How to Write Air, Water and Wind Requirements in a Window Specification

1. Write three performance subjects, not one marketing sentence

Air leakage, water penetration resistance and structural performance are related to a building envelope but are not interchangeable evidence. ASTM E283/E283M describes a laboratory method for air leakage through an exterior fenestration specimen at specified pressure differences. ASTM E331 addresses resistance to water penetration with water applied to the exterior face and exposed edges together with a uniform static air-pressure difference. ASTM E330/E330M addresses structural performance under a uniform static air-pressure difference in a test chamber. A result in one category cannot be used as a result in either of the other two.

The first drafting discipline is therefore grammatical as much as technical: give each requirement its own heading, named method, edition, unit convention, specimen description and result rule. Avoid writing ‘air/water/wind test: pass’. That phrase hides whether the project means laboratory or field evidence, whether water is being assessed at a manufactured assembly or installed perimeter, and whether ‘wind’ refers to a structural test, a design-load calculation, a cyclic protocol, impact exposure or another subject entirely.

FGIA’s NAFS performance-class overview also warns against reducing a Performance Grade to a single design-pressure number. Its public explanation describes PG as a group of requirements associated with a design-pressure range and the intended location. In other words, a named classification, a method result and a project design basis each have their own role. A clear specification states which one the project invokes and does not silently treat them as synonyms.

2. Use a requirement grammar that leaves no hidden comparison

A usable requirement has at least seven fields: the performance subject; the governing method and edition; the intended evidence context; the unit and test condition to be reported; the exact unit/configuration boundary; the project’s acceptance rule; and the authority for resolving a difference. These fields are more valuable than a long paragraph because a supplier, laboratory and reviewer can map them to a report without guessing what the writer meant.

Do not use a number without its method, unit basis and condition. ASTM E283/E283M explicitly notes that comparisons of air-leakage rates may not be valid unless tested components are essentially the same size, configuration and design. A concise requirement should therefore require the report to identify those facts. Likewise, ASTM E331 notes that actual service performance may depend on installation, adjustment, supporting construction and deterioration; a laboratory water requirement should not be worded as a blanket guarantee of the finished wall.

A method name alone is not enough either. The project must say whether it asks for a pre-award laboratory report, a project-specific mock-up, a field verification, or more than one evidence stream. A laboratory air test under ASTM E283/E283M does not automatically answer a field air-leakage question; ASTM identifies E783 for field measurement of installed exterior windows and doors. The specification should preserve that boundary rather than calling every check ‘ASTM tested’.

3. Define the proposed unit before asking for performance evidence

A report can be authentic and still be inapplicable. Require each supplier to submit a proposed-versus-tested comparison matrix. At minimum, identify the system, operation, dimensions, number of vents or panels, glazing make-up, hardware, reinforcement, seals, mullions/transoms, accessories, mounting arrangement and supporting construction described by the report. Then compare each fact with the scheduled project unit rather than with a product-family name.

The matrix should not pretend that every difference is fatal. Its job is to expose differences for the project’s technical review. A larger unit, altered panel arrangement, another glass build-up, a different mullion, changed hardware or different support can matter to one or more performance subjects. If the writer cannot state why an evidence transfer is acceptable, the correct status is ‘open for responsible review’, not ‘same series’.

4. Do not let the word ‘wind’ hide a missing project load basis

‘Wind resistance’ is often used loosely. ASTM E330/E330M explains that real loading depends on factors including direction, height, building shape, terrain and surrounding structures, and that the method evaluates the specified specimen rather than adjacent construction. Its public scope is a strong reason not to turn a general test report into a project wind design. The responsible project process must establish the actual design basis and translate it into the required evidence route.

When drafting, distinguish: the building/project load basis; the evidence method requested; the test specimen and its support; and the acceptance decision. If the project also requires cyclic pressure, impact, debris, seismic movement or another hazard, name it as a separate requirement with its own governing documents. Do not imply that a static structural test answers every exposure. ASTM’s public building-standards index lists separate methods for cyclic pressure and other performance subjects, which is a useful reminder that one method title cannot absorb another hazard.

The same care applies to pressure language. A project value belongs beside its source, units, duration/sequence where required by the selected method, and responsible interpretation. This article intentionally leaves those values blank: they arise from jurisdiction, building geometry, exposure, code, project design and the applicable current standard—not from a reusable web template.

• Identify the document and professional that establish the project load basis.
• Name the requested method rather than using ‘wind test’ as a generic label.
• State whether the evidence is laboratory, mock-up, field or calculation-based.
• Record the test mounting and the boundary to the supporting construction.
• List other hazards or performance subjects separately rather than implying coverage.
• Keep all project values, units and acceptance interpretations in the controlled specification register.

5. Make the performance submittal a decision gate

A useful submission package contains the controlled project requirement, original reports, laboratory identity, full method designations, specimen descriptions, proposed-versus-tested matrix, relevant drawings, any stated limitations, and a reviewer disposition. A one-page declaration may point to evidence but is not a substitute for it. Keep air, water and structural records separately indexed so a reviewer can identify a missing stream without treating the other two as compensation.

Close each row with one of three plain statuses: accepted for the named evidence purpose; difference requiring authorised disposition; or evidence not supplied/not traceable. None of those labels is a product certification. They describe the state of a project record at a point in time and make it possible to control substitutions before fabrication or installation changes the configuration again.

Compare the testing allowance, not just the headline pressure

Consider a hypothetical tender where one offer says pressure tested, a second includes an existing water-penetration report and a third prices new project testing. Return the same opening schedule to each bidder. Ask which air, water and structural subjects are included, which documents already exist and which activities require additional work. These are different deliverables even if each quotation calls them testing.

Keep every pressure with its original subject, units, method and specimen. A water-test pressure is not a structural design-pressure declaration. Leave a subject unresolved when its supporting evidence is absent instead of copying the strongest figure across three columns. Ask the project reviewer whether each result addresses the specified requirement.

Separate specimen manufacture, laboratory charges, transport, witnessing, reporting and agreed repeat work in the commercial return. Identify responsibility for the installation interface in a planned test. An existing report and a commissioned project test should not share one unqualified allowance.

Close the scope before production release

Link the reviewed evidence to the opening groups and configuration revision that were assessed. A later change from fixed to operable panels, or to another layout, needs a recorded review rather than an assumption that the earlier decision still applies. Ask how additional evidence, cost and programme implications will be communicated before authorising that change.

Keep a named owner and disposition for each subject: reviewed for the stated purpose, further submission required, or alternative awaiting approval. Preserve the original documents with this decision record. It is a procurement tool, not a new certificate or a substitute for the project's acceptance authority.

Use the performance report guide together with the window specification checklist to prepare the return. Discuss your performance brief with the opening schedule and the evidence you want compared.

FAQ

Continue the project review

Connect this guide with the proposed window and door systems, then review how Cavorlen approaches manufacturing and quality control. When your schedule is ready, send it through the project review form.

Related buyer guides

• How to Read Window Air, Water and Structural Test Reports for Supplier Technical Review
• An Aluminum Window Specification Checklist for International Projects
• Window and Door Installation Checklist: From Opening Review to Final Adjustment