This is an exciting project to develop a tanker which will be far superior to double bottoms in reducing spill volume in a grounding. The system can used to replace the dangerous double bottom; or if that proves politically infeasible, incorporated in double bottom ships. The concept is:
  1. Beef up the top of tank structure to take a full vacuum, about 10 M water under-pressure.
  2. Attached a vacuum pump or ejector to the IGS piping, exhausting up the stack. Keep the IGS piping very close to the deck. After loading pull all cargo tanks down to about the vapor pressure of the cargo, e.g. about -.7 bar or -7 meters water gage for Arab Light.
This will put the Neutral Level at about the main deck. 1 In other words, there will be no hydrostatic outflow from any damage that does not break the vacuum in the top of the tank. The resulting ship is equal to the Mid-Depth Bulkhead as far as side damage is concerned and superior to the Double Bottom because the side tanks are wider. 2

The resulting ship is far superior to the Double Bottom in bottom damage:
  1. Unlike the Double Bottom it doesn't depend on the bottom damage not penetrating the inner hull which as a practical matter will happen in any major grounding. The double bottom is only two to three meters deep and the connecting structure nearly guarantees that any substantial damage to the outer hull will involve the inner hull.
  2. Unlike the Mid-Depth it doesn't depend on the bottom damage not penetrating the Mid-Depth Bulkhead. This of course in extremely unlikely unless the ship is totally lost in which case all systems are the same.
Much more importantly, unlike the Mid-Depth or the Double Bottom, bottom damaged tanks will automatically pump themselves out to undamaged tanks. This transfer will begin immediately upon damage with absolutely no intervention from anybody. Please re-read the last two sentences. The increase in pressure at the bottom of the damaged tank will push cargo thru the IG/vacuum lines to the other tanks. 3 On a typical big V carrying Arabian light (density 0.85 and vapor pressure 0.25 bar absolute), the seawater will push an amazing 23 meters into the tank, i.e. just about to the loaded waterline. Assuming there is sufficient volume elsewhere in the system and the vessel draft doesn't change, the equation for equilibrium is
Draft * 1.02 = (Vapor pressure Suction Head) + (Height - x) * SG + 1.02 x.
where vapor pressure is in meters gage which will be negative, e.g for Arabian light, -10*(1 - 0.25), Height is the height of the cross over to the other tanks above keel (about 30 M for a U), SG is the density of the cargo, and x is the equilibrium oil/water interface above keel. In reality the vessel will sink as it floods which will help matters assuming it is not overdone. Obviously, it's essential that the damage doesn't break the vacuum, i.e., the damage is below the waterline. I've also assumed the vacuum doesn't decrease which requires either a lot of unfilled volume or the vacuum pump being still operational.

Not only does this automatic redistribution of cargo from the damaged tank(s) eliminate the dynamic spillage argument (Live Bottoms just above the bottom will generate some spillage from waves, current and vessel motion); but it will have a tremendous impact on the Exchange Flow that would have occured in either the Mid-Depth or Double Bottom from side damage. It also goes a long way toward handling the one extremely unlikely situation where the Double Bottom beats the Mid-Depth Bulkhead: a grounding at high tide where there's lots of tide which grounding penetrates only the outer hull of a double hull.

The Full Vacuum Tanker may be cheaper to build than either the Mid-Depth or the Double Bottom. The mid-depth bulkhead on a VLCC is about 16 M above the keel in a 26 or 27 M depth vessel. Therefore, the mid-depth bulkhead already has to be built to take a 10 or 12 meter pressure differential. Basically, what we are doing is moving the mid-depth bulkhead up to the main deck, eliminating the structure that is already there. The result is a more efficient structure than either the Mid-Depth or the Double Bottom. The Double Bottom results in a very unbalanced structure due to the lowering of the neutral axis. In the Mid-Depth Bulkhead all the mid-depth structure contributes nearly nothing to section modulus because it is very close to the neutral axis. For the same steel, we will have a stiffer ship. And the pressure/vacuum valves will be eliminated.

The Full Vacuum Tanker will be drastically superior to the Double Bottom and significantly superior to the Mid-Depth from the point of maintainability and safety. Ballast tank coated area will be less than one-third that of the Double Bottom. Like the Mid-Depth, the nightmare of maintaining and inspecting the inner bottom space is eliminated. We also eliminate the very substantial complexities associated with inspecting and COW-ing the lower tank in the Mid-Depth.

The resulting ship is even superior to existing single hulls from an operational point of view. It's important to recognize that this vacuum system does not depend on monitoring and controlling the amount of vacuum to avoid overstressing the structure. 4 Since the structure can take zero pressure, there's no way of pulling too much vacuum. Even existing old ships suffer from being very sensitive to overfilling or a P/V valve failing closed. There is simply no way that this ship can be damaged by underpressure (i.e. P/V valve fails during discharge) and it is much less sensitive to overfilling (the overfilled cargo simply goes to the wrong tank). Moreover, since there will be much less tank breathing (normally none), by leading the vacuum pump exhaust up the stack, we can create the same kind of deck we have on the LPG ships, a totally closed system. This means we can have pumps on deck, electric valve actuators, etc, etc. 5

There will be at least five objections to the Full Vacuum Tanker:
  1. A leak in the vacuum piping could expose the tanks to an explosive atmosphere. It will take a big leak to move us out of the too-rich zone but, at a minimum, this will require careful monitoring. At a maximum, it will require a double pipe system in which the annulus is inerted. Running some of the piping inside the tanks is another possibility. One thought is to inert the ballast tanks and then run the vacuum piping inside the ballast tanks. This is a legitimate concern and we need careful, conservative design here.
  2. Damaged stability could be compromised by flow thru the IG/vacuum lines to the low side especially in the case of raking damage. This also needs detailed study but one obvious, if inelegant, fix is list actuated one-way valves. At worst, it implies slightly larger or more compartmentalized ballast tanks to meet the same floodable length requirement. If the IG lines are the same height as they are now, the situation will be no different than what we have now.
  3. The system depends critically on the density and vapor pressure of the cargo. High vapor pressure and high density hurt. For a U in which the IG/vacuum line is 8 meters above sea level, the worst cargos I've looked at so far are Saharan Blend (very high vapor pressure but light) and Mayan (heavy with a surprisingly high vapor pressure). In both these cases, equilibrium occurs after the water has pushed about 5 meters into the tank. Of course, this is already much better than Double Bottom or Mid-Depth, and the situation can be improved drastically by lowering the IG line slightly at a cost of possibly exacerbating objection (2). 6
  4. Sooner or later during salvage you will to break the vacuum. True, but by then you can have salvage pumps in place ready to go, containment and collection equipment deployed, etc. You also will have had a chance to plug the damage in a zero or negative pressure differential situation. Finally, if the system really caught on, devising a salvage pump system that doesn't break the vacuum would not be difficult.
  5. The most difficult point for self-transfer will be at the beginning of loading and at the end of discharge, where the ship is high in the water. This issue too will require carefull study and may force additional ballast capability. On the other hand, few major spills occur at the begining of loading and at the end of discharge. The Full Vacuum Tanker functions best when the ship is loaded, which is exactly when you want it to.
  6. In reality, the environmentalists' primary objection will probably be it depends on the owners not overloading the tanks. This shows that they do not understand the commercial realities of tanker operation. The amount of cargo is carefully checked and thoroughly documented both at load port and discharge port by the owner, the charterer, the receiver and the terminal. Big dollars are at stake. This involves not only measuring the volume in each tank to (an attempted) five significant figures by an independent surveyor which measurement is witnessed by all the parties involved; but also comparing the results with those of the discharging or receiving terminal Some of these parties want the numbers high; some want them low. So there are natural checks and balances. Any mistakes/discrepancies cause at least one of these parties real money in either cargo revenue or freight. There is no way an owner can "over-load" a cargo tank without it being caught at both ends of the voyage. And of course, for those who cannot understand these realities, we could easily implement sealed continuous, tank level recorders. Same thing to document vacuum.
This project will undertake a number of design studies relating to the Full Vacuum Tanker.
  1. First and foremost, we have to determine how much extra steel will be required by the requirement to take a 10 meter under-pressure. Once the CTX has an operational hull finite element capability per the HULL project. This will by a fairly straightforward exercise.
  2. Assuming the results are favorable, then the project will have to address some of the above concerns. The stability issues can be analyzed by CTX_MATE. with possibly some modest modifications. The same thing is true of the start/load, end/discharge question
  3. We will also have to size and design the vacuum system and address the issue of how close to the cargo's vapor pressure we can actually achieve and maintain. This will almost certainly involve some kind of system for recovering and reinjecting the petroleum in the vacuum system, using the reliquification technology used on refrigerated LPG ships. Several big tankers already have vacuum recovery devices fitted. 7

    We will also have to confront the issue of where will we find room for the transferred cargo. One obvious place is the ballast tanks, but this raises a number of safety and regulatory issues which will have to be addressed.
  4. Finally, we will have to address the safety problems associated with air leaking into the tanks.

Email about this project should be sent to fvt@c4tx.org.


1. The use of vacuum for spill reduction is not a new idea. It goes back at least to Stenstrom in the mid 1980's. See Measures on Board Ships to Minimize the Escape of Pollutants, IMO MEPC 26/18, 1988-06-29. Perhaps the variant on which the most work has been done is the American Under-Pressure System (AUPS). Both AUPS and the Swedish proposal were aimed at being retro-fitted to existing ships with little or no structural modification. Therefore a relatively mild vacuum of -0.15 to -0.20 bar was used. The AUPS system was successfully tested on a modified US Navy oiler in 2001. But the USCG refuses to accept it, although in order to do so under OPA90, it had to argue that the system was less effective than a double hull. The only way it could do so was to focus entirely on the probability of no spill occuring, the so-called zero spill probability, completely ignoring the volume of oil released if a spill occurred. Basically, the USCG put itself into a position of defending big spills, arguing that two little spills are more damaging than a single immense spill. The AUPS effort has generated quite a bit of valuable data on crude oil partial pressures. See Crude Oil Under Negative Pressures and Hydrocarbons Emission Containment.

2. There is one awfully unusual case where the Mid-Depth is better. If the side damage penetrates into a cargo tank below the waterline but above the mid-depth bulkhead, this ship will be subject to Exchange Flow from the portion of the cargo tank that would be the lower tank in the Mid-Depth. On the other hand, if vacuum is not broken, there are a number of side damage situations where this ship will have only Exchange Flow while the Mid-Depth will have Hydrostatic.

3. The hull structure will have to be beefy enough so that its longitudinal strength can take the new cargo distribution. With some thought to pipe diameters, this shouldn't be too tough. Right now neither old ships nor new are checked for flooded damage strength,

4. Vacuum will be monitored and every once in a while the vacuum pump will have to be turned on to keep up with leakage. But whether on not this is done, there is no risk of structural failure.

5. The cargo tanks can only be inspected on ballast passages during which the Full Vacuum Ship will operate exactly the way we do now. As a practical matter, cargo tanks are nearly never inspected on loaded passages now.

6. As the above equation shows, the draft plus the vacuum pressure suction head must be greater than the height of the crossover piping times the cargo density. Otherwise, transfer will not occur.

7. An obvious variation on the FVT is to dispense with the vacuum producing system, and rely on the vacuum that will automatically be created when the tank is breached and attempts to empty itself. This is essentially the system originally suggested by Stenstrom in the late 1970's with the important difference that there are no P/V valves, since the hull structure doesn't require any. We will no longer have the self-transfer capability. In fact, each tank will have to be segregated from all the other tanks to prevent self-transfer the wrong way, which complicates inerting. But it is an alternative well worth studying, and can be studied with almost exactly the same tools as the FVT.