pumping system was configured based upon a cen- 

 trifugal type submersible pump. The system included 

 a control valve at the farthest downstream point 

 where sample integrity was required. The back 

 pressure in the lines caused by this control valve 

 would maintain a high equilibrium gas solubility 

 and thus reduce the tendency toward bubble forma- 

 tion. This valve would also serve to maintain control 

 over the flow rate of water continuously supplied 

 by the pump. 



Stripper Unit 



While attempts were being made to devise a 

 system to remove dissolved gases quantitatively 

 from a continuously supplied water sample, a device 

 developed by Dr. D. Williams and R. R. Miller (1962) 

 at the Naval Research Laboratory was discovered. 

 This device was an adaptation of the spinning disc 

 oxygenator used in the medical field to oxygenate 

 blood during open heart surgery. A device con- 

 figured for medical application was obtained from 

 Mary Hitchcock Hospital, Hanover, New Hampshire, 

 and tested. The results were so encouraging that a 

 stripper device specific to our application was con- 

 structed based on the design parameters published 

 by Williams and Miller (1962) (Fig. 1). 



Internal Magne 



0-Rmg 



RESULTS AND DISCUSSION 



Following construction, preliminary testing of 

 the stripper was conducted at CRREL. Tap water 

 containing 10 to 12 ppm of dissolved oxygen was 

 continuously passed through the device. Water at 

 the outflow was collected under an inert atmosphere 

 and analyzed for residual dissolved oxygen content 

 by the classical Winkler method. Levels consistently 

 less than 0.5 ppm were found. While this trace of 

 dissolved oxygen was always present, it could have 

 been introduced by dissolved oxygen within the 

 reagents or explained by problems in maintaining 

 a totally oxygen-free environment during the 10-min 

 period required to collect the sample. 



The pumping system was also configured within 

 the laboratory and under visual inspection did not 

 seem to produce bubbles within the pump lines. 

 Complete checkout was impossible, however, since 

 a large supply of water supersaturated with gases 

 was not readily available. 



Final checkout of both systems was conducted 

 in the field at the National Marine Fisheries Service 

 Environmental Research Laboratory at Prescott, 

 Oregon. This facility (Fig. 2) is located on a barge in 

 the Columbia River, 35 miles downstream from 

 Portland, with easy access to river water having 

 high gas levels. The field configuration included 

 a submersible pump which was maintained about 

 1 ft beneath the water, the stripping chamber, and 

 a portable gas chromatograph to be used as a detec- 

 tion system (Fig. 3 and 4). 



While the system was in operation, water sam- 

 ples were withdrawn with a liquid syringe from a 

 septum cap (Point A, Fig. 3) located on the system 

 just upstream from the water control valve, and 

 from the river near the location of the submersible 

 pump. The dissolved gas content of these samples 



FIG. 1 Diagram of stripper unit. 



The principle of operation is as follows. A con- 

 tinuous water sample is introduced to the device 

 (Fig. 1). A water level approximately 1/3 of capacity 

 is maintained in the device prior to outflow. A con- 

 tinuous supply of helium carrier gas circulates 

 through the stripper. During operation, the motor, 

 magnetically coupled to the internal shaft, rotates 

 the internal Mylar discs at several hundred rpm. 

 As the discs rotate through the water, a thin film 

 is spread over the surface and carried into the helium 

 headspace. This thin film rapidly equilibrates with 

 the headspace, which is essentially devoid of air. 

 Since this thin film and the air involved are con- 

 tinuously removed, rapid and efficient gas stripping 

 is achieved. 



FIG. 2. Prescott environmental field station. 



Continuous Monitoring 703 



