AUTOMATED PROCESS FOR MONITORING GROUND - WATER QUALITY USING ESTABLISHED MANUAL SAMPLING PROTOCOLS
Purge and Verification
The prototype used the purge criterion specified in the ground- water sampling protocol developed for the USGS National Water- Quality Assessment Program. Before the pump was activated, the system measured the water level and calculated the volume of water standing in the well. The flowmeter monitored the pumping rate during purging and sampling. Water temperature, pH, specific conductance, dissolved oxygen, and ammonium were measured and recorded every 3 minutes during the purge. The well was considered purged
(fig. 6) Real-time purge data collected on October 6, 1996
when the values of five successive measurements of these properties and constituents fell within specified criteria. The prototype operated successfully from May 10 to November 13, 1996, and sufficient data were collected to demonstrate that the data obtained by the automated method was equivalent to data obtained by manual sampling methods using the same protocols.
(fig. 7) Automated and manual measurements of groundwater quality from wells SDW 479-0028 and SDW 479-0033
The performance of the prototype was determined by a quality-
assurance/quality-control program based on periodic comparative
measurements using instrument calibration readings, and
measurements by independent field probes, as well as manual
sampling to make field measurements and collect duplicate and
equipment-blank samples for analysis at the USGS National Water
Quality Laboratory. Automated water-quality measurements and
manual field and laboratory measurements correlated closely for all
properties and constituents (fig. 7).
Automated measurements of pH were slightly but consistently lower than laboratory and field check measurements because the pH membrane was affected by elevated water pressure in the flow cell. Close correlations between automated and manual measurements were facilitated by the remote communication capability through the modem. System measurements could be examined at any time from the office via the modem, and any unexpected changes in water quality prompted a site visit for manual calibration and testing. For example, a field visit confirmed that the ammonium probe membrane had failed when the automated system indicated substantial increases in ammonium concentrations in mid-July. The probe was replaced and the period during which the probe was malfunctioning was shown as one of "no record" (fig. 7).
The automated monitoring system successfully documented the rapid and short-term changes in hydrologic and geochemical conditions resulting from the discharge of the sewage- plant effluent.
Back to: