by Mark Bowen
Drilling was always the key. Francis, the theorist, estimates that over the two decades that it took to build AMANDA and IceCube, he spent more than half his time thinking about drilling, directly or indirectly.
* * *
“Everybody knows that hot water drilling got started by a bunch of glaciologists standing around drinking beer, going to the Ice and putting two and two together,” Bruce once told me. “Something like that happens after about a six pack of beer.”
The idea is to use a hose with a nozzle on the end to spray a parallel stream of hot water into the ice, and let the nozzle free-fall as it melts its way down. In practice, this is a lot more complicated than it sounds. High technology has been applied to every detail, and the nozzle or drill head is especially sophisticated. Calipers are attached to the head to measure the diameter of the hole as it goes down; inclinometers are attached, so that the path of the head can be reconstructed as it works its way down and the straightness of the hole can be determined; etc., etc.
The technique was first employed in temperate glaciers in Alaska, where it wasn’t difficult to reach depths of a thousand meters. Since the temperature of the ice wasn’t that far below the freezing point, it wasn’t that hard to melt, and the water used for the drilling didn’t lose a lot of heat through the walls of the hose as it streamed down to the depths. The method was first attempted in the polar regions on the Ross Ice Shelf, where the ice is also relatively warm, to depths of about four hundred meters, and then at Antarctica’s truly frigid, 10,700-foot Dome C, but only to about sixty meters. The ice at Pole sits at minus fifty Celsius, so the Amandroids were taking a step into the unknown by trying to drill to a thousand meters.
What you want in a hot water drill is a lot of very hot water at high pressure and a hose with a wide diameter and heavily insulated sidewalls, so that the water will stay hot as it goes down and carry as much heat as possible out through the nozzle and into the ice. (There’s a bit of a balancing act here, because you do want some heat to escape through the sidewalls in order to keep the water in the upper portions of the hole from re-freezing as the drilling proceeds.) Bruce once pointed out that in the same sense that a neutrino telescope can never be too big, “you can’t have too big a hot water drill either, ’cuz the best hot water drill is the one that drills the hole instantly; it’s the most efficient.” This is a mental construct, obviously—a drill that will deliver an infinite amount of heat in zero time—but it gets the point across: you want a lot of heat and a huge hose.
Bruce knew that Bucky-1 represented a serious compromise in both respects. The hose was only an inch in diameter, and the heating plant, which consisted of a gang of hot water boilers standing in the open out on the Ice, produced only half a megawatt of power. The first hot water drill he had used on the Ross had produced two megawatts. “We knew [Bucky-1] was limited, but we didn’t know how limited, exactly, because it was the first time anybody really tried to go deep in ice this cold.” He calculated that by the time they got to a thousand meters the heat that the water would lose on its way down would exceed the output of the heating plant. In other words, the bottom of the hose stood a good chance of freezing. He could play various games to try to keep it from freezing, such as raising and lowering the drill in order to reheat the water in the newly opened hole, but this would cost fuel, and it was a tricky business in any case.
“Hot water drilling is not for the faint of heart either,” he says. “We’re trying to keep water in some place that’s minus fifty degrees, and that’s not a good thing to do.”
They had taken along a combination of detector strings, some of which had been constructed in Berkeley and the others jointly by Madison and Irvine. The phototubes on the Berkeley strings were only three inches in diameter, while the ones on Madison/Irvine strings were not only larger to begin with, they were encased in the same pressure vessels that DUMAND was using at the time: ten-inch glass spheres, designed to withstand the high pressures at depth. They were made by a company named Benthos and called Benthospheres.
The team took the conservative approach and deployed the smaller detectors first. It is unclear exactly how deep they drilled, but the process went well enough. Unfortunately, however, when they lowered the string into the hole, it got stuck at about eight hundred meters. According to Bruce, “they weren’t jigging the thing, and the holes were pretty irregular, so it froze to the side.” He had suggested to the physicists that they raise and re-lower the cable at least once every five minutes. They had let it sit for half an hour at one point, and that’s when it froze in. (Bob Morse observes that Bruce “was a taciturn man, who would only offer his advice up to the listener once, and you ignored it at your peril.”)
The string didn’t work very well, either. Of the four detectors, only one ended up functioning, the suspicion being that water had leaked into the other three. This first string ever deployed for AMANDA was the last without Benthospheres.
* * *
After a few days at Pole in summer, once you’ve adjusted to the altitude, the brightness, and the surroundings, you begin to realize that everything you learned in grade school about the spinning of the Earth and its motion about the Sun is probably true. It’s obvious that there’s only one day a year down there, as the Sun circles counterclockwise overhead, at a constant angle above the horizon that depends on how recently it has risen or how soon it will set, in other words, how far you are from an equinox. At the solstice, around December 21, it will be twenty-three-and-a-half degrees above the horizon, which is the angle between Earth’s spin axis and the ecliptic, the plane in which we rotate about the Sun. This is the latitude of the Tropics of Capricorn and Cancer.
Since summer is so short, three eight-hour shifts work round the clock all through the season, everyone works six days a week, and if your task allows, Sunday is your day off. This means that the place is on pretty much all the time. On top of that, it’s hard to turn yourself off anyway, since you begin to buzz on the dazzling amount of light pouring constantly into your eyes. It’s easy to get frazzled, and it can be difficult to get to sleep. There are those who fall into a pattern of staying awake all through the work week and sleeping all day Sunday—until they completely burn out. Furthermore, since it’s always “nighttime” for one of the three shifts, there’s usually a party going on somewhere, and people find creative ways to celebrate in what is truthfully, for most, a disconcertingly blank landscape.
The AMANDA team was keeper of a stellar attraction in the tubs that they used to heat water for their drill. Until NSF caught on and made them illegal, there were hot tub parties out at the drill site (see photograph 2). One time, the pilot of a Twin Otter, a relatively small plane that is ubiquitous on the continent, sighted a party in progress as he flew into the base. He landed on the snow nearby, taxied up, stepped out of the plane, stripped off his clothes, and jumped in. There were also naked rolling-in-the-snow contests. An Italian scientist set the record with twelve complete rolls.
It’s good they were having some fun, because the drilling was not going well.
On the second hole they got greedy and tried to go to a thousand meters. This led to about the worst thing that can happen in hot water drilling.
“We stuck Bucky-1,” Bob later recalled. “He’s still there.”
“Yup, radar marker,” said Bruce.
Bob was asleep in his cubical in one of the Jamesways in Summer Camp when a PICO driller named Dave Kestor poked his head through the curtain and whispered the bad news. (Since one of the three shifts is always trying to sleep, silence is observed in Summer Camp twenty-four hours a day.)
“What should we do?” Dave asked. “Are we going to put the instrument down or not?”
“God, I don’t know,” Bob replied, “but I think we probably ought to. We’ve invested this much in the hole. We ought to stuff whatever we can down there.”
He got out of bed and went to find Steve Barwick. “I said, ‘Steve, ah, they’ve stuck the drill
.’ And Steve went ballistic at this point and went screaming—he didn’t know where Tim Miller was, so he went walking into every Jamesway at about three in the morning, screaming at the top of his lungs, ‘Tim Miller? Tim Miller? Where the hell are you? Where the hell are you?’ … And I thought that there was gonna be some construction worker, some six-foot-four guy, was gonna get up and just kill Steve. I thought Steve was going to die.”
PICO had detailed only two shifts to the drilling, each working twelve hours a day, and it was the night shift that had stuck the drill. Bruce was on the day shift, and although he didn’t sleep much during drilling operations—he’d hang around to keep his finger on the pulse even when he wasn’t on duty—he happened to be asleep when the disaster occurred. Not only had the drill gotten stuck, the flow in the hose had stopped as well. This is about the only situation you can’t get out of with a hot water drill.
“Everybody who’s ever done hot water drilling has made the mistake of going with not enough heat once in their life,” Bruce later observed. He thought they were lucky to have gotten as far as they had.
They tried to pull it out with Caterpillar D7 bulldozer.
According to Bob, “That goddamn hose was like a violin string.… It necked down to about half its original diameter.… Bill Barber was the only one that had the courage to go over there and stand while the hose was down the hole and take a hacksaw and cut through it. And we saw this hose, or heard this thing, disappearing at the speed of sound down this hole, like phewwwwww.”
And so, with the drill head and a significant portion of the hose down the hole, they acted on Bob’s first sleepy thought and deployed one of the Madison/Irvine strings anyway. For whatever reason—the hose in the way, the hole too narrow—it only went down 150 meters. Then they began worrying that light from the surface might wick down to the detectors and flood out any muon signals they might possibly detect. The photomultipliers were exquisitely sensitive; they were operating at what is known as the single photon level, which meant that they could detect individual particles of light.
“I just looked around and said, ‘We’ve got to plug that hole. What can we do to plug it?’” Bob recounts. “There was asbestos and some green garbage bags around—the eco people would shit a brick if they heard this: I started stuffing asbestos insulation into the bags to give them bulk and flicking them down this hole, trying to make a light seal.… I think I threw down three or four. I threw down as many as I had.”
Experimental physics, blemishes and all.
Bob remembers the date of the disaster quite accurately, since it was his wedding anniversary: January 17, 1992. He sent an e-mail to Francis back in Madison.
“It said, ‘After a particularly difficult deployment, we’ve managed to put the phototubes exactly where we wanted them’—that was a touch of irony.… ‘Oh, by the way, buy my wife some flowers.’ Hah! Which he did! Francis can always be counted upon to do the right thing in that sort of situation. He sent her a beautiful bouquet of roses.”
9. The Crossover
There is no such thing as a disaster in basic research. Whatever happens, you learn from it. They had stuck their collective nose into it and learned what to do next time. They had even made a significant, though not unexpected discovery: that Antarctic ice is optically inert; it doesn’t give off light on its own. The few detectors they had succeeded in deploying proved to be extremely quiet when they weren’t picking up light from the occasional muon zooming by. This demonstrated that the ice did not contain potassium-40, the radioactive isotope of potassium that is present in all natural bodies of water. The background light given off by dissolved potassium-40 limits the ultimate sensitivity of all water-based neutrino telescopes and always will. Ice, they had now confirmed, does not have that limitation.
From a practical standpoint, the drillers had learned the necessity of more heat and a larger hose, and the physicists had learned that phototubes housed in Benthospheres could survive the refreezing of the water in a relatively deep, hot-water-drilled hole.
The collaboration had also come face-to-face with a management issue that would plague them for the rest of the decade: that PICO, the Polar Ice Coring Office, was fundamentally incapable of doing the job it had contracted to do. The problem was structural in nature; it had nothing to do, necessarily, with the competence or good will of its employees. Since PICO was a separate organization from AMANDA and answered to NSF rather than the physicists, it had no institutional stake in the outcome; there was no pressing need to go the extra mile. The AMANDA contract was one of several in its portfolio, and a small one at that. As long as the organization stuck to the letter of the contract, it would win it again the following year, whether the project succeeded or not.
Bruce Koci and Bill Barber were exceptions to this bureaucratic mentality. Bruce was totally committed to the project and intrigued by the science, and he understood the importance of working with the scientists to understand their needs—even if they couldn’t articulate those needs explicitly on their own. Research is when you don’t know what you’re doing. It also helped that Bruce and Bob Morse had become friends over the course of the season, and that Bruce was the most inventive engineer at PICO.
* * *
John Lynch and Peter Wilkness at NSF hardly batted an eye. They were used to failure the first time around, and the project had flown under the radar anyhow. Near the beginning of 1992, the AMANDA collaboration submitted a three-year grant proposal for $2.6 million that was greatly buoyed by the letter they had published in Nature. (This is a tried-and-true tactic in the grantsmanship game: publish in Nature or Science, get your grant. They would use it again at the next crucial juncture, ten years later.)
Funding commenced in July, and a few months later the collaboration picked up its first new members: a team from Stockholm and Uppsala Universities in Sweden, led by the distinguished high-energy physicist Per Olof Hulth.
Peo, as he was known to his friends (pronounced “Pay-oh”), was a modest and friendly man with a warm sense of humor, who didn’t see a lot of difference between work and play. He would come up with several inventive and even entertaining ideas for AMANDA and IceCube over the years, and students or professors who took sabbaticals in Stockholm have fond memories of excursions with him in his beloved sailboat in the Stockholm archipelago. He supplied a steady hand experimentally and a wise, behind-the-scenes approach to resolving conflicts in the collaboration, of which there would be more than enough over the years. Like most converts to the new field of astroparticle physics, he was seeking refuge from the accelerator world at CERN, with its huge collaborations, rigid management structures, and brutal politics. Astroparticle physics seemed “new and exciting,” he once told me. Its very existence demonstrated that physics was “dynamical”; it could change.
He and several of his countrymen had considered building a neutrino telescope in Torneträsk, a lake in northern Sweden that is reputed to be the clearest in Europe, but it turned out not to be as clear as they’d hoped, nor deep enough to provide the necessary shielding. They had then looked into DUMAND, but decided against joining, partially because it seemed too high tech. Peo told me that when he first heard of AMANDA, it seemed “like a joke … absurd.” Evidently, the more he looked into it the less absurd it became. The Swedes were a perfect match and began contributing right away.
* * *
After a bit of back and forth, the collaboration decided against another deployment attempt the following season. Instead, they took a deep breath, redesigned the drill, and began designing a sizeable array, which they hoped to begin deploying over the 1993–94 Antarctic summer.
In retrospect, Bruce concluded that it would have been impossible to improve much upon Bucky-1, no matter how hard PICO had tried, because the drill was limited by the hose, which was the best available at the time. This would be a recurring theme, all the way through IceCube. As prosaic as it may sound, one could say that the success of this project, that is to say,
the invention of neutrino astronomy, hinged upon the development of an adequate hose for hot water drilling.
Irrespective of its insulation within the bureaucracy, the first year’s failure had been an embarrassment for PICO, and this gave Bruce the leeway to develop a larger hose, which he outsourced to a company that could make it to his specifications. It was an inch and a quarter in diameter, only a quarter of an inch wider than Bucky-1’s, but since flow rate grows rapidly with diameter, he was confident it would allow them to reach a thousand meters the second time around. A team at Wisconsin’s Physical Sciences Laboratory, led by a superb engineer named Bill Mason, built the first dedicated drill for AMANDA, again to Bruce’s specifications (and again probably written on napkins). The Swedes paid for it.
In a vivid demonstration of the debt that AMANDA owed DUMAND, the first true array envisioned for the South Pole was almost a direct copy of the array that DUMAND was hoping to deploy at the same time. DUMAND-II, which had been funded in 1989, was to consist of eight strings arranged in a regular octagon, with a ninth in the center, while AMANDA-A, as it came to be called, was to have nine strings on the perimeter and a tenth in the center.
Bob Morse points out that the tenth string was added to break the high symmetry of the octagon, because computer simulations had shown that too much symmetry can confuse the reconstruction of muon tracks. It doesn’t make it difficult to reconstruct the direction of the track; however, surprisingly, it can make it difficult to figure out which way the particle was traveling—up or down, for example, which seems kind of crucial. AMANDA-A would have twenty modules per string, spaced ten meters apart. And they would be deployed in thousand-meter holes, so they’d end up precisely in glaciologist Tony Gow’s magical depth range of eight hundred to a thousand meters. The plan was to deploy six strings the first season and the remaining four the next.