Instrument Development, Model 512) was used. 

 Quantitative results were obtained using a sampling 

 loop and a thermal conductivity detector and com- 

 paring the output to that obtained when running 

 a calibration gas. 



The major problem encountered in obtaining 

 accurate measurements involved maintaining pre- 

 cise knowledge of the water and gas flow rates in 

 the stripper unit. Specifically the value obtained 

 from the output of the gas chromatograph must be 

 adjusted using a factor based upon the ratio of the 

 water and gas flow rates. If the gas and water flow 

 rates can be maintained at an identical value, the 

 concentration of dissolved gas found in the carrier 

 stream is equivalent to that in the river water 

 sample. 



The difficulty encountered maintaining con- 

 stant flow rates resulted from changes of resistance 

 in the gas flow system. These changes occurred as 

 the drying tube picked up water, resulting in a 

 higher gas head pressure in the stripper and a lower 

 water flow rate through the system. The second 

 problem was encountered when the valve on the 

 gas chromatograph sampling loop was switched, 

 sending a gas sample into the GC column. When 

 this was done, the resistance on the gas flow system 

 was also changed, resulting in a different head 

 pressure in the stripper. This caused a change in 

 water and gas flow rates as well as a change in the 

 water level in the stripper. 



In spite of these difficulties, when the system 

 was manually operated with extreme care, reason- 

 ably constant flow rates could be maintained. Values 

 obtained under these conditions were compared 

 with those found by injecting river water samples 

 into the Swinnerton chamber of the gas chromato- 

 graph. The values obtained for dissolved nitrogen 

 were 21.1 ppm by the stripper system and 20.7 ppm 

 by the Swinnerton method. While we were unable 

 to compare these methodologies over an extended 

 period, the results are quite encouraging. 



RECOMMENDATIONS 



The results of this investigation confirm the 

 feasibility of developing an unattended monitoring 

 system for dissolved gases. The next step is to 

 actually configure a prototype system for field 

 validation. A diagram of a specific system, which 

 utilizes a thermal conductivity detector for total 

 gas measurement, is shown in Fig. 6. Prior to fabri- 



Submersible Pump 



FIG. 6 Proposed monitor configuration. 



cation of this system the following areas must be 

 investigated: 1) The detector design must be 

 refined, a prototype assembled, and the system 

 validated under field conditions; 2) The electronic 

 systems required for detector operation, tempera- 

 ture control, flow cycling, and signal processing 

 must be designed and configured; and 3) Systems to 

 maintain a higher degree of control on flow rates 

 within the stripper unit must be designed and 

 implemented. 



Should a system be required to quantitate a 

 single component of the gas matrix, a more specific 

 detector would be required but the remainder of 

 the system would be largely applicable. 



REFERENCES 



Gorham, F. P. 1901. The gas bubble disease of fish and its 



cause. In: Bull. U.S. Fish. Commission, 1899, vol 19, pp. 33-37. 

 Marsh, M. C. and F. P. Gorham. 1905. The gas disease in 



fishes. In: Report of the Bureau of Fisheries, 1904, pp. 343-376. 

 Plehn, M. 1924. Praktikum der Fischkrankheiten. Handbuch 



der Binnen Fisherei Mittel Europas, 7:301-479. 

 Proposed Criteria for Water Quality, vol 1, October 1973, U.S. 



Environmental Protection Agency, pp. 102-103. 

 Swinnerton, ). W., A. J. Linnenbom, and C. H. Cheek. 1962. 



Determination of dissolved gases in aqueous solutions by gas 



chromatography, Anal. Chem., 34:483-485. 

 Williams, D. D. and R. R. Miller. 1962. An instrument for 



on-stream stripping and gas chromatographic determination 



of gases in liquids, Anal. Chem. 34:657-659. 

 Woodbury, L. A. 1941. A sudden mortality of fishes accom- 

 panying a supersaturation of oxygen in Lake Waubesa, Wise, 



Trans. Am. Fish. Soc, 77:112-117. 



Continuous Monitoring 105 



