Safety and Energy Efficient Marine Operations 19 th November 2015 Investigation into the Sinking of the Ro-Ro Passenger Ferry EXPRESS SAMINA Reference: A. Papanikolaou, D. Spanos, E. Boulougouris, E. Eliopoulou, A. Alissafaki, Investigation into the Sinking of the Ro-Ro Passenger Ferry EXPRESS SAMINA, 8 th International Conference on the Stability of Ships and Ocean Vehicles,STAB 2003, Madrid, Spain, September 15-19, 2003. Prof. Apostolos Papanikolaou, NTUA-SDL National Technical University of Athens Ship Design Laboratory (NTUA-SDL) GREECE Email: papa@deslab.ntua.gr, URL: http://www.naval.ntua.gr/sdl
Presentation Outline The accident Investigation set up Investigation analysis Conclusions and recommendations 2
SHIP PARTICULARS The Passenger/Ro-Ro Ferry EXPRESS SAMINA Ship Name EXPRESS SAMINA Ship Type Passenger / Car Ferry Builders CHANTIERS DE L' ATLANTIQUE - St. NAZAIRE Date of Built 01/01/1966 Area of Operation AEGEAN SEA Type of Trip Day / Night Length over all, [m] 115 Breadth moulded, [m] 18.1 Depth to the upper deck, [m] 10.7 Depth to the main deck, [m] 6.01 Subdivision Draught, [m] 4.55 Service Speed, [kn] 18.0 Gross Register Tonnes 4407 Deadweight, in tonnes 1089 Passengers according to summer 1038 certificate Passengers according to winter 335 certificate Number of persons in Boats 674 Number of crew 62 1 2 3 4 5 6 7 8 9 10 11 12 Compliance with the provisions of EUROSOLAS regulatory framework, disposing an A/Amax over 0.98, according to MSC/ Circ. 574 & Circ. 649 for one compartment standard. 3
The Passenger/Ro-Ro Ferry EXPRESS SAMINA 4
The Accident Accident time: Tuesday September 26, 2000, 22:10 local time The ship: passenger/ro-ro ferry EXPRESS SAMINA Trip conditions: Average speed of 18.5 knots and moderate weather conditions of 5-6 Bf People onboard: 533 people onboard, 472 passengers and 61 crewmembers Cargo payload: 17 trucks and 34 cars on car deck Approaching the port of the Greek island Paros, she collides with the rocky islet Portes, at a distance 3 nm from the destination port The ship sank after abt. half an hour as the result of flooding of the main engine room and further progressive flooding of the entire buoyant spaces The accident caused the loss of 80 people 5
The Accident 6
The Damage Openings Damage A : due to stabilizer fin penetration, abt. 3,00 m long Damage B : set of three openings (raking cracks) with the longest one 0,60 m Damage C : raking damage above WL, more than 6,00m long Damage D : cracks of secondary importance 7
The Main Damage Opening at E.R. The crucial underwater damage A due to the penetration of the right stabilizer fin into the main engine room (one compartment damage) 8
Flooding Scenario Accident s events reconstruction was based on the systematic analysis and matching between the information obtained from survivors testimonies and divers findings, the performed hydrostatic calculations and finally the numerical time domain simulation of ship s foundering. In this way, the most probable scenario was identified as the scenario that matched best the information and facts in hand and also satisfied the conditions resulting from the numerical tools simulation. Testimonies Divers Findings Flooding Scenario Simulation Hydrostatic Calculation 9
Testimonies Analysis One hundred and ninety two (192) testimonies of surviving passengers and seventy-five (75) testimonies of crew (from whom twenty-nine (29) testified more than once) were captured in a developed database and systematically analyzed. 10
Testimonies Analysis Time of Collision and Foundering Time of Hitting the Rock Time of Sinking 23:05 23:02 22:59 22:56 22:53 22:50 22:48 22:45 22:42 22:39 22:36 22:33 22:30 22:27 22:24 22:22 22:19 22:16 22:13 22:10 22:07 22:04 Time of Accident & Time of Sinking Average Time of Accident = 22:10 (Sample 92 passengers) Average Time of Sinking= 22:42 (Sample 43 passengers) 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101 106 111 116 121 126 131 136 141 146 151 156 161 166 171 176 Passengers 11
Testimonies Analysis Access to Life Jackets Τρόπος Παραλαβής Σωσιβίου How passengers got life jackets Άλλος From επιβάτης other passenger 23% 28 % Never Δεν έλαβε got σωσίβιο 2 % 4% On Με his Ιδία own Πρωτοβουλία 62 % 66% From Πλήρωμα crew 7% 8 % Sample Δείγμα of 181 221 passengers Επιβαινόντων 12
The Hydrostatic Damage Stability Analysis 3 Possible Damage Scenarios the ship proved satisfactory in terms of reserve buoyancy and relevant stability criteria to withstand the particular (as other) one compartment flooding. it proved that for the particular loading condition that the ship could survive this two compartment damage as well. For the indicated 3 compartment damage, the bulkhead deck line immerses at the equilibrium position, however the ship would have adequate reserve buoyancy and stability as well as adequate margin of heel to avoid an immersion of the damage opening located at platform deck. She would float at least for a prolonged time to allow safe evacuation. 13
The Watertight Doors The found most probable scenario considers 9 out of 10 watertight doors open. This scenario is practically also confirmed by the testimonies and divers findings. 14
NTUA-SDL Simulation Method Investigation analysis employed the CAPSIM time domain numerical simulation code of NTUA-SDL for the damaged ship motion and flooding simulation. In particular it was possible To simulate the flooding process of the damaged compartments and the subsequent progressive flooding and To estimate the sequence of a series of characteristic events that took place during the ship s sinking and are of importance for the legal process. 15
NTUA-SDL Simulation Method excitation X(t) Damaged Ship motion Y(t) f(x(t),y(t),t)=0 Simulation method calculates the nonlinear 6 DOF motion of an (intact or) damaged ship in time domain under the excitation of external forces (gravity, sea waves, wind etc) and of internal interaction effects due to floodwater. Ship to wave hydrodynamic interaction is treated within linear potential theory (panel code NEWDRIFT). Ship to floodwater interaction is treated by a simplified formulation of the coupling forces. Ship Particulars CAPSIM + NEWDRIF T Environmental Parameters CAPSIM Code 3D non-linear time domain code for the 6 DOF ship motion at zero speed of intact or damaged ship in regular and irregular seaways, considering variable flooded water (inflow-outflow hydraulic model) Computing time at ALPHA WS 1:4 (to real simulation) RAM requirement < 10 Mb NEWDRIFT Code Seakeeping code based on 3D panel method (pulsating source) - 6 DOF Linear and quasi-nonlinear frequency domain analysis (drift force effects included) Regular waves excitation irregular seaways through spectral analysis procedure Intact vessel 16
Geometric Modeling Damage openings, Doors, Staircases, Ventilation openings, etc. are geometrically modeled permitting the simulation of water flow through them. Spacious objects, like main engines and generators, were modeled as 3D geometric spaces. 17
Heeling [deg] The Estimated Heeling Scenario Comments on characteristic stages of ship s sinking 40 35 0.0 3.3 6.7 10.0 13.3 16.7 20.0 23.3 Time [min] Immersion of Embarkation Deck Immersion of Embarkation Deck 30 25 Immersion Immersion of of Promenade Promenade Deck 20 15 10 5 Collision Collision Max Transient Heeling Transient Heeling Immersion of of Damage 'C' damage C Shift of Cargo Shift of Cargo 0 0 200 400 600 800 1000 1200 1400 Time 1600 [sec] 18
Progressive Flooding of Spaces Floodwater Free Surface below Still Water Plane [m] Water pours into the main engine room through damage opening A and flows through the open left watertight doors into the entire ship. Through the numerical simulation, the time series of water elevation in every compartment is obtained, enabling the estimation of the time during which the particular compartment was accessible or not by people onboard. 0.0 0,0-0,5 3.3 6.7 10.0 13.3 16.7 20.0 23.3 Time [min] Main E.R. Aux Fwd E.R. -1,0-1,5-2,0-2,5-3,0-3,5-4,0 0 200 400 600 800 1000 1200 1400 Time 1600 [sec] 19
Heeling [deg] Sea Waves Effects Prevailing weather conditions during the ship s sinking were moderate up to 6 Bf with a significant wave height 1.5-2.0 m Numerical simulation proved that wave effects have limited effect on the events evolution with the main effect arising at the later stages of flooding 40 0.0 3.3 6.7 10.0 13.3 16.7 20.0 23.3 Time [min] 35 30 25 Heeling in Still Water 20 15 10 5 Heeling in Seaway Hs = 2.0 m 0 0 200 400 600 800 1000 1200 1400 Time 1600 [sec] 20
Floodwater Free Surface into E.R. below Still Water Plane [m] Sea Waves Effects Evolution of free surface of floodwater inside the E.R. 0.0 0,0-0,5 3.3 6.7 10.0 13.3 16.7 20.0 23.3 Time [min] Seaway Still Water -1,0-1,5-2,0-2,5-3,0-3,5-4,0 0 200 400 600 800 1000 1200 1400 Time 1600 [sec] 21
Heeling [deg] Alternative Scenarios Alternative scenarios considering watertight doors closed were also investigated and proved that ship would have survived in these cases or would have behaved in different way. 40 0.0 3.3 6.7 10.0 13.3 16.7 20.0 23.3 Time [min] 35 30 25 Heeling in Still Water 20 15 10 Scenario of three compartments 5 0 0 200 400 600 800 1000 1200 1400 Time 1600 [sec] 22
Ship Drift The ship wreck was found 1.4 nm away from the collision spot, just 0.3 nm away from the island s coastline. 23
Conclusions The passenger/ro-ro ferry EXPRESS SAMINA sank after collision with a rocky islet due to the lack of watertight subdivision resulting from the fact that the water tight doors were left open. Ship watertight subdivision should be recorded and be controllable at any time The employment of a time domain simulation method in the investigation of the ship s loss proved particularly useful, enabling the reconstruction of history and the estimation of the events sequence independently of the quality of available testimonies of survivors. Ship evacuation studies, considering the time of a ship to sink, should exploit modern analysis tools to rationalize and improve the evacuation procedures, minimizing the risk of loss of lives. An increased research effort in this field appears advisable. Time to sink and stability during the flooding process are two significant parameters that need further investigation 24
The Sinking Scenario The collision on the rocks at 22:10 (+0 min) 25
The Sinking Scenario The transient heeling of 5 deg at 22:13 (+3 min) 26
The Sinking Scenario Immersion of upper damage opening at 22:23 (+13 min) 27
The Sinking Scenario Immersion of promenade deck, heeling 23 deg, at 22:27 (+17 min) 28
The Sinking Scenario Immersion of embarkation deck, heeling 33 deg, at 22:30 (+20 min) 29
THANK YOU VERY MUCH FOR YOUR KIND ATTENTION Prof. Apostolos Papanikolaou, NTUA-SDL National Technical University of Athens Ship Design Laboratory (NTUA-SDL) GREECE Email: papa@deslab.ntua.gr, URL: http://www.naval.ntua.gr/sdl