Drilling!

It's a big day for the summer 2009/2010 southpole season: the first hole has been drilled, and the first string deployed into the hole. The IceCube detector will be made up of 86 individual strings (cables) each holding 60 light detecting modules. Installation of a string involves drilling a hole 2.5km into the ice, and then attaching the light detectors toa cable which is lowered into the hole.

The drill itself is a simple device to understand. Heat water, pump it through a long hose into the ice. Easy. Of course, like so many things that are easy to understand, there is a lot of complexity in the actual device itself. In fact, it has taken 30+ people working around the clock since November 11th to reach this point. The drill camp is a group of buildings (basically shipping containers) housing the heaters, pumps, generators, water tanks, and control systems necessary to drive hot water through the drill head. The water is heated to 190F and pumped at 800psi through a 3/4 inch nozzel. The power output is about 5 mega-watts or roughly equivalent to the output of a diesel locomotive at full power.

From the drill camp, the hot water is pumped to the tower operation site (TOS). The TOS is positioned over what is essentially a pilot hole drilled through the top layer of snow (called the firn - about 30 meters deep). The drill head and weight stack (a lot of weight that helps keep the hole straight) are lowered into the hole and away we go. The hose is taped to a support cable as it descends. Of course, it's all more complicated than that for the folks that have to do it. The TOS itself is quiet; nothing more than a hose going down into a hole. About every four meters, a sophisticated timing device goes off telling the person doing the taping to wrap another few rounds of tape around the hose and support cable. The action is all back at the drill camp where the pumps and generators are howling.

It takes anywhere from 24 to 30 hours to drill a hole. The slower you go, the larger the diameter of the hole. For the most part, 24 hours produces a hole large enough for the light sensors. Occasionally, however, we send different devices into the holes, some of which are larger than the typical sensors. To accomodate those devices, a slower drilling speed is used to produce the bigger holes. When the hole is done the drill is pulled out of the hole (a process that takes many hours) leaving a 2.5km column of water in the ice.

At that point, the hose and drill are moved to the second TOS (pain, I tell you), and the deployment team moves in. Deployment is the process of attaching the detectors to a cable which is lowered into the hole. I have not seen that process in person, so I can't comment much more than to say that it takes about 12 hours. There's usually an informal race between the drillers and the deployment team. The drillers try very hard to start the next hole before the deployment team finishes lowering the string into the hole. It's usually close. The string is anchored, and the water slowly freezes around the modules. In 100,000 years, they will reach the ocean.

As always, the first hole of the year has its little glitches. This year was no exception, but for the most part, things went smoothly. In fact, the official schedule says that the start of drilling should be 7 December, so we are already three days or so ahead.