December 7, 2009, South Pole Station –
Editor's Note: To see Hermann's neat pictures of IceCube/IceTop, visit this Web page.
First Filling of Water into the IceTop Tanks
IceTop, the surface component of the Neutrino Oberservatory IceCube, is made of 160 tanks, grouped in pairs near each IceCube hole. With IceTop we want to study the radiation that constantly hits our atmosphere. These cosmic rays generate a shower of particles when hitting the atmosphere. If the primary energy is sufficiently high, shower particles will hit many tanks, where their arrival time and deposited energy will be measured. From these measurements, the energy and direction of a primary cosmic particle can be reconstructed.
In IceCube, particles passing the detector are observed by detecting
the light they generate in the clear ice of the South Pole glacier.
The principle of detecting the shower particles with IceTop is
similar, the difference being that for the IceTop tanks, clear
ice has to be generated artificially.
Right: The two DOMs seen from inside the tank where later the ice will be. Looking into the tank, Casey signals the successful DOM installation. Photo by Casey O'Hara. Right: A view into the tank now filled with water, which covers the lower, light-sensitive sphere of a DOM. At the bottom of the tank, one sees the pump-filter system which extracts gases from the water. Photo by Chris Elliot
The light emitted by fast particles is detected by so-called Digital Optical Modules (DOM). Two of these DOMs are hung into an IceTop tank with the light-sensitive half sphere pointing downward (see picture). Then the tank will be filled with water such that the sensitive surface of the DOM is just covered (picture). The water will then be frozen in a very slow, controlled way to keep bubbles out of the water and to make it glass-clear.
After having prepared and installed the tanks into trenches, we started on December 2 to fill the tanks with water. The first filling of the tanks is a major step in the process of tank installation. In preparation for this step, we had to install the DOMs into the tanks (see picture above) and the devices and electronics which control the freezing process to make clear, bubble-free ice (left picture above).
We get the water from the drill camp where a lot of water for the hot-water drill of IceCube is produced. At the beginning of the season, the water is contaminated with glycol with which the whole system was filled over winter. To get the water to an acceptable level (less than 10 ppm) we run it through a special filter system and perform a chemical test on the glycol content.
Once the water has reached a good purity and the tanks are prepared, we can start the tank filling. First, a transport tank, called a ‘water buffalo’ is filled with about 1,700 gallons of filtered water and is then pulled by a bulldozer to a station (see picture). We fill the two tanks of one station with the same water charge. Since each tank will be filled with 660 gallons, the water buffalo has still somewhat more than 300 gallons of water when it is pulled back for refill. The remaining warm water is important to keep the tank and the inlet/outlet valve from freezing. Frozen valves and hoses are our largest problem in the tank-filling procedure. We have developed techniques to avoid freezing which include careful draining of a hose after use and electrical heating of all valves. Still – as we experienced on the next day attempting our second fill – freezing of hoses and valves can happen.
Left: The heavy “water buffalo’ being pulled by a bulldozer
to a station for filling IceTop tanks. Photo by Hermann Kolanoski.
Right: Filling of a tank via a hose from the ‘water buffalo’
to the tank. Hermann controls the flow speed (left) and Chris
watches the water level and communicates it to James who controls
the whole filling procedure via computer from our lab tent, the
‘purple palace.’ Photo by Casey O'Hara
Photos: Chris Elliott, Casey O’Hara, Hermann Kolanoski