The Oil Drum

Don't know about the rocket, but the liquid would instantly gasify on contact with the water.  The water would eventually freeze into an ever-expanding ice floe, and the liquid methane would run out across the floe.  The flowrate through the tanker hole would essentially be gravity driven; the tanker would depressurize quickly.

What happens next depends on the wind speed and direction.  The cold methane will hug the ground; on land it will follow low points like valleys or streets.  As it starts to mix with the surrounding air an opaque cloud will form from the condensation of water vapor.  How fast it mixes depends on the wind turbulence.  You would first get a severe asphyxiation danger from the cloud and then an explosion danger.  If the cloud reaches the explosive limit then something like a automobile distributor could set it off.

The tops of the LNG tanks are protrude about 30% above the deck.  Just shoot high and no hull to slow it down.  The spherical tanks are mostly well inside the hull, only getting near the hull at 3 tangent points, so a hit on a tank after hull impact would need accurate placement on the tangents.  The LNG tank itself is an outer steel shell, about 3 ? feet of expanded perlite type insulation, and an inner steel shell.  Shell thickness is about 50mm (2") steel. No way near like hardened armor.  I'm  no explosive expert, but seems like anything that can get through the first layer <rocket qualifies>, would still have enough momentum to make it the rest of the way, as the inner tank will have high specified Sharpy V-notch toughness, but at low temps could be pretty brittle for a head on impact from hi vel chunks.  I'd think you'd want to keep all the little rubber boats away from these things.    

I have not seen the movie Syriana yet, and I am a little dubious of Hollywood's representation of weaponry.  (These are the same guys who are trying to figure out how many shots are in a six-shooter.)

That being said, a SRAW (or Short Range Assault Weapon) would be an excellent choice, but a bit of an overkill (and kind of expensive given the price of RPG-7s).  The SRAW is American and RPG-7s are cheap, cheap, cheap.

Most hand-held rockets of this type use a shaped charged warhead.  If you think of the pointy end of an RPG rocket, this is actually a hollow, copper cone.  The explosive behind it is also hollowed out, so there is a cavity.  When the warhead explodes, this crushes the copper into a bit of plasma and shoots it forward (This is called the Monroe effect for those still reading).

The average RPG-7 will burn through about 330 mm of steel.  The good news is, it will burn through about 330mm of air too once the round explodes.  A common defense is to put a grate or something up to make the round explode before it hits the tank.  The latest upgrades to the US M-1s have a louver across the back to protect the engine.  So long as you have a foot or so for the plasma to burn up, you are good.  Reactive armor works in a similar fashion by using an explosion to disrupt the plasma jet.  (Friendly infantry don't care for this though).

The bad news is, a rocket can have 'stacked' warheads, so one explodes, disrupts the louvers, and the second one goes through.  Another solution is using multiple rockets.

Would an anti-armor rocket work against a LNG tank?  I don't know.  The insulation between the two tanks would certainly disrupt the plasma jet.  However, the explosion will make one hell of a hot spot on the tank and surely crack the tank.  I just know I don't want to be there when it happens.