The Economist has published an excellent article (link below) on the subject of deep-sea drilling. The first offshore oil well completely out of sight of land was drilled in 1947, 17 km off the Louisiana coast in the Gulf of Mexico. The platform of the drilling rig was no larger than a tennis court and was supplemented by several refurbished naval barges remaining from WWII which provided both space for storage and for the crew to sleep. The seabed was a mere 15 feet below, but now far greater depths of water are being fathomed.
In 2008, Shell's 22,000 tonne Perdido spar was towed from Finland where it was constructed to a spot 320 km off the coast of Texas, where it was chained to the seabed 2,400 metres below. This is connected to subsea wells in yet deeper water at 2,900 metres. Such offshore drilling operations are attended with daunting physical challenges. Since a longer and heavier drill string is required in deeper water, the supporting platform must be more heavily engineered. The interlocking sections of pipe are heavy, at around 30 kg/metre.
Since conventional onshore oil is held in countries that are unstable in regard to their political amenability to the West, offshore drilling is forcing private companies to look for oil further afield, in more inhospitable locations, and increasingly in deep water. The pressure of water increases by one atmosphere for each 10 meters of depth, and at almost 3 km, is close to 300 times normal atmospheric pressure, or around 4,500 pounds (2 tonnes) per square inch. The consequent pressure on the seafloor makes it more difficult to pump the oil up to the surface.
Hitting the right spot is a feat in its own right, and Robin Walker, of the oil-services company WesternGeco, uses the following analogy. "Imagine a large offshore oil rig as a matchbox. Next, imagine the matchbox on top of a two-storey building, with the upper floor filled with water and the lower floor filled with rock, sand and, in some cases, salt. Striking an oil reservoir with a drill pipe is then like hitting a coin at the base of the building with a strand of human hair." If the hit is in the wrong place, the costs are enormous, and an industry rule of thumb is that drilling a deepwater "dry hole" - a well with no oil - is around $100 million. BP says it can be as much as $200 million.
To avoid any errors, complex geophysical measurements are necessary including multi-dimensional seismic imaging, where a rough 3-D image is created of the subsurface rocks. Despite the difficulties, a number off major deepwater finds have been identified recently: Tupi, off the coast of Rio de Janieiro is thought to hold 8 billion barrels of oil, but this lies underneath 2,000 metres of water, 3,000 metres of sand and rock and a 2,000 metre layer of salt. Other "ultra-deepwater" discoveries, defined as those under 1,500 metres or more of water, have been made off the coasts of Angola, Sierra Leone and Nigeria, along with several more in the Gulf of Mexico.
Obtaining images of commingled salt and rock poses difficulties because the waves emitted from seismic sources travel faster through salt than in rock. When there is a mixture of reflected and refracted waves present, constructing an image of the subsurface from a normal sonar survey is not readily accomplished. Thus, rather than collecting seismic data in two dimensions, using streamers and them using computational processing to get a 3-D image, an actual 3-D data acquisition was done, using hydrophones and multiple seismic sources from three of four vessels moving in parallel, called a "wide-azimuth" survey. The accuracy can be honed further by passing over the same region a number of times from different angles ("multi-azimuth" survey).
The greatest challenge, however, is processing the data. When the surveys indicate there is a high probability that oil is down there, an exploratory well is drilled. This involves pumping a liquid called "mud" through the drill string to remove borehole cuttings and to cool the drill-bit and maintain pressure at the base of the well. As the drill cuts through the rock and sand under the seabed, the pressure of the "mud" in the drill must be kept within defined limits: if it is too low, the pressure of underground fluids and gases ("pore pressure") on the well wall will drive it to collapse, but if it is too high, the mud can accentuate and expand existing fractures in the surrounding rock which causes a loss in circulation as the mud leaks out into the newly formed fissures.
This is a simple overview, but all in all, the discovery and successful production of oil from these reservoirs offshore will depend on continual advances in technology and computation.
"Plumbing the Depths." http://www.economist.com/science-technology/technology-quarterly/displaystory.cfm?story_id=15582301