The smoke grenade release in Uppsala, Sweden, August 25, 1993 and The chlorine release at the Vanadisbadet, Stockholm, August 2, 1993 – KAMEDO-report 65

The Kamedo-reports are published by the Swedish Disaster Medicine Study Organisation (Kamedo), at the National Board of Health and Welfare. Observers study the medical, psychological, organisational and social aspects of disasters. The results, with a focus on experiences gained, are presented in the reports.

This report is entirely in Swedish. Only summary in English.


The smoke grenade release in Uppsala

On Wednesday August 25, 1993 just before 10.00 hrs it was reported to the Emergency Services Centre (SOS-Alarm) in Uppsala (SOS-C) that smoke was being emitted at a high school (Katedralskolan) with 1,100 pupils, approximately 80 teachers and 15 other personnel. The school was rapidly evacuated and all the pupils were assembled at the assembly point outside the school. Very soon smoke divers noticed that the smoke was coming from a smoke grenade, which had been defused in one of the pupil toilets. A thin white smoke including zinc chloride caused stinging and irritation from the respiratory tract. Zinc chloride is an irritant gas which can cause severe pulmonary damage including pulmonary oedema. A medical team (a nurse from the Emergency Department, the first-line on-call physician from the Department of Internal Medicine and a second-line on-call physician from the Department of Anaesthesiology and Intensive Care) soon left the University Hospital for the accident site. Approximately ten pupils presented symptoms from the respiratory tract and they were primarily taken care of by the ambulance personnel until the medical team arrived. The county council mobile medical equipment unit (MMEU) from the ambulance service and all unoccupied ambulances were directed to the accident site. To transport those with minor injuries, a large bus was ordered from the adjacent air force base. It took two hours before the last pupil needing hospital care had left the accident site. In all 24 persons were transported by ambulance to the University Hospital and approximately 30 by bus. Apart from the task of localizing and identifying the smoke grenade, the police were responsible for cordoning off the area. They did not register exposed persons at the accident site but did this later during transport to, and at the hospital. The police concentrated on crime scene investigation.
The medical team organization at the accident site was the following: the nurse did the first sorting of the victims into two groups: (i) unaffected people and (ii) people with symptoms such as coughing, respiratory distress, malaise or people without symptoms but with a history of respiratory tract diseases. The internist examined all with symptoms before they were transported to hospital. The anaesthesiologist did a preliminary triage of those with respiratory symptoms and thereafter accompanied one of them to the hospital where he remained. Five youngsters with moderate-to-severe symptoms were immediately transported to hospital by ambulance. Those with mild symptoms were taken to hospital in the bus ordered from the air force base.
The mass media were very offensive. The local radio station broadcast information to the general public within five minutes, which resulted in a great number of phone calls from worried parents, initially to the school but later to the hospital and the police.
Within two hours 107 persons affected by this accident had arrived at the hospital. Another 40 came later the same day to the Emergency Department. The next day a few came to the hospital, four of whom returned due to recurrent symptoms. In total 36 patients were admitted, 48 were sent to the Pneumology Outpatient's Clinic, 11 were sent home directly from the Emergency Department and one patient discharged himself. On the 17-bed-observation ward totally 77 patients were taken care of the first day. The next morning there were still 26 patients in the hospital for observation, but in the afternoon there were only two left (at the Department of Pneumology), and both were discharged within two days.
Red alert (disaster alarm) was never declared because initially only 30 or at least not more than 50 patients were expected. Nevertheless. the improvised organization and command ended up very close to the existing disaster plan. The second-line on-call surgeon and the Emergency Department head nurse took the overall command at the Emergency Department, directing the flow of patients, reviewing the available resources etc. The second-line on-call pneumologist and internist jointly took the overall medical policy decisions, for example deciding what patients should stay for observation and who could be sent home. The Hospital Medical Director jointly with the Clinical Medical Director of the Department of Anaesthesiology were in contact with all other hospital departments and the administration to arrange for necessary hospital beds. The Medical Director and the Clinical Medical Director of the Emergency Department handled contacts with the mass media.
To manage the great flow of patients, those with mild symptoms were kept outside the Emergency Department in the fresh air, being registered there and subsequently directed to the stretcher area, which functioned as a waiting room. There were chairs there and food and drink was served. The victims were handed their medical records. They were then taken in threes to the examination room for examination and treatment. This was possible as many were only mildly affected, and all with the same type of exposure. A centralized admission point was early set up for all patients, ensuring that all were registered before being directed to other parts of the hospital. Apart from using the observation ward, the postoperative room, the pneumology outpatient clinic and department, further beds were "borrowed" from other departments. The hospital pharmacy was informed at an early stage. Drugs were ordered by telephone and delivered immediately. All formalities such as signing the orders etc were done later. The hospital's computerized registration routines were too slow to manage this mass registration situation, so manual routines were soon used to record e.g. the time of admission and where the patients were transferred. The lists were photocopied as soon as each page was full and sent to the provisional information centre and to the Emergency Department office.
The informally established management group met once an hour in the Emergency Department office for updating and an exchange of information and experience. During the evening the emergency and pneumology departments were reinforced with one person each to handle inquiries from relatives and patients.
Just under 70% (99) of all patients coming to hospital the first day were sent home after six hours' observation. Another 30 patients left after 12 hours' observation. Only 17 patients were kept for observation for 24 hours and only two for 48 hours. 10-15 persons presented symptoms that were of psychogenic origin such as worry, anxiety, and hyperventilation. Only about 10 patients were estimated to have been exposed heavily to the smoke.
About an hour after the accident became known the command of the psychiatric disaster management group (PKL group) was alerted. A support centre for all those involved, including relatives, was set up at the Children's Hospital close to the Emergency Department. During the first day about 20 persons were engaged in the activities here. A well-attended information centre was opened in the lecture hall of the Children's Hospital and coffee and sandwiches were served. A few resource persons went to the Emergency Department and talked to people in need. In all 30 patients got crisis support. Information on the support centre was announced on the local radio, which also gave information on normal reactions to severe stress and practical advice on how to cope with these. The day after the accident about 20 youngsters were visited by the crisis intervention team and they also received written information. Later during the day another 10 youngsters were followed-up individually. Co-operation was established with the school authorities and, the day after the accident, 10 representatives from the PKL group reinforced the school health care staff. Four information meetings were arranged in the school auditorium for 1,100 pupils. General information on stress reaction was given and the classes were prepared for later visits by resource persons. In 15 classes a modified debriefing was held. These debriefing meetings were led by PKL group members together with a member of the school health care staff.
A questionnaire answered by the pupils showed that approximately 80% thought they had got too little or no information about what had happened in the acute phase of the accident. Suggestions for improvements were improvement of the fire alarm and better information about emergency exits. A loudspeaker system should be installed outdoors as well as indoors for better information about what was going on and for instructions on emergency evacuation etc. The teachers should have a more leading and organizational role and teachers and pupils should practice emergency evacuation of the school regularly.

Experience and conclusions

When local radio gives information on an accident, it should also give telephone numbers to which relatives can ring for additional information.
The police should register all exposed persons as soon as possible and keeping records on who are taken to hospital. This will then facilitate the information given to relatives, the incident commander and the hospital, but registration must not delay transport of patients to hospital. Persons with mild symptoms can preferably be transported by, for example, bus. During transport to hospital the police can complete their registration if they have not had time to do so earlier.
Medical team members should only exceptionally leave the accident site at an early stage. Special personnel should be used for accompanying patients during transport to hospital.
One should never hesitate to declare red alert (disaster alarm) when there is a major accident. It is important that all persons being taken to hospital pass a centralized registration point before they are directed to other parts of the hospital or sent home.
It is important for the hospitals) to receive regular information from the accident site.
Cordless telephones at the emergency department facilitate the work and allow flexibility.
Registration in a computer network should be possible also for mass registration so that all persons needing information on patients easily and rapidly can get this regardless of where in the hospital they are. A bar-code system might be useful for registration.
The psychiatric disaster management group (PKL group) must be alerted as soon as possible.
Having an observation ward closely linked to the emergency department is practical, as this ward can be used both for short-term care and observation and as an assembly ward before patients are directed to other parts of the hospital.
It is necessary that the activities of the PKL group are done in a flexible way. When appropriate supports given to victims, this should be done jointly with the medical staff. Rapid access to patient lists and easy access to telefax and cellular telephones are necessary for efficient PKL activity. It is essential to give affected people early and comprehensive information.
The disaster management group at the hospital should not be disbanded too soon as there might be a further need for its activities up to several days after the accident.

The chlorine release at the Vanadisbadet (Vanadis outdoor swimming pool) in Stockholm

On August 2, 1993 an accident involving an irritant gas occurred at Vanadisbadet in Stockholm. Vanadisbadet is a big outdoor swimming pool in the central Stockholm, in the Vanadislunden park. The swimming pool has up to 2,000 visitors on a sunny summer day. This particular early August day was sunny and rather warm (approx 20°C), but at the time of the accident there was a limited number of visitors (approx 100), mainly children and youngsters.
To clean the pool water sodium hypochlorite (NAOCl) is used. By the reaction between sodium hypochlorite and phosphoric acid, chlorine is released to clean the water. The sodium hypochlorite is stored in a 1,500 litre tank located below ground at the northern part of the pool, and covered by a one-metre-high wooden structure. Sodium hypochlorite for the tank is delivered about every three weeks. This time around 800 litres was to be filled into the tank. The transporter arrived at around 10.00 hrs carrying two semicontainers containing 13% sodium hypochlorite and one semicontainer containing 75% phosphoric acid. The containers looked the same and the labels faced in the "wrong direction". The driver took the wrong container and started filling phosphoric acid into the tank. He noticed almost immediately that a chemical reaction had started and he stopped filling the tank. Approximately 10-15 litres of phosphoric acid had then entered the tank. As the driver was experienced in handling chemicals he was aware of chlorine gas could be formed and he alarmed the personnel, who immediately evacuated the swimming pool area. The fire brigade was alarmed. In total 14 units from the Stockholm Fire Department were involved in the rescue work.
Experts from the Swedish National Rescue Services Agency estimated that around 14 kilograms of chlorine gas had been formed, corresponding to around 5,000 litres (5 m3). If all the phosphoric acid had been filled into the tank around 8 times more chlorine gas would have been formed i.e. around 40,000 litres (40 m3). The lorry transported 1,600 litres of sodium hypochlorite and 400 litres of phosphoric acid. If all the hypochlorite and phosphoric acid were mixed, around 80,000 litres (80 m3) of chlorine gas would have been formed.
A large area around the swimming pool was roped off. High concentrations of chlorine gas were measured close to the tank, but otherwise only low concentrations of chlorine gas within the roped-off area. After neutralisation of the contents of the tank (with sodium carbonate), which was finished about two hours after the fire brigade had been alarmed, the rescue work was ended.
It appeared that nobody had been exposed to high chlorine concentrations. In all 33 persons were taken to hospital for examination and observation. Of 15 patients (14 children and one adult) coming to S:t Göran's Children's Hospital, nobody had any severe symptoms from the respiratory tract. Eighteen persons came to the Karolinska Hospital, including five children. (One child and his mother were sent to S:t Göran's Children's Hospital.) None had any severe symptoms from the respiratory tract so there were no problems at the two hospitals taking care of the patients.
Most of the patients were transported to hospital by the police or in private cars, only a few by ambulance.

Experience and conclusions

Although this chlorine accident did not result in any severe injuries and only a limited number of persons went to hospitals for examination and observation, some valuable lessons were learnt.


From the fire brigade one could state that the breaking point and control point were in the same place, which was not advisable. Further it appeared that this place was too close to the chlorine release. No proper casualty assembly point was established and decontamination was not arranged for. The police had their own breaking point which became a preliminary casualty assembly point. Victims were transported partly in a disorganized way by the police without prior consultation with the incident commander or the control ambulance. However, very soon the police informed the incident commander about the number of persons they were transporting. Several persons were transported to hospital in private cars. This meant that the Emergency Services Centre (SOS-Alarm) did not get an overview of patient distribution to hospitals. The registration of victims was done during transport to hospital and at the hospitals by the police. The Poison Information Centre received many telephone inquiries without having been informed from the accident site or the Emergency Services Centre (SOS-Alarm).
At the Karolinska Hospital it was realised that there was a need to be able to administer oxygen to more than one patient per oxygen outlet. There were no oxygen masks for children. The plain clothes policemen registering patients or investigating the accident had identified themselves on arrival at the hospital, but many of the personnel did not know who they were or what their job was.


The breaking point and the control point should not be in the same place and they should be well away from any chemical leak.
Casualty assembly points must always be decided by the incident commander.
Transportation of victims should always be co-ordinated by the medical incident commander via the control ambulance. The Emergency Services Centre must be continuously informed, so its staff may maintain a good overview of the situation and hence can direct patients to the different hospitals.
The registration of victims by the police must not delay transportation of victims to hospitals. Registration can be done during transport or at hospital.
The Poison Information Centre must always be informed at an early stage so that adequate information can be given to the general public and to hospital staffs.
It must be possible to administer oxygen to more than one patient per oxygen gas outlet.
Oxygen masks for children should be available at all emergency departments, even if these primarily do not take care of children.
Policemen in plain clothes working (registering etc) in the hospital should wear armlets or similar so they can easily be identified by personnel and patients.

Read the full Summary