9624 (2) - Laserfiche WebLink

Home Browse Search

APPENDIXES 47 Quality-Control Methods and Analysis Quality-assurance and quality-control (hereinafter referred to as QC) samples were used to assess the variability and bias of water- quality data that may be introduced by sample collection, processing, storage, and laboratory analysis. Replicate and field-blank samples were collected and processed as QC samples by using the same equipment and procedures as environmental samples. Replicate samples for ground water were analyzed for major ions and trace elements, nutrients (nitrogen and phosphorus), dissolved organic carbon, total coliform bacteria, MBAS, and radon-222. Blank samples were analyzed for major ions, trace elements, nutrients, and dissolved organic carbon (ground-water samples only). Additional information on QC is given in Shelton (1994) and Koterba and others (1995). Ground water QC data will be discussed first, followed by surface water QC data. Replicate samples were used to test for data variability, which is the degree of random error in independent measurements of the same quantity. Replicate samples were collected and processed in sequence with environmental samples at a site to yield samples of presumably identical composition. By comparing the analytical results of the replicate pair (environmental sample and QC sample), information is obtained on the precision of the measured concentration values and consistency in identifying the constituents of interest. Replicate samples were first assessed with a unit called the lowest rounding unit, which is the magnitude of the least significant figure reported by the USGS NWQL. A nitrate concentration of 0.923 mg/L, for example, would be rounded to the nearest 0.001 mg/L, whereas a concentration of 4.73 mg/L would be rounded to the nearest 0.01 mg/L; a lowest rounding unit of 1 would represent 0.001 mg/L for the 0.923-mg/L sample and 0.01 for the 4.73-mg/L sample. For either sample of a replicate pair, concentrations below the laboratory reporting level (less than, or <, concentrations) were given a value of zero for replicate analysis. A replicate pair with estimated (E) concentrations was included in the replicate analysis when both the environmental and QC samples had an estimated value but were excluded when only one sample of the replicate pair had an estimated value. When analysis of lowest rounding unit data identified a constituent with 50 percent or more rounding units above 2, the mean relative standard deviation was used to estimate variability and uncertainty in concentration for that constituent. For a given constituent, the relative standard deviation was determined for each replicate pair, and then the mean relative standard deviation was computed for the constituent to determine variability and uncertainty in concentration. For major ions and trace elements, 204 replicate pairs for ground water were evaluated. Values of the lowest rounding unit difference ranged between 0 and 197 (tables 6-7, Appendix II). Most replicate pairs for bromide, chloride, fluoride, sodium, sulfate, and iron had a lowest rounding unit of 2 or less. These differences probably are well within instrument variation and do not seem to be related to a particular constituent concentration. Variability associated with sample collection, processing, storage, and laboratory analysis did not appreciably affect the results of environmental data for these constituents. Most ground-water replicate pairs for calcium, magnesium, potassium, silica, and manganese had a lowest rounding unit difference of 2 or greater (tables 6-7, Appendix II). Variability and uncertainty in concentration values were not determined for calcium, magnesium, potassium, and silica because of the small environmental significance of the concentration differences for the respective replicate pairs. The maximum dissolved-solids concentration of 269 mg/L in the ground-water samples was below the USEPA SDWR of 500 mg/L (table 2). The maximum concentration difference between environmental and QC samples for calcium, magnesium, potassium, and silica (constituents of dissolved solids) was 3.0 mg/L for calcium and was insufficient to significantly affect calculated dissolved-solids concen- trations. Variability in manganese concentrations, however, may affect the interpretation of environmental data because manganese concentrations in some samples exceeded the USEPA SDWR for manganese. Most differences in manganese concentrations in replicate pairs for sites 7, 8, and 9 were greater than 10 rounding units, whereas concentrations in replicate pairs for other sites varied insignificantly (tables 6-7, Appendix II). Variability in manganese concentrations for site 9 was 15.3 percent (±31.8 percent) of the concentration value at the 95-percent confidence level (table 8, Appendix II). Variability in manganese concentrations for sites 7 and 8 was not determined because of few data (only one replicate pair per well). For nutrients and dissolved organic carbon, 122 ground-water replicate pairs were evaluated. Lowest rounding unit differences ranged from 0 to 137 (tables 6-7, Appendix II). The latter value was an outlier, however, as the next highest difference was 19. Most replicate pairs for ammonia plus organic nitrogen, nitrite, orthophosphate, and dissolved organic carbon had lowest rounding unit differences of 2 or less. Interpretation of environmental data for these constituents was not affected by variability in reported concentrations. Lowest round- ing unit differences greater than 2 were most common for ammonia, phosphorus, and nitrate. For ammonia and phosphorus, all differences in lowest rounding unit were 6 or less. Interpretation of environmental data for ammonia was affected by the variability in reported con- centrations, as the variability was within the concentration range of the environmental data (table 8, Appendix II). Variability in phosphorus data did not affect the interpretation of environmental data for phosphorus. For nitrate, some differences in concentrations between the environmental and quality-control samples were large, ranging between 12 and 137 lowest rounding unit (tables 6-7, Appendix II). Exclud- ing the replicate pair with the latter value, variability in nitrate concentrations was 1.79 percent (±3.72 percent) of the concentration value at the 95-percent confidence level (table 8, Appendix II). However, because the maximum nitrate concentration of 4.72 mg/L in the ground- water samples was well below the USEPA MCL of 10 mg/L (table 2), variability in nitrate concentrations did not affect the interpretation of environmental data for nitrate. All environmental and QC samples for the six ground-water replicate pairs of total coliform bacteria had concentrations less than the reporting level of 1 col/100 mL (tables 6-7, Appendix II); therefore, variability in the concentration of total coliform bacteria could not be determined. Variability in the concentration of MBAS also could not be determined because most (15 of 18) of the replicate pairs had all concentrations below the reporting level (tables 6-7, Appendix II). Differences in radon-222 concentrations in the environmental and QC samples ranged between 7 and 78 lowest rounding unit (tables 6-7, Appendix II), which did affect the interpretation of environmental data