B1/1996Y The steel tank downfall in Valkeakoski, 27 March 1996
A massive steel tank of 6000 cubic metres, having the total height of about 35 metres, suddenly fell over at the end of the test filling. The tank was being filled with water to test its water tightness. A little earlier before the tank actually tipped over it was noticed that the tank was tilting. The time between the first observation of tilting and the actual capsizing was around 12 minutes.
One electrician, who was just finishing his daily work nearby destroyed building, died in the accident.
A part of the factory's machine repairshop, the automation and electricity workshop, the forklift truck repairshop, restrooms and other facilities for the employees, the canteen as well as the site huts on the roof, were crushed by the falling tank. In addition to this, the walls of the nearby buildings and site huts, cars, a truck and a lorry in the yard were crushed by a water volume of 5000 cubic metres that flooded from the tank.
On a cable bridge leading to the roof of the repairshop there was an industrial piping through which electricity, district heating, steam water etc. were drawn. This piping partially fell down and was damaged. It then caused further damages in three other factories and in the district heating net of the city.
The Accident Investigation Board found in its preliminary investigations that personal injuries and other damages would have been more extensive. This would have happened if the accident had occurred in the daytime or the tank had tippet over in the opposite direction where dangerous material, e.g. fluid oxygen, was being stored. For this reason an independent investigation group was nominated to investigate the accident as a disaster situation and to give recommendations to prevent similar accidents.
The following causes of this disaster are based on the conclusions of the investigation group.
Mistakes in the design process
According to the Finnish design standards the designed concrete slab of the tank as well as foundation piles were not properly dimensioned. The designer had misused the Finnish guidelines for driven piles without experience for large steel pipe piles dimensioning. The calculations and plans of the slab (12m x 12m x 1,5m) and foundation piles (12 steel pipe piles filled with reinforced concrete, the lengths around 5 m, the diameter 0,5 m) made by the designing engineer were not checked by any other expert. The builder and the main contractor were in no position to check the dimensions and calculations, either. The builder relied on the competence of the consulting office, because the office was a well-known and of good reputation.
The tank itself was made of steel and it was made correct enough, although, its design and dimensions were drawn up by only a simple hand calculation method.
Shortages in driving of the foundation piles
The driving of the foundation piles was carried out simultaneously of the foundation piling for a new paper machine under construction, the latter being the most important work for the building project of the estate. The driving of the foundations piles for the tank was only "a little job". That was maybe one reason why separate instructions for pile driving in this tank was not given despite of very large designed loads of piles. After the disaster it was found that most of the foundation piles did not reach the rock as was recommended.
After driving of all the foundation piles, no feedback, not even in form of driving record, was given to the planning engineer once the construction of the piles had been completed.
Testing of the tank
The filling test was primarily a test of water-tightness of the tank, not at all a structural test of the tank and its foundations. The person, who supervised the filling of the tank with water was not present all the time and he had no-one to replace him, but this was by no means the cause for the fall of the tank. Nobody could foresee the capsizing of the tank and neither employees nor supervisors were expected to take precautions during the test.
Supervision of the project
The whole building project concentrated mainly on paper process manufacture. The project organisers had failed to consider the dangers from a separate tank. However, up and till then, no similar tank had tippet over in Finland.
The municipal building inspection official did not participate in the supervision, which was entrusted to a person who worked at the same time in the project team as a site leader of the building project. This is a common method to supervise large industrial projects in Finland.
The whole building project was divided into several contracts and subcontracts to ensure specialised knowledge and quality, and also to control of the total costs. In fact this is a common, modern, acceptable and very efficient way to control large building projects.
The worksafety group in the factory, as well as the safety group set up for the building period functioned well and co-operatively considering the circumstances. The company itself has a positive safety culture.
Traditionally it is not the duty of the safety group to discern possible mistakes in the planning process and to consider the consequences which may follow. In fact, in this case it was nobody's duty, because the engineers relied on right doing of everyone too much.
To prevent similar accidents the investigation group has made some recommendations for actions. The actions are focused on as following
- the need for classification of building objects for design and execution; classification of companies
- the guidelines for design and building of demanding tank constructions
- the need of risk assessment of demanding constructions
- surveying existing safety level on building projects and measures to enhance the safety level when using new contracting procedures
- defining responsible bodies of inspection measures when using new partition contracting procedures with very many subcontractors
- the need to prepare design guidelines for large steel tanks.