AWWA ACE61621

A study sponsored by the American Water Works Association Research Foundation (Letterman, et al., 2002, and Viswanathan, et al., 2004) evaluated the inter-instrument variability of a diverse assortment of 6 bench top and 4 portable turbidimeters. The results of the AwwaRF study divided the instruments into two groups, A and B. The instruments in group A do not automatically set a predetermined reading when a sample of low-particle (dilution) water is put in the instrument during the calibration procedure and the instruments in group B use the calibration procedure to automatically set the low-particle reading at either zero or 0.02 NTU. Both groups include portable-type instruments and different types of light sources. Over a turbidity range of 0.01 to approximately 0.5 NTU (measured by the group A instruments) the mean readings for group A instruments were about 0.1 to 0.3 NTU higher than those of the group B instruments. The calibration method, type of light source, and the use of an instrument's ratio measurement feature did not have a significant effect on the results. This paper describes a second study that was undertaken to improve inter-instrument agreement at low turbidity levels using low-level turbidity standardization suspensions. The study was conducted using three group A instruments and four group B instruments from the AwwaRF study discussed above. The group A set included one portable instrument and the group B set included two portable instruments; the others were bench top units. The eight test suspensions used to determine inter-instrument agreement included a filtered tap water, an unfiltered tap water, and six dilute suspensions of kaolin clay and a sub-micron size fumed silica particle. The turbidity of the test suspensions ranged from about 0.01 to 0.8 NTU. Each instrument was calibrated with a given suspension (e.g., laboratory diluted formazin) using the manufacturer's directions and then the same type of suspension was used to prepare a standardization curve for that instrument. Four to six suspension concentrations (including a value for the low-particle dilution water when it was available) were used to obtain points for the standardization curve. After the standardization measurements had been made and the points of each data set fitted with a least squares linear regression line each instrument was used to measure the turbidity of the eight test suspensions. Each combination of test suspension, instrument and calibration/standardization suspension produced two turbidity values, a mean measured value and an adjusted value. The data was analyzed to determine if the adjusted turbidity values give improved inter-instrument agreement compared to the unadjusted values. Includes 2 references, tables, figures.