EPBD Data and Databases

Energy performance certificates generate data, and the EPBD framework expects that data to be collected, stored and used. National databases typically hold certificate records, including the building type, the calculated performance, the recommendations and the assessor or certifier, and they support quality checks, policy analysis and, in some countries, public access to certificate information. The recast strengthens this data dimension, with attention to interoperability, to linking certificate data with other building information, and to making data available for renovation planning while respecting privacy.

That data serves several purposes at once. For policymakers, aggregated certificate data shows how the stock is performing, where the worst performers concentrate and whether renovation policies are shifting the distribution. For owners, a database record provides a durable copy of a certificate that can be retrieved for a sale, a letting or a renovation decision. For quality control, databases allow checks for implausible results, for assessors whose outputs diverge systematically, and for certificates that need correction.

What a database contains is narrower than many people assume. A certificate database holds the standardised assessment and its inputs as recorded, not a full survey of the building's condition, not its real energy bills and not its structural or damp history. A database entry therefore supports comparison and planning, but it does not replace a physical survey when work is planned. Treating a database record as a complete building history overstates what it is.

Access rules also vary. Some countries publish certificate data openly, searchable by address, while others restrict access to owners, to professionals with a legitimate purpose, or to aggregated statistics. Those differences reflect different balances between transparency, market information and personal data protection, and a reader should check the national access rules before assuming a record is public or private.

Before using database data, check its date, its method and its coverage. A certificate from many years ago may reflect an older method or a building that has since changed, and a database that covers only certificates issued after a certain date will miss older buildings. Where a record looks implausible, check the underlying certificate and its inputs before drawing conclusions about the building, because database errors usually originate in the assessment or in data entry, not in the database itself.

For owners, a practical check is to retrieve the record for your own building where the national system allows it, and compare it with the paper or digital certificate you hold. Discrepancies in address, building type or rating should be queried with the issuing body or the database operator, since an incorrect public record can affect a sale or a letting even where the underlying building is unchanged.

By place, database maturity differs widely. Some member states have operated certificate registers for many years with broad coverage and public search, while others are consolidating registers or linking previously separate systems under the recast. The EU level does not hold a single register of every building; it works with national data and with observatories and statistical collections that aggregate national figures. A European comparison therefore depends on national data quality and on harmonisation efforts that are still under way.

Privacy and data protection shape all of this. Certificate data about a home can reveal information about its occupants' circumstances in indirect ways, and national systems must comply with data protection law. That is one reason access rules differ, and why aggregated policy use is often preferred to individual disclosure for research purposes.

A common mistake is to treat database coverage as stock coverage. A database that holds certificates for buildings sold or rented recently over-represents transacted buildings and under-represents homes that have not changed hands, so its average rating is not the average rating of the whole stock. Another mistake is to use an old database entry as if it described the current building, ignoring intervening work or deterioration.

A further mistake is to assume that because data exists, it is accurate. Databases faithfully store what was entered, including errors, and quality checks catch only some of them. A database record is a lead for verification, not a substitute for it.

Keep a copy of the certificate itself, not just a database printout, and note the database reference, the issue date and the method. When a later certificate is issued, keep both, so the change over time can be traced and any error in either record can be identified.

Questions to ask include: what does this database actually hold, who can access this record, and how old is the assessment behind it. Those questions keep building data in its proper place: useful, limited and checkable.