Whole-Window U-Value vs Centre-of-Glass: How to Compare Thermal Evidence

1. Draw the thermal boundary before comparing any number

The phrase ‘U-value’ is incomplete until it says which object is being evaluated. Centre-of-glass data concerns a central glazing area away from the frame and edge region. A frame value concerns a frame profile or section. Edge-of-glass information concerns the glazing-frame junction. A whole-window value combines the relevant elements for a defined product configuration and size under a stated method. A building-envelope junction can add still another layer of thermal-bridge analysis beyond the product boundary.

ISO 10077-1:2017 specifies methods for calculating thermal transmittance of windows and pedestrian doors with glazed and/or opaque panels fitted in a frame. ISO 10077-2:2017 specifies numerical treatment of frame profiles and the linear thermal transmittance of their junction with glazing or opaque panels. Both public abstracts help explain why a glass-centre result cannot stand in for a complete framed window. ISO 10077-2 also excludes some effects, including thermal bridges between the frame and the building structure; that boundary should remain visible in any project claim.

NFRC’s Component Modeling Approach describes configuring glazing, frame and edge-seal components into a whole product to estimate whole-product energy performance. That is a second useful evidence model: components can be inputs, while the product rating is the result for the configured product. Do not use an approved component listing as though it were an approved rating for every completed window.

2. Record the calculation or rating route with the result

A thermal number without its method is not fully comparable. ISO 10077-1 provides a calculation route for windows and pedestrian doors; ISO 10077-2 provides numerical treatment for frames and frame-glazing junctions. NFRC identifies NFRC 100 as its procedure for determining fenestration product U-factor and publishes certified-product and component systems. Each route has defined inputs, boundary conditions and reporting conventions. The project should name the route it accepts before bids are normalised.

Do not mix results expressed under a different unit system or condition set in a spreadsheet and call them like-for-like. Keep the original declared value, units, method/version, environmental conditions where stated, product size, configuration and source document in every row. If an authorised conversion is required for a tender format, retain both the original value and the controlled conversion record; do not overwrite the source result.

Likewise, do not merge U-factor with solar heat-gain coefficient, visible transmittance, air leakage, condensation index or any acoustic metric. NFRC lists these as separate product ratings. They may all belong in a performance schedule, but they answer different questions. A better thermal value does not prove a solar, acoustic, water or structural conclusion.

3. Compare the configured product, not the IGU slogan

A thermal quotation often begins with phrases such as ‘double glazing’, ‘Low-E’ or ‘thermal break’. These are configuration clues, not a whole-product result. The whole-window route needs the complete framed configuration. NFRC’s CMA approach explicitly combines glazing, frame and edge-seal components into a whole product. Its certified-product directory also presents product ratings alongside product characteristics such as frame type, glazing layers, low-E, spacer and gas fill. The practical lesson is to request the entire configuration record, not merely the glass sales sheet.

For each scheduled opening group, capture system and operation, frame and sash sections, thermal-break route where applicable, glazing make-up, coatings and surface, gas/cavity/spacer, dividers or grids, panel layout, overall dimensions, opaque infill if present, and the evidence method. If any component changes, the evidence record must be checked again. A new spacer or grid can change the product configuration even when the centre glass remains the same.

This approach avoids overstating difference as well as similarity. The purpose is not to tell buyers that a particular frame or spacer is always decisive; the project method and result decide that. It is to stop a component-level claim being promoted without evidence for the actual delivered configuration.

4. Ask what geometry and area relationship the result represents

A compact fixed light and a large operable window may use the same glass but have different relative amounts of frame, edge and glass. That is why a centre-of-glass value is especially weak as a tender comparator for different product geometries. A whole-window result should identify the product type and configured size or rating basis, and the reviewer should compare it to the proposed opening group rather than assume that a single family value covers every shape.

ISO 10077-1 covers calculations for framed windows and doors, while ISO 10077-2 addresses frames and the linear junction with glazing. The standards’ public abstracts do not authorise a generic area-weighting calculation in this article, nor do they prove a result for another geometry. If the project needs different sizes assessed, ask for the applicable project calculation/rating route and preserve the product sizes and configurations actually evaluated.

Do not jump from a whole-window U-value to whole-building performance. ISO 10077-1 and ISO 10077-2 have declared boundaries, including exclusion of thermal bridges between the frame and building structure. Building-level energy use also depends on many inputs outside a product number. Keep product evidence and building modelling as separate decision records.

• Group openings by product type, operation, size/geometry and configuration before comparing evidence.
• Retain the original result together with the evaluated product/rating size.
• Ask whether frame fraction, edge condition, dividers or opaque areas differ between evidence and proposed units.
• Use the project’s accepted calculation or rating route for any additional sizes.
• Keep building-junction and whole-building analysis outside the product-rating claim.
• Reopen the evidence review when a configuration substitution is proposed.

5. Turn thermal evidence into a controlled tender decision

At bid stage, ask every supplier for the same evidence pack: the project-required metric and method; original source report, calculation or certification reference; product configuration; evaluated size; declared conditions and units; configuration comparison; and a statement of items not covered by the evidence. This lets the reviewer distinguish a complete whole-window product result from a glass component quotation without dismissing either as useless.

At approval stage, status each opening group as: evidence matches the named project comparison purpose; evidence differs and awaits authorised review; or insufficient/non-comparable evidence supplied. Avoid writing ‘U-value approved’ without the boundary. The buyer needs to know whether the approval concerns a centre-glass component, a frame component, a calculated whole product or a certified product configuration.

This creates a durable replacement record too. Years later, a replacement request can retrieve the frame, glazing and spacer inputs rather than relying on the phrase ‘same energy glass’. That traceability is often more valuable than a headline number because it permits a qualified reviewer to see exactly what would change.

Untangle a glass value, a whole-window calculation and a label claim

Imagine three offers for the same opening schedule: glazing supplier centre-of-glass data, a whole-window calculation, and a certified product reference. Keep those evidence types visible. The first can inform glazing selection but does not complete the whole-window comparison. The other two still need the proposed configuration and the project's accepted method checked.

Where a bidder relies on an NFRC listing, request the exact product reference rather than a logo or the manufacturer's homepage. Inspect the identified directory record and retain the reference consulted. Do not extend a listed configuration to an unlisted alternative or call a product certified merely because one component appears in a system.

Keep U-factor, SHGC and visible transmittance in separate fields. If the tender expects different units, preserve the original entry and a checked conversion separately. A smaller printed number is not meaningful until units, metric, product boundary and configuration are aligned.

Review a cost-saving substitution against the whole order

Suppose a proposed saving retains the same centre glass but changes a spacer, frame arrangement or divider. Ask the supplier to identify the changed input and which calculation or rating covers the revised product. Do not carry the old whole-window figure into the revised quotation just because the glass description is unchanged.

Link accepted evidence to the opening schedule and revision, then retain that reference for replacements. Separate the offered product result from building-energy or wall-junction assessment required from the design team. Record the revised price and evidence decision together so procurement does not accept a saving while unknowingly keeping an obsolete performance description.

Use the thermal-break buyer guide together with the Low-E configuration guide to prepare the return. Request a configuration comparison 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

• Thermal Break Aluminum Windows: What Project Buyers Actually Need to Check
• Low-E Coating Surface Position: A Window Glass Review Guide
• How to Read Window Air, Water and Structural Test Reports for Supplier Technical Review