Laserfiche WebLink
WATER QUALITY E...._,RATIONS AT STATION A72 ON THE AN,&MS RIVER USING <br /> CONTROL CHARTS <br /> • Low Flow (Q < 100 cfs) <br /> • High Flow (Q > 300 cfs) <br /> • Intermediate Flow (100 cfs < Q < 300 cfs) <br /> For the purpose of this evaluation, the single flow value greater than 2000 cfs is <br /> treated together with the other high flow values, as there is no way to develop a <br /> distribution related to a single sample value. <br /> 11. Because each of these classes of flow conditions represents a physical response of <br /> the upper basin to the complex factors that may govern interactions between <br /> meteoric water and the geology (and mining), it is appropriate to evaluate water <br /> quality separately for each of the three classes. <br /> Method for Evaluating Future Water Quality with Respect to Reference Conditions - <br /> Control Charts <br /> 1. Control charts are a standard tool for comparing continuing measurements from a <br /> process (including from a measurement process) to a stable probability distribution <br /> (NBS91). <br /> 2. Control charts are designed to detect specified departures in the process from the <br /> "assumed conditions" (NBS91). With respect to water-quality monitoring at A72, <br /> the "assumed conditions" are the probability distribution of Zn values in the <br /> baseline set, taken to represent "reference water quality". The "specified <br /> departures" would be values that are (statistically) discernibly higher than those <br /> expected for "reference water quality", based on the probability distribution that <br /> describes the baseline data set. <br /> 3. An X-chart for Zn at A72, which provides a running, graphical record of Zn <br /> concentration with respect to sampling event (time), allows reviewers to evaluate <br /> (a) gross errors in analysis; (b) shifts in average values; (c) trends in Zn <br /> concentrations (NBS91). The X-chart uses the mean and standard deviation of the <br /> baseline data set (for each flow class, in the case of the Animas River at A72) to <br /> establish the range of values within which there is a 95% probability that samples <br /> from the same population as the baseline set will fall (i.e., the level of the test is <br /> 0.05, and the test is one-sided because we are concerned primarily with water <br /> quality that exceeds "reference water quality"). <br /> 4. Table 1 presents and analyzes Zn data for the Low-Flow (Q< 100cfs) data set. <br /> Given that the population of Zn concentrations in the Animas River at A72 can be <br /> described by Student's-t distribution, the upper confidence limit for the 95 <br /> envelope and six degrees of freedom has a critical t value of 2.447. For a mean of <br /> 517 ug/1 and a standard deviation of 87 ug/l, the upper control limit for Q< 100 cfs <br /> would be 731 ug/l. <br /> 5. Table 2 presents and analyzes Zn data for the Intermediate-Flow (100 <br /> cfs<Q<300cfs) data set. Given that the population of Zn concentrations in the <br /> Animas River at A72 can be described by Student's-t distribution, the upper <br /> Geochimica, Inc. 4 ANIMAS RIVER/9-Jun-95 <br />