Technical traceability in asset portfolios: data governance and audit trail

The technical due diligence of a portfolio of renewable assets is, in part, a substantial traceability exercise. The independent technical advisor (ITA) not only assesses whether the estimated P50 is reasonable — it checks whether it is possible to reconstruct the complete path of how that number was produced: which data were used, who validated them, which version of the software applied which methodology and whether any treatment decision was documented with justification.

When this path cannot be reconstructed, the ITA compensates for the undocumented uncertainty by increasing the uncertainty factors applied to P90 — which degrades debt service coverage (DSCR) and can render financial closure impossible or impose onerous previous conditions.

This note presents the technical traceability structure that Renewene applies in renewable resource portfolios: the four levels of documentation, which each protects and how the absence of each component results in additional quantifiable uncertainty.

The four levels of traceability

Technical traceability in renewable energy operates on four interdependent levels. A strong but weak chain in the lower levels does not compensate — the ITA assesses the complete chain:

1. Metrological traceability: Each measuring instrument shall be traceable to a national (INMETRO) or international (BIPM) standard by an uninterrupted chain of calibrations with declared uncertainty in each link;
2. Data custody: Each raw data file is linked to its origin, operator, timestamp and instrument, and each subsequent transformation is documented with the software, version and parameters applied;
3. Analysis version control: Each analytical result (MCP, Weibull, AEP) is versioned, with a record of which inputs generated which outputs and which analyst approved each version;
4. Audit trail: Any decision to process — excluding a month of data, choosing a clearsky model, applying a correction factor — is recorded with technical justification, normative reference and name of the person responsible.

Level 1 — Metric traceability

The ISO/IEC 17025:2017 defines metrological traceability as "the property of a measurement result by which the result can be related to a reference through an uninterrupted and documented calibration chain, each contributing to measurement uncertainty". For renewable resource campaign instruments, this means:

• Anemometers and calibrated laboratory piranometers accredited by INMETRO (primary or secondary standard);
• – Calibration certificate with instrument serial number, date, accredited laboratory, expanded uncertainty (k=2) and declared traceability to BIPM;
• History of calibrations: when the instrument was calibrated, what sensitivity before and after, and what drift accumulated since the last calibration;
• For instruments replaced in the field: two certificates — of the instrument removed and of the installed — with verification that the data in the transition are homogeneous.

The most common failure at this level is not the absence of calibration, but the Unlinking the certificate to the dataThe certificate exists on paper, but the data system does not record which instrument (by serial number) collected which lines of the file. In due due diligence process, the ITA requests exactly that link — and if it does not exist, it applies additional uncertainty over the whole series.

Level 2 — Data custody and lineage

Each raw data file must be sealed at the time of ingestion with unchangeable metadata: receiving timestamp, cryptographic hash (SHA-256 or equivalent), origin (instrument, site, height) and operator responsible for receiving. Any subsequent transformation — temporal aggregation, unit correction, QA/QC filter — generates a new derived file with reference to the father file and the software script/version that produced the transformation.

This data line model allows you to answer, at any future point, the question: "This number of P50 that I see in the report was calculated with the original data or with some manually modified intermediate file?" The correct answer should be: not only that the original data were used, but that no intermediate file was manually edited without registration.

Level 3 — Analysis version control

Renewable resource analyses evolve over a project: an initial MCP with 12 months of data is refined after 24 months; a power curve is updated with a new version of the manufacturer; a wake factor is recalculated with final layout. Each new version must coexist with the previous ones in the versioning system — not replace them — and each version must record:

• What input data were used (by hash and version);
• Which software and version executed the calculation;
• – what has changed compared to the previous version and why;
• Who approved the new version and when.

The most problematic scenario in due diligence is the existence of multiple versions of an analysis without clear hierarchy: Excel spreadsheets numbered as "final_v2", "final_v3_revised", "final_v3_revised_USE_THIS" in the same directory, with no record of which was effectively used in the submitted report. This pattern invalidates the chain of custody and forces the ITA to request further full analysis — unplanned cost and term.

– Portfolio traceability matrix

In portfolios with multiple assets, the heterogeneity of documentation is the rule, not the exception. Older assets tend to have fragmented documentation or in proprietary formats without digital custody chain; more recent assets follow stricter standards. The traceability matrix allows mapping this state and prioritizing remedial actions before a due diligence process:

Level 4 — Audit trail and decision log

The most demanding level of traceability is the record of decisions — not only what was done, but why it was done. Each treatment decision affecting the outcome of AEP should be documented in a structured log:

Decisions not registered in this format — even when technically correct — become invisible to the ITA. A month excluded without a decision log is equivalent, for due diligence purposes, to an arbitrary exclusion: the analysis is equally valid in both cases, but only one is auditable.

Impact on P90's uncertainty

The relationship between traceability and uncertainty is direct: each missing or untraceable documentation component forces the ITA to apply an additional uncertainty about the result. Not because the data is necessarily wrong, but because it cannot be verified — and what cannot be verified cannot be excluded as a source of error.

Minimum viable data governance system

Implementation of the four levels of traceability does not require complex systems from the beginning of a portfolio — but requires structural discipline from the first data collected. The indispensable elements of a minimum viable system:

1. Register of instruments: Database of field equipment with serial number, installation date, calibrations and replacement dates, linked by instrument and period to the corresponding data files;
2. Automated intake with hash: Every input of raw data into the system must automatically generate a content hash (SHA-256) and an immutable metadata record — eliminating the possibility of subsequent silent modification;
3. QA/QC Pipeline versioned: The data validation code must be controlled by version (git or equivalent), with each execution recording which version of the pipeline generated which set of validated data;
4. Structured decision log: Each exclusion, marking or correction of data shall be recorded in the format described in the previous section — not in spreadsheet comment, but in an auditable system;
5. Analysis version control: Each calculation of MCP, AEP or P50 must be reproducible from the record of which validated data + which software version + which parameters generated the result.

Documentation for due diligence

When a portfolio enters the process of due diligence — whether for refinancing, sale or extension of debt term — the ITA typically requests the following documents related to traceability:

• Complete registration of instruments with calibration certificates by serial number and period of validity;
• QA/QC log for each measurement campaign, with documented and justified decisions;
• Hash SHA-256 (or equivalent) of the raw and validated data files delivered to ITA — so that ITA can independently verify integrity;
• History of analysis versions with record of changes between versions and approver of each version;
• For portfolios with acquired assets: due diligence original technique of the acquired asset or, in its absence, documental remediation analysis describing which elements could not be recovered and which additional uncertainties were applied accordingly.