Fault Detection in MRI Chillers

From its birth in 1895 with the discovery of x-rays by Roentgen, medical imaging has become essential to diagnosing and treating patients. With the subsequent invention of high-computation computers, followed by computed tomography and computer-assisted tomography, and eventually magnetic resonance imaging (MRI), clinicians now have the tools to produce three-dimensional images of patients without the need for surgery.

MRI machines use powerful magnetic fields to create images of internal organs such as bones, teeth, cartilage, and muscles by analyzing the protons that make up the proteins and water in these tissues. Because it does not use damaging ionizing radiation such as x-rays, as CT scans do, it is a powerful, non-invasive tool healthcare professionals can use to accurately diagnose medical conditions and disorders, allowing them to recommend the optimal treatment. 

There are disadvantages to MRI, however, one being cost. MRI is considerably more expensive than x-ray imaging or other CT-type scanning. This additional cost is partly derived from the extensive cooling and power requirements used for MRI machines. When an MRI machine performs a scan, it consumes a large amount of energy and, in turn, generates considerable heat in its internal components. This is a problem, as the machine’s temperature must be kept within highly-specified, low temperature limits to prevent overheating, which can lead to malfunction. The reason these temperature limits are required is that MRI machines are based on superconducting magnets, and these magnets need to be kept at cryogenic temperatures. The HVAC system installed at the building or facility is not able to provide this low temperature cooling. 

An additional strategy that can be employed in these systems is to implement conductivity sensors to detect contamination in the coolant. While low flow or loss of flow is relatively easy to measure, identifying contamination is not as straightforward. Electrolytic conductivity (a measure of ionic content) is a widely used analytical parameter for water purity analysis, and is easy to implement to continuously test and validate your water-based solutions; avoiding faults, downtime, and possible catastrophic magnet quenching.

There are three commonly used techniques to measure the ionic content of a solution: 2-electrode, 4-electrode, and inductive conductivity. For this discussion, we will only be considering 2-electrode sensors, since they are best suited for lower conductivity ranges typical of high-purity cooling loops. In 2-electrode sensors, an AC voltage is applied across the two electrodes, and the resistance between them is measured. METTLER TOLEDO Thornton has an advanced conductivity sensor design in the UniCond™ conductivity/resistivity sensor, which integrates the measuring circuit and the physical sensor into a single unit.

UniCond conductivity/resistivity sensors provide exceptionally wide measurement ranges due to their advanced, built-in measuring circuit. This measuring circuit eliminates interference from lead wire resistance and capacitance, allowing a robust digital signal to be sent to the transmitter. A single UniCond 2-electrode sensor can accurately measure from ultrapure water to brackish water (up to 50,000µS/ cm), allowing for a wide breadth of applications with just one sensor type. Various digital or analog sensor options exist, as do various transmitters that can be used to connect the sensors to a PLC for decision making, such as an emergency shutdown or switching to a backup system in the event of fluid failure or contamination.

METTLER TOLEDO Thornton UniCond conductivity/resistivity sensors are the ideal solution for pure, ultrapure water, aqueous mixture (i.e., water:glycol) systems, and aqueous processes which require accurate conductivity and temperature measurement.