Fast, rigorous thematic synthesis matters when a single decision rests on tens of thousands of public comments. Western Australia's EPA re-opened consultation on Woodside's Browse Basin project in May 2025 and collected more than 17, 000 additional submissions; the EPA accepted amendments on 21 August 2025 and will finalise its assessment (source: www.abc.net.au/news/2025-08-21/epa-assessment-proceeding-browse-basin-woodside-gas-extraction/105681048). This post shows a repeatable, audit-ready approach (qualitative analysis of public submissions) and a 5-step Evidano workflow (www.evidano.com) you can run today to move from raw comments to stakeholder-ready themes.
Fast take: what changed and why it matters
The EPA has continued the assessment of Woodside's Browse Basin proposal after the company submitted amendments in May 2025; the agency judged the revisions to be of the "same character" and will finalise recommendations for the WA environment minister (21 Aug 2025).
- Why analysts care: reopened consultation generated >17, 000 submissions, a scale problem for manual coding and synthesis.
- Primary research task: convert high-volume, often short-format comments into rigorous themes, sentiment, and cross-segment comparisons for policy or legal review.
Findings snapshot
| Date | Metric | Value | Source / Note |
|---|---|---|---|
| 21 Aug 2025 | EPA action | Accepted Woodside amendments; assessment to be finalised | www.abc.net.au (EPA statement) |
| May 2025 | Amendments lodged | Reduced footprint; excluded shallow water near Scott Reef | Woodside submission (reported) |
| 2019 | Assessment start | EPA began assessing Torosa reservoir extraction | Assessment timeline |
| , | Public submissions (reopened) | More than 17, 000 additional submissions | Re-opened 4-week consultation |
| , | Original footprint | 1, 220 km² (reduced in amendments) | Project documents |
| , | Distance to coast | ≈430 km off Kimberley coast (Torosa reservoir) | Project location |
| 2024 | EPA preliminary finding | Found potential unacceptable impacts (Scott Reef highlighted) | Quoted in reporting |
What happened (plain English)
Woodside applied to extract gas from the Torosa reservoir (Browse Basin) in 2018; the EPA's assessment began in 2019. In May 2025 Woodside submitted five amendments, including a reduction of the project footprint (no shallow-water work near heritage-listed Scott Reef). The EPA reopened consultation for four weeks, receiving over 17, 000 submissions, and on 21 August 2025 the EPA accepted the amendments as being of the "same character" and will finalise its report for the environment minister.
- Scale problem: 17, 000+ free-text submissions are heterogeneous (short comments, technical letters, petitions).
- Analytic needs: code consistency, cross-segment counts (e.g., conservation vs local industry), evidence-linked quotes for reviews or appeals, transparent audit trail.
So what for researchers and policy teams: implications
Policy & legal analysts
You need verifiable linkages between themes and source comments during appeals or ministerial briefings. Automated thematic extraction plus quote indexing preserves provenance and speeds drafting.
Track change over time (2019 → 2025 amendments) by comparing earlier submissions to reopened consultation using cross-cohort frequency analysis.
Environmental researchers
Quickly surface recurring ecological concerns (Scott Reef, marine turtles, 'unproven technologies') and quantify how often each appears across stakeholder groups.
Extract and compare technical submissions (e.g., scientific reports) vs lay comments to guide targeted follow-up studies.
UX / stakeholder researchers
Segment sentiment and themes by geography or stakeholder type (traditional owners, local businesses, conservation groups) to inform engagement strategy.
Use co-occurrence networks to find latent concerns (e.g., 'revenue' often paired with 'jobs' and 'mitigation').
How Evidano maps to the job: from 17k comments to an audit-ready dossier
Ingest & prepare
Problem: submissions arrive as PDFs, emails, CSV exports and scanned petitions.
Evidano fit: bulk document ingestion + OCR and metadata capture so every comment is searchable with source, timestamp, and submitter metadata.
Scaleable thematic coding
Problem: inconsistent manual coding and slow inter-coder checks.
Evidano fit: import or build a codebook, apply AI-assisted coding to the corpus, then validate with spot checks; generates hierarchical codes → subcodes for policy granularity.
Quantify & compare
Problem: stakeholders ask “how many” and “who said what”.
Evidano fit: frequency tables, cross-segment analysis, and co-occurrence networks that show which themes cluster by stakeholder type or date.
Evidence trail & reporting
Problem: decision-makers need traceable quotes and exportable exhibits.
Evidano fit: clickable quotes, exportable visualizations (word clouds, networks), and downloadable codebook + audit log for appeals or ministerial briefings.
Security & repeatability
Problem: sensitive submissions and legal risk require strict data controls.
Evidano fit: end-to-end encryption, PII redaction options, and a clear policy, your data is never used to train third-party models.
Five-step workflow: qualitative analysis of public submissions
Step 1; Intake & metadata harmonisation
Collect exports (PDFs, CSVs, emails). Tag each item with source, date, and stakeholder type. Output: searchable corpus.
Step 2; Seeded codebook & AI-assisted coding
Create an initial codebook from policy goals (e.g., 'Scott Reef impact', 'mitigation', 'economic benefits'). Run AI-assisted coding at scale, then confirm with 200-300 manual checks.
Step 3; Cross-segment frequency & co-occurrence checks
Run frequency analysis by segment (conservation groups, traditional owners, local industry). Generate co-occurrence networks to surface compound concerns.
Step 4; Evidence packs & traceability
For each key theme prepare a pack: top 10 representative quotes, metadata, and linked source documents for legal or ministerial review.
Step 5; Report, iterate, and store
Produce visual summaries and an executive memo. Store the coded corpus and versioned codebook so the analysis can be re-run if new submissions arrive or appeals are filed.
Common questions about qualitative analysis of public submissions
How do I ensure code reliability at scale?
Use AI-assisted initial coding, then validate with random manual checks and compute inter-coder agreement on a holdout set.
Can I compare submissions across consultation rounds?
Yes, harmonise metadata fields and run parallel frequency/co-occurrence analyses to detect shifts in concerns between 2019 and 2025.
Is this approach defensible in appeals or legal contexts?
Auditability is essential: preserve originals, log every coding pass, and attach direct quote links to maintain provenance.
Wrapping up & next steps
If you're facing a public consultation like the Browse Basin reopening (May–Aug 2025, >17, 000 submissions), prioritize a repeatable, auditable workflow: intake → AI-assisted coding → cross-segment analysis → evidence packs.
- Ready to try it? Start a pilot: ingest a 1, 000-comment sample, run the 5-step workflow above, and compare manual vs Evidano-coded outputs for accuracy and time saved.
- Sign up or learn more at www.evidano.com and run a pilot to convert submissions into defensible themes and stakeholder-ready reports.
Keep reading
- Commentary on NewsTwo Definitions: Climate Change Acceptance for UndergradsHow a PLoS One Delphi study (Aug 25, 2026) defined climate change acceptance for undergraduate science students, and how AI-enabled qualitative analysis applies it.
- Commentary on NewsResearcher-in-the-loop: AI-enabled UX researchHow the researcher-in-the-loop model governs AI-enabled UX research. Learn practical governance, stats from the August 2026 piece, and how Evidano supports this workflow.
- Commentary on NewsResearcher-in-the-Loop: Governance for AI UX ResearchGovern AI in qualitative UX research with the researcher-in-the-loop model from Jennifer L. Bowie (Aug 25, 2026): practical rules, risks, and tool mappings.