Article - Issue 17, October/November 2003

Life safety in extreme events

Faith Wainwright FREng

Download the article (218 KB)

Engineers are involved in an infinitely wide range of challenges in designing for life safety in extreme events, whether it be natural hazards such as wind and earthquakes, or making our roads safer. This article is based on a presentation at the BA Festival of Science in September 2002, which addressed the issues raised by the events of 11 September 2001. Faith Wainwright writes on extreme engineering disasters from the Titanic to the Tacoma Narrows and how we can prepare for the unexpected.

Mention ‘September 11th’ and we all think back to 2001 – over two years ago, when New York City was plunged into the darkest days it has ever known as it suffered the unspeakable attack on the World Trade Centers. Think back and recall the day – what were you doing when you heard the news? And if you saw the buildings engulfed in smoke and fire, did you imagine for one moment that you would shortly witness the complete collapse of these very fine and beautifully engineered buildings?

For me, the news came as I was driving through Boston, having just landed at Logan airport on a flight from John F Kennedy airport. I will never forget the days that followed: the ghastly feeling of being separated from family, not knowing when we would be together again. The impact on those more closely involved is beyond what most of us can imagine.

The events in America that day (both in New York and at the Pentagon) have given engineers, building owners and all those involved in designing, owning, running and occupying buildings a challenge that sometimes seems insurmountable. Those who are construction professionals view safety as a top priority. Buildings should be safe places: at the most basic level we build in order to give people a safe environment – to protect them from winds, floods, parching sun. Before we cross the road we expect to look, listen and be very aware of the dangers that can be caused by vehicles on the road; we are not supposed to have to think about any kind of danger before deciding to work or live in a particular building.

In the days after the attacks, some said, ‘this will change the world’. When I returned to the UK not everyone agreed with that: ‘this will all settle down in time and people will carry on designing buildings just as they used to’, some said. Well, clearly things have not settled down. We may have recovered from the shock and horror, but fear still lingers about how our buildings would protect us from a terrorist attack perpetrated by people who have no fear for their own lives.

Learning from disasters

Of course this is not the first disaster to have far-reaching implications for engineers. Other famous disasters may not have any connection with terrorism, but there are similarities in how engineers have learned from them.

Arguably the most famous disaster involving huge loss of life was the sinking of the Titanic. A ship built to survive the worst sea crossing imaginable, it embodied the pride of its owners and visions of a very profitable business. The Titanic sank on 15 April, 1912, after striking ice in the North Atlantic Ocean. The official British inquiry into the Titanic disaster, Lord Mersey’s report, made 24 recommendations ‘with a view to promoting the safety of vessels and persons at sea’. Returning to this, it is salutary to note the following recommendations that were made:

  • Better watertight compartmenting schemes, both to reduce the likelihood of a ship's sinking, and to keep the ship on an even keel if watertight compartments filled.

  • Lifeboats for everybody. (The Titanic's boats had capacity for only 1178 of the 2201 persons aboard. Although 18 of the 20 boats were launched, many were not filled to capacity, owing to inadequate organisation and training of the boat crews, unwillingness of passengers to leave the ship, and other causes.)

  • Proper staffing and training of boat crews, and frequent drills.

  • Wireless installations on all passenger ships, with 24-hour monitoring of emergency radio traffic.

  • Prudent navigation in the vicinity of ice.

These are about the anticipation of disasters and effective implementation of measures to manage them and, as with the Titanic, the most significant lessons that have been learned from the World Trade Center disaster have been in the areas of evacuation and communication.

Another disaster to have far-reaching implications was the Tacoma Narrows Bridge collapse. The bridge opened in 1940. One day, with the wind blowing at 42 m/h the deck began heaving in vertical ‘waves’ up to 9 m deep. After three hours the motion changed to a torsional motion, with the deck twisting through 90°. A series of hangers broke and a 300 m long section of the deck fell into the water.

Later, laboratory tests showed that the deck of a suspension bridge can act like an aeroplane wing, with wind causing it to lift and fall. The research into this phenomenon and understanding achieved has changed forever our appreciation of wind forces and the design of slender structures which are vulnerable to such large motions being set up in the cross-wind direction. In this case, the wind had generated an extreme event – one that could have been prevented, if only the phenomenon had been understood.

Ah yes, but now we understand what the wind can do, what icebergs can do – but what extreme events ought a bridge, a ship, a building be able to withstand if we are talking about terrorist events – unpredictable in time and nature?

The engineers’ global role

Let us consider for a moment the phenomenal growth in our built environment – the extent of man-made structures that our normal lives depend upon. Long-span bridges, tunnels, highways and high-speed railways are providing new and stronger connections between communities than ever before. The Øresund Bridge connects Denmark to Sweden (Figure 1); the Channel Tunnel connects Britain with France; Italy is building a bridge to Sicily; and so it goes on. Not only is electronic communication bringing us close together, but so is our infrastructure.

Along with this consider the impact of our increasingly developed and densely populated world (Figure 2) on our environment and the imbalance between the quality of life for those in rich and poor countries. Issues of water, energy, health, agriculture and biodiversity are crucial – we are told that if everyone is to enjoy the same lifestyle that we enjoy in the West we would need four planet Earths to meet our requirements.

We live in a global age, where these issues of how we live, communicate and enable lives to flourish is the context in which we engineers practise. So when we think about our buildings and how to handle safety in the face of terrorist activity, how do we go about it?

Modelling impact events

Before considering buildings, let’s look at a few areas where engineers get to grips with complex phenomena and design to make life safer. The first example is a collision between a lorry and pier on one of our motorways (Figure 3). What happens if the lorry hits a bridge pier? What force should the pier be designed to withstand and what force should the lorry be designed to absorb safely? The most onerous event for the bridge is the lorry impacting the pier head-on – but how likely is that? The Highways Agency has to determine which design scenario to adopt to protect bridges at an affordable cost. The lorry designers have to consider which impact scenarios should be used for the design of the lorry itself – how to build in maximum protection against the most likely damage event.

A similar impact study involves a car travelling at 70 m/h on a motorway and driving into the barrier (Figure 4). What happens here is a function of the design of the barrier and the car. There are some 25 million miles of verge on major roads in the country. Where there are barriers, they need to be economic and maintainable: the cost of the barriers is balanced by the number of lives potentially saved through good design.

The interaction between the car and the steel barrier is complex – there are many types of vehicle just to begin with. One aim in barrier design is that, on impact, the vehicle should not overturn. It is desirable to have deformation in the barrier, but not to the extent that it encroaches on the lane of traffic going in the opposite direction.

Another aim of such design is that this extreme crash event should have a predictable outcome, obviously with lives saved. And when the extreme event strikes, protective measures come into place (the inflatable airbag is an effective emergency response in many cases: Figure 5).

Wherever there are people there are issues of life safety. In earthquake zones huge strides forward have been made in the past 30 years to improve the performance of buildings in an earthquake.

The Cypress Street viaduct in the San Francisco Bay area collapsed in 1989 in the Loma Prieta earthquake. The viaduct had been partially retrofitted with design upgrades as a result of improved knowledge about designing for earthquakes, but a planned second phase of the upgrade had not been implemented when this earthquake struck. The earthquake caused US$12 billion of damage and claimed 64 lives, half of which were on this 1.4 km section of the viaduct. Before the viaduct was rebuilt, it was essential to understand exactly how it had failed – how the earthquake had shaken it, and why it had proved to be inadequate.

Figure 6 shows a model of the viaduct responding to the earthquake and Figure 7 shows the damaged structure. Analysis of the old design demonstrated exactly the failure that had been witnessed in the earthquake, and therefore it was possible to be confident that the model of the new design demonstrated a simulation that would prove to be accurate.

One of the lessons from the earthquake was an understanding of how soft soils can magnify the force of an earthquake. The example shows the importance of disaster investigations in order to learn and improve future designs.

In the process of modelling one has to make assumptions about natural phenomena that cannot be totally defined. Figure 8 shows the potentially extreme event of an iceberg hitting an oil platform. In such models various scenarios have to be tested – a small iceberg travelling fast, a large iceberg travelling slowly – and the worst credible event needs to be decided on. Equally, a decision has to be taken on the performance: under what circumstances is this platform expected to survive without severe damage to the facility?

The role of modelling

Why use models to illustrate design concepts? It is because the key weapon in our armoury in battling against extreme events is our ability to experiment with ‘what ifs’. A future accidental or malicious event cannot be defined precisely, so many design scenarios have to be considered in order to determine a good course of action.

However, as engineers we are not simply there to test ideas. Our role is to create, to invent, to produce new machines, systems, buildings. They will need to be functional, but also beautiful (see Figure 9). A sense of proportion is as important as an appreciation of forces. Of course anything you touch might be made safer by being more protected, more chunky, less accessible. But an engineer considers everything in design, including delight. Delight is a vital factor in design and often excellent engineering and delightful products are synonymous.

Therefore the question is: how can we use our modelling ability when it comes to safety in tall buildings?

We already take considerable precautions against structural collapse. In the UK, Ronan Point, a block of flats in East London, was a turning point for us. A gas explosion on the 18th floor of this building took place at about 5 am. It lifted the floors, pushed out the walls and, as one floor fell onto the other, the tower progressively collapsed, resulting in four deaths.

This led to a fundamental change in the UK Building Regulations which now include requirements for tying structures together and making provision for the loss of a beam or a column. The requirements are prescriptive and intended to be simple to apply consistently.

These days we can use our explicit modelling techniques to look at exactly what happens when a building is damaged. Figure 10 shows part of a floor which has been overloaded by debris falling from above. The floor beams must be capable of withstanding significant distortion as they carry this extra load. This idea of distortion without breaking when overstressed is the concept of ductility. It is not new, but ductility has not to date been a consideration generally for building design in the UK.

Modelling is also vital for helping our understanding of what a biological or chemical attack might do.

Figure 11 shows a 14-storey building modelled using 280 000 volumes or cells representing air flows with 1 kg of sarin gas introduced to determine how the agent disperses through the building over time. Sarin is a dangerous nerve agent that is mainly taken up through the lungs and at a concentration of 100 mg.min/m3 is likely to be fatal in 50% of cases. The figure is taken from a study of this event and shows that such a dose could be lethal over a period of minutes before it is diluted.

Studies like this can help to determine measures that might be adopted to prevent such contamination of air intakes, and also to give a better awareness of what is likely to be happening in the building which can help the emergency response.

Many owners and designers have used modelling of evacuation to understand how quickly a building could be emptied if it was necessary to get everyone out at once. We can test evacuation plans theoretically through simulation – Figure 12 shows people evacuating a 50-storey tower and the model can be used to anticipate how long people would have to wait for lifts, how much longer it takes to evacuate if one or two of the staircases are not passable and so on.

Such tools are not a substitute for real practice drills, but they can be used to test the what-ifs and as part of the training for staff who are involved in managing an emergency response.

Conclusion – the tortoise or the cat?

We have to make judgements and choices about our built environment. This article makes reference to models that allow us to understand the actual performance of people and buildings in events that we choose to consider. We can throw everything into protecting ourselves from physical terror, by hiding behind fortresses or going underground into bunkers. Or we can do more to understand what is actually happening in a building – what exactly is going on – and be prepared to take different courses of action depending on the knowledge we have. In the former case we could consider ourselves like tortoises: cover everything with a hard shell and cope with the restrictions in movement that brings. The tortoise has little freedom but a great deal of protection. Or we can choose to be more like cats: with strong senses and the ability to react so quickly that we can escape danger before it overtakes us. This requires us to be nimble and alert, and to have thought through what responses are available to us when disaster strikes. We, as engineers, are intimately involved with the decisions that are made to make our world safer and also a delight to live in.

Faith Wainwright FREng

Director, ARUP

Faith Wainwright has led the structural design for a number of distinguished projects with leading architects. She has been a Director of Arup since October 1998 and currently is a leader in the Advanced Technology Group. After 11 September 2001, Faith became part of the Arup Extreme Events Mitigation Task Force, moving forward the thinking as to how this event impacts on the design of buildings in the future.

[Top of the page]