CWCT Test Certificates: Keeping Façade Evidence Findable

The email arrives on a Thursday, four years after practical completion. Water is getting into a flat on level nine, the building owner has appointed a consultant, and the request is one line long: please provide the test evidence for the curtain wall on the north elevation. Somewhere in your project files is a test report on a laboratory letterhead with a system name across the top. The question the consultant asks next is the one nobody prepared for, and it is not whether you have a test report. It is whether that report describes the thing that is actually on the wall.

What's Actually Being Asked

CWCT standards are the reference most UK façade specifications use to define how a building envelope must perform. The Centre for Window and Cladding Technology is an independent, membership-based body that publishes standards and technical guidance for building envelopes, routinely called up in façade specifications: weathertightness, structural performance under wind load, air permeability and impact resistance. Where a specification names a CWCT standard and a performance class, it is setting the bar the envelope has to clear.

Testing against that bar happens on a specimen. A representative sample of the system is built at an accredited test laboratory, put through a defined sequence, and the laboratory reports what that specimen was and how it behaved. That report is a statement about a specimen built under laboratory conditions on a particular date. It is genuinely valuable evidence, and it is not, on its own, evidence about your building. It is worth being precise about a word the industry uses loosely: an accredited laboratory issues a test report on a specimen tested against a named standard. That is not a certification of a product, a system or an installer, and a specimen passing certifies nothing about the wall you built.

The gap between the two is where façade contractors get caught. The tested configuration and the installed configuration are rarely identical. Bracket types get substituted, sealant is changed, a transom moves to suit a slab edge, a corner condition on level nine does not appear anywhere in the specimen. Some of those differences are immaterial, and somebody with the right expertise can say why. But "somebody said it was fine at the time" is not a record. The written engineering justification for a departure from the tested configuration is evidence in its own right, and in practice it is one of the pieces most often missing from a façade evidence pack.

So a request for test evidence is really four questions arriving together:

• Which standard and performance requirement did the specification call up?
• What specimen was tested against it, by which laboratory, and what did the report actually say, including any deviations the laboratory recorded?
• How does the installed system differ from that specimen, and on what basis was each difference accepted?
• What site testing confirmed the installation, as opposed to the design?

None of this is new to façade work. What has changed is who asks and how long afterwards. For higher-risk buildings in England, meaning buildings at least 18 metres or seven storeys with two or more residential units, the Building Safety Act regime requires the dutyholders and the accountable person to keep building safety information accurate, accessible and up to date. In practice that pushes down the chain: envelope information has to exist and be producible on request rather than reconstructed, and what you owe depends on your appointment and scope. But you do not need a higher-risk building to receive that Thursday email. A leak, an insurance claim, a fire risk assessment, a refinancing or a sale is enough.

The Evidence That Answers It

A façade evidence pack that survives a specific question is a chain, not a folder. Each link points at the same part of the building:

• The specification requirement. The clause, standard and performance class the envelope was sold against, kept with the revision it was at when you priced and built.
• The test report. The laboratory's report for the system as tested: specimen description, test date, standard applied, performance achieved, and anything the laboratory noted as a departure or a limitation.
• The tested configuration. Drawings or a specimen definition showing what was physically built for the test. Without this, nobody can assess whether the installed detail is covered.
• The justification for departures. For every difference between the specimen and the installation, a written assessment, signed and dated, saying who accepted it and on what technical basis.
• Materials traceability. Delivery records, batch references and declarations of performance connecting what was tested to what arrived on site and went into the wall.
• Installation evidence. Fixing and bracket records, torque checks, membrane laps and fire-stopping photographed before they were covered up, ITP sign-offs and benchmark approvals, each tied to an elevation, level and grid reference.
• Site test records. Hose tests and other site checks, where the specification calls for them and in line with the relevant CWCT technical guidance: location, date, operative, witness, result, and where a test failed, the remedial work and the re-test that closed it.
• Competence records. Who installed it and what qualified them, including third-party certification where the specification requires it, as it commonly does where fire-stopping sits in your scope.
• The change trail. RFIs, technical submittals, approved deviations, and the decision behind each one.

Notice what that list is doing. It is not proving you are a careful contractor. It is proving that a specific panel, in a specific position, was built from a specified system, by a qualified person, in a way the test evidence covers.

Why the Spreadsheet Version Fails

On a well-run façade package, most of these records already get generated. The failure is not diligence, it is geography. The test reports sit in a shared drive named after the system supplier. The hose test results are in a spreadsheet one person maintains. The installation photographs are spread across four phones and a WhatsApp group that has since been archived. The justification for the bracket substitution was agreed on a call and confirmed in an email thread that left the business with the project manager who ran it. Every one of those records is real. Collectively they cannot answer the question that actually gets asked, because the question is rarely "send me everything". It is "which panels did that failed hose test cover, what system are they, and does the test evidence for that system include the corner detail". Answering that from a folder structure means a person, a fortnight, and a set of assumptions nobody will write down. A pack assembled after the question was asked invites a second argument about the pack itself, on top of the argument about the building.

Keeping It Provable

The fix is not a better folder convention. It is deciding that the evidence is created once, as the work happens, and anchored to the building rather than to the filing system.

Anchored is the load-bearing word. A record filed by date, by document type or by the supplier's name can only be found by somebody who remembers the filing logic. A record attached to elevation C, level nine, panel type UC-04 can be found by the thing it describes, which is how the question usually arrives. When a hose test is attached to the panels it covered, and those panels carry the system reference, and the system reference carries the test report and the accepted departures, one question retrieves the whole chain. Nobody has to remember anything.

The chain also has to hold its history, so a failed test, the remedial work and the successful re-test read as one traceable sequence rather than three unconnected files and a gap, and a deviation shows the request, the assessment and the acceptance with names and dates that were not written retrospectively.

The test of whether you have this is simple. Pick a panel. Ask somebody who was not on the project to produce, from records alone, the evidence that it performs. If they can do it in an afternoon, you have a living record. If it takes a fortnight and three phone calls, you have a folder graveyard, and the only thing standing between you and that Thursday email is that nobody has asked yet.

The Monday Checklist

1. Pick one elevation on a live job and list, from records alone, the CWCT standard and performance class the specification calls up, or confirm that it calls up something else. Note how long it takes to find.
2. For that elevation, produce the test report and the specimen description together. If you have the report but not the tested configuration, request it from the system supplier now, while the relationship is live.
3. Write down every known departure from the tested configuration on that elevation, and check whether each one has a dated, signed technical justification. Chase the ones that do not.
4. Check that your site test records carry location, date, operative, witness and result, and that failures are linked to their remedial and re-test rather than sitting as separate entries.
5. Audit last month's installation photographs: can each one be tied to an elevation, level and grid reference without somebody standing there explaining it?
6. Confirm competence records for the operatives who installed that elevation are retrievable by name and date, not just held somewhere in HR.
7. Agree who owns the handover pack from day one of the next project, and give them the authority to reject records that arrive without a location.

Where BrieXO Fits

The BrieXO bundle that covers the site end of this chain is FIELD, and it is live today: daily site records and ITP sign-offs, governed photo evidence, operative competence records, record-linked communication and programme visibility, all captured as the work happens and anchored to the building, elevation, level and element they describe through the spatial spine every BrieXO tenant runs on, with approvals recorded in a tamper-evident ledger, and records that stay exportable. Linking test evidence directly to individual specification requirements is planned for XO Spec in the DESIGN bundle, and is on the roadmap rather than available today. BrieXO is not affiliated with or certified by CWCT: CWCT sets the standards, and the platform organises the evidence against them. It supports your golden-thread and envelope evidence duties; it doesn't discharge them, and nothing in it decides whether your envelope performs. The wider picture is set out in our CWCT testing and evidence guide.

Related reading:who evidences what, by trade, the as-built record that survives scrutiny, and why façade contractors are now in the evidence business.

George Sfica is a senior design manager at a specialist façade contractor, with 23 years in manufacturing and construction, eleven of them in façades and external envelopes, spanning fabrication, installation and design management, an MSc in Façade Engineering and an IFE Level 2 Certificate in Passive Fire Protection. He is the founder of BrieXO.