Examples of major disruptions include extreme weather, application failure, and criminal activity.

There has been extensive research into people's reactions in crisis situations going back to Prince's [7] study of a ship collision in 1917. Drabek and Evans [8] summarised the major contributions of American sociologists to an understanding of human responses to disaster. The first issue relates to labelling panic, looting and other anti-social behavior as myths [[9], [10], [11], [12], [13], [14], [15]]. Quarantelli [16] suggested that people as a whole do not panic and rather than reacting in an anti-social manner, people typically become more cohesive and unified during situations of collective stress. Tierney [17] writes that the incidence of panic behavior is vanishingly rare in crisis situations and despite fear and the entirely reasonable desire to flee, people in severe danger try to make rational decisions and work collectively. In an analysis of the World Trade Centre disaster 2001 extreme behavioral action, including panic, was noted in less than 1% of cases [18]. [19] suggest that anti-social behaviors during disaster scenarios, is, at best, over-exaggerated and that human behavior in disaster scenarios can only frequently be described as “pro-social”. In contrast to panic behavior, the typical human response to disaster warnings is disbelief, denial, and reinterpretation to reduce or eliminate the threat potential [20].

On the other hand, there are examples of looting and crime in a small minority of disasters such as New York blackout 1977 [[21], [22]], Hurricane Katrina [23]. Despite widespread fears of looting during the Kista blackout in a suburb of Stockholm, Sweden there was slightly more criminal activity than usual [24].

The second issue of relevance is that of emergent behavior – emergent groups of people working together in pursuit of collective goals relevant to the disaster but whose organization has yet to become institutionalized. Emergent phenomena have been a prevalent feature in disasters [25] and may on occasion lead to conflict with formal emergency management [26]. This is one aspect of the extensive literature on social resilience and adaptive behavior. Hutter et al. [27]; introducing a special edition on resilience, conclude that there are a multiplicity of perspectives on resilience and the lack of a clear definition poses a challenge to operationalising its use in disaster management. This is perhaps not surprising since resilience is a multi-dimensional concept encompassing social, economic, physical, technological and natural dimensions [28]. In social science, resilience is considered as a complex web of social connections and capacities that enable a community to cope with hazards [29,30]. Obrist [31] says resilience is a dynamic process, not as a static state, and social resilience enables people not only to cope with adverse conditions but also to create options and responds proactively to adversity. Keck and Sakdapolrak [32] also suggest that resilience goes beyond coping and encompasses adaptive and transformation processes.

Rogers and Nehnevajsa [33] report on behavior and attitudes under crisis conditions and suggest that people respond to crises in remarkably similar ways. The disruption created by disasters creates stress, yet individuals and organizations seek to maintain the continuity of behavior. The first response to impending danger reflects a strong commitment to protecting themselves, their loved ones and even others with whom they may become associated. The second type of response is information seeking. As people become alerted to the potential for danger they tend to seek information about the hazard, from their own senses and to get confirmation from others, particularly family and friends. Finally, there is a propensity to respond in groups, particularly families, by consolidating resources. These three basic behavior patterns were observed in the scenario workshops reported in this paper.

Electricity is an enabling technology for a host of quotidian activities [34], so its disruption affects normal daily life and has a cascading effect on other infrastructure, for example water supply [35]. Electric power systems are part of the critical infrastructure of modern societies, therefore is essential to boost their resilience [36]. Electricity infrastructure is necessary to sustain human and economic wellbeing [37]. Many countries have, therefore, introduced mandatory requirements for operators to maintain high reliability and stability of supply, for example, The U.S. Department of Energy, Federal Energy Regulatory Commission (FERC) [38] is working with power system operators to ensure reliability of the U.S. power system. In Germany, the transition from fossil fuels to renewables is pushing the industry towards greater flexibility and resilience through various innovations in market design and pricing mechanisms [39]. In France, RTE, the electricity transmission operator, introduced emergency response arrangements in the wake of the storms in December 1999. It is tackling current reliability challenges by leveraging power generation flexibility, stronger cooperation between transmission system operators and coordination centers and adherence to European grid codes [6]. However, power systems are highly nonlinear and it is challenging and uneconomical to make power systems totally resilient [40].

Drawing on examples from around the world, Wood et al. [41] argue that a crucial factor in people ability to cope with adversity is social capital: the extent to which social networks are able to support collective action, and promote psychological wellbeing. One of the earliest detailed accounts of people's reactions comes from the blackout across the eastern seaboard of the United States on 9 November 1965, affecting 25 million people. Six days after the breakdown, the National Opinion Research Centre (NORC) interviewed over a thousand people about how their reactions to the blackout. They found that most people were not in the least alarmed, there was no panic and the majority of people reported that strangers were unusually helpful and friendly to each other. Above all, they found that the blackout affected rhythms of communication and social interaction. For example, the New York Telephone Company reported most calls were to relatives and instead of turning for information to City Hall or formal organizations, people turned to neighbors and to local leaders in the neighborhood [42].

On August 14, 2003 a massive power outage caused blackouts throughout most of Ontario, along with parts of the US. This power failure was the largest in North American history, spanning 24,086 square kilometres and costing the Canadian economy $1–2 billion [43]. Brenda Murphy [44] conducted a survey of 1200 Ontario residents six months after the outage to discover how they had managed to deal with the emergency. She found that families relied on networks of family, friends, neighbors and community associations to help them get through a crisis. 89% of respondents stated that their neighbors would provide assistance and 37% said they provided assistance during the blackout.

Although rare, Europe has experienced major power failures and maintaining stability of power supply is especially challenging during natural catastrophes [45]. In France and Germany storms and floods are the natural hazards cause the biggest problems. On September 23, 2003, nearly four million customers lost power in eastern Denmark and southern Sweden following a cascading outage that struck Scandinavia [46]. Days later, a cascading outage between Italy and the rest of central Europe left most of Italy in darkness on September 28 [47]. In France and Germany, there have been 10 serious examples of storms causing temporary power outages in the past 10 years (Table 1).

Electricity outages can be both an effect and a cause of cascading effects [58]. For example, the 2013 Germany floods caused widespread disruptions in power supply, which in turn affected water treatment plants and public transport amongst other services and activities [59]. In February 2014 several storms again affected UK and 80,000 homes were left without power for several days. The impacts of power outages on households, the description of the challenges faced and the strategies adopted by them are described by Ghanem et al. [60]. The capacity of households to deal with power outages depends not only on the household itself, but also on the social resilience of the whole community. Cascading effects are common in disaster events and a chain of interaction can amplify the impact on society where the initial power failure can lead to fires, an increase in accidents and fatalities, stresses to the health sector, water shortages and hygiene issues, disruption of refrigeration and food shortages, environmental contamination, carbon monoxide poisoning from disruption of underground ventilation, traffic disruption, disruption to temperature control, people becoming trapped, increased crime, stress on police and emergency services, business continuity and logistics failure and last, but not least communications system and telecoms failure [61].

The 2011 Tohoku earthquake, in particular to the Fukushima Daiichi nuclear disaster, was a wake-up call to the risk of “cascading disasters” [62]. A paper by Pescaroli and Alexander [63] addresses the vulnerability of critical infrastructure to cascading effects of disasters. They suggest a shift attention from risk scenarios based on hazard to vulnerability scenarios based on potential cascading effects in which they identify nodes that are capable of generating secondary events. Early warning, the behavior of operators, communication failures and bureaucratic conflicts are, they suggest, the areas in which further research could improve understanding of how cascading disasters propagate in critical infrastructure, and how their progress can be arrested. This approach to systemic risk is being actively pursued by the Centre for Risk Studies in Cambridge University [64].

Four scenario role-playing exercise workshops were held with university students in Germany and France. These involved getting the students to dynamically ‘play’ through an imagined future. The game collapsed real time to highlight significant issues and to focus on people's decision-making, emotional responses and imagined actions [65]. The scenario exercises were designed with input from 5 international experts with experience of using scenario planning in risk management, social science, economics and urban management. The design also drew on an extensive literature review, summarised as follows.

Various authors have suggested the use of scenario planning as a way of exploring the critical role of decision-making and spontaneous behavior in the stressful time of an emergency like a blackout [[66], [67], [68]]. Moats et al. [69] describe how, after major disasters, people have to make high consequence decisions with incomplete or inaccurate information under time pressure and suggest scenario planning as a way to prepare. Alexander [70] described how scenarios are used in emergency management training and how outcomes are used as the basis of crisis planning. Bradfield et al. [71] say public policy makers are increasingly using scenarios to involve agencies and stakeholders in policy decisions.

In the disaster management field, the scenario approach has been applied in ‘ShakeOut’ drills to raise public awareness [72,73] and various authors describe how they are used in emergency management [67,71,[74], [75], [76], [77], [78]]. In Germany, scenario planning was used to explore the impact of a blackout on health care [79]. An article by Envision Tomorrow [80] suggested scenarios could be of two basic types – ‘normative’, which describe a preferred outcome, and ‘exploratory’, which describe unknown futures. The scenario could be timed for pre, during or post-disaster. The blackout scenario in this research was an exploratory scenario timed for during the event.

In summary, in terms of a conceptual framework to guide the research, this brief literature review points to the following issues. How do people imagine events might cascade after a major blackout and produce a chain of consequences? How do people's feelings change over the 72-h scenario? Are their signs of emergent groupings, coping strategies and adaptive behavior? How might communication failure impact on individual and communities?