Experimental study on total dissolved gas supersaturation in water

ArgAEWater Science and Engineering, 2011, 4(4): 396-404doi:10.3882/.issn. 1674-2370.2011.04.004http://www.waterjournal.cne-mail: wse2008@vip.163.comExperimental study on total dissolvedgas supersaturation in waterLu Qu1'2, Ran LI*2, Jia LI, Ke-feng Lr, Lin WANG21. Zhejiang Institute of Hydraulics and Estuary, Hangzhou 310020, P. R. China2. State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University,Chengdu 610065, P. R. ChinaAbstract: More and more high dams have been constructed and operated in China. The totaldissolved gas (TDG) supersaturation caused by dam discharge leads to gas bubble discase or evendeath of fish. Through a series of experiments, the conditions and requirements of supersaturatedTDG generation were examined in this study. The results show that pressure (water depth), aeration,and bubble dissolution time are required for supersaturated TDG generation, and the air-watercontact area and turbulence intensity are the main factors that affect the generation rate ofsupersaturated TDG The TDG supersaturation levels can be reduced by discharging water toshallow shoals downstream of the dam or using negative pressure pipelines. Futhermore, the TDGsupersaturation levels in silling basins have no direct relationship with those in reservoirs. Theseresults are of great importance for further rescarch on the prediction of supersaturated TDGgeneration caused by dam discharge and aquatic protection.Key words: total dissolved gas supersaturation; dissolved gas in water; experimental study;dam discharge1 IntroductionThe supersaturation of total dissolved gas (TDG) is caused by dam discharge, and it maycause gas bubble disease in fish and ultimately endanger their existence (Tan et al. 2006).Overseas research on TDG supersaturation has been conducted since many years ago. In the1960s, studies focused mainly on the effects of TDG supersaturation resulting from spilldischarge of hydraulic engineering structures in the Columbia River, which harmed fish(USACE 2005; Orlins and Gulliver 2000). Weitkamp and Katz (1980) observed the death offish caused by TDG supersaturation in the Saint John River of Canada. Huang (2002), Urbanet al. (2008), and Politano et al. (2007, 2009) established different two-phase flow models ofTDG supersaturation. The calculated TDG supersaturation levels were in agreement with theobserved data. Since more and more dams are constructed and operated in China, the problemof TDG supersaturation has become increasingly prominent. In recent years, the TDGThis work was supported by the National Natural Science Foundation of China (Grant No. 50979063).*Corresponding author (e-mail: liran@scu.edu.cn)Received May 1, 2011; accepted Oct. 8, 2011中国煤化工MHCNMHGsupersaturation problem and its negative effects on fish have captured the attention ofresearchers in China. The research includes the supersaturated TDG generation (Li et al. 2009;Chen et al.2009a) and release processes (Feng et al. 2010), the impacts of supersaturated TDGon fish (Tan 2006), and TDG abatement measures. Studies on supersaturated TDG generationare of great importance because they can allow for the prediction of TDG supersaturationlevels dowstream of high dams, and can help to explore improved methods for mitigating theeffects of supersaturated TDGCurrently, studies of the conditions for and the theory of supersaturated TDG generationdownstream of high dams are not systematic. The preliminary study of Jiang et al. (2008a)indicated that the aeration and bubble size were directly related to the TDG supersaturation in astilling basin. Furthermore, they established a relation between the TDG supersaturation leveland the water depth (pressure) of the sillig basin. In addition, field observations of TDGdownstream of high dams were conducted (Jiang et al. 2008b; Qu et al. 2011). Chen et al.(2009b) studied the supersaturated DO generation caused by the Three Gorges Dam discharge,and discussed the effects of supersaturated reoxygenation. Qin and Li (2008) performedpreliminary numerical simulations of supersaturated DO and reoxygenation when the damdischarges water. Based on comprehensive analysis of previous studies, it is clear that a betterand more universal forecasting model for supersaturated TDG generation is needed. In thisstudy, an experimental study was conducted to investigate the generation conditions andinfluencing factors of the supersaturatedTDG2 Definition of TDG saturation levelGas solubility is the volume of gas that can dissolve in a solvent (yielding a saturatedsolution) at a specific pressure and a specific temperature. If the volume of the dissolved gas isgreater than the gas solubility, it is called TDG supersaturation. The TDG saturation level, G, isdefined as follows:G=二x 100% .(1)Cwhere C is the concentration of TDG (mL/L), and C, is the solubility of TDG at the localatmospheric pressure and temperature (mL/L).TDG includes nitogen, oxygen, carbon dioxide, and rare gases. Due to the difficulty ofmeasuring certain dissolved gases, such as nitrogen and argon (Watson et al. 1998), the TDGsaturation level is obtained by calculating the total dissolved gas pressure based on theprinciple that the concentration of dissolved gas is proportional to the gas partial pressure. TheTDG saturation level can be written as follows:G= B+OP-x 100%(2)Pwhere P is the local atmospheric pressure (mmHg), and AP is the dfference between the totaldissolved gas pressure and the local atmospheric pressure (mmHg).Lu QU et al. Water Science and Engineering, Dec. 2011, Vol. 4, No.4, 396-404397中国煤化工MHCNMH G.3 Study of gas bubble dissolution and releaseIn order to explore the generation conditions of the supersaturated TDG and itsinfluencing factors, a series of self- designed experimental facilities were used to study theprocess of gas bubble dissolution and release.3.1 Gas bubble dissolution experimentThe experimental device of gas bubble dissolution is shown in Fig. 1. The main devicewas a plexiglass cylinder with a height of 2.0 m and a diameter of 0.4 m. The experimentalwater depth was 1.38 m. Compressed air generated by an air compressor entered the waterthrough a 1 mm diameter pinhole in the main device. Gas flow was regulated by a valve in theintake manifold. The TDG pressure in water was measured using a YSI 5200 TGP probe madeby the Yellow Springs Instrument (YSI) Company. The TGP probe was placed at a depth of 1.2m underwater. The TDG saturation level was obtained by dividing the TDG pressure by thelocal atmospheric pressure. Calibration showed that the instruments' mecasurement errors wereabout士0.1%. We used pretreated TDG unsaturated water. By raising the water temperature tocause part of the dissolved gases to release and then rapidly cooling the water to roomtemperature, we brought the TDG saturation level to less than 100%.Free suface、F号H不|YSI 5200Air conpressorGas bubbleFlowmeter -Fig. 1 Sketch of experimental device of gas bubble dissolutionExperiments were carried out under two conditions. Under condition 1, gas was mixedwith water through a single pinhole at a flow rate of 0.20 L/b. Under condition 2, gas wasmixed with water through three pinholes at a flow rate of 0.61 L/h. The change process ofTDGsaturation level is shown in Table 1.Table 1 Change of TDG saturation level in gas bubble dissolution experimentTDG saturation level (%)Condition .0.25 h0.5h224 h48h184.885.185.686.894.497.88.797.4100.3100.6100.5These results show that for the single-pinhole aeration (condition 1), the aeration intensityis low, and the gas can not quickly dissolve in the water because the air-water contact area is398Lu QU et al. Water Science and Engineering, Dec. 2011, Vol. 4, No.4, 396-404中国煤化工MHCNMH G.extremely small and turbulence is weak. For the three-pinbole aeration (condition 2), thecollision between bubbles is more intense, and the gas dissolves more quickly because theair-water contact area is larger. The TDG level reaches 100% of saturation in a relatively shortperiod of time. These experiments demonstrate that the gas dissolution rate is .associated withthe air-water contact area and turbulence intensity.3.2 Impact of gas bubble sizeIn order to further verify that the gas bubble size plays a role in TDG supersaturation, wedesigned an experimental apparatus for studying the impact of the gas bubble size (Fig. 2). Themain device included a 2.0 m-high, 0.4 m-diameter outer plexiglass cylinder, a 0.5 m-high,0.2 m-diameter inner plexiglass cylinder fixed on the inside of the outer plexiglass cylinder,and two cylindrical metal sieves with apertures of 150 um. The lower sieve (sieve 1) was fixedon the upper surface of the inner cylinder, and the upper sieve (sieve 2) was mounted on arotating shaft. The two sieves were close to each other but were not in contact with each other.An air compressor was linked to the bottom of the inner cylinder. The experimental water depthwas 1.8 m. The probe was placed at a depth of 1.1 m underwater.The experimental gas was provided by the air compressor at a working pressure of 3 atm.Gas generated by the air compressor formed bubbles after passing through the lower sieve onthe inner cylinder. Then, the bubbles overflowed outwardly, and were changed into tiny airbubbles when they encountered the high-speed rotary shear of the upper sieve. In order to studythe effects of bubble shear on the TDG saturation level, we also conducted a test in which theupper rotating sieve (sieve 2) was not installed (Fig. 3).Rotating shaftFree surface、各十不。106 ASieve1■Sieves 1 and 2YSI 5200104Sievei嘉10Air compressore 100Howlele二古古10Time (mia)Fig. 2 Sketch of experimental device for studyingFig. 3 Changes of TDG saturation level throughimpact of gas bubble sizeexperiment for studying impact of gas bubble sizeThe results of these two group experiments show that when there is no sieve 2, the gasbubble diameter is larger, turbulence is weaker, and the generation rate of TDGsupersaturation is lower. This indicates that the gas bubbles of smaller size dissolve morequickly because the air-water contact area is larger.Lu QU et al. Water Science and Engineering, Dec. 2011, Vol. 4, No. 4, 396 404399中国煤化工MYHCNMHG3.3 TDG release experimentThe experimental device for TDG release is shown in Fig. 4. In order to maintain aconstant depth in the water container, an overflow port was set at a depth of 1.08 m. A pumpwas used to maintain the circulation of water. Near the mouth of the pipe, there was aT-junction that enabled a high- velocity flow to carry the gas. A free-jet flow formed at 0.50 mover the water surface. The velocity of the jet flow was about 70 m/s.The change of TDG saturation level is shown in Fig. 5. The results show that the TDGsaturation level decreases continuously. The reason is that the locally negative pressuregenerated by the structural style of the T-junction leads to the release of TDG in the circularflow. This indicates that a certain pressure is required to maintain the TDG saturation level.T-juction←FlowE 10096988YSI 5200Pump94Time (min)Fig. 4 Sketch of experimental device for TDG releaseFig. 5 Change ofTDG saturation level in TDGrelease experiment4 Study of generation of supersaturated TDG4.1 High-speed air-water jet flow experimentA high-speed air water jet flow experiment was designed to test the effect of the jet in airand water. The experimental device is shown in Fig. 6. The main device was composed of twocylindrical water containers. In order to maintain a constant depth, the overflow port was set ata depth of 1.08 m.The experimental gas was provided by an air compressor at a working pressure of 3 atm.A jet flow formed in the water container on the right side after the airflow generated by the aircompressor and the water flow generated by the pump were mixed. The experiments werecarried out under two conditions: the water surface free jet and underwater submerged jet. Thedistance between the free jet point and the water surface was 0.5 m. The submerged jet pointwas located 0.5 m below the water surface.Fig. 7 shows the changes of the TDG saturation level with time. The TDG saturation levelin the free jet flow was lower. The gas in the free jet flow lacked sufficient power to enter thewater, so it was unable to form high-degree supersaturated water. For the submerged jet, the gasin the jet flow had full access to water and therefore dissolved rapidly. The TDG saturationlevel in the water was dependent on the water depth (pressure). The experimental results400Lu QU et al. Water Science and Engineering, Dec. 2011, Vol. 4, No. 4, 396-404中国煤化工MYHCNMHGindicate that sfficient aeration and pressure (depth) are required for supersaturated TDGgeneration. At the same time, the retention time of bubbles in the water is a main factor thatinfluences the supersaturated TDG generation.Flow. I Trjunction4- AiPump06「105上Freejet-Free surface103 t一一Submerged jetYSI 5200Submeged jet-首Air compressor2030Time (min)Fig. 6 Sketch of experimental device for high-speedFig 7Changes ofTDG saturation level in high-speedair-water jet flowair-water jet flow experiment4.2 High-speed aerated flow experimentA high-speed aerated flow experiment was designed to test the effect of the high-speedaerated flow on supersaturated TDG generation. The experimental apparatus is shown in Fig. 8.The height, diameter, and depth of the experimental water container were 6.0 m, 0.2 m, and 4.0 m,respectively. A high-speed circulating jet flow was generated by a pump. The jet exit waslocated at a depth of 3.0 m underwater, and the jet velocity was 70 m/s. The experimental gaswas provided by an air compressor at a working pressure of 3 atm. The outlet of the gas flowwas close to the outlet of the high-speed jet flow. The TDG test point was located at a depth of2.5 m underwater.The experiment was conducted under three different conditions. Under condition 1, therewas high-speed circulating water flow and no air compressor aeration. Under condition 2,there was air compressor aeration and no high-speed circulating water flow. Under condition 3,there were simultaneous high-speed circulating water flow and air compressor aeration. Thechanges of the TDG saturation level in the experiments are shown in Fig.9.'Air” TFlow。Condition I”Condition 2号30 r0 Condition 3| YSI 5200|Pumo二11000 AFig. 8 Sketch of experimental device forFig. 9 Changes ofTDG saturation level inhigh-speed aerated flowhigh-speed aerated flow experimentLu QU et al. Water Science and Engineering, Dec. 2011. Vol. 4, No.4, 396-404401中国煤化工MHCNMH G.Under condition 1, despite of the strong turbulence caused by the high-speed water flow,the TDG saturation level at the test point was only 101.7%. This was possible due to the factthat water was mixed with a stmall amount of air on the free surface, and there was only a smallamount of gas dissolved in the water. The experimental results indicate that aeration is anecessary component for the . supersaturated TDG generation. Under condition 2, bubblescollided with each other because of a large amount of aeration. Thus, the air-water contact areabecame larger. At the same time, the hydrostatic pressure and turbulence caused the bubbles todissolve faster. The TDG saturation level rapidly came to the maximum TDG saturation level,125.4%. Under condition 3, because the high-speed water flow cut the aerated bubbles, thebubble sizes were smaller than those under condition 2, the air-water contact area increased,and the bubbles dissolved even faster. The TDG saturation level reached 125. 4% more rapidlythan under condition 2. The experimental results from conditions 2 and 3 indicate that theair-water contact area (bubble size) and turbulence intensity are the main factors affecting thegeneration rate of TDG supersaturation.The degree of TDG supersaturation for the experimental circulating water had no effect onthe maximum degree of TDG supersaturation that the water ultimately achieved. Therefore, weconclude that the TDG saturation level in the siling basin has no direct relationship with thatin the reservoir.In the case of the circulating jet, the TDG saturation level in the water gradually increasesbeyond the initial TDG saturation level. When the TDG saturation level reaches the maximumsaturation level, it will not further increase in the continuous aeration. For any TDG saturationlevel upriver, when the dam discharges water, the maximum TDG. saturation level ofdownstream flood water is not affected.5 ConclusionsIn this study, the generation conditions of TDG supersaturation downstream of dams wereexamined through a series of experiments. The results indicate that pressure (water depth),aeration intensity, and bubble dissolution time are important factors that influence thegeneration of supersaturated TDG The turbulence intensity and water-air contact area (bubblesize) are the main factors that affect the generation rate of supersatured TDG The smaller thegas bubble diameter is, the more quickly it dissolves because of the larger air-water contactarea, and the faster the generation rate of supersatured TDG is. The stronger the turbulence is,the higher the generation rate of supersaturated TDG is. This indicates that sufficient aerationand pressure (water depth) are required for supersaturated TDG generation when the high damdischarges water. Larger water depth and higher pressure can generate a higher TDGsupersaturation level. The circulated jet experiment designed in this study can not effectivelyraise the maximum TDG supersaturation level. This indicates that, during the spill period of thedam, the TDG saturation level in the stilling basin has no direct relationship with that in theLu QU et al: Water Science and Engineeing, Dec. 2011, Vol. 4. No. 4, 396 404中国煤化工MYHCNMHGreservoir. As the maximum TDG supersaturation level is closely related with pressure, the TDGsupersaturation can be reduced by reducing the water pressure. The supersaturated TDG levelcan be reduced by discharging water to shallow shoals downstream of the dam or usingnegative pressure pipelines in practical engineering.The results are of great importance for further research on the prediction of supersaturedTDG generation and aquatic protection. The factors influencing supersatured TDG generationare very complicated because the air-water two-phase flow affects the supersatured TDGgeneration when the high dam discharges water. In addition, the conditions for the supersaturedTDG generation in the experiments. and prototypes were not all the same. The physicalconditions of the TDG dissolved in water were analyzed through the experiments. The resultsof qualitative analysis are reliable, but for quantitative study, these experimental conditions arenot sufficient. In the future study, we need to consider how to build a hydraulic model ofsupersaturated TDG The effects of the energy dissipation patterm, spill rates, and operationmodes need to be considered in the model when the dam discharges water. This suggests thatthe experimental apparatus need to be further improved. Furthermore, the theory and numericalsimulation of supersatured TDG generation are important issues that require further researchand discussion.ReferencesChen, Y. B., Peng, Q. D., and Liao, W. G 2009a. The evolvement study on supersaturation of dissolved gas inthe middle reaches of Yangtze River after the Thrtee Gorges Project runping. Journal of Hydroecology,2(5), 1-5. (in Chinese)Chen, Y. C.. Fu, J,, Liu, Z. W., Cheng, X. J., and Zhu, D. J.2009b. Analysis of the variety and impact factorsof dissolved oxygen downstream of Three Gorges Dam after the impoundment. Advances in WaterScience, 20(4), 526- 530. (in Chinese)Feng, J. J, Li, R., Li, K. F, Li, J, and Qu, L. 2010. Study on release process of supersaturated total dissovledgas downstream of high dam. Journal of Hydroelectric Engineering, 29(1), 7-12. (in Chinese)Huang, H. Q. 2002. Computational Model of Total Dissolved Gas Downstream of a Spilway. Ph. D.Dissertation. lowa: University of lowa.Jiang, L, Li, J, Li, R., and Li, K. F. 2008a. A study of dissolved gas supersaturation downstream of Zipingpudam. Advances in Water Science, 19(3), 367-371. (in Chinese)Jiang, L, Li, J, Li, R., and Li, K. F 2008b. The supersaturation of dissolved gas in downstream of high dam.Journal of Sichuan University (Engineering Science Edition), 40(5), 69-73. (in Chinese)Li, R., Li,J, Li, K. F, Deng, Y, and Feng, J. J. 2009. Prediction for supersaturated total dissolved gas inhigh-dam hydropower projects. Science in China, Series E: Technological Sciences, 52(12), 3661-3667.[doi:10.1007/s11431-009-0337-4]Orlins, J. J.. and Guliver, J. S. 2000. Dissolved gas supersaturation downstream of a spillway IHI:Computational model. Journal of Hydraulic Research, 38(2), 151-159.Politano, M, Carrica, P, Turan, C. and Weber, L. 2007. A multidimensional two-phase flow model for thetotal dissolved gas downstream of spillways. Jourmal of Hydraulic Research, 45(2), 165-177.{doi:10. 1080/00221686.2007.9521757]Politano, M, Carrica, P., and Weber, L. 2009. A multiphase model for the hydrodynamics and total dissolvedgas in tailraces. International Journal of Multiphase Flow, 35(11), 1036 1050.Qin, C. L., and Li, L. 2008. Numerical model of dissolved gas supersaturation flow over sillway of GezhouLu QU et al. Water Science and Engineeing, Dec. 2011, Vol. 4, No. 4, 396404403中国煤化工MYHCNM HG.Dam Projcct. Science and Technology Review, 26(18), 45 48. (in Chinese)Qu, L, Li, R, Li, J., Li, K. F, and Deng, Y. 2011. Field observation of total dissolved gas supersaturation ofhigh-dams. Science in China, Series E: Technological Sciences, 54(1), 156 162. [doi:10.1007/s11431-010-4217-8.]Tan, D. C. 2006. Research on the Lethal Effect of the Dissolved Gas Supersaturation Resuted from ThreeGorges Project t0 Fish. Chongqing: Southwest University. (in Chinese)Tan, D. C., Ni, z. H, Zheng, Y. H., Li, L. Y, and Li, Y. F. 2006. Dissolved gas supersaturation downstream ofdam and its efects on fish. Freshwater Fisheries, 36(3), 56-59. (in Chinese)Urban, A. L, Gulliver, J. S, and Johnson, D. W. 2008. Modeling total dissolved gas concentrationdownstream of sillways. Journal of Hydraulic Engineering, 134(5), 550-561. [doi:10.1061/(ASCE)0733-9429(2008)134:5(550)]U.S. Army Corps' of Engineers (USACE). 2005. Technical Analysis of TDG Processes. Jacksoville: u.s.Army Corps of Engineers-Northwest Division, Environmental Resources and Fish Planning ffices.Watson, C. C., Walters, R. W, and Hogan, s. A. 1998. Aeration performance of low drop weirs. Journal ofHydraulic Engineering, 124(1), 65-71. [doi:10.1061/(ASCE)0733- 9429(1998)124:1(65)]Weitkamp, D. E., and Katz, M.1980. A review of dissolved gas supersaturation literature. Transactions of theAmerican Fisheries Society， 109(6)， 659-702. [doi:10.1577/1548-8659(1980)109<659:ARODGS>2.0.CO;2]404Lu QU et al. Water Science and Enginoering, Dec. 2011, Vol. 4, No. 4, 396 404中国煤化工MHCNMH G