Issue 02: Disproportionate Collapse in the UK
- shingi tarirah
- Aug 31
- 16 min read
Updated: Sep 15
I knew that some aspects of my work would change when moving back to the UK. For example, I would need to (re)learn structural design to Eurocodes. Projects are also structured differently here, so my workflow would need to adapt. There were, however, a couple of notable unknown unknowns regarding materiality that became evident and important early on.
My first tasks were doing carbon counts for colleagues’ projects, but I had never heard of a carbon count before, nor did I understand how it was used to inform design choices. Any given project, and consequently the materials used in it, is a confluence of various objectives including client goals, architectural intent, safety, funding, and in some cases, a desire to experiment with form and performance. While sustainability had also been a consideration in my Japanese projects, the focus was on efficient energy and water use; structural considerations were cursory. If a project is to pass through a series of filters to appraise its performance, for a long time the key filters in my designs were beauty, safety, cost, and yarigai, the concept of whether something is worth doing. Sustainability in its multivalency turned out to be an important introduction to that list. How fine a filter would this be for me? How would my perspective sit next to that of the client and the rest of the design team?
Later when I was assigned a design project, I was quickly confronted with the phenomenon of disproportionate collapse. The project was an all-timber, multi-storey educational building in America. At the time, the first American standard for disproportionate collapse was still under production, so we used local guidance to provide proof of concept. I learned that disproportionate collapse requirements are based on usage, size and material. Steel and concrete structures are typically easier to detail while timber and masonry structures can become more involved, particularly for taller structures. Here was the second new metric added to the toolbox. The feasibility of adequate disproportionate collapse detailing became a stronger consideration when I made design suggestions.
This Part looks at what I have learned so far about disproportionate collapse safety measures, the commonalities I see with Japanese design, and how I relate this to other issues such as sustainability in my practice.
PART 2: Dispropotionate Collapse in the UK
The story of disproportionate collapse design in the UK begins with the Ronan Point collapse of May 1968. The disaster was triggered by a gas explosion in an 18th-floor apartment which set in motion the progressive collapse of one corner of the building. The building was of LPS (large panel system) construction where concrete walls and floor panels stack onto one another and support each other. The construction is like a tower of cards which is stable under its self-weight. The explosion caused the failure of critical load-bearing elements which left upper floors unsupported. These fell onto lower floors and in turn caused their failure until the whole corner of the 22-storey building collapsed, resulting in the death of 4 people and the injury of 17 others.
The disaster was an example of both progressive and disproportionate collapse. The failure of one section of the building triggered the collapse of other sections which in turn triggered the collapse of others further still - progressive collapse. The extent of the observed failure was also deemed to be disproportionately large compared to its cause - disproportionate collapse. These terms are often used interchangeably but we are separating the terms here for clarity.
Tower blocks received higher subsidies compared to houses. Taller tower blocks received higher subsidies compared to shorter ones.
The 1968 report from the investigation conducted in the wake of the collapse, known colloquially as the Griffiths Inquiry, became a major public document which laid out the determined causes for the collapse. It also recommended ways to increase safety and the robustness of precast concrete buildings like Ronan Point. The Inquiry found that there were both design and construction inadequacies. Though considered relatively small, the blast load was unforeseen and therefore not accounted for in the design.
The tower was also designed to codes that did not reflect the wind loads associated with the trend of increasing building heights of post-war construction. As a result, the design wind load was underestimated, and the panel system may not have been able to withstand the expected wind loads. Both these points are impactful for LPS systems as they are easier to destabilise under lateral loads compared to other forms of construction.
The Inquiry also highlighted the extent of construction defects including the inappropriate filling out of joints using paper rather than concrete. Poor construction remained an issue after the collapse. The work of high-rise safety campaigner Sam Webb showed that defects still impacted the tower even when residents were moved back in after its reinstatement.
The building was one of many other similar prefabricated panelised buildings that were being erected across Europe as part of post-war regeneration. The systems developed by the likes of Raymond Camus and the Larsen & Nielsen Company were popular for their quick construction and were also favoured outside of Europe.
The demand for public housing in the UK was pressing. The Housing Act of 1956 incentivised councils to construct taller apartment blocks. Tower blocks received higher subsidies compared to houses. Taller tower blocks received higher subsidies compared to shorter ones. Significant investment was made into social housing and hundreds of LPS tower blocks went up across the country in the latter half of the 20th Century.
Ronan Point constituted a critical turning point for the prefabricated industry; eventually LPS buildings fell out of favour. Successful campaigning led to its demolition 18 years after its construction having served only 16 years of its assumed 60-year design life. Other tower blocks later followed suit as fears about their integrity heightened, and in 2023 the demolition of the tallest remaining tower, the 23-storey Goscote House in Leicester, was completed.
The Grenfell Tower fire of 2017 renewed attention on the safety of high-rise residential towers which has fuelled concerns about their safety and the transparency about possible risks. Like Ronan Point, it also highlighted the problematic culture of prioritising profit and ease over quality that is now associated with the construction industry. The need for public housing feels ever present and the political, economic and social pressure to provide people with safe and affordable homes continues.
Tragedies such as these have far-reaching impact for society broadly, but for designers this usually comes in the form of changes to codes and standards. Other impacts can also include fundamental changes in practice, professional certification and insurance structures. An increased appetite for litigation is also a common side effect.
Development of Design Codes
Ronan Point elicited the first design codes for disproportionate collapse in the UK. The first guidance appeared in the 1970 amendment of the Building Regulations for England and Wales (Fifth Amendment) and much of what was established in this first iteration is still use today. The new clauses in this Amendment are taken from the guidance in the Griffiths Report which sought to remedy the lack of continuity and ductility in precast panel systems. The advice in the report and in other guidance following the collapse was specific to the nature of precast panel construction. Though some behaviour and failure mechanisms may be unique to LPS buildings, the Report’s advice applied to all construction types in the Regulations.
There are several methods for designing for disproportionate collapse and they are typically rooted in providing and enhancing structural robustness. The definitions of structural robustness vary somewhat but most focus on highlighting a structure’s ability to resist collapse despite incurring damage. Events like earthquakes, fire or impact are some of the considered triggering events. Some of the considered events may be classified as accidental or abnormal. Robustness is also defined as the load-redistribution capacity for a structure. Remember in Part 1 we saw how Route 3 seismic design relies on this principle when considering the potentially devastating impact of earthquakes.
There are several ways to provide and enhance robustness and a variety of methods are considered internationally. The methods stated in the Building Regulations for England and Wales (Regs) are tie-force design, element removal, key element design and systematic risk assessment.
Design Methods
Tie-force Design
Early tie-force design was first seen in the Fifth Amendment where it appeared as a requirement to provide alternative load paths. More specific guidance was proposed in the Institution of Structural Engineers (IStructE) report RP/68/05. Tying was originally proposed to improve continuity between precast members and prevent brittle collapse.
Though the nuance is lost in modern codes, ties were intended to provide resilience through specific survival mechanisms such as catenary, membrane or cantilever action. It is crucial that connections must meet a minimum level of performance to allow these actions to develop. Once mobilised, the mechanisms redistribute load from failed primary load paths to elsewhere in the building. These assumed alternative load paths must have enough capacity to withstand the newly distributed load to prevent the propagation of failure.
Initial guidance aimed to prevent brittle collapse but there were no associated quantitative checks. This means the rotational and ductility requirements of joints are not explicitly assessed to determine if they are sufficient for survival mechanisms to develop. As discussed in Part 1, ductility allows for increased energy absorption without collapse or strength loss. Improved ductility can increase the time available for evacuation and improve the conditions under which this occurs.
Ties were intended to provide resilience through specific survival mechanisms such as catenary, membrane or cantilever action.
Additionally, advice was specific to precast construction and was developed during a time when steel and RC structures were not considered to be susceptible to progressive collapse. As a result, little attention was given to the behaviour of these structures and the suitability of tying methods in such buildings. Even today, most modern codes still lack provisions addressing rotational or ductility requirements.
Member connections are to be designed for a tie force that is separate from the forces that arise from typical design loading. The prescriptive nature of the tie force inherently ignores the particularities of any given structure, making it difficult to assess the method's efficacy. Additionally, since the initial development of tying methods, construction design and architectural tastes have evolved. Buildings have increasingly large column spans, particularly those for commercial use. With large spans, load redistribution over larger distances becomes less probable and this should be considered when assessing the suitability of tying methods.
Despite these limitations, tying is understood to have a positive contribution to robustness and is widely (though not universally) specified as an effective method for increasing robustness. However, the predominant view among researchers in the field is that tying’s contribution should be deemed minimal making it more appropriate for low-risk buildings.
It became clear for me while working in the UK that modern British construction typically separates the vertical load-bearing and lateral stability functions in structural planning. This is different from Japanese design where the whole structure can fulfil both functions and rigid frame construction is commonplace. Separating structural functions in this way can allow for the simplification of connections: they typically only resist shear and axial forces. However, simple connections like these have reduced inherent ductility compared to moment-resisting connections like those in Figure 4 that are common in seismic design.
These connections are designed so they do not fail before the building reaches its ultimate state i.e. failure. It is unlikely that moment-resisting connections will regain popularity for everyday use in UK construction without a significant shift in norms, but it is important to understand the impact of the design thinking (and its alternatives) on the design.
The prototypical details in the Standards... were not explicitly designed with collapse loads in mind.
As many buildings in the UK are of masonry construction, it is important to touch on the measures for these structures. Most load-bearing masonry structures are houses which typically fall into CC1 where no additional measures are required, but larger-scale residential developments often fall in higher classes where additional measures may be necessary.
For CC2A buildings, effective anchorage may be used as a viable alternative for horizontal ties. Lateral restraint requirements outlined in the British Standards predate disproportionate collapse guidance. The prototypical details in the Standards are intended for use where wind forces act. Though these details were not explicitly designed with collapse loads in mind, they were widely adapted in practice as adequate disproportionate collapse measures for CC1 and CC2 buildings, despite differences in requirements. This is presumably because major failures were not identified regularly or occurred to alarming enough an extent.
Current IStructE guidance moves away from this qualitative approach and advises the use of calculated horizontal tie force values. Additionally, as of 1 March 2025, the National House Building Council (NHBC), the country's largest provider of warranties and insurance for new homes, stated on its website that simply specifying prototypical details will no longer suffice as valid justifications of disproportionate collapse provisions.
For Class 2B buildings where both horizontal and vertical ties are required, it has long been acknowledged that vertical tying is not always practical. To resolve the detailing within the masonry, ties would need to be placed in voids in the masonry walls. This may not always be practical as voids rely on the bond pattern of the masonry. Steel columns used as wind posts that run along the height of the masonry wall are an alternative, though floor anchorage may still pose difficulties. As a result, early on the codification journey, researchers posited that alternative methods such as element removal may be more practical for both design and construction.
Element Removal & Key Element Design
Element removal and key element design also made their first appearance in the Fifth Amendment. Notional element removal is usually considered when effective tie design cannot be guaranteed. It ensures that even with notional removal of a load-bearing structural element like a column or wall, the whole structure remains stable and the damage caused by the removal is limited to a localised area. This method also aims to reduce the extent of damage following a failure. In Approved Document A, the area affected by the removed element should not exceed the lesser of 15% of the floor area and 100m2. Column spacings have increased over time meaning the affected area by the notional removal of a column has also increased; the area limit was likely increased to 100m2 to reflect this.
Where an element's removal would cause excessive damage to the structure or compromise its overall stability, the element is to be designed as a “key element”. Examples include transfer slabs or beams which support several storeys, and the columns and/or walls that may support these. Key elements are required to resist a 34kPa notional static blast load considered in conjunction with other accidental loadings. This notional load is first seen in the Ministry of Housing and Local Government circular 62/68 and is derived from a 5psi approximation of the force during the blast event at Ronan Point.
This method is essentially scenario- or threat-dependant. Critical hazards, and their associated loads, may vary from structure to structure. Some loads may also differ from the prescribed load and potentially be failure-inducing. Without understanding failure capacity, it is difficult to understand whether a building is truly robust against specific loads.
Key element design is a relatively straightforward method, but some older codes emphasise that it should be considered as a last recourse only employed if alternative load paths cannot be established. This is because the failure of a key element would present an intolerable risk to the structure. It is therefore up to us as designers to judge whether additional checks beyond those outlined in the codes are required to ensure adequate robustness.
Systematic Risk Assessment
Systematic risk assessment, which applies to CC3 structures, has the largest breadth of scope among the stipulated methods. Because specific checks are not prescribed, there is increased latitude for design solutions. There is also an onus on the designer to thoroughly identify relevant risks and evaluate the proposed structural response. This method encourages us to look more closely at the particularities of any given design and apply sound engineering judgement. Systematic risk assessments do not have to be limited to CC3 buildings and may also be conducted for other buildings of significant complexity or importance.
In England and Wales, the assessment means each structure must be evaluated for hazards that may impact it. Where possible, hazards should be eliminated. Should hazards remain, all risks inherent to them must be reduced so far as is reasonably practicable. This is known as the ALARP principle; ALARP meaning as low as reasonably practicable. Whether a structural solution is deemed reasonable may also be influenced by buildability, budget and programme, so input from others may be necessary.
How much is too much?
Whilst in Japan, I viewed materials as inherently neutral and the skill I needed to cultivate was that of choosing the right material for the right job. Move 6,000 miles west however, and a shift occurred. It seemed materials were now on a hierarchy based on their carbon values. It felt concrete was villainised for its resource- and energy-intensive production process. Alternately, timber took on a messianic quality and was often talked about as a kind of carbon emission panacea. With my instincts primarily based on seismic design, I struggled to accept the recommendations for timber’s widespread use.
We have a dual responsibility as designers. One is to nature...the other is to society.
Timber can be a good solution for small- to medium-rise buildings and where lateral forces are relatively small. However, there are limitations around connection detailing for large forces from events like earthquakes. Fire design in high-rises also seem challenging to overcome without using framing with large proportions. Over time, the conversation about materials and their role in the climate crisis has become more nuanced, but I at times return to the tension I initially felt between sustainability and structural safety.
We have a dual responsibility as designers. One is to nature: responsible resource management is needed to avoid overstressing natural systems. The other is to society: safe, well-designed physical infrastructure is essential to support social and economic progress. The interplay between these is best exemplified by looking at emerging economies, where providing urgent, vital services may not look as “clean” as is aspired to in global climate targets. However, designers in these contexts are still encouraged to view climate-conscientious design as a moral and ethical obligation. Indeed, a recent opinion piece in The Structural Engineer used off-the-cuff calculations to illustrate that “for every 200 000t of GHGs [greenhouse gases] emitted into the atmosphere (or 160 000t of carbon dioxide if you prefer to count it that way), someone, somewhere will die prematurely as an indirect consequence.” According to the US Environmental Protection Agency, that is the equivalent of running approximately 48 wind turbines for a year.
These issues are intertwined and can be difficult to reconcile at large scales, but we are going to zoom in on the impact on structural design. Let’s start with material use. The essential aim is to use only as much as necessary.
One of the pleasant discoveries I made working in the UK was the renewed celebration of refurbishment and retrofitting. Aside from the unique opportunities to intimately interact with history, such projects can reduce the amount of new material required. Expenditure can also be reduced by designing so the capacity of frames and systems is more fully utilised. The energy and resources needed for virgin materials can be minimised by using recycling or reusing. Additionally, using imposed loads that more accurately reflect actual intensity is also an option. I will subsume these approaches under the term “lean design”.
Next, we will look at the impact of lean design on robustness and safety. Alternative load paths, which underlie most methods explained above, are only viable if they can sustain redistributed loads. To absorb extra loads, there should be enough strength and stiffness in the rest of the system which may need be fleshier as elements become larger and connections are made more robust. How, then, do we ensure structural safety alongside careful use of resources?
The suggestion here is that life preservation should be the principal objective. In the long run, unsafe buildings ultimately introduce unacceptable risks and are more likely to be demolished if their safety issues cannot be easily corrected. Resource management and other critical priorities are then planned once design has percolated through safety considerations.
When designing for alternative load paths, utilisations may be maximised when designed including redistributed loads rather than only those used for typical design. Designing members under typical loads with fractional capacity margin will not support the assumptions of alternative load path design. Using code-prescribed imposed loads as they are, can afford some degree of redundancy if the loads are assumed to be rarely fully realised. This creates margins in loading, though this would need to be quantified to determine the effectiveness of this approach.
Where risks are assessed as part of the redundancy design, it is important to clarify the tolerable level of risk and the required performance implied by this. Excluding unreasonable hazards can also help avoid overdesign which may lead to excessive material use. Additionally circular principles such as design for disassembly (DfD) can be applied. It is also important to consider the availability of materials in your location. In some cases, the most pragmatic option is to use what is readily available to avoid mortally threatening an essential project.
I still believe the thoughtful and appropriate dressing of structural designs is more impactful than the indiscriminate use of low-carbon materials. I also think that taking a practical approach that appreciates my design environment is important. I am regularly refining and rethinking the way I work. As I design more, I will continue to learn what to put first in each situation. The ledger of priorities might look very different depending on where in the world I find myself designing. My hope that I get better at contributing to outcomes that duly serve the people and places involved.
How are you managing the priorities on your projects? How much are you design really costing the earth? What practical compromises are you making to make sure people get what they need?
Selected Bibliography & Further Reading
Buksh, A., & Jackson, L. (2025, January 18). Over 100 London tower blocks may have safety issues. Retrieved from BBC News: https://www.bbc.com/news/articles/c3vppz1q4e1o
Fire Industry Association. (2025, January 21). Concerns Mount Over Safety of Large Panel System Tower Block. Retrieved from Fire Industry Association: https://www.fia.uk.com/news/concerns-mount-over-safety-of-large-panel-system-lps-tower-block.html
Hawkins, W., Peters, A., & Mander, T. (2021, May). A weight off your mind: floor loadings and the climate emergency. The Structural Engineer, 18-20.
Jessel, E. (2018, April 10). Demolition planned for tallest remaining Ronan Point-style tower block. Retrieved from Architectural Journal: https://www.architectsjournal.co.uk/news/demolition-planned-for-tallest-remaining-ronan-point-style-tower-block
Kokot, S., & Solomos, G. (2012). Progressive collapse risk analysis: literature survey, relevant construction standards and guidelines. Luxembourg: Publications Office of the European Union.
Vogel, B. (2024, January 23). Why is the Building Safety Regulator targeting large panel systems? Retrieved from Construction News: https://www.constructionnews.co.uk/health-and-safety/why-is-the-building-safety-regulator-acting-on-large-panel-systems-23-01-2024/
Webb, S. (2019, May 17). The government is not doing enough to prevent another tragedy in a high-rise tower. Retrieved from Inside Housing: https://www.insidehousing.co.uk/comment/the-government-is-not-doing-enough-to-prevent-another-tragedy-in-a-high-rise-tower-61453
Image Credits
Thumbnail image: Tasos Katapodis/Getty
Figures 4, 7: © analogs | this~that
Disclaimer
This article is for general information purposes only and may rely on the author's personal experience. It does not constitute the practice of structural engineering including the giving of structural engineering design advice. The use of this information and the materials linked or referred to in this article is at the user's own discretion and risk. The content in this article is not intended to be a substitute for professional structural engineering advice. The blog accepts no responsibility for the accuracy of materials linked to or referred to in this article and the materials are subject to change outside of the blog's control. Materials linked to or referred to in this article may not be the current at time of reading. This article is the copyright of the analogs | this~that blog. This article must not be altered, reproduced, or distributed without the consent of the blog. The author accepts no responsibility for any discrepancies arising as a result of the original information being altered by others. No analogs | this~that blog content may be used for training may be used for training, testing or developing machine learning or AI systems without prior written permission from analogs | this~that.