Serious Errors in the SSPA Reports

spotted by Anders Björkman


About us


Contact info


Order books


Serious Errors in the SSPA Reports (20 October 2008)

The Vinnova/SSPA report no. 4006 4100-4 - Foundering tests - contains serious errors that are repeated in the final Vinnova/SSPA report. The errors concern what happens when M/S Estonia in full scale and model scale capsizes and floats upside down on the compressed air enclosed or trapped in the hull. Evidently the pressure on the enclosed air differs in full scale and model scale as explained with the below figures 1-4. When the scale is 1/40 the water pressure on the air inside a capsized hull is abt. 40 times smaller than in fullscale. But the vessel does not only float on compressed air upside down. The total volume of the ship's superstructure and deck house below water is then about 59 190 m3 and it contains a fair amount of buoyancy. With an average permeability of 0.934 (space full of water) there is still 3 906 m3 of material (incl. cargo) in the superstructure and deck house that provides buoyancy.

Full Scale

In figure 1. below Estonia (full scale) floats in the water. Her displacement, i.e. the volume of water below water line pushed aside by the hull, is abt. 11 930 m3 (10 666 m3 air, 1 264 m3 solids) and she has abt. 6 886 m3 (6 156 m3 air, 730 m3 solids) reserve buoyancy between waterline (WL) and main deck (UD). Her draught is abt. 5.2 metres. The vessel floats due to this displacement/buoyancy; compare Arkimedes! The total volume in the hull (displacement + reserve buoyancy) is then abt. 18 816 m3 of which 16 822 m3 is air and 1 994 m3 is solid material.

Above the hull is the car deck superstructure and the deck house but they do not contain any air providing buoyancy. However, the superstructure and deck house, below water after capsize, contain material that occupies volume which in turn provides buoyancy; e.g. if there is 3 910 tons of steel (specific gravity 7.82) in the superstructure and deck house below water, it occupies 500 m3 and thus provides 500 m3 extra buoyancy or, if there were 5 000 tons of other objects in the superstructure and deck house with specific gravity 1.47 (wall and ceiling panels, carpets, furniture, cargo, etc), it occupies 3 406 m3 and provides another 3 406 m3 extra buoyancy. Note that this extra non-air buoyancy is not subject to compression. Thus Estonia probably had about 3 906 m3 of buoyancy in superstructure and deck house that would assist her floating upside down after capsize.

When Estonia has capsized, figure 2, she floats upside down due to buoyancy of solid material now below water + air trapped inside the hull. The air in the hull is compressed due to external water pressure at a new equilibrium. The bottom of the compressed air bubble is say 5 metres below WL and the air has been compressed to abt. 11 215 m3.

As you still need 11 930 m3 buoyancy to float on and you have 3 906 m3 in superstructure and deckhouse + 11 215 m3 compressed air at 1.5 bar in the hull, there is abt. 3 191 m3 volume of the hull that remains above WL. That's where the reserve buoyancy ended up. Estonia thus floats upside down with abt. <1-2 metres of the hull above WL. 5 607 m3 of original air volume in the hull is replaced by sea water.

Model Scale 

In model scale, say 1/40, the total volume of the hull is only 18816/64000 = 0.294 m3 or say 294 litres of which 186 litres is buoyancy (the model weighs 186 kgs and displaces abt. 13 cms) and you have 108 litres of reserve buoyancy - see figure 3. The model also has an unknown amount of buoyancy in the superstructure/deck house which should correspond to 61 litres of air. When the model turns upside down, it floats on 61 liters of solid material below water and 125 litres of compressed air inside the model hull. But then the outside water pressure on the air is only about 10 cms of water or 1.01 bar, so the volume of compressed air inside the model is only reduced by abt. 3 litres! As the model requires 186 litres buoyancy to float, there remains 105 litres reserve buoyancy in the hull above WL after model capsize - see figure 4. Thus the model will float with about 10 cms of hull (4 metres full scale) above WL! Reason is simply that the pressure to compress air in the hull is abt. 40 times smaller in scale 1/40.

To adjust that height to full scale you have to allow abt. 55 litres (3 529 m3 full scale) of air to escape from the model. The model will then still float with the remaining 50 litres of air (3 191 m3 full scale) above WL.

SSPA model of Estonia floats after capsize!

If Estonia full scale would have floated upside down, Estonia model scale would have floated upside down. No sinking. Apparently it was not the case as in the model tests the model sinks slowly (after air in the hull is slowly being released).

How does Vinnova/SSPA describe this in their report?

"A number of tests were carried out where the model capsized, trapped air and remained floating upside down. The volume of this trapped air was measured, and a mean value was found to be around 40 liters (2 560 m3 fullscale) . Also the pressure of the trapped air was measured. The scaling laws give for the present situation that about 20% of the trapped air should be evacuated to give a proper remaining amount of trapped air in the model, see Appendix 1. In this case around 8 liters (512 m3 fullscale) could be let out in order to fulfill the scale laws. The two valves in the bottom of the model were calibrated giving a flow of 6.7 liters (429 m3 fullscale) each per minute at the actual pressure. This means that one valve could be held open a little more than 1 minute during the test."

Not very scientific or convincing and not in accordance with real full scale and model scale situations. SSPA ignores completely the extra, permanent buoyancy provided by the superstructure and deck house that are now below water. That buoyancy is around 3.906 m3 fullscale or 61 litres model scale. 105 litres air model scale is 3 529 m3 full scale and maybe this what Estonia had above waterline after capsize. There are about 125 litres air and 61 litres of solids below waterline to float on. 6.7 liters model scale is 429 m3 full scale! Why let that out per minute? SSPA knows that the full scale ferry could not have sunk at all after capsize.

The Vinnova/SSPA final report does not include any descriptions and calculations of buoyancy of solid material in hull, superstructure and deck house and (compressed) air in the hull of Estonia full scale and in model scale and available buoyancy after capsize. It is very serious. How can you explain sinking, if you do not calculate available buoyancy at every instance?

Condition for immediate Sinking

It should be noted that, if Estonia full scale did not have 6 886 m3 reserve buoyancy in the hull as assumed in figure 1 above, but only 4 000 m3 reserve buoyancy and much less buoyancy in the superstructure and deck house, she would have sunk immediately after capsize. The total air volume in the hull then would have been compressed to <10 000 m3 at capsize. The capsized hull floats deeper and the pressure would be higher on the air in the hull, >1.5 bar, and the remaining buoyancy provided by compressed air would be too small to allow floating. The ferry, upside down then, sinks at once and the air in the hull is compressed more and more - to say 6-7 bar when it touches bottom at 80 metres. The volume of the air in the hull is then compressed to less than 3 000 m3.

In model scale, the model will float with equivalent of 7 056 m3 reserve buoyancy, i.e. 105 litres, and you have to remove these 105 litres of air, at once, to allow the model to sink. If you only remove 6.7 litres each minute (429 m3/min full scale) it takes the model 16-17 minutes to sink ... as shown in model tests videos but it has nothing to do with reality. You must evidently remove all the excess air ... at once! And that's where the model test goes wrong! Apart from ignoring constant buoyancy in the hull, superstructure and deck house due to solid volumes there being submerged.

Full-scale Computer Simulations

The full-scale computer simulations done by Safety at Sea, Glasgow, strangely also copy the slow, 16-17 minutes, sinking as per SSPA model tests. But in full scale the air is compressed at once after capsize and the ship should sink immediately it that were the case ... so why does the computer animation exactly copy the model sinking? Air being let out? It is not possible! The Glasgow company Safety at Sea Ltd's (associated with Strathclyde University) computer simulation is full scale and there is no need to let out air. The computer simulations are also faked! One moment the simulated ship is seen floating upside down with bottom/keel high above waterline, the next is sinks ... slowly. But there is no way the air inside the hull can escape ... slowly. The authors of the simulation, Dracos Vassalos and Andrzej Jasionowski, suggest that water flows up from below?? But where does the air go? It is quite serious when scientists of a university starts to fake their work! Because that is what they are doing. Actually the work is done by underpaid students that are forced to manipulate the input to achieve the desired result of their teachers!


The model tests and theoretical calculations by Vinnova/SSPA of Estonia capsizing does not compare with or reflect full scale or reality.

As Estonia, full scale, would have floated after capsize, the model would also have floated, albeit much higher, and never sunk regardless if some adjustments were done by allowing trapped air to escape.

If Estonia full scale, would not have floated after capsize, she would have sunk immediately. To show this in model scale, you have to allow, say 105 litres of trapped air, to escape at once. To play around with one valve letting out only 6.7 litres each minute, first aft and later fwd, delays the sinking 16-17 minutes, actually seen in the model tests videos. But it has nothing to with reality. The computer simulations of the same thing by Safety at Sea, Glasgow, are also faked! This is very serious and has nothing to do with real safety at sea that Heiwa Co works for.

Anders Björkman, Heiwa Co, 20 October 2008

Heiwa Co start page

Heiwa Co English Estonia page

Heiwa Co Swedish Estoniasida