By Nicolas Bock (DRU), Cordula Dittmer (AKFS) & Peter Windsheimer-Kolla (DRU)

10. April, 2026

http://dx.doi.org/10.17169/refubium-52004

Background

On Saturday, the third of January 2026, a massive power outage occurred in the south-west of Berlin. The localities Nikolassee, Wannsee, Zehlendorf and Lichterfelde in the district of Steglitz-Zehlendorf were largely without power for over four and a half days in winter weather conditions. The grid operator Stromnetz Berlin reported that more than 45,000 households and over 2,000 commercial customers were affected. The cause was a fire on a cable bridge, which affected the functionality of a distribution substation in Lichterfelde. Investigators quickly concluded that the damage was caused by arson. A left-wing extremist group claimed responsibility for the attack in a letter posted online. Acts of sabotage on Berlin’s transport, energy and telecommunications infrastructure have become increasingly frequent over the last five years. One such incident occurred in September 2025, when two electricity pylons in south-east Berlin were deliberately damaged, leaving around 50,000 people without power for approximately two and a half days (rbb24).

The most recent outage not only led to a loss of power supply in private households and businesses, but also had cascading effects on heat supply, telecommunications (failure of the mobile network throughout the affected area – a total of 86 mobile phone masts from various providers were affected), transport infrastructure such as traffic lights and suburban rail services, emergency calls, lifts, and digital payment systems.

Electric roller shutters could no longer be opened, leaving some houses and flats without light even during the day, doorbells no longer worked, supermarkets remained closed and had to either transfer their goods to refrigerated trucks (provided by the companies themselves) or dispose of them. Although the water supply was not fundamentally at risk, it could not be guaranteed across the board in high-rise buildings due to their dependence on in-house water pumps.

The affected area also includes five hospitals, 74 nursing and retirement homes, three refugee shelters and people in need of outpatient care. By Saturday, it had become clear that due to the massive damage to the pipeline, repairs would take 5-6 days. The district of Steglitz-Zehlendorf set up a crisis management team and opened Zehlendorf Town Hall as a central emergency call centre, warming centre and care point. Aid organisations such as the German Red Cross (DRK) and the Arbeiter-Samariter-Bund (ASB) set up emergency shelters in sports halls, while shops and civil society organisations, such as church communities, set up care points and warming stations where those affected could charge their mobile phones and receive hot water or food. The police increased their presence in the affected neighbourhoods and alongside the Federal Agency for Technical Relief (THW), erected light poles at traffic hubs. These poles lit up dark intersections and simultaneously served as charging and information points, especially in the first 48 hours.

Gradually, various support structures were established. Information about them was primarily disseminated via the internet, social media and police loudspeaker announcements. On Sunday afternoon, Berlin’s Senator of the Interior declared a ‘major incident’ (“Großschadenslage”) for the first time, a category of emergency response that was only introduced in Berlin in mid-2021 as part of an amendment of the local law on disaster management (“Katastrophenschutzgesetz”). Declaring a major incident allows emergency response operations to be centrally coordinated, volunteer responders to be released from their regular duties, and assistance to be requested from the German Armed Forces. In addition to emergency shelters and private accommodation agencies, unoccupied hotel rooms were made available for a fee of 70 Euro. On Monday evening, it was furthermore decided that the district would cover the hotel costs in full (through the Senate Administration).

The state of North Rhine-Westphalia supplied five emergency generators to maintain power at various hubs. The German Armed Forces refuelled the generators and distributed food at a number of locations, especially on Wednesday. On Wednesday morning, the electricity grid operator announced that power would be restored by 11 a.m., but that people should continue to use electricity sparingly so as not to overload the grid. This was communicated in the media as well as via loudspeaker announcements and flyers distributed by the police in the emergency shelters and care centres. In addition, Cell Broadcast and the NINA app were used to communicate relevant instructions. The major incident status was officially lifted on Friday evening.

Why conduct quick-response research?

Before presenting the findings and conclusions from our quick-response research during the power outage, we will briefly explain and evaluate this method, as it’s still far too little known in the German context and should be used much more extensively. The Disaster Research Unit (DRU/KFS) at the Free University of Berlin and the Academy of the Disaster Research Unit (AKFS) have been conducting this type of research for many years, such as during the 2013 floods in Magdeburg, during the refugee situations in 2015/16 (Dittmer & Lorenz 2016) and 2022 (Dittmer & Lorenz 2022), in the first weeks of the SARS-CoV-2 pandemic in 2020 (Dittmer & Lorenz 2022; Schulze et al. 2020) and during the winter floods of 2023/24 (Bock et al. 2024). Quick-response research can be seen as a “first draft” of the rapidly changing “disaster narrative”. When researchers go into the field quickly after an extreme event, they have the opportunity to collect data that would otherwise disappear or change shortly after the end of a disaster. This includes attitudes, feelings, social interactions or experiences that are often forgotten over time, or that become part of an overarching narrative and interpretation of the disaster or are reframed (Oulahen et al. 2020; Johnson & Vindrola-Padros 2017; Díaz et al. 2023).

If quick-response research is established institutionally as part of disaster management, it can provide constructive support in coping with the situation, as it contributes additional information and perspectives that are often not (or cannot be) seen by crisis management. However, its implementation is very demanding. There is little time for preparation, the research process must be able to adapt dynamically to the situation, and the data situation can be difficult, as detailed interviews are often hardly possible due to the involvement of the actors in the management of the situation. Quick-response research hardly allows for the use of classic criteria of qualitative social research in terms of establishing contact, quantity and quality of data, etc.  Due to the extreme nature of the situation, ethical questions also arise, e.g. regarding the possible retraumatisation of those affected (and the researchers themselves), the disruption of operational processes, the danger of “disaster tourism”, or the instrumentalization of research in sometimes highly emotional contexts (Dittmer & Lorenz 2018).

In addition, particularly in the German context, third-party funded projects often lack the financial and human resources to respond to spontaneous events, so that researchers themselves have to bear the costs involved. Fortunately, in addition to AKFS resources, our quick-response research on the Berlin power outage was able to draw on human resources from the BMFTR security research project KRISENFIT, as this project explicitly provides for in-situ research. The project focuses on investigating the interfaces between local government, civil society actors, and disaster/crisis management.  As a starting point for investigating this topic in the context of the power outage, aid centres such as heating and information points provided by civil society actors, as well as state emergency shelters and “emergency lighthouses” were identified. Information about these offerings could be found via regional broadcasters and the press, as well as on notice boards, e.g. in Zehlendorf Town Hall.

The DRU and AKFS found a number of such offerings in media listings and went to visit them on Monday, 5 January 2026, Tuesday, 6 January 2026 and Wednesday, 7 January 2026. No advance notice was given to the locations visited. The focus in selecting locations to be visited was on ensuring as diverse a range of actors as possible (churches, social associations, cultural associations, outpatient health facilities, political parties, aid organisations, sports clubs, shopping centres). A total of nine civil society aid centres were visited, five of which were still active and responsive. At four of the aid centres visited, no corresponding services were found and no contact person was available, even though they were still listed in the accessible information channels. It appears that in such cases the relevant list could not be updated quickly enough to reflect the actual situation.

We conducted interviews with members of a Protestant church congregation, a cultural centre, a day centre for mentally ill and distressed people, the district association of a political party, and an emergency shelter in a sports hall run by an aid organisation involved in disaster management. Visits of critical locations (e.g. Mexikoplatz S-Bahn station as a central hub for emergency services, supermarkets serving as care points, Zehlendorf S-Bahn station) supplemented this approach. The aid offerings investigated were mainly located outside of the area in Steglitz-Zehlendorf affected by the power outage. With the exception of the aid organisation’s facility, it would not have been possible to operate the assistance centres in the same way in areas without power, as emergency power supplies or feed-in points were not available to them. Apart from Zehlendorf Town Hall, the aid centres visited were mainly located on the north-eastern edge of the affected area in Zehlendorf and around the Steglitzer Kreisel roundabout. In parallel, media coverage was closely monitored, in particular the live tickers of the Tagesspiegel and RBB (which had been quickly established on Saturday morning).

Following this quick-response research, we conducted a population survey and interviewed further experts and affected parties.

Civil society response and operational disaster management activities

Range of aid offerings provided by civil society aid centres

The majority of aid offerings for those affected and residents in need of help in the area affected by the power outage organized by civil society actors were already active from Saturday or Sunday. In most cases, they were available from morning to late afternoon. Civil society associations with regular opening hours on weekdays extended them for the duration of the situation.

The main material services offered by civil society aid centres were warm rooms, (hot) drinks, the opportunity to charge electrical devices and use the internet. Aid centres were also used as social meeting points and for exchanging information. In some cases, affected people also visited the aid centres to continue their daily work, e.g. for remote work or video meetings.

Civil society aid centres operating in the immediate vicinity of the power outage area reported a fairly consistent number of visitors throughout the power outage. Those located in the Steglitz area reported a steady decline in visitor numbers after a peak on the first day of the power outage. This is in part due to the high density of shopping centres, grocery stores and restaurants in the area not affected by the power outage, which opened normally on Monday. A person in charge of one of the aid centres in Steglitz attributed the declining number of people seeking help to the Senate’s announcement that it would cover hotel costs and the start of assistance provided by the German Armed Forces. In all facilities visited in the district, visitors consisted of a mix of regulars, clients, members and relatives of the organisations providing the aid, and people seeking help due to the situation. In some cases, it was not possible to clearly differentiate between these groups.

The emergency accommodation offered by the aid organisation in Dahlem saw consistent use throughout the crisis, though there was a slight increase in overnight guests as it progressed. The reason for this was the increasing cold in people’s homes. Shelters run by aid organisations played a special role as facilities of the state emergency network. People who could not be cared for elsewhere because they were too weak, sick, old or otherwise vulnerable could stay there temporarily until they could be placed in other care and support contexts or the situation was resolved. Their advantage over private or church-run services was the continuous provision of medical care on site.

Self-organisation of civil society actors

In most cases, civil society actors were able to set up the aid offerings they provided using their own resources and without significantly increased material needs. In many cases, organisations mainly adjusted opening hours, organised an internal surge of volunteers from their own structures and scaled up existing service structures (e.g. opening up more spaces, food also available from donations). Some actors paused their main activities and sought out community/association members who might need help at their home. In some cases, they referred those affected to other, more suitable sources of assistance. Many aid points also offered the possibility to shower and perform basic hygiene.

The volunteers at the locations we visited were highly committed and motivated to help. Due to school closures, young people also played an active role in running assistance centres. We found no evidence that the members of the associations, organisations and communities running the centres had been significantly affected by the situation, or have been affected to an extent that would have had a negative impact on their aid activities.

Civil society organisations and associations organised themselves along their regular internal hierarchies and based on the capacities of their members. In most cases, those in leadership positions within the organisational structure also became heads of crisis management when their organisation’s location became an aid centre. Organisations large enough to have formal subdivisions continued to organise according to them and distributed tasks among their substructures. In some cases, new tasks arose, such as setting up a telephone service or appointing someone to handle email correspondence and disseminate information. Helpers were mostly able to organize themselves using their usual communication channels, as they usually had a working mobile network connection.

In some cases, civil society actors had previous experience with crisis management or were able to apply knowledge gained through their day-to-day activities to the crisis. For example, some organisations are active in helping the homeless or were able to draw on their experience in humanitarian aid during the refugee crises of 2015 and 2022. As part of the Behörden und Organisationen mit Sicherheitsaufgaben (BOS) — the authorities and organisations with security tasks — the aid organisation’s unit that operated one of the aid stations was the only entity whose humanitarian activities was tasked and legitimised by the state. It received equipment and was officially assigned a specific role in managing the situation. All other organisations independently organised relief efforts using their own personnel and resources, driven by intrinsic humanitarian motivation and a sense of duty to civil society.

Information management between the crisis management team and aid centres

Up-to-date information on the overall situation or official information was only partially available at the civil society aid centres. Both aid centre staff/volunteers and affected people present at the centres were able to independently inform themselves via the internet without restrictions. Some exchange of information occurred between the individual aid centres, and one centre often knew of the other (e.g. through inter-organisational links, personal contacts or via the media or social neighbourhood networks). However, the district, the Senate and the BOS did not engage in dedicated information-sharing with the aid centres. All centres registered independently via the district of Steglitz-Zehlendorf’s disaster management email address and, in most cases, received confirmation of this registration. In some cases, the district referred to civil society aid centres in some of its information channels. Some organisations interpreted this as confirmation that they were registered as an aid centre at the relevant address.

Although the church-organised assistance centre was in regular contact with the BOS, including a police liaison officer and units of disaster management organisations passing by, the centre didn’t proactively receive inform about the current situation. The volunteers at the emergency shelter run by the aid organisation received updates on the situation mainly through the official structures of the operational command. According to the volunteers, however, these updates did not differ significantly in terms of content or timeliness from the information publicly available or broadcast to the public.

Some interviewees said that they would have liked greater clarity about how many people they would need to provide for (for their own resource planning and coordination between different civil society aid centres) and that more consistent and active communication on the part of the district would have been beneficial. According to one interviewee, it would have been beneficial, if the district had systematically enquired about possible needs or the situation at the aid centre. One aid centre reported that it had heated premises that would have been suitable as emergency overnight accommodation and had submitted a request to the district for cots, but this was rejected because priority was given to official emergency shelters. However, if those affected had actively requested overnight accommodation at the facility, the facility would have been able to create emergency overnight accommodation to a limited extent by procuring mattresses itself. In many cases, aid centres also matched emergency overnight accommodation offered by their own members and people outside the organisation with the reported needs of private individuals.

Some civil society actors suspected that requests and information provided to passing disaster management units were not being forwarded to the crisis management team or coordinating bodies. There was no feedback from official bodies. However, other organisations (with a public mandate and existing day-to-day contacts with the public administration) also praised what they perceived as relatively short response times from the district, particularly in email correspondence. One interviewee mentioned that, due to her day-to-day work, she had good contacts in the district’s social administration. She was therefore confident that she would receive quick and uncomplicated support if necessary. Overall, interviewees did not mention any specific negative effects or disadvantages resulting from the information deficits some of them criticise. All interviewees were able to obtain information themselves (e.g. via the internet) throughout the situation.

Some assistance centres found it challenging to coordinate offers of housing, as well as donations of goods, particularly food. Most of them had experience in processing food and also had the necessary kitchens and materials for it. However, in many cases, anticipating the demand was not possible. In some cases, aid centres had to impose a ban on food donations from the public because the amount of food being given out was no longer proportionate to the volume of donations received. The organisations and communities running the centres anticipated that redistributing food donations would be one of their tasks once the crisis was over.

Impact and affected people

Care workers interviewed at aid centres, who were in close contact with those affected by the situation, reported that many had moved to hotels or the homes of friends or relatives in unaffected areas. Apart from the cold, one of the biggest challenges for those who remained in the area was their fears, some of which were specific and others less so. While some were afraid for their possessions, others were unsettled by the darkness itself or by neighbours walking around with pocket torches.

Here, too, it became apparent that crises and disasters always also promote social cohesion (at least temporarily): many neighbours who had little or no contact in everyday life came into contact during the crisis through neighbourhood and self-help initiatives. The aid centres also served as social spaces where neighbours could come together, get to know each other, and nurture acquaintanceships. In some cases, these interactions reportedly led to the formation of new friendships. Across the board, assistance centres reported a strong sense of gratitude expressed by people seeking help as well as visitors, which contributed to aid workers’ and volunteers’ experience of self-efficacy.

Classification in the Sociology of Disasters

The results of our quick response research about how various actors organised themselves and provided assistance in the situation at hand can be analysed according to the DRC typology, which has been used and further developed in disaster sociology research for decades. This model, first proposed by the Disaster Research Centre (DRC) in the 1970s, classifies the various types of actors and/or organisations involved in disaster response on a spectrum according to whether the situation can be managed with existing resources or whether entirely new actors and tasks are required (emergent groups). When this is the case, it can be concluded that the existing structures were obviously insufficient to address the necessary needs and manage the situation appropriately.

DRC typology

	Existing structures	New structures
Routine activities	Established (Type I) – Actors responsible for fire protection, rescue services and disaster control. Example: the fire brigade extinguishes fires in a disaster.	Expanding (Type II) – actors such as aid organisations that continue their routine activities even during a disaster, but can only manage this with additional volunteer-based structures. Example: German Red Cross (DRK).
Non-routine activities	Extending (Type III) – existing organisations take on new functions. Example: churches operate emergency shelters.	Emergent (Type IV) – new actors, individuals or organisations arise and fill gaps in the formal system. Example: initiatives such as the Helfershuttle in the Ahr Valley.

If we classify the actors who were active during the five days of the power outage into different types, the following picture emerges:

Type 1: Established groups: These are actors who are officially responsible for fire protection, rescue services or disaster control. Volunteer fire brigades and especially the police took on their everyday routine tasks within existing structures. For example, fire brigades continued to extinguish fires and the police patrolled the affected areas or regulated traffic.

Type 2: Expanding groups: At the latest after a major incident had been declared, disaster management organisations were able to use additional volunteer resources to carry out routine tasks, such as caring for those in need of care or setting up emergency shelters.   

Type 3: Extending groups: Sports clubs, the Hellenic community, church congregations and supermarkets used their existing structures, members and resources to take on new tasks in providing emergency care to a large number of people.

Type 4: Emergent groups: In contrast to many other major crises and disasters, such as the refugee situation in 2015/16 or the flooding and heavy rainfall disaster in western Germany in July 2021, no new groups or actors emerged to fill the gaps left by established actors (types 1-3). This leads us to the conclusion that, at least on an operational and conceptual level, existing civil society structures were sufficient to address the situation in terms of meeting the needs of the population.

We acknowledge that the situation was difficult and challenging for many of those affected, and that, as after any such event, there have been many calls for improvements. However, our analysis based on the DRC typology shows that the kinds of developments that typically occur in complex crisis/disaster situations, which indicate significant gaps, did not occur in this case.

Conclusions

During and after the power outage, criticism was and continues to be expressed about the district and senate’s crisis management, crisis communication, disaster response equipment and the protection of critical infrastructure. Some of this criticism is entirely justified. Without questioning the need to objectively and critically reflect on the weaknesses in the preparation and management of this situation and to learn from it, the flood of media assessments by “experts” is neither constructive nor does it lead to differentiated assessments. This is the case because these “experts” are not familiar with disaster management in Berlin or disaster response in general, have not dealt with those affected or those responsible on the ground, and have no other expertise. Instead, it demotivates all the volunteers and civil society actors who, as in every disaster, have dealt with the situation as best they could.   

Public and social media also play a central role in this. During the initial response to the situation, the live tickers on tagesspiegel.de and rbb24.de provided important information and closely followed developments on social media. However, field research also showed that at times there were more members of the press present at the aid centres and emergency shelters than people seeking help. Those responsible at the aid centres reported a “desperate” search for those affected, who “in the best case” had also had negative experiences with the management of the situation and promised the highest possible number of clicks. Initially, media representatives were allowed to enter the shelters without restriction, but after numerous negative experiences with media representatives invading privacy, bans were imposed on visiting the shelters without prior notification. Dealing with politically extreme media also proved increasingly problematic, especially considering the very dominant debate about the letters claiming responsibility for the sabotage and political orientations. In addition, there were isolated cases of defamation of migrant-led aid initiatives on social media, as well as disinformation campaigns relating to the alleged accommodation of refugees in luxury hotels.

In summary, it can be said that civil society initiatives are visible and their services are also recognised by the state in times of crisis. At the same time, as in any disaster, there is room for improvement in the coordination and provision of information for these initiatives. In this case, an appropriate interface between civil society initiatives and the administration/crisis management team would be conceivable, especially where not many prior informal contacts between actors exist (e.g. in smaller communities, such as in rural areas). Horizontal networking in everyday life and in crises between civil society initiatives for the joint coordination of needs and services (“in der Krise Köpfe Kennen” – literal translation: “knowing heads in crisis”), potentially moderated and coordinated by local authorities where appropriate, could become part of an official local crisis management that provides structure and supports aid efforts without regulating or levelling civil society initiatives in their autonomy or their role as crisis response actors.

Civil society organisations are highly capable of self-organisation and providing assistance in a coordinated manner. Such efforts are fueled by the intrinsic motivation of those providing assistance, humanitarian, charitable and social organisational backgrounds and self-perceptions, as well as a strong sense of self-efficacy and regained autonomy through neighbourhood assistance. It is important to establish strong civil society structures in everyday life. Sufficient public funding and support, networking between initiatives and local authorities with a view to increasing resilience, and maintaining constant, appreciative and solution-oriented communication between civil society initiatives and the state are basic prerequisites for the feasibility of effective civil society self-help and the integration of civil society into the official response system in crises and disasters. Civil society self-help cannot be taken for granted; it must be nurtured and developed. The role of civil society, and in particular civil society organisations and institutions, in reducing the burden on disaster response organisations as well as for society as a whole, should not be underestimated by the state and emergency services. Every person who does not find themselves in an acute emergency situation thanks to civil society efforts and who thereby does not require help from official disaster management structures allows the latter to focus specifically on the most vulnerable members of society and on managing the overall situation at hand. Therefore, civil society self-organisation and public funding and support for social infrastructure and associations — such as social institutions, community centres, neighbourhood associations, sports clubs, and other civil society organisations — can also be seen as a contribution to societal resilience and to overall civil protection and disaster preparedness efforts.