Early on May 16, 1968, a tenant on the 18th floor of Ronan Point, a new council tower in east London, struck a match to light her cooker. Gas had leaked into the flat. The explosion that followed was modest, and she survived it. It did push out a load-bearing concrete wall panel, though, and one corner of the 22-story block came down from the roof to the ground. Four people died. This guide covers how the building was put together and how a small blast became a full-height collapse. It also covers the inquiry and regulations that followed, and what the case means for a responder sizing up a building today.

May 16, 1968: a match, a cooker and flat 90

Ronan Point stood in Canning Town, in the London Borough of Newham, and it was one of a group of residential towers the council had commissioned during the postwar push to replace bomb-damaged and substandard housing. The contractor, Taylor Woodrow-Anglian, built it from precast concrete panels using a Danish method known as the Larsen-Nielsen system. Most accounts say residents began moving in during March 1968, so the building had been occupied for only about two months when it failed.

The tenant of flat 90 on the 18th floor was Ivy Hodge. Accounts place the explosion at around a quarter to six in the morning, which was a Thursday. She went into her kitchen and struck a match to light the gas cooker, and the gas that had leaked into the flat ignited. The leak was traced to a faulty connection on the supply to the cooker. The specific fitting and its defect are set out in the inquiry report, and anyone who needs that detail should read it there rather than rely on the condensed versions that circulate.

The pressure from the explosion pushed out the external flank wall of the flat, a storey-high concrete panel at the corner of the building. That panel was carrying the corner panels above it. Without it, the floors and walls of that corner from the 19th floor to the roof fell, and when they landed on the 18th floor the weight and impact carried the failure down through every floor below to the ground. Mrs. Hodge was injured and survived. The photographs taken that morning show a tower with one corner gone along its entire height and the rest of the building still standing beside it. Those photographs have been reproduced in structural engineering textbooks ever since.

Large panel construction: concrete walls carrying concrete floors

A large panel building has no frame in the sense most responders picture, with columns and beams holding up the floors and walls hung or infilled between them. The walls are the structure. In Ronan Point the panels were cast in a factory and lifted into place by crane. Storey-height wall panels sat on the edges of the floor slabs below, and each floor slab rested on the wall panels that supported it. The building went up as a stack of these alternating layers, with the joints between panels filled and connected on site.

The weakness was in those joints. In this type of construction the connections between a wall panel and the slab beneath it rely largely on gravity and friction, which works well while every load pushes straight down. A wall panel held in place mainly by the weight of the building above it has little to resist a force pushing it sideways, and it has nothing to hold it if the panel beneath it disappears. The general point the Ronan Point inquiry made about the system was that it gave very little tying between the elements and therefore very little capacity to find another route for the load if one panel was lost. I am describing the system in general here. The inquiry report contains the joint details for Ronan Point itself.

Large panel systems were attractive in Britain in the 1960s for reasons that made sense at the time. They promised speed, a factory-controlled product and less dependence on skilled site labor during a period when councils were under heavy pressure to build homes quickly. Industrialized building methods had government encouragement, and a number of proprietary systems were in use. None of that made the panels themselves weak, because a concrete wall panel in a large panel block is a strong component. The problem was that the building as a whole depended on every one of them staying where it had been placed.

The corner: how one panel brought down the full height

A gas explosion in a closed room produces pressure that pushes outward on every surface at once. The walls and floor of a flat in a framed building can take some of that pressure and vent the rest through windows. In flat 90 the flank wall was a load-bearing panel held in place mostly by the weight resting on it, so the explosion pressure pushed it out of the building. Some accounts suggest the pressure on the ceiling also briefly lifted the panels above, which would have reduced the weight holding the flank wall. That is a plausible mechanism and it appears in engineering discussions of the case, but readers should take the precise sequence from the inquiry report and not from summaries.

Once the flank wall at the 18th floor was gone, the corner wall panels of the 19th floor had nothing beneath them. They fell, and so did the floor slabs they had been supporting at that corner, and so on up to the 22nd floor and the roof. That debris then landed on the 18th floor slab at the corner. A floor slab is designed to carry the furniture and people of one flat, with a safety margin, and it cannot carry several floors of falling concrete arriving at speed. The 18th floor corner gave way, the load passed to the 17th, and the failure ran down the building to the ground.

The collapse stopped at the corner because the failure was confined to the bay of rooms whose support chain had been broken. The panels elsewhere in the tower still had their own supports below them and were not tied strongly enough to the falling corner to be pulled down with it. The same lack of continuity that let the corner fall also kept the rest of the building standing. That is why the photographs show a clean vertical gap down one corner of an otherwise intact tower.

Four dead: the count, the hour and the myth of a big blast

Four people died at Ronan Point, and most accounts give the number injured as seventeen. The death toll in particular is consistent across the published record, including the inquiry and the engineering literature, and I have not found a reputable source that gives a different figure. Some retellings imply that dozens or hundreds died, sometimes because Ronan Point gets folded into general accounts of failed British tower blocks, and those larger figures are wrong. Anyone writing about the case should use four dead and attribute the injury figure to the accounts that publish it.

The toll was low for a collapse running the full height of a 22-story building, and accounts commonly attribute that to the hour and to the layout. At a quarter to six in the morning most residents were still in bed, and the rooms in the collapsed corner are usually described as living rooms rather than bedrooms, so the people asleep were not in the part of each flat that fell. I treat that as the common explanation and not as something I have verified flat by flat. Different timing or a different room arrangement in that corner could have produced a much larger number.

The second error to correct is the idea that the building was destroyed by a catastrophic blast, which is a myth. The explosion was a domestic gas explosion in one kitchen, and the person standing closest to it survived. As the inquiry’s conclusions are generally reported, the explosion fell within the range a domestic gas leak could produce, and a building with adequate robustness should have been able to lose that one wall panel without losing the corner above and below it. That finding is the engineering significance of the case, because it located the fault in a design with no reserve capacity, and it explains why the lasting response came through the building regulations, which were changed to require tall buildings to tolerate local damage of that kind.

Progressive collapse: what the term means and why proportion matters

Progressive collapse is the spread of an initial local failure from element to element until a disproportionately large part of the structure has failed. The important word is disproportionate. Every building can be damaged, and a large enough explosion or impact will bring down part of any structure. Engineers worry about collapse that is out of scale with its cause, where the loss of one wall, column or slab leads to the loss of a whole bay or a whole building. British practice since Ronan Point has used the term disproportionate collapse for this, and American practice more often says progressive collapse. The two describe essentially the same concern.

Engineers generally describe three ways of controlling it. The first is to tie the structure together, horizontally and vertically, so that elements hold onto each other and a floor whose support is removed can hang from or span across its neighbors instead of falling free. The second is to design an alternative load path by checking that the structure can survive the notional removal of a supporting element, with the remaining members redistributing the load. The third, used where an element cannot practically be bridged over, is to design that element as a key element strong enough to resist a specified accidental pressure. Ronan Point had little of any of these, because its joints were not designed to tie, its corner had no alternative path, and its flank wall was not designed as a key element.

Ductility matters here as well. A steel frame or a well-reinforced and well-tied concrete frame can deform a great deal before it separates, which buys time and allows loads to find new routes. Stacked panels joined mainly by gravity behave much more brittlely, because once a panel moves off its bearing it is simply gone. The site’s article on the Murrah Building collapse in Oklahoma City in 1995 covers a different structure and a vastly larger and deliberate blast, but it is often discussed in the same engineering literature for the same reason, which is that the extent of the collapse depended on how the building carried its loads after the initial damage.

The Griffiths inquiry: weak joints and the gas decision

The government appointed an inquiry chaired by Hugh Griffiths QC, sitting with senior engineers as assessors, and its report was published later in 1968. The report should be read in the original by anyone who needs the findings at full precision. I am paraphrasing its general conclusions as they are consistently reported in the engineering literature and not quoting from it.

The central finding, as generally summarized, was that the explosion was within the range a domestic gas leak could produce and that the building’s failure came from its lack of resistance to the loss of a single load-bearing panel. The inquiry looked beyond gas to the other accidental loads a tall large panel building might meet, and it is commonly reported to have raised concerns about the system’s behavior under wind and fire as well. Its recommendations pushed toward designing tall buildings so that local damage would not spread, and toward appraising the existing large panel blocks that had already been built.

The immediate practical response concerned piped gas. Large panel blocks across Britain were appraised, and many had their gas supply removed and were converted to electric cooking and heating, while others were strengthened so that they could keep their gas. Removing the gas reduced the likelihood of the triggering event without changing the structure, which explains why both strengthening and gas removal were pursued. Ronan Point itself was repaired and strengthened and people lived in it again.

It is worth being careful about the larger story that sometimes gets attached to the inquiry. Ronan Point is often described as the event that ended high-rise council housing in Britain. It certainly damaged public confidence in system-built towers, and it is a landmark in that history. The decline of tower building had several causes, though, including cost, social and planning factors and changes in government policy, so the collapse should be treated as one major influence among several rather than the single cause.

Disproportionate collapse in the regulations, and the block taken apart

The Building Regulations for England and Wales were amended in the early 1970s to introduce a requirement that buildings resist disproportionate collapse, and as originally framed it applied to buildings of five or more stories. The supporting design guidance set out the approaches described above, which were tying, notional removal of elements and key element design. The key element provision used a specified accidental design pressure. Current UK guidance gives that pressure as 34 kilonewtons per square metre, and engineers commonly describe it as the metric form of the 5 pounds per square inch figure that came out of the post-Ronan Point work. Anyone applying it to a real building should work from the current edition of the guidance and not from this summary.

In England the requirement now sits in Schedule 1 of the Building Regulations 2010 as Requirement A3, on disproportionate collapse, with the detailed guidance in Approved Document A. Later editions of that guidance replaced the simple five-story threshold with a classification of buildings by use, height and occupancy, and each class carries its own level of required robustness. Scotland, Wales and Northern Ireland have their own regulatory documents, so a reader outside England should check the regime that applies. In the United States, the general structural integrity provisions of the ASCE standard on minimum design loads address the same concern, and some federal owners, including the General Services Administration and the Department of Defense, publish their own progressive collapse design criteria. Verify the current versions with the publisher or your building official before relying on any of them.

Ronan Point stayed in use after its repair until concerns resurfaced in the 1980s. Sam Webb, an architect who examined the Newham blocks, argued that the building remained unsafe, and his concerns covered the behavior of the joints and the effects of fire and wind. The block was emptied in the mid-1980s, and most accounts date its removal to 1986. It was dismantled in a controlled way rather than demolished with explosives, so that the joints could be examined as the panels came apart. Reports of that examination described joints that had not been properly filled or made up as designed, which suggested that workmanship as well as design had reduced the building’s margin. I report that as what was published about the dismantling, and the original survey records are the place to confirm it.

Construction type on the fireground: what Ronan Point tells a size-up

American fire and building codes classify construction into Types I through V according to the combustibility of the structural materials and their fire-resistance ratings. Those classifications describe how a building behaves in a fire, and they say nothing directly about how it behaves when a structural element is removed. A precast concrete panel building can be fully noncombustible and still fall into the category of buildings that depend on every element staying in place. A responder who reads Type I or Type II on a preplan and concludes the building is robust has confused fire resistance with structural redundancy. The fire service literature on building construction, with Francis Brannigan’s book Building Construction for the Fire Service the best known example, has been making that point for decades.

True large panel residential towers are uncommon in the United States, but several construction types that responders see routinely raise the same question of what holds an element in place. Tilt-up concrete warehouses have tall wall panels that depend on their connection to the roof structure for stability, and when the roof fails in a fire those walls can fall outward. Precast parking structures are assembled from beams, double tees and columns joined at connections that may have limited capacity for unexpected loads. Older unreinforced masonry, wood frame buildings with lightweight engineered members, and buildings altered over decades by removing walls each carry their own version of the problem. In each case the size-up question is whether the loss of one element would be contained or would spread.

For a gas explosion call specifically, Ronan Point shows that the explosion itself may be over in a fraction of a second while the structure is still deciding whether to stand. Displaced or bowed wall panels, cracks running along joints, slabs that have dropped at an edge, doors and windows racked out of square, and visible gaps between a wall and the floor above it all indicate that loads may have shifted. Those are reasons to set a collapse zone, limit the number of people inside, and request a structural engineer early. That engineer might be a building official, an engineer on contract to the jurisdiction, or a structures specialist from a FEMA Urban Search and Rescue task force or a state team, depending on what the local plan provides. Controlling the gas supply with the utility, so that a second ignition cannot occur in an already damaged structure, belongs in the same early decisions.

What to do at your agency

• The pre-incident planning officer pulls the preplans for every tilt-up, precast concrete and panelized building in the response area and adds a one-line note to each stating which elements the building depends on for stability, such as roof-to-wall connections or precast beam bearings, and finishes when every such preplan carries that line.
• The training officer adds a session to the existing drill calendar this month in which each company walks one precast or tilt-up building in its district and identifies the visible connections and the likely collapse zone, recording the building address and findings on the drill sheet.
• The communications center supervisor confirms with the fire prevention office that CAD premise notes for multistory residential buildings record the construction type and any known structural concerns, and enters the missing entries for buildings already identified in preplans.
• The emergency manager writes one paragraph into the existing emergency operations plan naming who provides structural engineering assessment after an explosion or partial collapse, whether that is the local building official, an engineer on contract or a state or federal urban search and rescue structures specialist, with an after-hours contact path.
• The fire marshal or operations chief obtains the gas utility’s current emergency contact procedure and confirms that the department’s standard operating procedure for gas odor and gas explosion calls names who requests utility shutoff and how the request is made.
• The fire marshal asks the building official for any list or records identifying local buildings of panelized or precast load-bearing construction, and adds the answer to the agenda of the next regular fire prevention and building department meeting.

Takeaways

• On May 16, 1968, a domestic gas explosion in flat 90 on the 18th floor of Ronan Point pushed out a load-bearing flank wall panel, and the corner of the 22-story tower collapsed from roof to ground.
• Four people died and most accounts give seventeen injured, and retellings that suggest a much larger toll are wrong.
• The explosion was modest and the tenant who struck the match survived, so the idea that a catastrophic blast destroyed the building is a myth.
• The Larsen-Nielsen large panel system relied largely on gravity and friction at its joints, which left the corner with no alternative load path once one panel was removed.
• The Griffiths inquiry led to the appraisal of existing large panel blocks, many of which were strengthened or had their piped gas removed.
• Disproportionate collapse became a requirement of the Building Regulations in the early 1970s and now sits in England as Requirement A3, with guidance on tying, alternative load paths and key elements in Approved Document A.
• Ronan Point was dismantled in 1986 by most accounts, and reports of that examination described joints that had not been made up as designed.
• A building’s fire-resistance construction type does not describe its resistance to progressive collapse, and a size-up after an explosion has to ask what each damaged element was holding up.