Skip to main content
ARS Home » Northeast Area » Leetown, West Virginia » Cool and Cold Water Aquaculture Research » Research » Publications at this Location » Publication #154517


item Brazil, Brian
item Rodriguez, Felicia

Submitted to: Book of Abstracts Aquaculture America
Publication Type: Abstract Only
Publication Acceptance Date: 9/1/2002
Publication Date: 2/20/2003
Citation: Watten, B., Brazil, B.L., Summerfelt, S., Rodriguez, M.F., Sharrer, M. 2003. Monitoring dissolved gas concentration based on head-space partial pressures. Book of Abstracts Aquaculture America.

Interpretive Summary:

Technical Abstract: A multi-station dissolved gas monitoring system was developed that eliminates the need for submerged analytical components so as to circumvent problems with biological fouling. Dissolved gas concentrations are calculated using Henry's Law, water temperature, and the partial pressure of a gas that develops within the headspace of a vertical gas-liquid contacting chamber. Water enters the chamber as a spray, and then exits into a receiving basin through a cone diffuser designed to minimize bubble carryover. Headspace gas composition changes as equilibrium is established between gas-phase partial pressures and dissolved gas tensions. In early work, calculated dissolved oxygen (DO) concentrations agreed well with those obtained by Winkler analysis (N-67) over a range of DO (0-18 mg/l), water temperature (11.5-27.5C) and dissolved nitrogen levels (73-107% of saturation). Differences between the two analytical methods averaged 0.25 mg/l (range -0.5-0.9 mg/l) with precision, as measured by the coefficient of variation, averaging 0.9% at 10.2 and 1.2% at 25 C. Dissolved carbon dioxide (DC) concentrations were also determined using either a gas phase infrared detector or by measuring voltage developed by a pH electrode immersed in an isolated sodium carbonate solution sparged with head-space gas. Calculated DC concentrations were compared to those obtained by titration (N=96) over a range of DC (2-28 mg/l), total alkalinity (35-250 mg/l) and DO concentrations (7-18 mg/l). Here, the absolute errors associated with use of the infrared detector and pH electrode averaged 0.9 mg/l and 0.5 mg/l, respectively. The coefficient of variation was 1% for headspace calculations and 1.7% for titration-derived values. In the present study a single infrared detector (CEA Instruments Model GD44) was coupled with 4 replicate headspace units using 3 solenoid valves (3 way, 1.6 mm bore) activated by a time-based controller. Gas Samples were obtained sequentially, at 15 minute intervals, by the meter's internal sampling pump. Readings were logged automatically over a 21-day test period to test for instrument drift as indicated by errors in calibration and DC calculations. Calculated DC concentrations were compared to titration-derived values once or twice daily (N-36). The infrared detector's (error) calibration did not drift above or below a preset limit of 0.1% CO² during the test period. Absolute error over the DC concentration range tested (36-40 mg/l) averaged 2.9, 3.1, 3.7, and 2.7 mg/l for replicate headspace units 1-4, respectively. Use of a singe detector to measure four stations reduced the capital cost per sample site from $2070 to $720. Cost per site was reduced further in a second test series by exploiting the infrared detector's ability to monitor gas phase concentrations of oxygen concurrently with carbon dioxide measurements, i.e. two gas species were measured at each of four stations in a large recirculating water fish production system