Inviting Disaster

James R. Chiles

• Publisher: HarperCollins e-books
• Year: 2008
• ISBN:
  9780061734588
• Topic: Design
• Subtopic: Industrial Design

Combining captivating storytelling with eye-opening findings, Inviting Disaster delves inside some of history's worst catastrophes in order to show how increasingly "smart" systems leave us wide open to human tragedy.

Weaving a dramatic narrative that explains how breakdowns in these systems result in such disasters as the chain reaction crash of the Air France Concorde to the meltdown at the Chernobyl Nuclear Power Station, Chiles vividly demonstrates how the battle between man and machine may be escalating beyond manageable limits -- and why we all have a stake in its outcome.

Included in this edition is a special introduction providing a behind-the-scenes look at the World Trade Center catastrophe. Combining firsthand accounts of employees' escapes with an in-depth look at the structural reasons behind the towers' collapse, Chiles addresses the question, Were the towers "two tall heroes" or structures with a fatal flaw?

1: SHOCK WAVE

HIGH TECH ON THE HIGH SEAS

On the morning of February 14, 1982, the crew of the drill rig Ocean Ranger began squaring away for a storm predicted to arrive later that day. The men weren’t worried; the Ranger had been through at least fifty storms in six years of offshore work. It was the biggest floating drill rig in the world when built six years before and was designed to survive hurricane weather of 110-foot waves. Weather reports for the Grand Banks put the coming North Atlantic gale at well below the design limits of 100-knot winds, so no extraordinary measures were called for.

Take a rectangular end table and glue a toy oil well on top; now nail on two additional table legs on two of the longer sides, making a total of eight legs; now glue the legs on top of little pontoons and float the table on a lake with the pontoons sitting just underwater and the rest of the table rising above the surface (see Figure 2). That was the general layout of the gunmetal-gray Ocean Ranger. The flat top of the table, the double-decked “drilling platform” on board the rig, was home and workplace for eighty-four men. The drilling crew worked in alternating twelve-hour shifts that changed at noon and midnight. They worked seven days a week for two to three weeks, then rotated to shore via helicopter for a paid leave.

FIGURE 2: OCEAN RANGER

PROBABLE SEQUENCE

• 1. Wave breaks out “portlight” window to ballast control room, dousing electronics with seawater.
• 2. Short-circuited valves open and close in ballast tanks.
• 3. Crew shuts off power, later restores power.
• 4. Opened valves allow seawater ballast to shift to bow.
• 5. Crew attempts to empty bow tanks with suction pumps, but rig goes further off balance.
• 6. Rig’s list allows storm waves to reach openings into chain lockers in bow.

Adapted from Royal Commission report

Held in position by three strong anchor cables at each corner, the fourteen-thousand-ton Ranger took the waves so easily that on a calm day a visitor might be fooled into thinking that its corner legs ran all the way down to rest on the seafloor, 260 feet below, instead of riding on pontoons. Most of the time the decks showed a slow, shallow roll that was barely enough to affect games on the rec-room pool tables. A rig superintendent once called it unsinkable.

Viewed from over the steel railings, the waters below didn’t look like much of a menace to men living so high above the water. But the appearance was deceptive, and the men of the Ocean Ranger had fresh proof of that. Compared to the twenty-seven years of the offshore oil industry, major rig mishaps had been hitting historic highs in recent years, with twenty-two rigs reporting fires, blowouts, capsizes, or sinkings in 1980 alone. The reason, said industry sources, was the worldwide acceleration in drilling activity. Although shipyards could build rigs quickly enough, finding expert crews to man them was a major difficulty.

In that troubled year of 1980, the Alexander Keilland, a floating dormitory for North Sea oil workers, had rolled over in a winter storm, killing 123 men. It happened after a cracked strut in the steel framework finally broke all the way through under the constant pounding of waves. A shipyard painter had painted over it during construction (known because some of the crack length contained paint) but apparently had not reported it. The strut’s failure allowed rough seas to break off one of the Keilland’s five legs, throwing it out of balance. The Keilland took twenty minutes to roll over, spilling the men into the frigid sea or trapping them below.

The Ranger had fifty-four Newfoundlanders aboard, a nod to Canadian regulations that encouraged local hires. Two of them were Stephen and Robert Winsor, brothers aged eighteen and twenty-three. A mishap on board the Ranger the previous summer had crushed Stephen’s hand, but he went back to the Ranger as soon as he recovered. Though the brothers had started as roustabouts without any oil field experience, the rig operator offered pay much better than any land job open to them. An older brother, Gordon, worked on board the SEDCO 706, a smaller floating rig stationed nine miles east. The Ocean Ranger, the SEDCO 706, and a third rig, the Zapata Ugland, were all within sight of each other.

According to the schedule made out for February 14, evening was Gordon’s time to rest, but the rolling of the SEDCO rig in the rising storm kept him awake in his bunk. In almost six years on offshore rigs, it was the worst North Atlantic storm he’d ever experienced. At 7:00 P.M. he got up to walk around. At about that time the rig jolted heavily with the impact of an oversize wave striking the steel legs. Two more big waves followed immediately.

Outside, seawater was “shipping green” over the upper deck, meaning it was so deep that no foam showed. Water even flowed down ventilator shafts into the engine room, so the wave crests must have topped eighty feet. Pumps started dumping ballast water out of the pontoons and the SEDCO 706’s drilling platform slowly rose another fifteen feet above sea level. This extra height would make the drilling platform less vulnerable to wave damage; already the waves had bent thick steel I-beams.

Gordon joined his shift on the derrick at midnight. Soon he heard through friends that the Ranger, where his brothers were working, was transmitting distress calls over the Telex machine. He hustled to his rig’s radar room. Sometime around 3:00 A.M., as Gordon watched, the blip that was the Ranger disappeared from the SEDCO 706 ’s radar screen. He asked for sedatives.

When the first reports went out that the state-of-the-art Ranger had overturned and sunk in relatively shallow waters, maritime experts were baffled. One theory was that it had succumbed to a strange sort of undersea blowout called a “gas-loaded ocean,” where the seafloor would erupt with a geyser of gas that so foamed the water under a rig that the vessel would lose buoyancy and sink. Another theory was that a rogue wave had taken the Ranger down, the kind of killer wave that arises out of nowhere, towering ninety feet high or more, the sort that had bashed in the bridge of the Queen Mary in 1942, seven hundred miles off the coast of Scotland. Rogue waves are a known hazard off the east coast of South Africa.

In fact, one of the waves that had rocked Gordon Winsor seven hours earlier had done in the Ranger, though it was no rogue wave. Divers would find that the wave, by breaking a simple sheet of glass and throwing hundreds of pounds of seawater into the ballast room, had tripped off a long sequence of malfunctions and errors that killed the entire crew.

The Ranger disaster lays a good foundation for the subject of clashes between man and machine because it shows how failure ripples through a system like a slow-motion shock wave. Consider how a person can lose his or her emotional equilibrium for a full day or more after an argument or a minor car crash. Similarly, the crew of the Ranger lost its poise that night and never regained it.

While following the stormy saga of the Ocean Ranger and other conflicts between machines and men, we will look for lessons beyond the obvious errors pointed out by the official boards of inquiry. Observe the fixation on daily routines at these sites right up until the culminating event, without any serious planning to survive an emergency that was well within the bounds of foreseeability.

STATE OF THE ART

The number 2 Eba shipyard of Mitsubishi Heavy Industries finished the Ranger in 1976 according to plans drawn up by New Orleans–based Ocean Drilling and Exploration Company (ODECO). ODECO was a seasoned offshore operator. Its owner, Murphy Oil Company, had built one of the first transportable, submersible platform rigs, called Mr. Charlie, back in 1953.

Since 1976, the Ranger had drilled off the coasts of Alaska, New Jersey, and Spain before parking in Canada’s offshore Hibernia Field, where Mobil Oil of Canada leased it to put down test holes. Three floating rigs worked the Hibernia Field at the time. This was all exploratory work and very expensive, but in 1982 the oil companies and the Newfoundland government hoped that mighty shiploads of oil would start flowing out of Hibernia within five years.

Registration papers filed with the Coast Guard listed the Ocean Ranger as a “mobile offshore drilling unit.” A MODU like the Ranger was paradoxically both more stable and less stable than your typical ship. It was more stable because its deep draft, long legs, and multiple anchors allowed it to ride above big waves more smoothly than any ship, which pitches, heaves, and rolls in heavy weather. But the Ocean Ranger was less stable than a ship if it went out of balance, because it was top-heavy. The Ranger was 151 feet high from decks to keel, not counting the derrick. Including the derrick, more of the rig’s structure was above the water than below it. Furthermore, shifting heavy loads around on the platform could throw the ship seriously out of balance.

Begin from the bottom up. The Ranger sat on two giant submarine-size pontoons, each 400 feet long and set 240 feet apart, like catamaran hulls. Inside each pontoon, steel walls divided the space into tanks that held ballast, fuel, and fresh water. When the rig was cruising from place to place the pontoons rode up on the surface, the rig moving under the power of diesel-electric motors and thrusters at the stern. The Ranger could and did cross oceans on its own power. When the rig arrived on a station, the crew opened sea-chest valves and let seawater into pontoon tanks as ballast so the pontoons would submerge to about eighty feet below the surface. This made the rig less top-heavy. Waves had little effect when the Ranger sat so deep in the water, doing its drilling work. This depth was called the “drilling draft,” and it put the decks fifty feet over the ocean’s surface, enough for high waves to pass underneath. If a big storm came up, with higher waves possible, the crew pumped out enough ballast to lighten the rig and bring the superstructure another ten feet out of the water. Called the “survival draft,” this configuration gave the platform more clearance against damage from giant waves.

Welded on top of the pontoons were eight massive steel legs, rising vertically to support the drilling platform, like the piers of a bridge. The platform consisted of two horizontal decks. The platform was home to the crew and the base of all drilling operations. For an emergency evacuation, the platform held three working lifeboats, ten life rafts, and a helipad.

The support legs were so roomy inside that the designers also used them for storage and working compartments. This economical use of space would contribute to the Ranger’s demise in several ways: it put a critical electronic control room within wave-splashing range of the ocean, and it made holes in the legs that could allow them to flood in a storm.

The legs placed at the corners of the platform were massive, each almost forty feet in diameter. Each corner leg housed a tall compartment called a chain locker, which had enough space to store thousands of feet of wire rope and anchor chain when the rig was moving around and not secured to the seafloor. At the top of each corner leg, seventy feet above sea level, were three big holes for the wire ropes and chains to come aboard. This was a serious but unrecognized problem. Each hole into the chain locker was a gaping five feet across. Nobody who designed the Ranger thought to provide a method of shutting those openings when the rig was on station, nor did they think it necessary to provide some kind of warning to the crew if the chain lockers began to fill with seawater through the holes. Flooding of the chain lockers through the holes wouldn’t happen unless the rig was drastically tilted in bad weather, and apparently everyone in charge thought this impossible.

Although the pontoons’ machinery compartments could be reached via ladder ways from the surface, normally no one worked down in those dark spaces. One of the smaller, middle legs on the starboard side held the ballast control room. Seated comfortably in a round, metal-walled control room twenty-seven feet above the water’s surface, the ballast operator could fill or empty all the ballast tanks by remote control. Dials and lights on a display board told him what was happening down in the pontoons.

On this rig, ballast operators also needed to see the painted draft marks on the legs and what was happening with supply boats outside, so the control room had four round glass windows, called “port lights,” set into the steel walls of the leg. These port lights, each eighteen inches across, could not be opened, but they were only made of glass and could break under stress. Part of the ballast operator’s duties involved pushing red and green buttons that operated electric solenoid switches. The solenoids controlled a flow of compressed air that ran in copper pipes, down the legs, to the seawater valves in the pontoons. Powered by compressed air, those big seawater valves actually did the work of connecting the pumps in the stern, through pipes, to the tanks spaced along the pontoons. There were sixteen tanks in each pontoon and only three pumps, so to work the ballast system properly, the operator had to set the valves to link the proper pump with the proper tank. When connected through the valves and turned on, the pumps could empty ballast tanks by sucking seawater out of the tanks and dumping it into the ocean. They could also shift water from one tank to another. These pumps were “powerful and capable,” an inquiry board said later. But unbeknownst to the operators they had peculiarities that would emerge in mysterious fashion when things began going wrong.

Though the rig seemed stable with its pontoons submerged and all twelve anchor cables rigged out to the seafloor, keeping it level was a balancing act because weight and balance shifted constantly with the busy pace of offshore work and resupply efforts. The Ranger could hold almost four thousand tons of supplies and drilling gear, but even that was not enough for weeks of continuous work, so work boats resupplied it every few days. It needed regular deliveries of drilling mud mix, diesel fuel, tools, fresh water, and supplies for the crew. Even between supply runs, the rig could lean off balance as the crew drained tanks and shifted thick steel pipe from the storage rack to the drill string. If a ballast operator failed to make timely corrections to the ballast tanks as new loads came aboard or as the roustabouts worked, the rig’s angle would go off center. Drilling work might have to stop, or something worse could happen. Even five degrees off plumb was a serious matter.

One demonstration that the Ocean Ranger was dangerously vulnerable to ballasting mistakes came just a week before the disaster. On February 6, as a supply ship pumped fresh water through a hose to the rig, the ballast operator on duty, Bruce Porter, wasn’t in the control room. He was down in one of the pontoons working on a mechanical problem. Trying to help out in Porter’s absence, the rig’s master, Clarence Hauss, sat down at the controls and attempted to correct a slight tilt caused by the water-loading operation. But Hauss unknowingly left one of the inlet valves open to the sea. Tons of unwanted water flooded into two port ballast tanks. The rig leaned almos...