Skip to content

NC-METH-001 v1.1 — Annex J: Audit Register v0.3 → v1.0 Closure Status

Live Tracking Document for the 47 R3 Register Items

12 May 2026 · Internal · Author: AI-assisted, M. Forni / S. Rubinyi / B. Signer review


J.0 Overview

Annex J is the live tracking document for the audit register first established as NC-MN-001-R3 v0.3 (47 findings) and folded into the v1.1.0 methodology lock process.

The annex tracks every register item from open status through closure (or retirement). Each item carries: - ID (AUDIT-001 through AUDIT-047) - v0.3 status code (8 classification states; see J.0.2) - Materiality (basis-points IRR impact at 70% LTV, where applicable) - Direction (+/−/±) - Resolution path (the work that closes the item) - Owner (named person responsible for closure) - v1.1.0 status (live update column — NEW for this annex) - Closure target (WP number and target date, or deferred)

J.0.1 Why the annex exists

Without a live audit register, methodology drift becomes invisible. The R3 cycle found 23 active gaps where the financial model, the IC paper, the segment register, and the memory had silently diverged from each other. Some of those gaps were material (±220 bps IRR at the largest); some were governance-only; some were retired as speculative.

Annex J is the methodology's standing answer to "what do we know is broken, what's the plan, who owns it, when does it close?" It must be live — updated as WPs progress — not static.

J.0.2 Classification scheme

Eight v0.3 status codes:

Status Meaning Action
ACTIVE-IRR Real gap with material IRR impact (≥ 30 bps) Close via WP; gating for v1.1.0 lock
ACTIVE-METHODOLOGY Real methodology gap, may have IRR impact Close via WP; address in v1.1.0 or v1.2
ACTIVE-GOVERNANCE Process / structural gap, no IRR impact Close via Part F or annex
RETIRED-SPECULATIVE Finding shown to be unfounded or speculation Retire; no action
NOT-APPLICABLE-LC Real for some estates but not LC No action for LC; carry to estate where applicable
MEMORY-HYGIENE Memory update needed Memory edit (M. Forni); not methodology document edit
DUPLICATE Same as another item Retire; reference primary
MODEL-HANDLED Model addresses the concern; methodology only needs documentation Document in methodology; no model change

J.0.3 v1.1.0 status semantics

Each item gets a v1.1.0 status (live):

v1.1.0 status Meaning
OPEN Not yet started
IN-PROGRESS Work underway
STAGED Work complete; awaiting review/lock
CLOSED Resolved and locked in v1.1.0
DEFERRED-v1.2 Carried to v1.2 sprint
RETIRED No longer applicable (finding rejected or duplicated)
NOT-APPL Not applicable to current scope

J.0.4 Quick summary

As of 12 May 2026 (Annex J initial issue):

Category Count Status
ACTIVE-IRR 3 All OPEN; WP1 / WP3 close
ACTIVE-METHODOLOGY 11 Mix of OPEN and IN-PROGRESS; WP1-WP4 close
ACTIVE-GOVERNANCE 9 Mix of CLOSED-via-Parts (8) and OPEN (1; Annex I)
RETIRED-SPECULATIVE 8 All RETIRED
NOT-APPLICABLE-LC 6 All NOT-APPL for LC; flagged for other estates
MEMORY-HYGIENE 5 All CLOSED via memory edits 12 May 2026
DUPLICATE 2 All RETIRED
MODEL-HANDLED 3 All CLOSED via Part C/D documentation
Total 47

J.1 Active register — high-materiality items

J.1.1 ACTIVE-IRR (3 items, gating for v1.1.0 lock)

ID Title Materiality Direction Resolution path Owner v1.1.0 status Closure target
AUDIT-001 LC yield 1380 (model) vs 1485 (methodology RES-001) +220 bps if convention reconciles + Solcast/Solargis multi-source validation; document P50 convention (AC delivery vs DC generation; pre/post soiling; pre/post Y1 degradation) Steven + analyst OPEN WP1 — week 4
AUDIT-014 Interest rate 6.0% (LC model) vs 5.75% EXIM (methodology FIN-003) +30 bps if applied + Confirm EXIM 5.75% is committable for LC; update model OR document why 6.0% blended Jin Yong + Steven OPEN WP3 — week 7
AUDIT-016 Per-typology O&M build-up not applied; model uses flat \(20/kWp/yr vs methodology weighted ~\)33/kWp/yr −230 bps if applied at full magnitude Build LC-specific O&M case via engineering review of structural context (T6W monthly cleaning vs others quarterly; insurance; IEAT JV fee) Steven + EPC partner OPEN WP3 — week 7

Net IRR impact at v1.1.0 lock: AUDIT-001 (+220 if reconciles toward 1,485) plus AUDIT-016 (−230 if methodology O&M is applied at full magnitude) plus AUDIT-014 (+30 if 5.75% applied) — could net to anywhere from +20 bps (best case all WP closures favor LC) to −180 bps (worst case methodology fully applied). Practical expected outcome based on WP scopes: ±50 bps net direction.

J.1.2 ACTIVE-METHODOLOGY (11 items, mix of WP closures)

ID Title Materiality Direction Resolution path Owner v1.1.0 status Closure target
AUDIT-002 C4 corrosion BOS premium in unit costs (coastal estates: LC, MTP, MTP Port, Songkhla) −50 to −80 bps if not included Verify segment register source for C4 inclusion in unit cost build-up; engineering review of marine-grade specs at LC Steven OPEN WP4 — week 8
AUDIT-017 Per-segment yield modeling (currently estate-canonical, not per-segment) ±20 bps ± Solcast Phase 4 work; per-segment yield from satellite irradiance + Digital Surface Model Steven OPEN WP1 — week 4
AUDIT-021 OPEX inflation simplified (2.5%/yr flat vs per-cost-layer refinement) ±10 bps ± Refine inflation framework (modules deflation -2%/yr vs labor CPI+ +3%/yr vs services CPI +2%/yr) Steven OPEN WP3 — week 7
AUDIT-023 Tax structure (BOI Section 8yr / EEC 13yr enhancement under evaluation) ±50 bps ± Phase 0 evaluation of EEC 13yr CIT enhancement; legal/tax opinion Jin Yong + Chandra IN-PROGRESS WP5 — week 8
AUDIT-025 T1 DSM/shading per-segment −10 to 0 bps LiDAR/DSM via Phase 4 (T1 is 6% of LC envelope so minor) Steven DEFERRED-v1.2 v1.2
AUDIT-026 Multi-source yield validation (enables AUDIT-001 resolution) 0 directly Solcast + PVGIS + Solargis cross-check; bankable PV yield study Steven OPEN WP1 — week 4
AUDIT-027 Per-tenant credit framework (Tier-1 doesn't quantify credit risk; concentration is tier-blind) 0 directly Annex G with industry-tier proxy table; supports tenant ESA risk-grading Benedikt OPEN WP4 — week 8 / Annex G
AUDIT-036 T4B-DC undefined in canonical methodology 0 New canonical entry; document arterial canopy typology (closed via Part C C.4) Steven CLOSED Part C-revA
AUDIT-038 T6W undefined in canonical methodology 0 New canonical entry; document wide canal canopy typology (closed via Part C C.5) Steven CLOSED Part C-revA
AUDIT-040 LC-specific top-down cost framework documentation 0 Annex H — document LC-specific cost drivers explicitly (concrete-banked canals; large parking; IEAT-owned structures; single-mob procurement) Steven OPEN WP2 + WP4 — week 8 / Annex H. GATING for v1.1.0 lock
AUDIT-041 Bifacial uplift not typology-specific (model uses estate-canonical only) −10 to 0 bps Typology bifacial framework (parking ~5% / road ~5% / water ~1-2% / gravel ~7%); apply in model Steven OPEN WP4 — week 8

v1.1.0 closure status: 2 of 11 CLOSED via Part C-revA (T4B-DC and T6W typology definitions); 8 OPEN pending WP1-WP5; 1 DEFERRED-v1.2 (AUDIT-025 T1 DSM, minor impact).

J.1.3 ACTIVE-GOVERNANCE (9 items, mostly closed via Parts F + Annex I)

ID Title Materiality Direction Resolution path Owner v1.1.0 status Closure target
AUDIT-005 No IRR walk in IC paper 0 (presentation) Add explicit IRR walk to IC deck v2.0 (post Phase 0 closure) M. Forni + Steven DEFERRED-v1.2 v1.2 IC deck refresh
AUDIT-015 Discount rate 8% real canonicalisation 0 (presentation) Standardise discount rate convention across methodology and model (closed via Part A § A.4.7) Jin Yong CLOSED Part A
AUDIT-030 Regulatory monitoring framework (6 sources, named monitors) 0 Named monitors for TGO/BOI/EEC/ERC/ICC/MEA; quarterly review (closed via Part F § F.3) Kwan CLOSED Part F
AUDIT-031 Anomaly detection / QA framework 0 Paired with pytest acceptance test suite (Annex I) Steven OPEN Annex I
AUDIT-032 Duplicate methodology folder (methodology-package 2) 0 Retire methodology-package 2 (closed via Part F § F.7.5) Steven CLOSED Part F
AUDIT-033 No change-control governance 0 Semver tagging + change log + approval workflow + Git repo (closed via Part F § F.1 + F.7) M. Forni + Steven CLOSED Part F
AUDIT-034 No acceptance test suite 0 Annex I — pytest suite with replication tests (LC IRR ±50 bps; BP IRR ±50 bps); cross-estate ranking; high-materiality findings; governance Steven + analyst OPEN Annex I
AUDIT-035 Part A-F vs 12-step framing inconsistency 0 Lock Part A-F as primary framing in v1.1 (closed via the structural document family) M. Forni CLOSED Parts A-F
AUDIT-044 Segment-ID consolidation (V3 to v1.0) 0 Retire V3 segment IDs in v1.1; v1.0 register canonical (closed via Part F § F.5) Steven CLOSED Part F

v1.1.0 closure status: 6 of 9 CLOSED via Parts A + F structural lock; 2 OPEN (AUDIT-031 and AUDIT-034) both closing via Annex I; 1 DEFERRED-v1.2 (AUDIT-005 IRR walk in IC deck).


J.2 Closed register — items not requiring further action

J.2.1 RETIRED-SPECULATIVE (8 items)

ID Title Why retired
AUDIT-006 Flat PPA vs TOU misframed Inverted finding; LC 3.85 IS contractual flat at 90% MEA TOU peak (per memory edit 16)
AUDIT-020 Monte Carlo independence Model uses P50/P90 deterministic — industry-standard for project finance
AUDIT-024 Engineering levers cost-benefit Unsupported claim; not part of LC investment thesis
AUDIT-028 Probabilistic envelope P50/P90 deterministic is industry-standard; probabilistic refinement beyond scope
AUDIT-037 LC T4A unit cost $742 vs $950 v1.1 claim Phase 3B desk bottom-up unvalidated; LC model $742 may be LC-defensible (engineering review in WP2 confirms or revises)
AUDIT-039 LC T6W unit cost $732 vs $1150 v1.1 claim Phase 3B desk bottom-up unvalidated; LC model $732 may reflect LC-specific structural context (engineering review in WP2 confirms)
AUDIT-045 Pre-rigor LC IRRs / mix optimisation framework Based on phantom IRR walk; v1.0 model output is correct
AUDIT-047 LC CAPEX +18% rollup claim Rollup of speculative AUDIT-037 + AUDIT-039 + bottom-up; all retired

Status note: AUDIT-037 and AUDIT-039 are retired as specific dollar-claim findings, but the underlying question (whether LC unit costs are LC-defensible or below-range) is addressed in Annex H via WP2 engineering review. The retirement here is of the speculative dollar claim, not the underlying engineering question.

J.2.2 NOT-APPLICABLE-LC (6 items, flagged for other estates)

ID Title LC status Where applicable
AUDIT-003 Carbon EF 0.475 vs 0.4091 TGO CM NOT-APPL (LC has no carbon by design) BP and other estates with carbon revenue
AUDIT-004 BESS reframe $400 → $175 + Phase 0A scope-in NOT-APPL (LC no-BESS) BP and other estates with BESS
AUDIT-013 BESS THB pricing NOT-APPL (LC no BESS) BP and other estates with BESS
AUDIT-018 BTM curtailment per-segment NOT-APPL (LC self-consumption; no export) Export-capable estates
AUDIT-022 PPA escalation NOT-APPL (LC uses 0% nominal flat) Estates with CPI-linked escalation negotiated
AUDIT-029 Per-segment BESS sizing NOT-APPL (LC no BESS) BP and other estates with BESS

Status note: These items are CLOSED for LC's v1.1.0 lock. They re-open as ACTIVE items for the relevant estates (BP first; in their respective audit registers).

J.2.3 MEMORY-HYGIENE (5 items, all closed via memory edits 12 May 2026)

ID Title Memory action
AUDIT-007 FX 32 (memory) vs 35 (model) Memory edit 13: "FX=35 THB/USD main program / 32 IET-only"
AUDIT-008 MEA tariff context (BMA 4.20 vs LC 3.85) Memory edit 16: "LC tariff 3.85 = 90% MEA TOU peak 4.28; BMA 4.20 user-locked"
AUDIT-009 BESS $400 (memory) vs $175 (canonical) Memory edit 16: "BESS $175/kWh LFP + thermal premium $10-15/kWh; LC no-BESS"
AUDIT-010 Grid EF 0.475 (memory) vs 0.4091 (canonical) Memory edit 16: "Grid EF 0.4091 tCO2/MWh TGO CM (was 0.475 Scope-2)"
AUDIT-011 Carbon $15 GS vs ITMO/CORSIA Already in memory critical standing corrections (ICC Aug 2025); confirmed complete

Status note: All 5 closed via memory edits on 12 May 2026. Memory _recent_updates reflects the changes. No methodology document edits required beyond reference (Parts A and D cite these conventions).

J.2.4 DUPLICATE (2 items, retired)

ID Title Primary reference
AUDIT-012 Solar yield 1350 in older artefacts Rolled into AUDIT-001 (yield convention reconciliation)
AUDIT-046 T6W bottom-up cost gap Same as AUDIT-039 (retired-speculative)

J.2.5 MODEL-HANDLED (3 items, methodology documents but no model change required)

ID Title Where documented
AUDIT-019 Envelope reconciliation (47 vs 39.1 MWp vs 50 V3 segments) Resolved Phase 3A; 47 segments / 34.59 MWp is canonical per LC v1.0 register; documented in Part B § B.2 and Part F § F.6.3 deprecation list
AUDIT-042 Thai Summit Group concentration Tenant_Consent_Sensitivity tab in LC model quantifies cluster-out scenario (1.04× DSCR / 11.4% IRR / BREACH); documented in Part C § C.5.6 and Part E § E.7.1
AUDIT-043 LC2 Power Plant ESA legal Model classifies as special case; $2.40M replaceable from T6X reserve; methodology should formalise as "power-producer ESA sub-class" (Part D § D.4.6 documents the gap; full formalisation deferred to v1.2)

J.3 WP closure mapping

Mapping of active register items to the WP that closes them:

WP1: Bankable PV Yield Study (Weeks 1-4)

  • AUDIT-001 (yield convention reconciliation) — gating
  • AUDIT-017 (per-segment yield modeling)
  • AUDIT-026 (multi-source yield validation)

Closure deliverable: Solcast + PVGIS + Solargis cross-check report; per-segment yield calibration; documented P50 convention reconciliation. ~$10-15K commissioned study + analyst time.

WP2: LC-specific Cost Driver Engineering Review (Weeks 1-4)

  • AUDIT-040 (LC-specific top-down cost framework) → Annex H gating
  • AUDIT-002 (C4 corrosion BOS premium verification)

Closure deliverable: Engineering review of LC unit costs by typology; T1 \(412/kWp priority review; documented LC-specific cost drivers per typology. ~\)20-30K engineering firm scope.

WP3: Financial Model Methodology Alignment (Weeks 5-8)

  • AUDIT-014 (interest rate 5.75% vs 6.0%)
  • AUDIT-016 (per-typology O&M build-up)
  • AUDIT-021 (OPEX inflation framework)
  • AUDIT-041 (bifacial uplift typology-specific)

Closure deliverable: Updated LC model with methodology-aligned parameters; re-run IRR/DSCR; documented per-typology O&M and bifacial uplift. ~$15-20K analyst + Steven time.

WP4: Per-Tenant Credit Framework (Weeks 5-8)

  • AUDIT-027 (per-tenant credit framework) → Annex G

Closure deliverable: Annex G with credit grading methodology; LC's 12 Tier-1 anchors graded; concentration risk treatment formalised. ~$10-15K analyst + Benedikt time.

  • AUDIT-023 (BOI / EEC 13yr enhancement evaluation; partly closes via this WP)
  • Plus residual governance / ERPA structure items

Closure deliverable: KTB IE pre-engagement; Baker McKenzie ERPA structure review; documented EEC 13yr eligibility per estate. ~$25-35K legal scope.

Annex closures (post-WPs, Weeks 9-10)

  • Annex G (per-tenant credit): WP4 output
  • Annex H (LC-specific cost drivers): WP2 + WP4 output → gating for v1.1.0 lock
  • Annex I (pytest suite): closes AUDIT-031 + AUDIT-034
  • Annex J (this document): live update through lock

J.4 v1.1.0 lock items

The specific register items that must close for v1.1.0 lock:

# Register item Closure path Status
1 AUDIT-001 (yield convention) WP1 OPEN
2 AUDIT-014 (debt cost) WP3 OPEN
3 AUDIT-016 (per-typology O&M) WP3 OPEN
4 AUDIT-040 (Annex H — gating) WP2 + WP4 OPEN
5 AUDIT-041 (bifacial uplift) WP4 OPEN
6 AUDIT-002 (C4 corrosion) WP4 OPEN
7 AUDIT-017 (per-segment yield) WP1 OPEN
8 AUDIT-021 (OPEX inflation) WP3 OPEN
9 AUDIT-023 (BOI/EEC tax) WP5 IN-PROGRESS
10 AUDIT-026 (multi-source yield validation) WP1 OPEN
11 AUDIT-027 (per-tenant credit → Annex G) WP4 OPEN
12 AUDIT-031 (anomaly detection — Annex I) Annex I OPEN
13 AUDIT-034 (pytest suite — Annex I) Annex I OPEN

13 items gating v1.1.0 lock. Target lock date: 22 July 2026 (10 weeks from 12 May).

Items not gating: AUDIT-025 (T1 DSM, deferred v1.2); AUDIT-005 (IRR walk in IC deck, deferred v1.2); CLOSED items (AUDIT-015, AUDIT-019, AUDIT-030, AUDIT-032, AUDIT-033, AUDIT-035, AUDIT-036, AUDIT-038, AUDIT-042, AUDIT-043, AUDIT-044).


J.5 Items deferred to v1.2

The following active items are explicitly deferred from v1.1.0 to v1.2:

ID Title Deferral rationale
AUDIT-025 T1 DSM/shading per-segment LiDAR/DSM scope; T1 is 6% of LC envelope so impact minor (−10 to 0 bps); fold into v1.2 cross-estate yield refinement
AUDIT-005 No IRR walk in IC paper Presentation refinement; doesn't gate model accuracy; address with IC deck v2.0 post-Phase 0 closure
AUDIT-043 formalisation LC2 Power Plant power-producer ESA sub-class Edge case; model handles as special; full sub-class formalisation requires legal opinion deferred to v1.2

Plus the two structurally deferred annexes from Part A § A.6: - Annex K (Vietnamese / Indian canal-canopy operational benchmarks) - Annex L (Bottom-up cost stack methodology)


J.6 Live update protocol

J.6.1 Update frequency

Annex J updates on the following triggers: - Weekly during the WP1-WP5 sprint (Weeks 1-10): status changes from OPEN → IN-PROGRESS → STAGED → CLOSED - On WP closure: each WP completion generates an update batch - On lock: full re-scrub of all 47 items at v1.1.0 lock - Quarterly after v1.1.0 lock: standing review

J.6.2 Update process

  1. WP owner reports closure to Steven (process discipline)
  2. Steven updates Annex J: status column changes; evidence linked
  3. Marc reviews weekly during sprint; quarterly thereafter
  4. Annex J version increments per Part F semver (PATCH for status updates; MINOR for register addition)

J.6.3 Adding new items

If new findings emerge during WP1-WP5 or post-lock, they get added to the register as AUDIT-048 onward. Adding requires: - Title and cluster - Status code per J.0.2 - Materiality assessment - Resolution path proposal - Owner assignment - Steven approval before registration

J.6.4 Status changes that escalate

The following status changes require IC notification: - Any item moving from ACTIVE-METHODOLOGY to ACTIVE-IRR (materiality elevation) - Any item OPEN → RETIRED-SPECULATIVE (finding rejected) - Any item OPEN → DEFERRED-v1.2 (delay) - Any item DEFERRED-v1.2 → ACTIVE-IRR (escalation)


J.7 Closing summary

J.7.1 State of the register at v1.1.0 structural lock (12 May 2026)

  • 47 items total
  • 24 CLOSED or retired in v0.3 → v1.1.0 transition (51%)
  • 8 CLOSED in v1.1.0 structural lock via Parts A-F documentation (17%)
  • 13 OPEN gating v1.1.0 lock via WP1-WP5 + Annexes G/H/I (28%)
  • 2 DEFERRED-v1.2 explicitly carried (4%)

J.7.2 Concentration of remaining work

The 13 OPEN items concentrate in WP1 (3 items), WP2 (2 items via Annex H), WP3 (4 items), WP4 (4 items, two via Annex G), WP5 (1 item), and Annex I (2 items closure path).

Per WP outputs, the lock criteria narrow to: Annex H (gating) + pytest suite operational + all 13 items moved to CLOSED.

J.7.3 What the register tells us about methodology health

Of the 47 findings raised in the R3 audit, only 3 had material IRR impact (≥30 bps): yield convention, debt cost, and per-typology O&M. The remaining 44 were either: - Methodology completeness gaps (descriptive holes; close via documentation) - Governance gaps (process; close via Parts A and F) - Speculative (retired) - Not applicable to LC (other estates) - Memory hygiene (corrected via memory edits) - Duplicates (rolled up) - Model-handled (documentation only)

The methodology was structurally sound at v1.0 baseline. The R3 audit's contribution was making the 3 ACTIVE-IRR gaps visible (rather than letting them silently propagate) and producing the closure plan that becomes v1.1.0.

This is the diagnostic answer: the methodology had drift but not fundamental flaws.


J.8 References

  • NC-MN-001-R3 v0.3 audit register CSV (Drive ID 14yGDTxmZMONNTCRZ20weyyPzcRwz3HNSw4Ubyrt9oLs) — original 47-item register
  • NC-MN-001-R3_v1_1_close.md work plan (Drive ID 1b25nINIWWvuUpa7Wj9PZaNWbaWaQovRf) — 5 WPs over 10 weeks
  • NC-MN-001-CLOSEOUT_MEMO.md (Drive ID 1cMeFCS2NVjWw7nJ-hortfsadCF24YckJ) — audit cycle close-out
  • NC-METH-001 v1.1.0 Part F § F.8 — v1.1.0 lock gating events
  • Memory edits 1-30 (per memory_user_edits view 12 May 2026)

End of Annex J v1.1.0 initial issue.

Live update target: weekly during WP1-WP5 sprint; full re-scrub at v1.1.0 lock 22 July 2026.