No abstract

No abstract

No abstract

Ground‐water‐sampling protocols generally recommend that a well be purged prior to sampling. This recommendation is based on the assumption that the water quality of the water standing in the casing is not the same as that in the aquifer. Three criteria that have been used to determine when a well has been purged sufficiently to yield “representative” water‐quality samples are (1) flushing an arbitrary number of casing volumes, usually a minimum of three, (2) flushing the well until field water‐quality characteristics‐temperature, pH, specific conductance, and dissolved oxygen‐in the purge water are stable, and (3) flushing until hydraulic equilibrium between casing water and aquifer water is achieved. The primary purpose of this study was to evaluate the effectiveness of these criteria in sampling for purgeable organic compounds. Wells, screened in unconsolidated sand and gravel aquifers at six sites, were sampled during a total of 10 purgings. The discharge from each well and the field characteristics were monitored as a function of time and number of casing volumes flushed. Grab samples for purge‐able organic compounds, chloride, and ultraviolet absorbance at 254 nanometers wavelength‐a gross indicator of the presence of aromatic organic compounds‐were taken each time a set of field measurements was taken. The variation of field characteristics, purgeable organic compounds, chloride, and ultraviolet absorbance at 254 nanometers was compared with time and number of casing volumes flushed. The results indicate that (1) purgeable organic compound concentrations sbailized when three casing volumes were purged in only 55 percent of the cases evaluated in this study, (2) purgeable organic compounds concentrations did not consistently follow the temporal variation of, nor stabilize at the same time as, the measured field characteristics, and (3) purging to achieve hydraulic equilibrium between casing and aquifer water consistently underestimated the time and casing volumes needed to achieve stable values of water‐quality measurements in highly transmissive aquifers. The conclusion from these data is that none of the previously recommended criteria for purging a well can be applied reliably to collecting a “representative” sample of purgeable organic compounds. These results indicate that the criteria for purging a well prior to sampling for purgeable organic compounds must take into account other factors, such as the unique hydrogeologic characteristics of a site, the nature and extent of purgeable organic compounds present, the areal extent of the contamination, the well construction, and the sampling objectives of the investigation.

Factors that affect the mass of chemical constituents entering a well include the distributions of flow rate and chemical concentrations along and near the screened or open section of the well. Assuming a layered porous medium (with each layer being characterized by a uniform hydraulic conductivity and chemical concentration), a knowledge of the flow from each layer along the screened zone and of the chemical concentrations in each layer enables the total mass entering the well to be determined. Analyses of hypothetical systems and a site at Galloway, NJ, provide insight into the temporal variation of water‐quality data observed when withdrawing water from screened wells in heterogeneous ground‐water systems. The analyses of hypothetical systems quantitatively indicate the cause‐and‐effect relations that cause temporal variability in water samples obtained from wells. Chemical constituents that have relatively uniform concentrations with depth may not show variations in concentrations in the water discharged from a well after the well is purged (evacuation of standing water in the well casing). However, chemical constituents that do not have uniform concentrations near the screened interval of the well may show variations in concentrations in the well discharge water after purging because of the physics of ground‐water flow in the vicinity of the screen. Water‐quality samples were obtained through time over a 30 minute period from a site at Galloway, NJ. The water samples were analyzed for aromatic hydrocarbons, and the data for benzene, toluene, and meta+para xylene were evaluated for temporal variations. Samples were taken from seven discrete zones, and the flow‐weighted concentrations of benzene, toluene, and meta+para xylene all indicate an increase in concentration over time during pumping. These observed trends in time were reproduced numerically based on the estimated concentration distribution in the aquifer and the flow rates from each zone. The results of the hypothetical numerical experiments and the analysis of the field data both corroborate the impact of physical and chemical heterogeneity in the aquifer on water‐quality samples obtained from wells. If temporal variations in concentrations of chemical constituents are observed, they may indicate variability in the ground‐water system being sampled, which may give insight into the chemical distributions within the aquifer and provide guidance in the positioning of new sampling devices or wells.