Open data practices for Horizon 2020 water research

Data management and open access in Horizon 2020 research projects are practical research activities, not administrative afterthoughts. ElectroSAnMBR combines electrolysis, anaerobic digestion and membrane bioreactor technology to treat ship bilge water. Its datasets can therefore connect reactor operation, pollutant removal and membrane performance, making careful stewardship essential to the project’s scientific value.

The goal is to make data as open as possible while protecting information that genuinely cannot be shared. The project objectives show how each work package contributes to removing oils, chemicals and organic pollutants. Open-access publications, documented datasets and access-controlled records can form one coherent evidence base. Australian participants can meet these expectations without weakening European grant obligations or shipping partners’ confidentiality.

Why data stewardship matters

Horizon 2020 grant delivery is closely linked to responsible research data management. Beneficiaries maintain a data management plan, publish peer-reviewed findings in open access and manage digital outputs with FAIR principles in mind. FAIR means findable, accessible, interoperable and reusable; it does not require every file to be downloadable without restriction. A searchable metadata record remains valuable when sensitive raw data are governed by an embargo or agreement.

For Australian universities, engineering consultancies and maritime suppliers in Sydney, Melbourne and Perth, this creates a shared vocabulary with European partners. Local teams can align file formats, authorship rules and repository deposits with the grant agreement while commercial operators retain control over proprietary process information. A data plan maintained as a living responsibility document is more useful than a once-only compliance exercise, particularly when laboratory, analytical and operational teams work in different locations.

Designing a consistent experimental record

ElectroSAnMBR’s experimental record may span electrical signals such as current and voltage, reactor conditions, membrane flux, gas measurements and chemical assays for oils and other pollutants. Each observation needs a date and time, reactor identifier, operating condition and sampling point. Raw instrument exports should remain separate from cleaned datasets so that corrections stay traceable.

A common naming convention should distinguish sample, reactor, treatment stage and measurement method. Metadata should record units, detection limits, calibration procedures, software versions and the treatment of missing values. Controlled vocabularies can prevent equivalent terms for bilge-water composition, organic load or membrane fouling from fragmenting a dataset.

Consistency becomes particularly important when affiliated laboratories use different analytical protocols. Reference materials, blanks, duplicate samples and instrument identifiers provide evidence of quality beyond a final removal percentage. Data stewards should specify these checks before experiments begin and review them as equipment, personnel and methods change.

Choosing repositories and metadata

Different outputs require different deposit routes. Repositories should receive completed datasets, working analysis code, protocols and a plain-language description of restrictions. Persistent identifiers such as digital object identifiers make records citable, while metadata connects each output to project milestones, publications and contributor roles. Institutional, domain and general repositories such as Zenodo each have a role; selection should reflect audience, preservation policy and access controls.

For researchers new to open science, general open access resources can offer an initial orientation, but project-specific rules should guide the final choice. Australian institutions often provide local research-data support, and the Australian Research Data Commons offers pathways for describing and linking Australian outputs. Repository selection should be settled early enough to influence file naming, consent wording and documentation.

Protecting sensitive and regulated information

Transparency does not mean indiscriminate disclosure. Bilge-water records may reveal vessel routes, cargo characteristics, treatment performance or commercially sensitive operating conditions. Sampling locations can carry security implications, and personal information may appear in operational logs. The data management plan should classify these risks, name a decision owner and define review or embargo periods.

Access should be tailored to the sensitivity and reuse value of each data layer:

Physical samples raise a separate Australian compliance issue. Customs declarations and Australian biosecurity requirements may restrict the movement of biological or contaminated material; a digital open-access deposit does not override those controls. Organisations shipping samples from Australia should confirm permits, packaging and receiving approvals. Teams must also respect contractual confidentiality, third-party rights and Australian privacy obligations.

For environmental consultancies serving the maritime market around Port Botany, Port of Melbourne or Fremantle, tiered access can protect client details while sharing methods and aggregated results. Where Indigenous cultural interests or community knowledge are involved, relevant cultural and intellectual property expectations should guide consultation and attribution. Explicit boundaries keep those safeguards compatible with open science.

Practical data stewardship choices

Clear assignments translate the plan into daily research practice across laboratories, engineers, students and industry partners.

For laboratory and engineering teams:

• Assign a data steward and deputy at each laboratory to resolve naming, versioning and deposit questions.
• Create read-only raw-data folders with separate backups before processing begins.
• Record sample identifiers, blanks, duplicates and instrument calibration in a quality-control log.
• Keep analysis scripts in version control and link released datasets to the exact script version used.
• Use documented, non-proprietary export formats where practical, retaining original files.

Repository and coordination duties need equally explicit ownership:

• Map every output to a repository, named owner, licence and access category in the data management plan.
• Record participant consent, confidentiality and third-party restrictions before data collection.
• Check metadata completeness, file integrity and anonymisation before any public release.
• Publish a citable dataset record even when selected files remain under controlled access.
• Update the plan after major method changes, new work packages and final repository deposits.

Teams can apply these lists during laboratory meetings, periodic reviews and close-out audits. Immediate correction of an ambiguous filename prevents costly reconstruction later.

Demonstrating quality and research value

Evidence of good stewardship extends beyond the number of files uploaded. Dataset completeness, documented provenance, searchable metadata and timely corrections show that a record can support verification. Reuse citations, software use and documented adaptations indicate practical value. Coordinators should review these indicators collectively rather than treating them as a ranking of individual scientists.

Horizon 2020 reporting also depends on traceable links between grant deliverables, project outputs and peer-reviewed publications. ORCID identifiers, contributor roles and a shared project identifier help credit people who prepared samples, maintained equipment, analysed data and reviewed manuscripts. Open-access licensing should be applied where appropriate, while restricted datasets receive clear justification rather than an unexplained absence.

Sustaining a durable open research workflow

An open research workflow has three stages. During setup, teams agree responsibilities, consent language, metadata fields and repository routes. During experiments, automated backups, version-controlled processing and quality reviews keep records reliable. At project close, the coordinator checks each work package, resolves remaining access decisions and deposits a final dataset with a persistent identifier.

Workflows should survive staff turnover. Standard operating procedures, data dictionaries, archiving instructions and training material are valuable project outputs. Formats such as CSV, TIFF and NetCDF can improve long-term accessibility, while archived code environments help reproduce numerical analyses. Preservation checks should confirm that deposited files can actually be opened and understood.

ElectroSAnMBR can present its methods, performance data and publication record as an integrated resource for the wider ship-wastewater sector. Links between reactor design, analytical procedures and removal results make the evidence useful to researchers entering the field. Shared underlying data support independent scrutiny of treatment claims and adaptation to different bilge-water conditions.

Act on the project’s open research legacy

Begin with the next laboratory meeting: assign responsibility for the data management plan, inspect existing files against a shared template and nominate the repository for each output. Publish the first metadata record early so contributors can check descriptions while details remain familiar. Where access must be limited, document the reason and review date rather than leaving a file unexplained.

Support the project by using its published methods and datasets in future research, citing relevant outputs and crediting the full consortium. A complete, discoverable and well-governed record will give the ElectroSAnMBR team a stronger evidence base and the maritime research community a reusable foundation for improving ship bilge-water treatment.