Statistical analysis of auto dilution vs manual dilution process in inductively coupled plasma spectrometer tests

Abstract

Gregory Poole Equipment Company is the Caterpillar Dealer in eastern North Carolina which sells and services anything that Caterpillar Incorporated manufactures. One value added service provided by this company is fluid analysis. This service consists of 22 separate tests on engine oil, power train lubricants, hydraulic oils, coolants and fuels. One particular test performed in this suite of tests is a plasma spectrometer test which analyzes wear metals in the oil. This test is specifically used to help technicians troubleshoot problems or to assist in scheduling planned maintenance in construction and mining equipment fleets. This particular test is currently the bottleneck in the oil lab and more specifically the dilution sub-process in prepping the oil sample for the plasma spectrometer. To overcome this bottleneck an auto diluter was installed to complement the manual dilution process that is also still in process today. The lab manager questioned if the auto-diluter process is generating the same results as the manual process with regards to the quality control check performed on three different certified known value calibration solutions. This paper analyzes control data results from the two populations of different dilution methods to determine if they are producing the same results.