In 1996, Shell’s success turned sour. The company struck a succession of dry holes in the Gulf, as did their rivals at BP, Texaco and Amoco. After the seventh dry well, everyone stopped. Exxon’s explorers congratulated themselves for their refusal to risk millions of dollars just as oil prices were falling, and for waiting until others had neutralised the hazards. The failures coincided with the US government’s announcement of a second auction of leases for deep-water exploration in the Gulf. Shell’s breakthrough should have triggered a boom to buy new leases, but the rash of dry wells caused head-scratching across Houston.
The explorers gradually realised that a mile-thick layer of salt beneath the sea bed, below the silt that had poured out of the Mississippi river and above the oil-bearing rocks, was causing scientific mayhem. Finding oil relies on plotting formations of rock created up to 60 million years ago. Based on a century’s experience, geologists know which rocks are likely to contain oil. Their knowledge guarantees some predictability in the Middle Eastern deserts, the Siberian tundra and the North Sea. In those areas, the question was not whether oil would be found, but whether the quantity was sufficient to make its exploitation commercially viable.
Identifying rock formations 70,000 feet below the Gulf’s surface was technically feasible. Ships dragging seismic equipment were regularly criss-crossing the Gulf, firing sound bops to the sea bed and, every millisecond, recording the pattern of echoes zooming back from below. Old-timers recalled watching pallets of magnetic tape of seismic data being unloaded by forklift trucks: processing them through nine-track computers took three months. Twenty years later, all that information could be stored on an iPod and analysed by computer within two hours. But either way, the results in the Gulf were notoriously inaccurate. As the seismic soundwaves passed through the salt, the ricocheting bops from the rock strata were grossly warped. ‘Recording the sound through salt,’ Rainey realised, ‘is like photographing through frosted glass. The image and the sound is distorted.’ Identifying the location of oil through salt was impossible. Shell’s early successes had been due to nothing more than luck. ‘Don’t worry,’ David Jenkins, BP’s head of technology, assured John Browne. ‘You’ll find more Mars-like oilfields once we can see through the salt.’
Texaco and Amoco had developed computer programmes to show two-dimensional images of rocks, slightly reducing the risk of dry holes. During the 1990s the experts predicted that 3D, and even 4D, images would further reduce the risk but only drilling produced conclusive evidence. On the grapevine, BP’s executives heard Shell’s boasts about its success with Chevron at the Perdito field in the Gulf, which it claimed was the result of superior seismic processing. ‘It’s a strong indicator of our success,’ said Dean Malouta. Rainey was dismissive about Shell’s reliance on seismic evidence rather than ‘human experts’. In wild frontier areas, Rainey believed in geology. He could cure the salt problem, but the cost would be $100 million. BP could not commission any trials unless a rival corporation agreed to share the expenditure.
At the time, BP was a junior partner with Exxon in unsuccessfully exploring a block in the Gulf called Mickey. Faced with poor seismic images, Rainey tried to persuade BP’s richer associate to finance more expensive tests. The latest computers producing three-dimensional images of the rocks were being fed seismic data recorded by ships travelling half a mile apart. BP had financed the development of software using seismic echoes recorded from cables just 12 metres apart, considerably improving the 3D image. But gathering raw data across a 300-square-mile block would be hugely expensive. ‘We need to go back from geophysics to geology,’ Rainey explained to Exxon’s geologists. ‘We need to put everything back in its proper place.’ Renowned for their technical excellence, Exxon’s executives are also infamous for believing that anything not invented by Exxon is certainly wrong. Ideas offered by an enfeebled, recently denationalised British operator were thus automatically suspect. Unlike BP, Exxon had focused on finding oil in West Africa, especially Angola, and with its enormous spread of interests the corporation lacked the financial imperative to find oil in the Gulf of Mexico. However, Exxon’s technicians were eventually convinced to finance the experiment, and Rainey’s idea was proven to be correct. Other oil companies were spurred to adopt the enhanced seismic measurements, reducing the cost for BP.
By itself, the intense mapping of rocks was worthless. Identifying the location of oil depended upon producing accurate geological maps. The oil companies raced to recruit mathematicians and geophysicists to compose computer programmes based on algorithms to rectify the seismic data. Rainey’s challenge was to recruit better mathematicians than his rivals, especially Amoco, the masters in this field. The breakthrough coincided with BP leasing a nine-square-mile block called Mississippi Canyon 778 off Louisiana, recently abandoned by Conoco after a succession of dry wells.
The opportunity to buy the block arose after Conoco had failed to find oil at Milne Point in Alaska. As oil prices slid, the company needed to cut its losses, and BP agreed to trade Milne Point for acreage in the Gulf of Mexico. Nonchalantly, the BP negotiator said, ‘There’s a value gap in the deal. We’ll agree if you throw in Block 778.’ Conoco’s negotiator was happy to oblige. Conoco, the BP team believed, had committed a cardinal error by misreading the geology at an unexplored depth. Concealing BP’s calculations from its rivals across town, Rainey was confident of success, even though the whole Mississippi Canyon area covered 5,000 square miles.
‘Everyone in the Gulf is making the same mistake,’ Rainey said in 1996. ‘The model’s wrong. We’re focusing on the geophysics.’ Rainey was convinced that his unique understanding of the Gulf would enable him to pinpoint a reservoir: ‘Shell and Chevron are fixated by seismic tests. They’re too rigid. They’re forgetting about the geology.’ While Alaska’s rocks had taken three years to master, the complications in the Gulf took 40 years to understand. ‘Everyone in the Gulf is focused on “top down”, relying only on the seismic and forgetting the rocks! It should be “bottom up”.’ Rainey insisted that BP’s rivals were looking at seismic images corrected by computers, and not at the rocks themselves. In their quest to find the rocks which 10 to 20 million years ago had heated up and generated oil, they had ignored the key factor: less dense than rock, oil attempts to escape. ‘The deeper I go, I can see the traps, but I can’t see the hydrocarbons,’ said Rainey. ‘We need to find the plumbing’ – shorthand for the ‘migration pathway’ where the oil had flowed and become trapped.
Peering at the 3D images generated by the computers in the HIVE, Rainey reminded his team: ‘The Gulf is the most complex area on the planet. You’ve got to stay humble because you can never crack the Gulf. Just as you think you have mastered it, some rocks come up and kick you in the backside. Science is helpful but in the end success depends on human understanding.’ The team debated whether the white columns spiralling out of the rocks on the screen were salt or sand. If they were sand, the oil would have leaked away and a $100 million test drill would be wasted. ‘Follow the salt,’ Rainey urged. The salt was an obstacle, but also an asset. The secret was to find a lump or hill rising within the rock: that would be the trap where the oil would gather, unable to leak out, sealed by the impenetrable salt. ‘I need people who think like a molecule of oil – where will it go into the rock?’ said Rainey. In his efforts to resolve the problem he had abolished the demarcation between geologists and geophysicists. Working together, they could determine whether the rocks had ever contained oil and whether the oil was still trapped. Like the pioneers