Abstract |
The paper discusses important aspects of the conceptual design of evaluation of studies of the performance of proprietary analytical test kits for drinking water surveys and corrosion control studies. The Hach Lead-Trak test kit was tested for precision, accuracy, operator bias, and performance relative to a standard GFAAS methodology. A detection limit of 4 milligrams/L was observed, with generally accurate results for standards up to 100 milligrams/L. Best performance was found below 80 milligrams/L. Spike recoveries for 15 milligrams/L in six different drinking waters were good. Interferences from 13 important cations and eight important anions were tested for realistic concentrations. Calcium and sulfate caused positive bias; Zn, Fe, orthophosphate, tripolyphosphate and pyrophosphate caused significant negative bias. No operator bias was found. The test kit was extremely sensitive to the test pH, and failed on acidified samples. |