วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 Oxygen-Transfer Measurement in Clean Water Zhen He*, Anurak Petiraksakul** and Warawitya Meesapya** 1. Abstract Introduction This paper presents the experiments on The oxygen transfer capacity and the aeration measurement of oxygen transfer capacity in clean efficiency characterize the performance and economy water absorption of aeration installations in activated sludge plants up a [1]. From 1978, the guideline for the determination small-scale tank with a volume of 17L. Pure of the oxygen transfer capacity has been published. oxygen was used to increase the dissolved In 1984, the American Society of Civil Engineers oxygen concentration in clean water (desorption (ASCE), with international participation, published measurement). While sodium sulfite was added to the ASCE Standard “Measurement of Oxygen decrease the dissolved oxygen concentration in Transfer in Clean Water”. The test water shall be absorption measurement. Standard oxygen transfer equivalent in quality to a potable public water supply. coefficient ( KLa20 ) was calculated based on the Repetitive testing may be conducted in the same water, variation of dissolved oxygen concentration with provided that the TDS is not over 2000 mg/L [2]. time. For the absorption method, a mean value This article presents the methods of measurement of KLa20 can be obtained as 8.60 h-1 with a based on the second edition of ASCE Standard water standard and German ATV Standard, and verifies the oxygen transfer efficiency (SOTE) was shown influence of measurement condition. A diffuser was in the range of 4.5-4.9% with water depth 0.3m adopted to convey the oxygen into a batch tank. by correcting the airflow condition. Desorption Absorption and desorption measurements were used measurement was investigated to certify the to determine the oxygen transfer coefficient and influence of water depth on SOTE. All the data oxygen transfer efficiency. All the results were on dissolved oxygen concentration and tested analyzed based on standard condition, which is water temperature was read by an electronic DO defined as water temperature of 20°C and normal meter. atmospheric pressure (1013 hPa). A DO meter is the by techniques. using desorption Experiments volume of 14L. were and set Meanwhile, in main instrument to measure the dissolved oxygen Keywords : dissolved oxygen concentration, oxygen concentration, which can be used to calculate the transfer coefficient ( KLa20 ), standard transfer rate. Finally, the paper concludes the oxygen transfer efficiency (SOTE), characteristic of the diffuser by SOTE. Meanwhile, absorption various influence factors will be analyzed by the measurement, desorption data treatment. measurement and water depth. * Department of Environment & Resource, Technical University of Denmark. ** Department of Chemical Engineering, Faculty of Engineering, King Mongkut’s Institute of Technology North Bangkok. 14 วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 2. Theory 2.2 Summary of Test Methods 2.1 Definitions 2.2.1 Absorption Measurements 2.1.1 Oxygen Transfer Coefficient (KLaT , h-1 ) The oxygen transfer is determined from the KLaT is determined by evaluation of an oxygen increase of the previously, artificially lowered DO transfer test in clean water at a certain aeration concentration [1]. The depletion of oxygen can be setting and at a certain temperature. It is implemented by either adding chemicals or stripping ° converted to the standard temperature of T = 20 C with nitrogen gas. Normally, sodium sulphite is as follows: used to decrease the oxygen by the reaction: K L a 20 = K L a T * 1.024 (20 - T) 2Na2SO3 + O2 → 2Na2SO4 (1) (5) To remove 1 kg of dissolved oxygen 8 kg of 2.1.2 Standard Oxygen Transfer Rate (SOTR20, kg-O2.h- 1) Na2SO3 is required. In order to expedite the reaction, SOTR shows the amount of oxygen transferred the catalyst of cobalt should be introduced into a per hour at the standard condition. It can be test tank. After the DO reach zero, the aeration calculated by the following: system will be switched on. By dissolving the oxygen of the air into the water, the oxygen SOTR = K L a 20 * C s ,20 * V concentration increases according to the saturation (2) function: C s ,20 where is the steady-state ° saturation concentration at 20 C and V DO C t = C s - ( C s - C 0 ) * exp ( - K L a T * t) (6) is the liquid volume. where Ct , Cs and C0 is the DO concentration at time 2.1.3 Standard Oxygen Transfer Efficiency (SOTE, t, saturation point and initial point, respectively. %) KLa can be obtained by the relationship between SOTE refers to the fraction of oxygen in Ct and t. an input airflow dissolved under the standard condition. It can be computed by: 2.2.2 Desorption Measurements The oxygen transfer is determined from SOTE = SOTR / WO2 (3) the decrease of the previously increased DO concentration. Pure oxygen should be used to where WO2 (kg/s) is the mass flow of oxygen in air improve the DO concentration. Hydrogen peroxide stream, which is calculated by: can be an alternative of pure oxygen in most cases. An increase of DO of about 10 mg/L should be WO2 = 0.2765 Q s (4) achieved before the measurement. The decrease of the DO follows a reversed saturation function, Q s refers to air flow rate at standard condition which equals to the one in absorption measurements. ° defined as 0 C, 1.00 atm, and dry air (0% relative humidity) [1]. Therefore, data about humidity, 2.3 Correction Factors pressure and temperature are necessary for the Due to the difference between the test condi- conversion of the airflow rate. tion and the standard condition, the data from the 15 วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 measurement should be converted into standard 18 condition by some correction factors. 15 Absorp tion  1  C s ,20 = C s ,T    τΩ  12 DO (mg/l) (7) τ is the temperature correction factor equal to C st, T /Cst, 20 where C st, T and C st, 20 are the tabular value of DO surface saturation concentration Desorp tion 9 6 3 0 0 at test temperature and 20°C, standard total pressure 100 200 300 400 500 600 Time (s ) of 1.00 atm and 100% relative humidity. Ω is the pressure correction factor, which can be described Figure 2 DO variation in different processes as Pb / Ps where Pb and Ps are the barometric pressure at test site during test and at standard by the Figure 2. The variation of DO concentration state (1 atm). In this experiment, an assumption was is completed by either pure oxygen injection or made that Ω is 1. chemical addition. 3. Experimental Process 3.2.1 Calibration of DO Probe Calibration of DO probe is necessary for the 3.1 Set-up The equipment includes a 17 L of cylindrical precise measurement. The methods used in this tank with a diameter of 24 cm., a DO meter (YSI experiment consist of zero check and saturation model 550), a cylindrical ceramic diffuser, airflow check. The zero check is run every time after the meter and aerating device (Figure 1). The DO probe probe is turned on, by putting probe into a 1-L tube was installed at half water depth for absorption added by 1 g sodium sulphite and 1 mg cobalt. measurement. If more DO probes are applied, they can With excess chemical, the water in the tube should be installed at different water depth. Conductivity meter contain zero dissolved oxygen. Saturation check was used to measure the content of dissolved solids. can be done before and after measurement, according to the operation manual for the probe [3]. 3.2 Operation Process 3.2.2 Temperature Measurement Both absorption measurement and desorption measurement were applied in the experiment. The The temperature of test water can be read from difference between two methods can be explained the DO meter. The variation of temperature shall be with an accuracy of ± 0.5 ๐C at the beginning and the end of each test [1]. Probe M eter Batch tank 3.2.3 Airflow Rate Measurement DO Probe A laboratory flowmeter made by G.A. Platon Ltd is used to measure the airflow rate. The Diffuser Air flow meter accuracy of the flowmeter has been checked before Air pump application. The inverse measuring cylinder filled with water was used to calibrate air flowrate. The airflow of 1550 cm3/min was applied in the measurement. Figure 1 Sketch for experiment set-up 16 วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 3.2.4 Water quality 0.04 TDS (Total Dissolved Solid) and conductivity of the test water were measured by a conductivity AC (mg/L) 0.02 meter, when each test was finished. If TDS shows a value higher than 2000 mg/L, the test water (clean water) shall be changed. 0.00 0 100 200 300 400 500 -0.02 3.2.5 Adding Chemicals (Absorption Measurements) -0.04 All the chemicals should be dissolved before Time (s) they are added. The amount of sodium sulphite should be 10-15% more than the calculated result Figure 4 Plot of residues of DO from a test due to a lag time for admixture. Normally a cobalt AC: Difference between the measured concentration in water of 0.5 mg/L is sufficient. DO and the predicted DO If the test water is kept same, one time of addition of cobalt is enough. Cobalt should be added before should not exceed 0.8 Cs, T . At least 30 values should sodium sulphite. be collected for the data analysis. Desorption method demands a start point that is 10 mg/L higher than the saturation value. 3.2.6 Desorption Measurements Pure oxygen and air tubes were connected to a three-way valve fitting the other end to the diffuser. 4. Data Treatment Pure oxygen was fed from an oxygen cylinder by a Linear regression method is used to analyze pressure-regulator valve through the diffuser into the data from the experiment. This method is based the clean water. When the DO concentration was on the model through the DO-versus-time data. increased to about 17 mg/L, the regulator was shut However, equation (5) gives a nonlinear relationship off and the air pump was switched on. between DO concentration and time. Transformation of this equation is necessary before it is applied. A linear relationship can be obtained by equation (8). 3.2.7 Data Collection With absorption method, when the concentration  C − Ct ln s  C s − C0 of oxygen increases from zero, the test data can start at about C = 0.1 Cs, T . The end point of the test   = K L aT * t   (8) See in Figure 3, the slope of the plot between  C s − Ct  LnY = ln C − C  and time giving a value of is 0   s 0.0031 s-1. The best estimates of the parameters, y = 0.0031x + 0.0024 1.50 2 R = 0.9998 Ln Y 1.00 K L a T , is selected as the values that drive the model 0.50 equation through the prepared DO concentrationversus-time data points with a minimum residual 0.00 sum of squares [2]. The residual refers to the 0 100 200 300 400 difference in concentrations between a measured 500 Time (s) DO value at a given time and the DO value predicted by the model at the same time (Figure 4). Figure 3 Plot of LnY versus time 17 วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 If the residue shows a curve, normally the 4.9 4.8 the lagging sodium sulphite oxidation or unstable 4.7 SOTE (%) initial values of DO concentration are falsified by mixing conditions. Incorrect values at the end of the curve could influence the curved path of residues. 4.6 4.5 So, a new calculation is necessary, leaving out several 4.4 initial values or several final values [1]. 4.3 27.6 27.7 27.8 28.2 28.6 2 9 o 5. Result and Discussion T e s te d wa te r te mp e ra tu re ( C) 5.1 Determination of KLa20 in Clean Water Absorption method is adopted to measure the Figure 6 Results of SOTE in clean water oxygen transfer rate in clean water. Results have been converted into the standard condition by a temperature correction factor shown in equation (1). 4. 6 Figure 5 shows the results from the water sample SOT E( % ) with a depth of 0.30m. One thing should be noticed that the following results are not trying to explain the relationship between temperature and K L a 20 or SOTE. Theoretically, values of KLa20 should be 4. 4 4. 2 same due to the temperature correction. 4 0.2 The values of KLa20 are between 8.3 and 9 h-1 0.25 0.3 0.35 Wat er dept h ( m ) in the range of tested water temperature of 27.6-29°C. The mean K L a 20 is 8.60 h-1. Temperature, water Figure 7 Effect of water depth on SOTE depth, and airflow could be the influence factors. The results of SOTE values are within the readings generally require a liquid velocity of at range of 4.52 -4.81% as shown in Figure 6. The least 0.3 m/sec near the water interface [2]. In this values are much smaller than the reference numbers experiment, we avoid stagnation by rapidly moving because of the low water depth. Water turbulence the probe through the sample [3]. may affect the probe readings. Accurate probe -1 Desorption technique was use to verify the 9 influence of water depth in the clean water. Five 8.8 different water depths (0.24, 0.26, 0.27, 0.30, and 8.6 0.32m) are chosen as test objective. L K a 2 0 (h ) 5.2 Influence of Water Depth 9.2 8.4 Since the diffuser is installed at the bottom of 8.2 the tank, water depth can indicate the diffuser 8 depth. Figure 7 shows that SOTE increases with the 27.6 27.7 27.8 28.2 28.6 2 9 water depth because the contact time increases o T es ted w ater tem perature ( C) between the bubble and the water. The increase of Figure 5 Results of transfer rate in clean water by oxygen partial pressure could be another reason for absorption measurement SOTE’s increasing [5]. However, this relationship 18 วารสารวิชาการพระจอมเกล้าพระนครเหนือ ปีท่ี 13 ฉบับที่ 1 ม.ค. - มี.ค. 2546 The Journal of KMITNB., Vol. 13, No. 1, Jan. - Mar. 2003 only gives a tendency of variation of SOTE values, strictly, such as temperature, airflow rate and stir of instead of the precise values. Due to the small interval water. The precision of experiment can be improved of water depths chosen in this experiment, it is by applying more DO probes at the different test impossible to measure the exact change of SOTE points. with water depth. For a precise measurement of SOTE 7. Acknowledgement variation with water depth, a larger interval (0.5 m or The authors would like to acknowledge. The more) between water depths should be chosen. Thailand Research Fund for financial support. They 5.3 Comparison of Absorption and Desorption also would like to thank Department of Chemical Measurement Engineering of KMITNB and IAESTE Thailand for their kind help. Figure 7 gives a SOTE value of 4.56% at water depth 0.3m by desorption measurement, which is within the range of 4.52-4.81% (Figure 6) at the same References water depth from absorption measurement. Both 1. German ATV Standards : Measurement of the methods can meet the requirement for clean water Oxygen transfer in Activated Sludge Aeration measurement. Tanks with Clean Water and in Mixed Liquor. With addition of chemicals in pp.1-54. 1996. absorption measurement, TDS shall be controlled 2. under 2000 mg/L; while no consideration of TDS ASCE Standard : Measurement of Oxygen is necessary with desorption measurement. The Transfer in Clean Water. Second Edition. economic factor shall be taken into consideration pp.1-42. 1992. when choosing a method. The comparison of cost 3. Operations Manual for YSI 550 Handheld between pure oxygen and chemicals added can be Dissolved oxygen and Temperature System, investigated further. YSI Incorporated. pp.1-27. 4. pr-European Standard : Wastewater treatment plants - Part 15 : Measurement of the oxygen 6. Conclusion transfer in clean water in activated sludge The practice of absorption and desorption methods for measurement of oxygen transfer aeration tanks: pp.1-16. 1999. coefficient in the laboratory, have been presented in 5. this paper. The test condition should be controlled ASCE Water Environment Federation : Aeration A Wastewater Treatment Process. pp.21-71.1996. 19