Medical Metrology

Chronicle of a death foretold:
Intentional quench of a 3 T Bruker MRI scanner

Background

After 14 years of smooth operation PTB's 3 T whole-body Bruker scanner had reached the end of its lifecycle. Gradient performance, system electronics, or sequence portfolio: they all were no longer up to modern research demands. Thus it was decided to replace it by a new, state of the art 3 T scanner. In lack of free lab space the old system had to go and make room for its successor. Considering costs as well as total system-down time, it turned out that a controlled quench was the fasted and most economical way to initiate the replacement of the scanner. The date was set to October 9th, 2009.

Preparation

Seven weeks in advance the weekly He refills were interrupted until the He level had dropped to the safe-operation minimum set by the manufacturer. Also, at minus three weeks, the liquid nitrogen supply for the thermal shield had been stopped. On October 8th, the day of last measurements, we had 500 litres (50%) of liquid He in the cryostat but no more liquid nitrogen, exactly as intended. To get rid of the last drops of nitrogen, warm gas was passed through the tank. Also, the cooling system of the gradient coil was completely emptied to prevent any damage from freezing water.

A hall probe had been installed at the isocenter of the magnet and hooked up to a digital storage scope. A steel clip on a strap, holding a piece of paper for better visibility and halfway drawn into the magnet, served as a visual field indicator. From the outside, some LED lights and a simple web cam streaming to a laptop were inserted into the quench pipe, facing the burst disk.

The quench

On October 9th, 2009, 15 minutes past 3pm, we pressed the quench button in the scanner room. About five seconds later the burst disk fulfilled its very mission with a considerable bang. After one minute the field was down (< 80 mT), after two more minutes the He blow off expired: the show was over.

The whole process went absolutely smoothly. Quench button, burst disk, quench pipe: everything worked as supposed to. There was no detectable release of He into the scanner room; somebody lying in the magnet during the quench would have been perfectly safe. Not a bad record for a system after 14 years at 4.2 K.

Casualties

One hard drive: after we had installed the web cam into our (very short) quench pipe we brought a laptop to hook it up. We kind of forgot the magnet was still on field.

One bottle of champaign: we had placed it in the end piece of the quench pipe so we would have it well-chilled for the farewell toast. Not a good idea.

No, we did not forget to list the web cam. It survived.