About us


Contact info


Order books

Welcome to a chapter of the e-book Disaster Investigation.

2.3 The Cause of lost Stability. Leak in the Sewage Tank Room. Dangerous Routines on the Bridge

The proximate cause of the sudden loss of stability at 01.02 hrs was certainly a hull plate leak in the swimming pool compartment bilge structure on deck 0 forward or another similar compartment further aft, e.g. the stabilizer compartment or the sewage tank room four meters below the waterline -see figure 2.3.1 in this window.

(Fig. 2.3.1 - Swimming pool and stabilizer spaces on deck 0)

The pool compartment was a 'wet' space, where water always spilled out from the pool and ended up in the bilges. The pool itself was probably recessed into the double bottom, i.e. the regular height of the double bottom was reduced below the pool. The space between the outer shell and the swimming pool bottom was probably a cofferdam and you wonder about the condition of this very strange arrangement.

Flooding of the pool/sauna space had occurred previously, e.g. May 1994 - 20 centimetres of water in the sauna! - and been reported to the Commission (act F10) 7 October 1994.

It was evidently never investigated. The Commission investigated nothing but just blamed the accident on the visor.

The leak may alternatively have started in the sewage tank room aft of the pool room or at the stabilizers. A fracture may have occurred in the hull shell bilge plates. The two big bangs and the scraping noise heard by passengers prior to the listing indicate that the 'Estonia' may have collided with something.

The shell plate and the lose shell frames then deformed elastically inboard - like a ripped paper - due to the big pressure difference and a big opening >0,3 m² (say the fracture is 6 meters long and the plate was pushed in 5 centimetres and you have a very big opening 0,3 m²) developed and water started to fill the compartment - >50-100 tons/minute. Maybe the many surviving crewmembers 1.42 were called upon to stop the leakage and failed? Suspect areas with possible corroded hull plates have been observed on the ROV-video taken in December 1994 1.16 - see photos right and below. The German group of experts 3.13 believes the engine crew had been called upon to try to secure the visor or the leaking ramp, but there is no evidence that the visor was lose at this time. The water on deck 0 started to spill out in the centreline corridor on deck 1 above at about 00.56 hrs - it had risen up through the staircase - where several passengers observed it. The watertight doors to the flooded compartment on deck 0 must then have been closed.

Fig. 2.3.2

The water inflow had then been reduced due to pressure balance outside/inside, but the compartment was almost full to deck 1 level and spilled out when the ship rolled. Then the fracture in the bilge plate suddenly developed further - if the bilge plate and frame structure were corroded, so were the forward and aft bulkheads in the sauna compartment, and these were now under pressure - the compartment was filled with water. It is assumed that the fracture first developed aft into the conference room/sewage tank room at 00.57 hrs - and the room was flooded and filled up quickly - the water level in the other compartment may have dropped.

Fig. 2.3.3+4
Then the fracture developed aft into the empty starboard heeling tank - the bulkhead split with another bang at 01.00 hrs - and the heeling tank filled. This resulted in the sudden listing at 01.02 hrs! Then more water flowed in and spread everywhere through open watertight doors. Many passengers heard the 'bangs' - they thought that there was grounding or collision - but splitting plates also makes big noise - the ship hull is like a violin box.

The leakage mentioned above caused the ship to temporarily trim on the bow! Any water leaking in at the forward ramp on the car deck would then collect inside the ramp! The visor was also dipping deeper into the waves due to the bow trim! If, as the Germans suggest, the visor was never properly secured prior to departure, it could very well have been partly broken lose at this time - as a consequence of the major leakage in the sauna/pool compartment.

That the 'Estonia' would sink due to a major leakage in combination with open watertight doors is clear. That the visor may have been partly detached at the same time was a coincidence that the Commission used to cover up the real cause - the leakage.

Open watertight Doors

The 'Estonia' always sailed with many open watertight doors 1.23 in contradiction to SOLAS safety requirements. The five doors on deck 1 forward were always open. The many doors between the engine room compartments were also open. The doors to the leaking compartment were apparently closed (so that the compartment filled up and water spilled out on deck 1 above).

If a hull leak developed, the crew, as per Instruction 5.81.1 in the 'Estonia' safety manual (see Supplement no. 226 in the Final report (5)), on the bridge should close (sic) the watertight doors, which was a totally unsafe procedure. The doors should have been closed all the time! The correct instruction should have been to verify, if all watertight doors were closed, when a leak alarm was received.

There is no evidence that the watertight doors on the 'Estonia' were closed after the listing. In one sentence in a testimony is stated that the doors may have been closed after the listing, but then it was too late. The testimony confirms on the other hand, which all surviving passengers on deck 1 agree to, that the doors were open before the accident. No passenger has testified that the watertight doors were closed later.

Leakage detected. Bilge Pumps running

The crew (Linde or Treu/Sillaste) probably discovered the leak in the compartment prior to the sudden listing and started the bilge pumps. We know that the bilge pumps were running 1.3, so the ship was certainly leaking and sinking had begun. Bilge pumps are only started, when there is water in the bilges. The car deck was six meters above the bilges and above the waterline and was drained by gravity.

When only one hull compartment was leaking, the initial stability was still positive. When the adjacent compartment filled up, the initial stability was still positive, but reduced. The ship should survive with two hull compartments flooded. But when the leak developed aft and flooded the empty heeling tank, the initial stability was not only very small - you suddenly had 183 tons of water in the starboard heeling tank 2.17.

It caused the sudden rolling >30-40 degrees at 01.02 hrs and a stable condition at 15 degrees list at 01.05 hrs. Persons not holding on to fixed structure and lose objects were thrown down into the lee, persons broke arms and legs, and panic developed. Due to the small initial stability (at 15 degrees list) and small righting arm, the ship rolled deeply 20 degrees port and starboard, so when the ship rolled to port, it was almost upright, and persons could evacuate over rather flat decks and reach deck 7, 2.12 and 3.21. When the ship rolled to starboard the heel angle was >35 degrees.

It is remarkable that the divers have not examined the normally unmanned spaces on deck 0 of the ferry from inside to locate the hull damage in the shell plate and examine the suggestion that there was a leak. Some 'experts' have later commented upon the proposal that the 'Estonia' was leaking.

These 'experts' suggest that shell plate must be ripped off - 5 or 20 m² - to sink the ship and that it would be easy to spot the hole from outside. They talk nonsense. It is sufficient with a thin fracture damage splitting the plate and that the watertight doors on deck 0 are open. An opening 0.3 m² between split plates allows 100-200 m3/minute inflow!

The fracture opening has probably closed itself due to the elastic behaviour of the fractured plate. Only by very close examination of the hull plates from outside will the fracture be seen, but an experienced diver can probably find it in a few hours. At the only official dive inspection taking place the divers were not asked to inspect the hull.

The 'experts' have not commented upon the splitting of the shell plate and that one edge deforms elastically sideways to produce the required opening - only 0,2-0,3 m² - or about the alleged holes due to corrosion seen on the ROV-film above.


To 2.4 Back to index