Observations on centrifugation: application to centrifuge development.

This report outlines the background to the development of an automated, serial, discrete centrifuge, reporting on the criteria considered essential in such an instrument. We established the criteria by examining the detailed logistics of centrifuge operation in a hospital laboratory. The mean sample load per run, using six centrifuges, was 13.6 samples, and the user-selectable cycle time ranged from 00:01:10 to 00:12:33 (hours:minutes:seconds) with a fixed g value of 1050. During the laboratory working window, (0900-1700), only 50% of the centrifuge capacity was utilized and more than one-third of the sample workload was delayed for >5 min because the centrifuges were not emptied promptly. In addition, 35% of the sample workload was centrifuged for less than the time prescribed in the operational specifications. Based on these findings, we designed a new continuous, serial centrifuge to overcome some of the deficiencies noted in the logistics study. The centrifuge operates continuously, nominally treating 150 samples/h, with a cycle time of 5 min at 1,000 g. The cycle time and g value are variable between limits, and their selection governs the throughput rate. Each sample is centrifuged separately in individual rotors mounted in a sturdy carousel with a periphery that traverses a load/unload station. There is no sample delay because of operator absence, and the capacity is fully utilized. The centrifuge can operate in a stand-alone capacity or has the capability of being integrated into a sample preparation system or as a direct front end for high-throughput analyzers.