fix the right thing. Could you clarify what "Verra approach" refers to in this context?
Specifically:

• Is the concern about the test project data (I used a CDM-registered project rather than a VCS-registered one)?

• Or is there a specific parameter, formula, or VCS Program overlay requirement that the policy isn't handling correctly?

I want to resolve this accurately rather than make changes in the wrong direction. Happy to address whatever the specific gap is once I understand it.