Dalian Oppgate Glass Windows & Doors Co., Ltd. 路 Exporting WorldwideVerified Alibaba Storefront

Heat-Soaked Toughened Glass for Overhead and Structural Glazing: Nickel Sulphide Inclusions, EN 14179-1 and the Limits of the Test

Two black-framed aluminium roof skylights mounted side by side on a flat white roof, the nearer sash raised on a pair of electric actuators above a stepped aluminium upstand, with sealed glass units set into the frames
Two black-framed aluminium roof skylights mounted side by side on a flat white roof, the nearer sash raised on a pair of electric actuators above a stepped aluminium upstand, with sealed glass units set into the frames
Key Architectural Takeaways
  • Heat soaking is a probabilistic batch process, not a per-pane certificate: it lowers the rate of nickel sulphide failure, it cannot reduce it to zero.
  • EN 14179-1 heat soaking applies to thermally toughened glass only, before lamination and before insulating glass unit assembly.
  • A CE declaration for toughened glass under EN 12150-1 is not evidence that the panes were heat soaked.
  • Overhead glazing normally needs both defences: heat-soaked panes for probability, laminated construction for consequence.

1. The Breakage Nobody Specified: Why Toughened Glass Fails in Service

Toughened glass is not stronger because the material is different. It is stronger because the quenching process leaves the surfaces in a state of high compression and the core in tension. A crack can only propagate where the local stress is tensile, so a surface flaw has to work against the compression layer before it can open. That balance is what allows a 6 mm pane to resist wind load and impact that would destroy the same glass in the annealed state, and it is also why the pane disintegrates into small blunt fragments once the balance is lost. The energy stored in the tension zone is released all at once.

When a toughened pane breaks in service, the failure usually belongs to one of two families. The first is thermally induced. A temperature difference across the pane, or between a sunlit glazed area and an edge concealed behind a rebate, gasket or spandrel panel, generates tensile stress at the edge, and the break runs in from the boundary. The second has no external cause at all: a microscopic inclusion inside the glass, sitting in the tension zone, expands slowly until the local stress exceeds what the surrounding glass can carry. The fracture then starts at a point well inside the pane and radiates outward. Around that point the fragments form a characteristic pair of butterfly-shaped pieces, which is the closest thing to a signature the failure mode has.

Both failures end the same way on site, which is why they are so often confused in claims. The difference matters commercially because the remedy is different. Thermal stress is managed by edge clearance, glazing detail, heat-strengthening or shading; inclusion-driven breakage is managed at the glassworks, before the pane is ever assembled into a frame.

The reason this subject belongs in a procurement document rather than a maintenance manual is the consequence. A break in a ground-floor window is a replacement order. A break in a skylight over an atrium, a canopy over an entrance, or a structural glass wall over a public route is a closure, a temporary platform, a crane or a mobile elevating work platform, and an insurance conversation. On overhead glazing the cost of one failure routinely exceeds the price difference between ordinary toughened glass and heat-soaked toughened glass for the whole elevation.

2. What a Nickel Sulphide Inclusion Is, and Why It Waits

Nickel sulphide stones form in float glass when nickel-bearing material enters the batch or the melt. The usual sources are contaminated cullet, stainless steel scrap that reaches the furnace, nickel alloys in contact with the melt, and in some cases nickel compounds present in refractory or in the fuel system. At melting temperature the nickel and sulphur combine into a sulphide droplet, which is then frozen into the ribbon as a small opaque stone, typically a fraction of a millimetre across and invisible at normal inspection distance.

The behaviour that causes the problem is a phase change. At high temperature the sulphide exists in a phase that is stable in the melt. The rapid cooling of the glass, and of the toughening quench in particular, does not give the inclusion time to transform, so it is retained in a metastable high-temperature form. At ambient temperature the lower-temperature phase is the stable one. The transformation happens anyway, slowly, over months, and it is accompanied by a volumetric expansion of a few percent. A few percent does not sound like much until it is confined inside a solid: the inclusion pushes against the surrounding glass, and if it sits in the tensile core of the pane, the local stress it adds is tensile and it is added to stress that is already there.

Three consequences follow, and each one shapes how the risk should be specified. First, only inclusions above a critical size release enough strain energy to start a crack, which is why not every contaminated pane fails, and why the failure rate is expressed per tonne of glass rather than per pane. Second, the position of the inclusion in the thickness matters as much as its size. An inclusion in the compressive surface zone is harmless; the same inclusion in the core can be critical. Third, the failure is delayed. Breakage typically appears weeks to a few years after installation, which places it after practical completion, after the retention period has been negotiated, and often after the supplier considers the project closed.

Inspection does not solve this. The stones are not visible through the glass at any useful production speed, and no sampling regime can prove a batch free of them. Automated inclusion scanning of float glass exists and is a useful screening step, but it cannot certify a toughened batch, because the transformation and the stress state are what determine failure, not the mere presence of a particle. This is the gap that heat soaking was invented to close.

3. Where the Standards Draw the Line: EN 12150-1, EN 1863-1 and EN 14179-1

European practice separates three products that are routinely confused in schedules. EN 12150-1 defines thermally toughened soda lime silicate safety glass: glass that has been heated and then rapidly cooled so that the surface is in compression, and which must fragment into a specified minimum number of particles when tested, counted away from the edges and corners of the specimen. That fragmentation requirement is the practical definition of the product, and it is what distinguishes it from heat-strengthened glass.

EN 1863-1 defines heat-strengthened glass. It receives a milder quench, retains a lower level of residual stress, and breaks into larger, slower-moving fragments rather than dice. Because it is not a safety glass in the fragmentation sense, heat soaking as defined in EN 14179-1 does not apply to it. This is the single most common drafting error in glass schedules: asking for heat-soaked heat-strengthened glass, or accepting an offer of heat-strengthened glass described as heat treated when the specification asked for heat soaking.

EN 14179-1 defines heat-soaked thermally toughened soda lime silicate safety glass. Its scope is a toughened product that has been subjected to a defined heat soak cycle in order to reduce the risk of spontaneous breakage from nickel sulphide inclusions. Two boundaries in that scope matter on a project. The process applies to toughened glass, not to heat-strengthened glass, and it is applied to the individual pane before lamination and before the insulating glass unit is assembled. A finished double-glazed unit cannot be heat soaked: the interlayer or the sealed cavity would not survive the cycle. If a project requires heat-soaked laminated glass, the requirement is that the constituent panes are heat soaked, then laminated under EN 14449.

ProductGoverning standardFracture behaviourHeat soaking applicableTypical use
Annealed float glassEN 572-2Large sharp shards, low residual stressNoNon-safety locations, or as a component of laminated glass
Heat-strengthened glassEN 1863-1Larger fragments than toughened, higher residual stress than annealedNo, the product is outside the scope of EN 14179-1Spandrels and locations where flat fracture retention is wanted
Thermally toughened glassEN 12150-1Dice-like fragmentation, minimum particle count per unit areaNot unless separately specifiedDoors, sidelights, balustrades, overhead glazing
Heat-soaked thermally toughened glassEN 14179-1As toughened, with a reduced nickel sulphide riskThis is the heat-soaked productOverhead, structural and difficult-to-access glazing

4. What the Heat Soak Cycle Actually Does

There is also a marking issue that catches specifiers during submittal review. Toughened glass supplied under EN 12150-1 carries the CE marking and declaration of performance for that standard. Heat soaking is a separate process defined in a separate standard, and the declaration for the toughened product does not by itself state that the panes were heat soaked. The heat soak record has to be requested and read as its own document, with its own batch references.

The heat soak process is deliberately simple in principle. The toughened panes are loaded into an oven, brought up to a holding temperature of the order of 290 degrees Celsius, held there for approximately two hours, and then cooled at a controlled rate. The temperature is high enough to accelerate the phase transformation inside a nickel sulphide inclusion without softening the glass or relaxing the temper. An inclusion that would have transformed in the second year on a facade transforms in the oven instead, and the pane breaks there, in the plant, on a rack, where the loss is a manufacturing cost rather than a site incident.

What survives is glass in which the risky inclusions have already failed. What survives is not glass that is guaranteed free of them. A pane that passes has a materially lower probability of spontaneous breakage, and the industry treats the reduction as substantial, but the mechanism is statistical and the standard describes a process control, not a warranty of zero failures. This distinction is where specification language most often goes wrong: a schedule that says no spontaneous breakage will occur is not a specification, it is a promise no glass processor can make.

Because the process is applied to a batch, the records are batch records. A furnace cycle covers a defined load of panes, and the useful documentation identifies the cycle, the date, the furnace, the loading and the panes included. Panes that break during the cycle are removed, and the replacements must themselves be toughened and heat soaked before they can join the order. This is a detail worth writing down, because on a fast programme the replacement panes are exactly where the process gets quietly skipped.

Two limits deserve to be stated plainly. Heat soaking does not repair edge damage, and edge damage is a leading cause of toughened glass failure in service. Nor does it change the thermal stress behaviour of the pane: a heat-soaked pane installed with insufficient edge clearance, or in a spandrel that traps heat, will still break from thermal stress. The process addresses one failure mode, and a project that treats it as a general cure is buying reassurance rather than performance.

5. Why Heat Soaking Does Not Replace Laminated Glass

Heat soaking manages probability. Lamination manages consequence. On overhead glazing the two are not alternatives, and a schedule that specifies one while ignoring the other has left half the risk in place.

A heat-soaked toughened pane can still break, from edge damage during handling, from thermal stress, from impact, from a restraint that was tightened against the glass, or from a nickel sulphide inclusion that was below the critical size at the time of the cycle. When it breaks, it breaks completely and immediately, into fragments that leave the frame. Above a walkway, an atrium floor, a stair or a canopy, the difference between a pane that falls and a pane that stays in its opening is the interlayer.

Laminated glass under EN 14449 bonds two or more panes to a polymeric interlayer. After fracture the interlayer holds the fragments together and, depending on the interlayer and the support conditions, keeps the broken assembly in the frame long enough for the area below to be closed. The practical specification for overhead work is therefore laminated units manufactured from heat-soaked toughened panes, with the interlayer selected for the post-break condition the project requires rather than for acoustic or solar reasons alone.

In an insulating glass unit the laminated pane is normally positioned so that fragments are retained on the inboard face and the cavity remains sealed, and the structural design of the unit has to account for the fact that a broken laminated pane still carries load through the interlayer while a broken monolithic pane carries none. None of this is a substitute for heat soaking. Lamination without heat soaking converts a falling hazard into a retained broken pane, which is an improvement in safety but still a failure, a replacement and an insurance event. Heat soaking without lamination reduces how often that event happens but not what happens when it does.

6. Writing the Requirement into the Glass Schedule

The requirement has to survive translation from the architect's specification to the glazing subcontractor's order to the processor's work ticket. In practice it survives only if it is written as a product and a process at the same time, with the records named. The following clauses are the minimum that will still mean something three steps down the supply chain.

  • Name the product and the process separately: thermally toughened soda lime silicate safety glass to EN 12150-1, heat soaked in accordance with EN 14179-1. Do not use heat treated, HT, or HS as shorthand; those abbreviations are read differently by different suppliers.
  • State that heat soaking applies to each pane before lamination and before insulating glass unit assembly, and that the requirement therefore applies to the constituent panes of any laminated or double-glazed unit in the affected locations.
  • List the locations by elevation or by glass type reference rather than applying the clause to the whole order. Heat soaking adds a process step, a handling cycle and a breakage allowance, and blanket application where it is not needed is money spent on nothing.
  • Require a heat soak record per batch, identifying the cycle date, the furnace or plant, the quantity and the project reference, and require that the record be submitted with the glass release documents rather than at handover.
  • Require that replacement panes, including site breakages and spares, are heat soaked to the same standard. A replacement pane that skipped the cycle reintroduces exactly the risk the specification was written to remove.
  • Prohibit cutting, drilling, grinding or edge reworking after toughening, and state that any pane so modified is rejected. This is not a heat soak clause, but it is the clause that determines whether the heat soak still means anything when the pane reaches site.
  • Ask for the marking and the traceability that lets a pane on site be matched to its batch record. Without that link the record proves that some glass was heat soaked, not that this glass was.

During submittal review, the document to look for is the heat soak record, not the CE certificate. A supplier who responds to a heat soak clause with a CE declaration for EN 12150-1 has answered a different question.

7. Procurement and Programme: What Heat Soaking Changes in an Export Order

Heat soaking sits between toughening and assembly, so it lengthens the glass programme at exactly the point where a fenestration package is usually already compressed. The panes are toughened, loaded, soaked, cooled, inspected, and only then laminated or built into units. Each transfer is a handling step, and each handling step is an opportunity for the edge damage that the process cannot protect against. On an export order this matters more than on a domestic one, because the panes then travel several thousand kilometres before anyone sees them installed.

The scheduling consequence is that the glass release has to be treated as the critical path item it is. Frames can be fabricated against approved shop drawings while the glass is still being processed, but the units cannot be glazed until the heat-soaked panes are available, and the ship date cannot be protected by compressing the soak. A programme that assumes heat soaking can be absorbed at the end of the glass cycle is a programme that will either slip or quietly lose the requirement.

Order a spare set. The economics are straightforward: a small number of additional panes, processed in the same cycles as the main order, costs far less than a single air-freighted replacement unit manufactured after the project has closed. Spares should be recorded against the same batch references and stored with the project documentation, not absorbed into general site stock.

Packing has to be specified as if the glass were not heat soaked at all, because it is not more resistant to edge damage. Panes travel vertically in purpose-built crates or steel A-frames, separated so that glass never contacts glass, with the edges protected and the frame bearing on the correct face. Units should not be stored flat with weight bearing on the centre of the pane, and they should not be unpacked and leaned against a wall on site while the openings are prepared. Most of the breakages attributed to inclusions in the first weeks after delivery are the result of handling, and the fracture origin at the edge is what gives them away.

8. Handover, Diagnosis and the Failures That Get Misread

The handover file for overhead or structural glazing should contain three things that are almost never all present: the heat soak records linked to pane identification, a drawing or schedule that records which panes in which locations were heat soaked, and a stated replacement procedure including the spare pane stock. Without the second item, a future maintenance team cannot tell whether a pane in a failed location was specified as heat soaked at all.

When a break does occur, the fracture pattern is the first diagnostic tool. An origin well inside the pane, with a butterfly pair of fragments around it and no contact damage nearby, points to an inclusion. An origin at the edge, with the fracture running inward, points to edge damage, a hard contact, or thermal stress at a restrained or covered edge. An origin at a fixing or a setting block points to installation stress, restraint or movement. These are not subtle distinctions, but they are routinely skipped because the fragments have already been swept up by the time anyone with an interest in the cause arrives.

Preserve the evidence. Photograph the opening before clearing, keep the fragment containing the origin if it can be identified, record the pane reference, and report it to the supplier with the batch record. A pattern of breaks in one batch, in one elevation, or one size is information; a single break with no record is an anecdote, and it will be settled by whoever holds the retention.

The honest summary for a specifier is that heat-soaked toughened glass is a well-defined process with a measurable effect on one specific failure mode, sold sometimes as a guarantee it cannot provide. Specifying it correctly means naming the standard, naming the process, requiring the batch record, keeping lamination for consequence, and accepting that the remaining risk is managed on site through edge protection and correct glazing detail rather than eliminated in the furnace.

MATCHING ARCHITECTURAL SYSTEM

Explore OPPGATE Architectural Skylights

CE certified modern motorized sliding roof skylights and architectural roof windows with double-tempered Low-E glass.

View All Architectural Skylights Models
Architectural Skylights

Planning a Commercial or Villa Project?

Send your window schedules or architectural CAD drawings to the OPPGATE engineering team for a prompt, certified quotation and structural review.

Request an Engineered Quotation
Request a QuoteSend Drawings