Ship was carrying 77033 tons of HFO loaded in Ventspils
enroute to Singapore via Gibraltar.
This is a near-repeat of the Erika,
badly corroded, ballast tank, side plating failing in a heavy beam sea.
CTX has been told privately that it was the Master
that made the decision to counterflood
the port ballast tanks in order to reduce the list.
Apparently, one reason he did this was that there was so much list
that the cargo tanks were leaking out the tank lids.
Presumably he thought that it he could stop this leakage,
he might be allowed a port of refuge.
Also, the main engine would not operate at this list.
However, this decision probably increased the stress
on an already damaged hull;
but not nearly as much as ABS implies.
ABS correctly claims the bending moment went
from 125% of allowable to 163%,
implying the max stress increased by a similar amount.
However, the reduction in list also reduced
the the maximum distance to the neutral axis,
effectively increasing the section modulus.
As a rule of thumb, for a given bending moment,
a 15 degree list will increase the maximum stress
by about 15% (Principles of Naval Architecture, 2nd Ed, p 191).
In other words, in the listed condition,
the max stress was at least 20% higher than ABS implies.
The counter-flooding increased the bending moment by about 30%;
but also increased the effective section modulus by over 20%.
The actual increase in max stress was of the order of 10%.
In short, it is not at all clear, as ABS implies,
that the ship would not have sunk
if the counter-flooding had not been done.
The ship was already about 50% over design stress in the listed position
for a non-corroded, intact hull.
And this hull was badly corroded (in part because of lousy Class practice)
and far from intact.
Moreover, if he could get his engine back,
the Captain could maneuver the ship to reduce wave stress.
This, like the Erika, was a MARPOL 13G(4) induced spill.
Under the modified IMO rules introduced in 1992,
when a pre-Marpol (non-segregated ballast) tanker turned 25,
she had to either go hydrostatically balanced loading (HBL) or segregated ballast.
Most owners opted to go segregated ballast.
For one thing many charterers, such as BP, would not accept HBL ships.
Usually this meant using cargo tanks
that were not designed for seawater (uncoated) as ballast tanks.
Unless these tanks were coated when the ship converted to segregated ballast,
and few were (the 13G(4) regulation did not require it),
rapid corrosion was inevitable.
The Prestige as built had only three segregated ballast tanks:
the forepeak tank and 2A, port and starboard.
Presumably [need confirmation] these three tanks were coated at build and fitted with anodes,
which was the normal practice on pre-Marpol tankers.
When [we need date] the ship converted to segregated ballast per IMO 13G(4),
3P and 3S were designated the new segregated ballast tanks.
This gave the ship the required 30% side coverage.
However, the new segregated ballast tanks were not coated,
and thus rapid wastage was inevitable.
(See also Castor.)
It was this wastage that required 360 tons of new steel in 2001,
despite class's very flexible rules regarding renewals.
And it was this wastage that caused the 3P side plating
and/or the 2A/3P corrugated bulkhead to fail
as soon the weather got bad enough.
The post 13G(4) loading pattern with the 3P and S empty cut the loaded cargo capacity
and put more buoyancy amidships than the ship originally required.
It generated a rather mild overall stress pattern.
But when the 3P flooded, the ship took a big list.
Either this list or the or the fact that the original steel failure
extended forward into 2A caused the original segregated ballast tank 2A
also to flood.
Now the ship, with a 25 degree list, was in trouble,
even if the master had not counterflooded.
For one thing the computed bending moment was above design levels,
but more importantly ABS's use of overall bending moment
with a totally assymetric loading pattern is specious,
and ABS knows it.
As argued above, the max stress in the listed
condtion was not 25% above allowable but close to 50%.
The standard procedure for computing stress levels from
overall bending moment is only approximately accurate
and then only if the ship is at near zero list.
ABS should have used a finite element model to compute
the damaged stress patterns.
Either way the ship needed refuge as quickly as possible.
When that was denied, the loss of the ship becames inevitable.
But the primary cause of this loss was the conversion
of the 3 wings to segregated ballast without proper steel protection,
strongly abetted by the Classification Societies
understanding surveying practices.