California Water Science Center

The accuracy of the analytical method was assessed by spike recovery experiments using surface-water samples from Orange County, California. The water samples were diluted 1:5 with organic-free water prior to dosing (to lower the background concentration of HAA to appropriate levels). The samples were spiked with 8 μL of working standard solution B prior to extraction (table 6). The spiked concentrations were equal to the concentrations in the level-5 calibration standard (table 5). Background concentrations of HAA in these samples had been measured previously. Forty-one samples were used in the spike recovery experiments. Method accuracy is expressed as the mean percent recovery of the spike concentration. The number of samples used to calculate the mean percent recovery varied between 23 and 41 for the nine HAA (table 6) because the measured concentrations (spike plus background) for some analytes in some samples exceeded the concentration in the level-9 calibration standard and, therefore, could not be quantified.

The percent recovery for each analyte is calculated from the following equation.

where

• C_meas is the measured concentration,
• C_backgr is the background concentration of the sample, and
• C_fortified is the concentration of analyte added to the sample

Precision is expressed as the percent relative standard deviation (RSD), which is calculated from the mean and the standard deviation of the replicate analyses:

where

and

and where

• x is the value for an analysis,
• n is the number of replicate analyses,
• is the mean of the replicate analyses,
• s_x is the standard deviation of the replicate analyses, and
• RSD is the relative percent standard deviation for the replicate analyses.

The mean percent recovery ranged from 99 to 117 percent and the percent RSD ranged from 17 to 28 percent. Ongoing accuracy and precision determinations are made by analyzing a set of spiked sample duplicates for 10 percent of the water samples analyzed.

The MDL, as defined by the USEPA, is the minimum concentration that can be measured and reported as greater than zero at the 99-percent confidence level (U.S. Environmental Protection Agency, 1997). The MDL was calculated using the formula:

where

The MDLs were determined from eight replicate analyses of 40-mL samples of organic-free water spiked with 8 μL of working standard solution A. The spiked concentrations of the analytes (table 7) equaled the concentrations in the level-2 calibration standard (table 5). These eight replicate samples were extracted, methylated, and analyzed using the same procedures used for all samples. All eight were prepared on the same day. This spiked concentration was used because it contained all the analytes at concentrations less than five times the expected MDLs, but greater than the baseline noise and drift. The MDLs were calculated from the standard deviations about the mean concentrations for the nine HAA from the eight replicate samples using equation 5. The MDLs for the nine HAA range from 0.11 to 0.45 μg/L (table 7).