Category: Engineering

It’s Friday the 13th, a notorious day for bad luck. Speaking of bad luck, let’s look at the 29 men who died on the Edmund Fitzgerald on November 10, 1975. They’re the subject of an interesting new book: Gales of November by John Bacon. Not exactly a spoiler: everyone dies, just as Gordon Lightfoot explained in his enduring hit song (from the book I learned that Lightfoot was an experienced recreational Great Lakes sailor).

Some things that I didn’t know:

• it was the last trip of what was going to be the retirement year for the captain and many of his long-time partners amongst the crew
• the captain was considered the best on the Great Lakes and had 45 years of experience
• the weather was a 50-year event, perhaps, and not merely an ordinary “gale of November”

Perspective on the suffering of Jeffrey Epstein’s victims

Since almost everything in the news these days must be referenced to the Epstein Files…

The sinking was preventable, as we’ll see below, and quite a few mistakes were made.

Columbia offered some of the families the victims’ last paycheck ($568.25 for a deckhand) and $750 for the victim’s lost personal effects, an amount determined by their labor contract, but other families didn’t receive even those checks. When Columbia offered nothing else, the families had little recourse but to file wrongful death lawsuits, for which Columbia paid whatever the families’ individual attorneys could negotiate. Some families received about $35,000, or a little more than a year’s wages for a deckhand making overtime, and others marginally more.

The typical woman who alleged that she suffered by receiving cash, Gulfstream rides, free rent in Manhattan, luxury vacations, etc., in exchange for sex 10, 20, or 30 years earlier, was paid about $3.5 million from Epstein’s estate, a JP Morgan settlement fund, and a Deutsche Bank settlement fund. That’s 100X what an Edmund Fitzgerald crewmember’s survivors were paid. Thus, we can conclude that having sex on Jeffrey Epstein’s island was 100X worse than riding through 30-50′ waves for hours and ultimately being drowned in 42-degree water.

The bargain all the crewmen on the Fitzgerald had signed up for was a hard one, but straightforward: The work will be taxing, and you will miss most of your family’s best moments, but you will retire relatively young, with a good pension and nothing to do but hunt, fish, and play cards, pool, and golf. Best of all, you’ll have plenty of time and energy to spend with your grandchildren. The Fitzgerald’s twenty-nine crewmen and their families had all paid their deposits up front, but never got to enjoy the sweet side of the deal.

The Special Hazards of the Great Lakes

The Great Lakes can be more treacherous than the oceans. One reason is the distinct structure and frequency of the Great Lakes’ freshwater waves. In the oceans, salt weighs down the water, squashing the waves and spreading them out, so they typically form larger but smoother swells, similar to a roller coaster. On the Great Lakes there’s no salt to hold down the waves, so they rise more sharply and travel closer together, like jagged mountains of water coming at you in rapid succession. These waves don’t roll; they peak, crest, then crash down on whatever is unlucky enough to lie below them.

That’s another reason why Great Lakes waves are so steep and ragged and travel so fast. On the ocean the waves are usually about ten to sixteen seconds apart, so even a large container ship can fit between them. On Lake Superior the waves run four to eight seconds apart, which means that a seven-hundred-foot lake freighter can be riding atop two waves at once.

That problem produces more problems. On the Great Lakes a ship that long can impale its bow in one wave, which can lift it up thirty feet or more, while the ship’s stern can be simultaneously stuck in the wave coming right behind it, raising the ship’s back end in the air another thirty feet. That leaves its midsection, which could be loaded with 58 million pounds of iron ore—the equivalent of 4,200 adult elephants—suspended between the two waves, with nothing supporting it. That creates a phenomenon naval architects call “sagging,” in which the unsupported middle of the ship sags toward the water below it, exerting a tremendous strain on the hull.

After sagging between two waves, just seconds later the ship might face another threat: riding over the peak of a single colossal wave. This creates a condition known as “hogging,” the opposite of sagging, where the vessel drapes over the wave’s crest, with both the bow and stern drooping downward, again placing immense pressure on the center of the ship’s hull.

[Michigan Tech’s Guy] Meadows’s research shows that Lake Superior’s biggest storms occur every thirty years or so, but even in milder storms the waves on the Great Lakes can be alarming. In a pretty unremarkable 2020 storm, for example, two of their buoys anchored off Lake Superior’s southern shore, far from the path of Superior’s biggest waves, measured waves reaching 28.5 feet, almost as high as a three-story building.

If the biggest waves within a few minutes are at 30′, according to Guy Meadows, and they stay there for 24 hours, a ship should experience at least one 60′ wave (1 in 10,000).

Another bad feature of the Great Lakes is that it is easier for freshwater to freeze above the waterline on a ship than it is for saltwater to freeze, thus adding weight to a ship that is already in peril. A typical modern iron boat’s cargo of taconite also creates a hazard:

Because it’s two-thirds clay, a porous material, it can absorb up to 7 percent of its weight in water, and four times that can get trapped between the pellets when it’s piled high.

(i.e., the cargo hold can hold a lot of water weight that isn’t possible to pump out; separately, if you thought that math professors were useless, the book notes that U. Minnesota math prof Edward W. Davis is the person who figured out how to work with taconite, a low-grade iron ore)

Overloading as a Factor

Prior to the wreck of the Fitz, Great Lakes captains prided themselves on moving maximum tonnage per trip and per season. Everyone loaded up the ships to the painted Plimsoll line at which point the freeboard is the minimum required for safety. The Fitz was operated with far less freeboard than her original designers had intended:

Given how such incredibly small margins on the Plimsoll line could produce such prodigious gains, especially when multiplied by forty to fifty round trips each shipping season, the executives at Columbia Transportation must have been thrilled when the American Bureau of Shipping (ABS), working with the Coast Guard, reduced everyone’s required freeboard in 1969, and again in 1971, and again in 1973. For the Fitzgerald, that meant the original requirement of 14 feet, 9.25 inches of freeboard when sailing in November had been reduced to 11 feet, 6 inches—a total drop of 39.25 inches, or more than a yard, in just a few years.

Captains would then use various techniques to add a few inches of cargo beyond even this reduced freeboard, described in a chapter titled “Cheating the Plimsoll Line”.

The Edmund Fitzgerald was loaded to the absolute maximum, and a little beyond, on what was supposed to be her last voyage of the season:

The loading speed was all the more impressive because the dockworkers had filled the Fitzgerald’s belly with 26,112 long tons of taconite—far from the staggering amounts the new thousand—footers were carrying, but a fitting finale to McSorley’s celebrated career. The load that day was almost 4,000 long tons more than the 22,509 the Fitzgerald needed to set the Great Lakes record on its very first run just seventeen years earlier, a testament to the extra 39.25 inches the Coast Guard had allowed the Fitzgerald to sink in the water since 1973—plus the crew’s ability to cheat a few more inches on the Plimsoll line. This was such a common practice it would have been more remarkable if the crew had not cheated on the Plimsoll Line. Even the crew’s families knew about it. “They were overweight because they wanted to break their own record,” says Blaine Wilhelm’s daughter, Heidi Brabon. “So they cheated.” But what was unusual, according to former Fitzgerald deckhand Terry Sullivan, was that she was carrying a full load so late in the season, when even the USCG rules start scaling back the limits. While the Fitzgerald might not have been cheating by much, on the grand scale, when you combine those extra inches, plus the 39.25 inches the Coast Guard had already granted the Fitzgerald two years earlier, and the fact that all ships were supposed to scale back their loads for the rougher fall weather, any reasonable analysis can draw only one conclusion: The Fitzgerald had loaded thousands of tons more than what her architects had designed her to carry.

Primitive weather forecasts and distributions of forecasts

One of the blessings of the semiconductor revolution kicked off by William Shockley and carried forward by Jack Kilby is that computers have gotten vastly more powerful, thus enabling weather forecasts to become more reliable even without any advances in our understanding of meteorology. The National Weather Service was consistently late and consistently underestimating the strength of the storm that would sink the Fitz.

That afternoon [Sunday, November 9, 1975] the National Weather Service had posted a “gale warning,” a level of caution the NWS had created after the 1913 Storm of the Century showed it needed intermediary warnings, not just hurricane alerts. A gale warning predicts winds blowing thirty-nine to fifty-four miles per hour. But the NWS projected the wind would barely reach the gale range, which meant about forty miles per hour.

By 7 p.m. the National Weather Service noticed the storm system that started out of California had reached Iowa, and was gaining speed. It issued a gale warning for all of Lake Superior, correcting its earlier prediction that the storm would slip just below the big lake. Now, the NWS meteorologists said, the storm would cut diagonally across Lake Superior, producing waves from five to ten feet. That might not sound like much, but because the Fitzgerald had only 11.5 feet of freeboard, ten-foot waves wouldn’t give the ship much margin for error.

Late Sunday night the National Weather Service revised its forecast again, now predicting that waves Monday morning would reach ten to fifteen feet high. At 2 a.m. the NWS escalated its gale warning to a storm warning, reserved for winds expected to reach fifty-five to seventy-three miles per hour—strong enough to tear off roofs, uproot large trees, and knock over people attempting to walk outside. But the reality on the water was already starting to outpace the NWS’s forecasts. Winds on Lake Superior had already surged past fifty-eight miles per hour, and were still accelerating. The storm was gaining power like a boulder thundering down a steep hill.

[at 4 pm on Monday, November 10] Captain Cooper, now running an hour behind the Fitzgerald, later reported waves “up to twenty-five feet” when the Anderson neared Caribou Island. Since the Fitzgerald was now about twenty-five miles and an hour and forty minutes beyond the coordinates whence the Anderson had made that report—that much closer to the safety of Whitefish Bay, but also the storm’s epicenter—the waves the Fitzgerald was now experiencing were probably worse than the twenty-five-footers the Anderson had reported. Because the Fitzgerald had only 11.5 feet of freeboard, and probably less by the hour due to its compounding problems, those waves would be more than enough to wash green water—entire waves, not just the spray—over the Fitzgerald’s deck.

The Fitz sank at about 7:10 pm.

Get-there-itis

Some captains put in at Thunder Bay after realizing that the storm was