According to the CTX spill database,
by far the single most important cause of spills is structural failure,
much of which is caused by ballast tank corrosion.
Ballast tank area is the Achilles heel of the double hull.
Most owners did a lousy job of protecting less than 40,000 square meters
of ballast tank coated area in the old single hull VLCC's.
Now these same organizations are supposed
to properly protect more than 250,000 square meters in a double hull VLCC.
There is no magic coating that will solve this problem.
Using waterborne zinc silicate or a really well-designed solvent free epoxy
might be a substantial help,
but, unless the yards are forced to completely change their coating procedures,
-- a very good idea, by the way --
we will be forced to continue to use "application friendly" coatings
which in longevity are little better
than the coal tar epoxies of 25 years ago -- and in some cases worse.
No paint vendor will guarantee these coatings for more than ten years
and these guarantees are a bit of a joke.
They typically allow progressive breakdown which starts at 2% year at age 1
rising to 5 or 6% at age five.1.
Even 1% breakdown in a salt water tank will quickly lead
to severe localized corrosion and dangerous cracks and holes
between the cargo tanks and the double hull spaces.
There have been numerous reports of double hull tankers
less than five years old requiring massive coating repair.
The best that an owner of a double hull VLCC relying on coating can hope for is
to put off a 15 million dollar reblast and recoat for ten or so years.
The problem for the regulator of course is that most owners
will put off this kind of expenditure too long,
which will generate a series of casualties,
some of which may only involve spillage,
but some of which will involve the loss of a crew.
We have other steel protection tools at our disposal.
We can handle the ballast leg with anodes.
But this has to be done properly
and currently most tanker owners do a putrid job
of maintaining cathodic protection in ballast tank.
The method of choice is a superintendent periodically
inspects a tanks, kicks the anodes,
and pontificates that the anode is or is not still effective.
I don't know how many times I've gone into a tank and
watched somebody kick an anode and write down 30% wasted,
and then the next guy come along and write 60% wasted,
and then I write down something else.
Nearly useless.
Our superintendent may replace a few anodes, the amount depending on his budget
and the overall state of the tanker market and his reading of the owners mind.
He knows it's nearly impossible to justify an unusual anode expense
while there is no problem in justifying "necessary" steel replacement.
2.
The overall result is that the great majority of all ballast tanks
that are more than three or four years old are underprotected cathodically.
Many of them are essentially unprotected.
The proper approach is not anode kicking
but monitoring the tank's polarization.
Whenever a tank is ballasted,
the crew should measure the potentials on a daily basis.
If the tank is not up to 800 mV (against a Al-AlCl half-cell) in four days,
then more anodes should be added.
If this policy is followed, there will be nil corrosion on the ballasted leg
even where the coating is broken down.
This will be evident in the color of the tank.
When you go in an anoded tank, you want to see white, all white;
no red or brown where the coating is broken down.
The white is a calcium precipitate
which occurs if and only if the steel is acting as a cathode.
If and only if we see white everywhere, we know the tank is properly protected.
The issue here for regulators is
whether or not to enforce proper cathodic protection.
But anodes at best can only solve half the problem; the easy half.
When a ballast tank is empty,
we have ideal conditions for atmospheric corrosion: lots of moisture, lots of salt.
This is why uncoated SBT tanks don't work (witness Erika, Prestige, Castor, etc, etc),
why localized corrosion is so rapid in way of coating breakdown in coated tanks,
and why ballast tanks corrode from the top down (while cargo tanks corrode from the bottom up).
In an empty ballast tank, we can't control moisture or salt,
but we can control the third requirement for corrosion: oxygen.
No oxygen; no corrosion.
This suggest inerting the ballast tanks
but this in turns means solving the sulphur problem.
In 1990, Tankship Transport and then Hellespont Shipping
began experimenting with double scrubbing of inert gas.
To our surprise we found that by running two normal scrubbers in series,
we were able to reduce SO2 levels inthe inert gas from 50 ppm
(a good number for a well-operated single scrubber system)
to less than 2 ppm at cargo discharge rates and less than 0.3 ppm at deballast rates.
This is a largely a product of the fact that the extra scrubber cools the gas
to about 3C above ambient as opposed to the 17C plus at the outlet of the first scrubber.
Far more of the water vapor condenses and takes the sulphur with it.
Uncondensed water vapor cannot remove sulphur.
You will get more than 0.3 ppm on a winter day in New York with an inversion.
In fact, to measure these levels,
we had to use instrumentation developed for atmospheric pollution.
In 1993, we installed the first full scale ballast tank inerting system on a ULCC.
This involves treating the ballast tanks just like cargo tanks.
The ballast tanks have their own deck seal, P/V breaker, P/V valves etc.
We intentionally damaged 240 spots on the coating, measured the thickness in those spots,
and began inerting.
In the next seven years, we were not been able to measure any wastage
in way of the damaged spots.
Moreover, we have noted that where there had been existing rust in the tank
it has been converted from Fe3O4
back to black magnetite Fe3O2.
In short, we had created reducing conditions in the tank.3.
If you inspect a properly inerted, properly anoded ballast tanks,
in way of the coating breakdown,
you will see only black (magnetite) and white (calcareous deposits), no red or brown.
As long as that it is the case, there will be no loss in steel.
In short, the twin keys to protecting double hull ballast space
are proper anodes and proper inerting.
Inerting ballast tanks not only protects the steel on loaded legs
but also avoids the risk of an explosion
in the event of a leak from the cargo space.
It also allows us to purposely transfer cargo from a damaged tank
into a ballast tank safely.
Finally, an important by product of double scrubbing is that the
gas that is being injected into the cargo tanks is also very low in SO2.
The weak acid etching sometimes observed
in way of cargo tank IG inlets does not occur in these tanks.
However, in inerting the ballast tanks with very clean gas,
a difficult crew safety problem has been created.
This gas is so clean that you cannot see it -- there is no smoke --
you cannot smell it, you cannot taste it.
Yet tanks inerted with this deadly gas must be inspected and maintained.
The double hull ballast spaces are a warren of steel.
Attempting to blow them out using normal methods is a non-starter.
To this end, Hellespont developed a system for thoroughly and quickly
purging and inerting double hull ballast spaces.
The key is to treat the double hull space as a series
of longitudinal ducts formed by the stringers in the double sides,
and the longitudinal girders and bulkheads in the double bottom.
An injection pipe is fitted at one end of the tank.
This pipe runs vertically downward from the deck to the bottom
and then transversely to just inboard of the inboard most
longitudinal girder/bulkhead in the double bottom.
Orifices are drilled into this pipe in each of the ducts.
The idea is that the great bulk of the air/gas from each orifice
will flow longitudinally throught its duct to the other end
of the tank and then exhaust upward and outward.
To this end, openings in the longitudinal members in the double hull spaces
are kept to an absolute mininum
except in the forwardmost and aftmost web frames.
At the same time, openings in the transverse members must be
made large enough to direct most of the flow in each duct longitudinally.
To properly size the injection pipe, the injection pipe orifices,
and the opening in the double bottom structure, Mike Kennedy,
the Technical Director of Hellespont wrote a program called
VENT2D
to model gas flows in double hull ballast spaces.
VENT2D divides the ballast tank space into a number of cells
formed by the webs running tranversely
and the stringers and inner bottom girders longitudinally.
The program takes as input the size of the openings between each cell
and the pressure drop between the inlet and the outlet.
The program can also model an injection pipe
with outlets into one or more of the cells.
The output of VENT2D is the resulting flows into and out of each cell
for a given injection pipe and pressure drop.
This program was successfully used in 1999
in designing both a class of double hull VLCC's and a class of ULCC's.
The program indicated that with proper design
the worst case tanks in a double hull could be safely purged in four or five hours.
Full scale tests proved that the VENT2D results were conservative.
In fact, the worst cell in the worst tank had an O2 content of more than 20% in less than two hours.
The flow is essentially displacement.
These numbers are far better than those
that can be achieved in a conventional box-like tank.
You can actually feel the airflow on your face when you are in the tank.
It's very comforting.
Hellespont and Kennedy have generously made VENT2D available
to one and all under the Gnu Public License.
Notice that this system works when a tank is partially flooded.
This is an obvious requirement
but not true of some of the other systems we've seen.
The goal of this project will be to promulgate these results
both to owners and regulators.
To owners in the hope that it will induce better ballast tank protection.
To regulators in the hope that they will enforce better ballast tank protection.
CTX feels strongly that all double hull ballast spaces must be properly inerted
except when they are bing inspected.
Otherwise, the double hull experiment will turn out
to be a massive debacle as these ships age.
The outputs of the Ballast Tank Protection (BTP) project will be
A model newbuilding specification for double hull ballast tank inerting system.
An operating manual for maintaining both ballast tank cathodic systems
and double bull ballast tank inerting systems.
A better user interface and manual for VENT2D.
Currently, unless your name happens to be Mike Kennedy,
VENT2D is not real easy to use.
Footnotes
1. These solvent-borne epoxy coatings have
a Glass Transition Temperature (GTT) in the range of 5O to 60C.
This is the temperature at which the coating becomes partially plastic.
If an epoxy coating is cycled through the GTT,
it quickly loses its strength and flexibility and ages rapidly.
But the temperature on a red (or worse, green) tanker deck can reach 65C in the tropics.
This is why epoxy ballast tank coating start breaking down from the top down.
2. Most tanker owners these days hire third party ship managers
on the basis of which manager can offer the lowest operating budget and keep to it.
Anodes count against that budget; "non-routine" steel renewal does not.
In additions, most owners do not want to hear about bad news,
so most third party superintendents oblige
and report that the tank is OK regardless.
It is not until the tank gets in such desperate condition
that even Class on one of its five year visits can no longer stomach it
that the owner finds out he has a massive steel repair job.
The other alternative is that we all find out about it a la Erika.
3. We did see some brown and red and some corrosion
in the underdeck area in the very top of the tank
on this older ULCC after inerting.
This was caused by our failure to always maintain low O2
in this area due tank breathing.
There are two solutions to this problem:
Continuously moniter tank inert gas pressure, and top off whenever
that pressure gets below 500 mm WG (at 0200) or 200 mm WG (immediately).
Paint the decks and topsides pure white.
This effectively eliminates tank breathing
while at the same time keeping the tank steel temperatures
in this area below 45C on even the hottest June day in the Persian Gulf.
This is below the Glass Transition Temperature of the epoxy paint
greatly extending the life of the coating.
It also essentially eliminates cargo evaporative losses to the environment
and drastically reduces fuel used for topping off.
Both these measures were implemented on the Hellespont ULCC's
effectively eliminating underdeck corrosion.