The real burden of risk

 

                       A piece of the Sanriku coast at Minamisanriku, NE Japan.
                       In 2011 there was a 20.5-metre tsunami here.

In 1966 the eminent Californian risk analyst Chauncey Starr published a seminal paper in Science Magazine in which he stated that "a thing is safe if its risks are judged to be acceptable." In effect, he built his reputation on the premise that the acceptability of risk is arbitrary.

Before I travel on university business I am required to fill in a complicated on-line form called a risk assessment. Recently, I had to do this before attending a series of meetings in Japan. By and large, Japan is a very safe country in which to travel and sojourn. By contrast, where I live in north London, within a radius of 400 metres of my front door, there have been at least two murders, a major and lethal terrorist incident, an international terrorist conspiracy, a series of road accidents, some of which were fatal, episodes of chronic pollution and overcrowding, and a constant battle between the police and a well-organised, wide-reaching drug supply ring. In the 1950s the area was immortalised in the photographs of Don McCullen, who was attracted by the presence of the London mafia.

When I travel from north London to my university I use public transport, which is of variable reliability. I then have to cross a busy four-lane arterial road. At one point the green light allows five seconds for pedestrians to scoot across, while at another designated crossing place they are afforded no protection at all against the streams of roaring traffic. At my university a problem with vibrating equipment caused bouts of deafness and nausea, and for various reasons it was two years before something was done about it. We then discovered that our dilapidated teaching rooms were lined with asbestos and the university was not aware of its presence. For none of this was I ever required to fill in a risk assessment form.

According to the set procedure for funded travel, I need to assess the risks of being in Japan. In cities a car cannot even cross a pavement without the presence of one or two uniformed characters waving flags to alert passers-by. For crossing roads, Japanese urban designers afford pedestrians the same status as traffic, with ample margins of time, wide crossings and perfect signage. I was once in a coffee bar in Japan when there was a magnitude 6.8 earthquake. Elsewhere it might have caused major devastation; in Sendai, people in the coffee bar did not even stop reading their newspapers.

A colleague who intended to do fieldwork in Turkey was required to produce a personal evacuation plan to be used if there were a major earthquake. Clearly, such a plan would be immediately invalidated by disruption to normal transportation schedules. Now if a repeat of the 1923 Kanto earthquake were to occur while I happened to be in Tokyo, I doubt very much whether the risk assessment would help me. I would have to resort to awareness and common sense.

Despite these musings, risk assessment is, of course, not useless. The problem is that procedural rigidity constrains us to use methods for harmless travel that are the same as those that apply to dangerous experiments and surgical operations. No doubt if I were travelling in eastern Ukraine or Yemen I would dedicate myself more willingly to considering the risks, but not for places where risk assessment effectively cannot help.

The possible solution to this state of affairs would be to divide risk assessment into two. For genuinely risky enterprises the approach would be technical and scientific. The odds of a mishap would be calculated and, where possible, reduced. For un-risky work, a different approach is needed. In this case, the principal value of risk assessment is to ensure, as far as possible, that the university is not sued. Perhaps, then, we should leave it to the lawyers to fill in the forms.

If, gentle reader, this diatribe should strike you as being mere petty complaining, please consider the wider implications, those beyond the shadow of lawsuits and injuries (however faint that shadow may be). Our principal motivation for being academics is to exercise our creativity. However, before collecting data we need data protection registration, ethical approval, deposition of itineraries, risk assessments and more. As we all know, similar bureaucratic loads apply to teaching. The effect of all this form-filling-in is to sap our creativity. Colleagues complain to me that they lack the energy to do real academic work after a day of grappling with poorly designed software intended to collect information that no one really wants, or dealing with procedures that merely add another layer of complexity to what was once a simple, refreshingly human activity.

We watch with alarm at the way bureaucracy grows unstoppably in our universities, how processes that are ostensibly designed "to make our work easier" are instead piling on the burden. And, of course, no one in the university has ever conducted a risk assessment of the impact of the bureaucracy!

 

The United Kingdom's National Risk Register - 2023 Edition

 

At the time of writing this, the UK Government has just released the 2023 edition of the National Risk Register (NRR, HM Government 2023). This document was first published in 2008 and has been updated (somewhat irregularly) at roughly two-year intervals. The new version presents 89 major hazards and threats that could potentially disrupt life in the United Kingdom and possibly cause casualties and damage.

Over the years this document has acquired momentum based on a solid commitment to persist with it and create periodic revisions. It is the public face of the National Security Risk Assessment (NRSA), a document (and a process) that has various security classifications and is generally not available to citizens and organisations. The current version of the NRR draws more on the NRSA than did previous versions. In this, the UK Government is honouring its promise to promote greater transparency in risk assessment.

The first edition of the NRR was a pioneering document that has been emulated by a variety of other countries. It makes sense to enunciate the major risks that a country faces so that all citizens can be clear about what needs to be tackled in terms of threats to safety and security in the future. The 2023 NRR is clear and concise. It explains its own rationale and presents the 89 'risks' one by one.

Although the NRR is certainly a valuable--and many would say necessary--document, it has some drawbacks.

(a) As noted by the House of Lords Select Committee on Risk Assessment and Risk Planning (House of Lords 2021), the NRR is not very "user-friendly" and is not well-known. One hopes that the latest version will reach a wider audience of citizens and organisations that did the previous editions.

(b) In terms of its methodology, the NRR discusses vulnerability but does not accept the premise (Hewitt 1983) that it is the major component of risk. Hence, the risk register largely discusses hazards and threats, not risks sensu stricto.

(c) The register uses a two-year assessment period for malicious risks and a five-year period for others, but many risks that threaten the UK will be evolving over a longer period. It therefore does not consider how risks are likely to evolve in the future. This is particularly important for those hazards associated with climate change. The register is thus not well connected to the foresight programme run by the UK's Government’s own Office for Science.

(d) The NRR does not consider risks as ensembles, despite the fact that they frequently materialise in groups. For instance, the NRR presents widespread infrastructure failure as a risk, but  if it were to occur, it would probably be the result of another hazard or threat such as a major storm or a successful cyber attack. This is a simple example: others are more complex, but the intricacies do need to be confronted.

(e) The risks are prioritised by giving most weight to those associated with hostile activity. In reality, it is at least equally likely that the major burden the UK will have to bear will involve natural hazards such as storms, heatwaves, wildfire or cold and snow. In the new version of the NRR natural hazards are given shorter descriptions and less prominence than that attributed to hostile risks.

(f) As a result of the previous two points, it is difficult to turn the risks, as they are described, into planning scenarios. This is a pity as it could be the NRR’s greatest source of utility.

The UK National Risk Register is allied to a number of other documents. One of these is the National Resilience Framework (HM Government 2022). This document has the merit of setting goals and targets for the achievement of resilience in Britain. However, it has serious weaknesses. For example, it makes no mention of gender, ethnic minorities and people with disabilities. That is most unfortunate because it is here that the efforts to create resilience need to be concentrated.

The 2023 National Risk Register has made some progress in responding to criticisms of the previous versions, but it could have made much more. As risk is largely a function of vulnerability, this fact needed to be acknowledged, rather than concentrating entirely on hazards and threats. There is no geographical dimension, which avoids the question of what size events are likely to be and whether certain parts of the country, and certain groups of citizens, would be most at risk.

The scenarios of risks described in the register are mostly described in 100-200 words. They are restricted to the "plausible worst-case" (which is usually a highly debatable concept). One great paradox here is that the worst effects may not necessarily come out of the worst impact. More reflection is needed.

The United Kingdom does not have a proper civil protection system. What it does have is fragmentary, confusing, overcomplicated and in places amateurish. This is a great pity as there is no shortage of expertise in the country. As I said in the witness box of the UK Covid Inquiry earlier this year, given the question "within the limits of what a government can, and should, achieve, does the UK Government keep citizens safe?", my answer is "no".

References

Hewitt, K. (ed.) 1983. Interpretations of Calamity from the Viewpoint of Human Ecology. Unwin-Hyman, London: 304 pp.

HM Government 2022. The UK Government Resilience Framework, December 2022. UK Government, London, 79 pp.

HM Government 2023. National Risk Register 2023 Edition. UK Government, London, 191 pp.

House of Lords 2021. Preparing for Extreme Risks: Building a Resilient Society. Report of Session 2021-22. HL Paper no. 110. Select Committee on Risk Assessment and Risk Planning, House of Lords, London, 127 pp.

Prolonged, wide-area electrical power failure

What are the likely consequences of prolonged, wide-area electrical power failure?

• lifts [elevators] blocked: people possibly trapped in them
• trains stranded: people possibly stranded in them
• traffic control inoperable: possibility of accidents and queues at road junctions
  
• critical facilities (hospitals, police stations, etc.) dependent on their own generators
• food refrigeration stops: perished food needs to be disposed of
• food manufacture lines cease operation; food perishes
• no water or sewerage pumping: health and safety regulations put buildings out of use
• electric cars and electrical equipment cannot be recharged
• people with disabilities and illnesses who are dependent on home electrical equipment are at risk or in difficulty
• fuel cannot be pumped: vehicles are gradually immobilized
• there is no street illumination: safety issues at night
• mass communication ceases or is severely limited
• internet communication and commerce cease
• loss of heating systems (for people and indoor animals)
• cows cannot be milked electronically and milk cannot be refrigerated
• cellular telephone networks cease operation as masts and repeaters run out of charge: mobile phone calls are not possible
• people in need are isolated by the cessation of public and personal transportation
• global navigation and positioning systems cannot be used at ground level
• problems may occur with air traffic and navigation as a result of any unreliability of electrically controlled navigation systems (satellites and avionics)
  
• if conditions are cold, cases of hypothermia rise as a result of loss of heating systems; if the weather is very hot hyperthermia is the risk
  
• alarm systems are inoperative: cases of burglary and theft rise
• people are stranded away from home as a result of the cessation of public transport and difficulties with private transportation (lack of availability of fuel or vehicle recharging)
  
• the use of candles leads to an increase in home fires
• electronic transactions do not take place: home purchasing conveyancing fails
• stock market transactions are reduced or cease
  
• toll booths do not work: loss of revenue or access
• critical information is almost impossible to share (on closures, warnings, evacuation, critical needs and emergency situations)
• it is difficult to construct and share the common operating picture of the emergency, leading to reduced situational awareness by emergency services and crisis managers
• emergency response coordination is plagued by uncertainty, and possibly equipment failure
• it is difficult to buy food, drink and medicines as tills that operate electronically are locked shut and credit and debit cards do not work
  
• reduced hygiene
• social isolation occurs, with particular effect on vulnerable people, including those with cognitive disabilities
• workplaces are constrained to shut down, with losses of revenue and income
• potential increase in criminal activity (not to be overestimated)
• potential increase in stress-related violence (not to be overestimated)
• cessation of many public administration functions
• supply chains are interrupted or slowed down
• cold chains are out of action
• rubbish collection and disposal are affected by lack of fuel
• educational activities are suspended, with an effect on learning, examinations and childcare
• risk of unrest in prisons
• postponement or cancellation of major activities (sports, arts, religion)
• vehicles (and their occupants) are stranded when they run out of fuel
• failure of sensors, including CCTV
• medical and dental appointments are cancelled
• there may be an increase in the birth rate nine months later
• the functionality of electricity generators in critical facilities cannot be guaranteed and may deteriorate over time
  

No doubt there are many other potential consequences. The list needs additions and some form of classification (and cross-referencing) of the impacts.

Reflections on the Turkish-Syrian Earthquakes of 6th February 2023: Building Collapse and its Consequences

 

                                                        Source: Wikimedia Commons

An interesting map was published by the US Geological Survey shortly after the Turkish-Syrian earthquakes.[1] It showed (perhaps somewhat predictively) that there was only one tiny square of the vast affected area in which Modified Mercalli intensity (which is largely a measure of damage) reached 9.0, the 'violent' level.[2] This is--just about--enough to damage very significantly a well-engineered structure (but not necessarily enough to bring it crashing down). Although the disaster of 6^th February 2023 produced, in fact, stronger shaking than this (maximum 1.61g), it should not have caused 5,500 large buildings to collapse as large parts of the epicentral area had accelerations <0.6g, a design level for antiseismic construction in areas of known high seismicity. The disaster in Turkey and Syria is very obviously the result of poor construction. This is painfully visible in the video images of buildings collapsing. The patterns of collapse are also the same as those in the last dozen Turkish earthquakes, although they are doubtless more extensive this time around.

In 43 years of studying disasters I have seen few events that so clearly illustrate the primacy of vulnerability over hazard impact as does the Turkish-Syrian earthquake sequence of February 2023. Work at universities in Florida and Colorado strongly suggests that corruption is the principal cause of earthquake disaster, world-wide. The Turkish anti-seismic building codes have been revised five times in the last 55 years, including a thorough and intelligent upgrade in 2018. However, in 2016 and at nearly 20 other times there were amnesties that decriminalised those in the construction industry who ignored the laws, and those who modified buildings in ways that stopped them from being compliant with the regulations. Such practices were extremely widespread, the norm rather than the exception. This is also my experience from having spent extended periods in such buildings in Anatolia. 

Building codes in Turkey are now perfectly good enough. The tragedy lies in their non-observance and the paucity of retrofitting. It is a mixture of simple errors, lax procedures, ignorance, deliberate evasion, indifference to public safety, untenable architectural fashions, corruption and failure to enforce the codes. Many, perhaps most, people in Turkey live in multi-storey, multiple occupancy reinforced concrete frame buildings. It is these that collapse. Most of them are highly vulnerable to seismic forces. There is plenty of engineering literature on the typical seismic performance defects of such buildings in Turkey. Perhaps we can grant a small exception for Syria, although before the civil war it did have building codes and earthquake research. However, the comment by a leader of the Syrian Catholic Church that buildings had been weakened by bombardment was something of a red herring. This probably affected about 2-3% of those that collapsed.

To know whether a reinforced concrete building is safe to live in would require knowledge of:-

• the shear resistance (i.e., quality) of the concrete
• the presence or absence and connectivity of shear walls
• whether there are overhangs or other irregularities of plan that distribute the weight of the building unevenly or concentrate load on particular parts of it
• the presence or absence of a ‘soft-storey’ open ground floor which concentrates the load above columns that cannot support it during seismic deformation
• the connections between beams and columns, especially how the steel reinforcing bars are bent in
• whether there are proper hooks at the end of rebars on concrete joints
• whether the rebars were ribbed or smooth
• the quality of the foundations and the liquefaction, landslide or subsidence potential of the underlying ground
• the state of maintenance of the structural elements of the building
• any subsequent modifications to the original construction (e.g. superelevations).

An experienced civil engineer could evaluate some of that by eye, but much of the rest is hidden and only exposed once the building collapses. A short bibliography of sources that deal with common faults in Turkish R/C construction is appended at the end of this article.

Many of the news media that have reported the disaster have presented it as the result of inescapable terrestrial forces. While that cannot be negated, it is less than half of the story. The tragedy was largely the result of highly preventable construction errors. Vox clamantis in deserto: to examine this aspect of the disaster one would have to face up to difficult issues, such as corruption, political decision making, people's expectations of public safety, and fatalism versus activism. How much simpler to attribute it all to anonymous forces within the ground!

A well-engineered tall building that collapses will leave up to 15% void spaces in which there may be living trapped victims. It was notable that, in many buildings that pancaked in Turkey and Syria, the collapses left almost no voids at all, thanks to the complete fragmentation of the entire structure--i.e., total loss of structural integrity. This poses some serious challenges to search and rescue. In some cases the collapse was compounded by foundation failure, leading to sliding or rotation of the debris.

There was also an interesting dichotomy in the images on television between the "anthill" type of urban search and rescue, carried out by people with no training, no equipment and no idea what to do, and professional urban search and rescue (USAR), which sadly was in the minority of cases. Nevertheless, it remains true that the influx of foreign USAR teams is, sadly, both riotously expensive and highly inefficient, as they tend to arrive after the 'golden period' of about 12 hours in which people could be rescued in significant numbers.

Among the damage there is at least one classic example of the fall of a mosque and its minaret, the same as that which happened in the Düzce earthquake of 1999. Mosques are inherently susceptible to collapse in earthquakes: shallow arches, barrel vaults, rigid domes and slender minarets. The irony is that the great Turkish architect of the 16th century, Mimar Sinan (after whom a university in Istanbul is named) had the problem solved. He threaded iron bars through the well-cut stones of his minarets, endowing them with strength and flexibility. It is also singular that one of the first short, stubby minarets in Turkey (located in Izmir) was built 300 years after Sinan died in 1588.

The earthquakes chart a map of illegal and ineffective construction methods. Relatively few Turkish mass media openly discuss this (exceptions are KSL-NewsRadio and Bianet), and those that do are at risk of being treated as criminals. Nevertheless, the only way for reconstruction to succeed is for there to be a radical change in Turkish policy towards building practices. The issuance of a hundred prosecution notices to builders and engineers is a somewhat hypocritical response, given the amnesty they enjoyed. It shows that political responses to disasters depend on the electorate's short memory.

The President of Turkey has publicly vowed to "reconstruct thousands of houses within one year". This is not a good idea. It should take two or more years to conduct geotechnical survey (microzonation) and urban planning. More time is required for necessary public consultation on the plans. Failure to recognise that time is socially necessary in reconstruction risks marginalising the problems involved rather than facing up to them.

Finally, there is a seismic hazard map of the area affected by these earthquakes. It was made in 1967 and events have shown it to be substantively accurate. No one can say that the risk was not well known, or that the events were unexpected.

Select Bibliography of Sources on Turkish R/C Construction Practices

Cogurcu, M.T. 2015.Construction and design defects in the residential buildings and observed earthquake damage types in Turkey. Natural Hazards and Earth System Sciences 15: 931-945.

Dogan, G., A.S. Ecemis, S.Z. Korkmaz, M.H. Arslan and H.H. Korkmaz 2021. Buildings damages after Elazığ, Turkey earthquake on January 24, 2020. Natural Hazards 109: 161-200.

Dönmez, C. 2015. Seismic performance of wide-beam infill-joist block RC frames in Turkey. Journal of Performance of Constructed Facilities 29(1): 1-9.

Erdil, B. 2017. Why RC buildings failed in the 2011 Van, Turkey, earthquakes: construction versus design practices. Journal of Performance of Constructed Facilities 31(3):

Korkmaz, K.A. 2009. Earthquake disaster risk assessment and evaluation for Turkey. Environmental Geology 57: 307-320.

Ozmen, H.B. 2021. A view on how to mitigate earthquake damages in Turkey from a civil engineering perspective. Research on Engineering Structures and Materials 7(1): 1-11.

Sezen, H., A.S. Whittaker, K.J. Elwood and K.M. Mosalam 2003. Performance of reinforced concrete buildings during the August 17, 1999 Kocaeli, Turkey earthquake, and seismic design and construction practise in Turkey. Engineering Structures 25(1): 103-114.

Corruption and Earthquake Disasters

Ambraseys, N. and R. Bilham 2011. Corruption kills. Nature 469: 153-155.

Escaleras, M., N. Anbarci and C.A. Register 2007. Public sector corruption and major earthquakes: a potentially deadly interaction. Public Choice 132: 209-230.
 

---

[1] https://earthquake.usgs.gov/earthquakes/eventpage/us6000jllz/shakemap/intensity

[2] https://www.usgs.gov/programs/earthquake-hazards/modified-mercalli-intensity-scale