GERMAN ATV RULES AND STANDARDS W A S T E W A T E R
-
W A S T E
ADVISORY LEAFLET ATV M 209E Measurement of the Oxygen Transfer in Activated Sludge Aeration Tanks with Clean Water and in Mixed Liquor June 1996 ISBN 3-934984-50-9
Marketing: Gesellschaft zur Förderung der Abwassertechnik e.V. (GFA) Theodor-Heuß-Allee 17 D-53773 Hennef Postfach 11 65, 53758 Hennef
ATV M 209E The following members belong to the ATV Working Group 2.6.2 "Measurement of Oxygen Transfer under Process Conditions" which prepared this Advisory Leaflet: Prof. Dr.-Ing. R. Kayser, Braunschweig (Chairman) Dr. techn. W. Frey, Wien Prof. Dr.-Ing. H. Kapp, Stuttgart Dr.-Ing. B. Teichgräber, Essen Dr.-Ing. M. Wagner, Darmstadt The following have participated as guests: Dipl.-Ing. G. Fröse, Braunschweig Dipl.-Ing. T. Grimm, Scharbeutz Dipl.-Ing. J. Reichert, Darmstadt
The Advisory Leaflet presented here has been prepared within the framework of the ATV committee work, taking into account the ATV Standard A 400 "Principles for the Preparation of Rules and Standards" in the Rules and Standards Wastewater/Waste, in the January 1994 version. With regard to the application of the Rules and Standards, Para. 1 of Point 5 of A 400 includes the following statement “The Rules and Standards are freely available to everyone. An obligation to apply them can result for reasons of legal regulations, contracts or other legal grounds. Whosoever applies them is responsible for the correct application in specific cases. Through the application of the Rules and Standards no one avoids responsibility for his own actions. However, for the user, prima facie evidence shows that he has taken the necessary care.
All rights, in particular those of translation into other languages, are reserved. No part of this Advisory Leaflet may be reproduced in any form by photocopy, microfilm or any other process or transferred or translated into a language usable in machines, in particular data processing machines, without the written approval of the publisher. Gesellschaft zur Förderung der Abwassertechnik e.V. (GFA), Hennef 1996 Produced by: J. F. CARTHAUS GmbH & Co, Bonn
June 1996
2
ATV M 209E
Contents 1. 2.
Preamble
5
Important terms and relationships 2.1 Measurements in clean water and in mixed liquor 2.1.1 Measurements in clean water 2.1.2 Measurements in mixed liquor using the non-steady state method 2.1.3 Measurements in mixed liquor under steady state conditions 2.2 Oxygen transfer capacity, aeration efficiency 2.2.1 Oxygen transfer capacity in clean water 2.2.2 Oxygen transfer capacity in mixed liquor 2.2.3 Aeration efficiency in clean water 2.2.4 Aeration efficiency in mixed liquor 2.2.5 Oxygen transfer efficiency 2.2.6 Oxygen transfer coefficient 2.2.7 α-value (interfacial factor) 2.2.8 Standard oxygen saturation value 2.2.9 Oxygen saturation value 2.2.10 ß-value (salinity factor) 2.3 Methods of measurement and evaluation 2.3.1 Absorption measurements 2.3.2 Desorption measurements 2.3.3 Evaluation of absorption and desorption measurements 2.3.4 Off-gas measurements 2.4 Influences of tank geometry and mixing conditions 2.5 Selection of the procedure for the measurement of oxygen transfer
3. 3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 3.4
Execution of tests Measurement equipment DO probes Recording of DO concentration Technique of off-gas measurement Power input measurement Air flow rate measurement Temperature Chemicals Cobalt catalyser Sodium sulphite Hydrogen peroxide Pure oxygen or nitrogen gas Absorption and desorption measurements Determination of the test procedure Test preparation Arrangement of DO probes Warming up of the aeration installation Addition of chemicals Measurement and recording of data during a test Off-gas measurements
6 6 6 7 7 7 7 7 8 8 8 8 8 9 9 10 10 10 12 12 14 16 17 18 18 18 19 19 22 22 23 23 23 23 24 25 25 25 26 27 27 27 28 29
June 1996
3
ATV M 209E 4. 4.1 4.2 4.3
Test evaluation and report Absorption and desorption measurements Off-gas measurements Test report
31 31 31 32
5. 5.1 5.2
Measurement tolerances and guarantees Measurement tolerances Guarantees
33 33 33
6.
Symbols*
34
7.
Bibliography
36
Appendix 1: Table of standard oxygen saturation values
37
Appendix 2: Derivation of essential equations
38
Appendix 3: Desorption tests with the application of pure oxygen gas
42
Appendix 4: Evaluation of adsorption and desorption tests using a computer
46
*[Translator's note: In order not to complicate comparison between the original German text and this translation, suffixes used with symbols have not been changed from the original German language Standard although, in many cases, these suffixes are based on a German word, i.e. "L" for "Luft" has not been changed to "A" for "air"]
June 1996
4
ATV M 209E
1.
Preamble
In addition to the adaptability and controllability, the oxygen transfer capacity and the aeration efficiency characterise the performance and economy of aeration installations in activated sludge plants. In order to standardise the measurement technique for oxygen transfer tests a Guideline for the Determination of the Oxygen Transfer Capacity of Aeration Systems in Clean Water was published in 1978 [1]. In Austria, the ÖNORM (Austrian Standard Specification) M 5888 "Wastewater Treatment Plants, Oxygen Transfer Performance of Aeration Installations, Determination in Clean water" was published also in 1978 [2]. In the USA, in 1984, the American Society of Civil Engineers (ASCE), with international participation published the ASCE Standard (comparable with a Standard Specification) "Measurement of Oxygen Transfer in Clean water". Following minor amendments this was published in 1992 in a 2nd Edition [3]. Completely new is the "Standard Guideline for In-Process Oxygen Transfer Testing" [4] also elaborated by the ASCE with international participation. While both the measurement technique and the test evaluation are equal for the determination of the oxygen transfer in clean water and under process conditions (with the exception of off-gas measurements), both measurements in clean water and also measurements under process conditions are dealt with in this Advisory Leaflet. In Germany and Austria one still works with a reference temperature of T = 10° C for oxygen transfer [1,2]. As a rule the highest oxygen demand occurs with the highest reactor temperature (in ATV Standard A 131 it is, for example, recommended that the aeration system should be designed for the oxygen demand at T = 20° C). Internationally it has, for some considerable time, been usual to relate the oxygen transfer to T = 20° C [3, 4]. In this Advisory Leaflet T = 20° C is therefore also taken as the standard temperature. For conversion the following applies: OC (T=20° C) = 1.02 · OC (T=10° C) The difference between the values for T = 10° C and T = 20° C is, therefore, negligibly small. From clean water tests there is sufficient experience available on many types of aeration installation. The requirements for guarantee tests in clean water should therefore be limited to special cases; additionally, with small plants, the costs for clean water tests often bare no relation to the benefit. Much more important and interesting for the operating behaviour of aeration installations are measurements under process conditions with mixed liquor. Here, principally the maximum oxygen transfer capacity and the aeration efficiency with the current average loading are of interest.
June 1996
5
ATV M 209E In this Advisory Leaflet, due to past experiences, unclear situations and/or the existence of different possibilities often have the note "to be agreed beforehand". This agreement must be made between the customer and the supplier of aeration facilities and, if required, also with the institution which carries out the oxygen transfer tests. This Advisory Leaflet applies only for the determination of the oxygen transfer in activated sludge plants. It does not apply for aerated wastewater lagoons. The following are to be agreed and, as far as possible, already taken into account with the request for offers: • Which water, if not drinking water, will be used for clean water tests, comp. 2.1.1. • For which aeration settings and how many tests with each aeration setting are to be carried out, comp. 3.3.1. • Which measurement technique is to be applied, comp. 2.5 • Whether measurements are to be carried out even with very low water temperatures, comp. 3.1.6. • Whether sodium sulphite should be added in dry form, comp. 3.2.2. • How the power input of the aeration installation is to be measured, comp. 3.1.4. • How the air flow rate is to be measured, comp. 3.1.5. • With what deviations are given guarantees to count as being met, comp. 5.
2.
Important Terms and Relationships
2.1
Measurements in Clean Water and in Mixed Liquor
2.1.1
Measurements in Clean Water
For clean water tests the aeration tank is filled with clean water. Water with drinking water quality [5] should be used. With this water the highest reproducibility of the tests results is ensured. If, due to the high costs, no drinking water can be used then this is to be agreed on, i.e. already with the placing of the order for the aeration installation. A low content of neutral salts is important, therefore the concentration of dissolved solids should be ≤ 500 mg/L. By addition of sodium sulphite for the tests the salt content increases further. Tests can be carried out up to a salt concentration of 2000 mg/L dissolved solids; this corresponds with an electrolytic conductivity of ca. 3000 µS/cm. Organic substances in the test water lead to complexing of the cobalt applied as catalyst; they can in addition reduce the oxygen transfer. Biologically treated wastewater or clean water coloured by algae is therefore not usable. Salting as well as the use of cobalt are avoided if the dissolved oxygen (DO) concentration is lowered by stripping with nitrogen gas or if the DO is increased for desorption tests by gassing with pure oxygen. 2.1.2
Measurements in Mixed Liquor using the Non-Steady State Method
For application of the non-steady state method, the wastewater flow and the return sludge flow and, if installed, the internal circulation flow are interrupted before the start of June 1996
6
ATV M 209E testing. The aeration is set to a predetermined value. The test is started only when an almost constant respiration rate of the mixed liquor, also characterised by a constant DO, has been reached. As through this TOC, COD and surfactants are biodegraded to constant low concentrations, the reproducibility of tests at different times with the same aeration setting is higher than with measurements under steady state conditions. Both, the off-gas method as well as the absorption or desorption method are possible. 2.1.3
Measurements in Mixed Liquor under Steady State Conditions
For steady state conditions the aeration tank with the mixed liquor remains in full operation, however, the aeration is set to a predetermined value. As the oxygen transfer coefficient (αkLa20) is dependent on the water quality (salt content, TOC, COD surfactants etc.), measurements on different days and at different times may show different αkLa20 or αOC values. The off-gas method is ideal for measurements under steady state conditions. With adsorption or desorption tests a constant respiration over longer periods is a precondition for usable results. Such measurements are therefore only practical if, through constant loading (e.g. balancing tanks), the respiration can be kept constant. 2.2
Oxygen Transfer Capacity, Aeration Efficiency
2.2.1
Oxygen Transfer Capacity in Clean water, OC [kg/h]
OC is the mass [kg] of oxygen transferred by an aeration installation in one hour in a tank of certain size filled with clean water at DO of C = 0 mg/L, a water temperature of 20° C and normal atmospheric pressure (1013 hPa). The following applies: OC = 2.2.2
V ⋅ k L a 20 ⋅ C S,20
(1)
1000
Oxygen Transfer Capacity in Mixed Liquor, αOC [kg/h]
αOC is the mass [kg] of oxygen transferred by an aeration installation in one hour in a tank of certain size with mixed liquor at DO of C = 0 mg/L, a water temperature of 20° C and normal atmospheric pressure (1013 hPa). The following applies: αOC = 2.2.3
V ⋅ αk L a 20 ⋅ β ⋅ C S,20
(2)
1000
Aeration Efficiency in Clean Water, OP20 [kg/kWh]
OP is the oxygen transfer in clean water (OC) divided by the power input P [kW] of the aeration installation including associated mixing installations. The following applies:
OP = 2.2.4
OC P
(3)
Aeration Efficiency in Mixed Liquor, αOP20 [kg/kWh]
αOP is the oxygen transfer in mixed liquor (αOC) divided by the wire power uptake P [kW] of the aeration installation including associated mixing installations. The following applies: June 1996
7
ATV M 209E αOC P
αOP = 2.2.5
(4)
Oxygen Transfer Efficiency, OCL,h [g/(m3N·m)] or OAh, [%/m]
The oxygen transfer efficiency expressed as g oxygen per m³ of air and per m of diffuser depth (OCL,h) or as % oxygen absorbed per m of diffuser depth (OAh) is used for the comparison of diffused air aeration systems. Following many investigations OCL,h and OAh are, down to diffuser depths of hE = 6.0 to 8.0 m, to a large extent independent of the diffuser depth [6]. The following applies with QL [m3N/h] (dry intake air flow rate at 0° C and 1013 hPa) and the associated density of the oxygen in the air of 0.299 kg/m3: OC ⋅ 1000 Q L ⋅ hE
(5)
100 ⋅ OC hE ⋅ (Q L ⋅ 0.299)
(6)
OCL,h =
OA h =
Note: The temperature of the intake air is to be set at T = 20° C when designing the blower. In the case that the intake air is used for cooling the blower a correspondingly higher temperature is to be selected.
αOAh and αOCL,h for conditions with mixed liquor are calculated correspondingly. 2.2.6
Oxygen Transfer Coefficient, kLaT or αkLaT [h-1]
The oxygen transfer coefficient is the characteristic value for a certain aeration setting in a certain tank. It is calculated from the oxygen transfer test at temperature T [°C]. With the same aeration setting kLa (or αkLa) increases with increasing temperature. The following applies:
k L a 20 = k L a T ⋅ 1.024 ( 20 − T )
(7)
The value of θ = 1.024 has been selected, in the spirit of international agreement, to correspond with the ASCE Standard [3]. In the previous Advisory Leaflet [1] θ = 1.02 was used. Extrapolating a test temperature of T = 10° C to the standard temperature of T = 20° C kLa is approximately 4 % greater than with the application of θ = 1.02. 2.2.7
α-Value (Interfacial Factor),[-]
Through various substances of wastewater, primarily surface active substances (surfactants), kLa in wastewater and in mixed liquor is smaller than in clean water at otherwise similar conditions. The α-value serves for comparison and is defined as: k a in mixed liquor α = L 20 k L a 20 in clean water
(8)
Both kLa20 values must be determined in the same tank with the same aeration installation and with the same aeration intensity, as α is not only dependent on the
June 1996
8
ATV M 209E properties of the wastewater or mixed liquor but also to a considerable extent on the aeration system [7]. With diffused air aeration systems αkLa increases with increasing neutral salt content due to the reduction of the coalescence of bubbles [8]. 2.2.8
Standard Oxygen Saturation Value, CSS,T [mg/L]
The standard oxygen saturation value occurs in water with constant and same temperature of water and water saturated air at a pressure of 1013 hPa. Standard saturation values are shown, dependent on the temperature, in a table in DIN 38 408, Part 23 (see Appendix 1). 2.2.9
Oxygen Saturation Value, CS,T [mg/L]
The oxygen saturation value occurs in a tank with clean water at constant aeration and constant water temperature (T), i.e. the saturation value is achieved if the oxygen concentration in the tank no longer alters. The determination of the oxygen saturation value is possible in clean water tests only. It is determined either chemically in accordance with EN 25813 or by means of precisely calibrated DO probes in accordance with EN 25814. Alternatively the supersaturation can be measured [9]. The precise calibration of the DO probes is not necessary for the determination of kLa. To convert the saturation value CS,T measured at T° C, to the standard conditions (T = 20° C, p0 = 1013 hPa), one uses the standard saturation values and takes account of the actual atmospheric pressure p [hPa]. The following applies: C S,20 = C S,T ⋅
C SS,20 1013 ⋅ C SS,T p
(9)
With surface aeration systems the following can be applied as approximation: C S,T = C SS,T
(10)
For diffused air aeration the pressure at half the diffuser depth can, based on considerable experience, be used as an approximation for the calculation of the saturation value. The following applies: h C S,T = C SS,T ⋅ 1 + E 20.7
(11)
Using Eqns. 10 or 11 one saves carrying out the precise calibration of the DO probes. If, in one tank, there are both surface aerators as well as diffused air aeration operating then the determination of the oxygen saturation value in situ is necessary at clean water tests. 2.2.10
ß-Value (Salinity Factor)
The oxygen saturation value is reduced by neutral salts. This is expressed by the ßvalue. The following applies:
June 1996
9
ATV M 209E C S,20 in wastewater β = C S,20 in clean water
(12)
In municipal wastewater usually ß = 1.0. One can calculate the ß value approximately with the aid of the total dissolved solids content TDS [mg/L] [4]: β = 1.00 − 0.01⋅
TDS 1000
(13)
More precise details can be found in EN 25 814. 2.3
Methods of Measurement and Evaluation
2.3.1
Absorption Measurements
With absorption measurements the oxygen transfer is determined from the increase of the previously, artificially lowered DO concentration. With tests using clean water the DO in the water is removed by the addition of a certain quantity of sodium sulphite. The DO concentration can also be lowered by stripping with nitrogen gas. It must be ensured that, after switching on of the aeration system, the mixing conditions become stable at C = 0 mg/L. Through absorption (dissolving) of the oxygen of the air in the water, the oxygen concentration increases according to the saturation function (Fig. 1). The following equations are explained in more detail in Appendix 2. C t = C S − (C S − C 0 ) ⋅ exp ( −k L a ⋅ t)
(14)
At measurements with mixed liquor, after a constant oxygen uptake rate was reached, the aeration is switched off and the tank is mixed or the aeration is throttled. As a result of respiration the DO drops to zero following a linear function. After adjustment of the aeration to the setting be investigated the DO increases in the mixed liquor also in accordance with a saturation function until the apparent saturation value C* is reached (Fig. 2). C t = C * −(C * −C 0 ) ⋅ exp [− (αk L a + q) ⋅ t ]
(15)
At non-steady state tests the water flow becomes zero (q = 0)
June 1996 10
ATV M 209E
Fig. 1:
Course of DO at an absorption test in clean water
If the apparent saturation value C* is smaller than approx. 0.5·CS then αkLa can become uncertain. If the respiration rate of the mixed liquor drops during the test, then a too low value of αkLa is determined; conversely with increasing respiration rate one obtains a too great value of αkLa. Measurements with C* < 0.5 · Cs should, as far as possible, be avoided.
Fig. 2:
Course of DO at an absorption test in mixed liquor
June 1996 11
ATV M 209E
Fig. 3:
Course of DO at a desorption test in mixed liquor
2.3.2
Desorption Measurements
With desorption measurements the oxygen transfer is determined from the decline of the previously increased DO concentration. The DO can be increased both in clean water and in mixed liquor by gassing with pure oxygen. In most cases hydrogen peroxide is used to increase the DO of the mixed liquor. Then a 35 % solution of H2O2 is dosed directly into the aeration tank with aeration operating at the setting to be investigated, whereby C* should be > 2.0 mg/L [10]. The DO concentration in the tank increases with a jump due to the released oxygen (Fig. 3). Enough peroxide should be dosed so that an increase of DO of about 10 mg/L is achieved. The decrease of the DO follows a reversed saturation function which, in principle, equals Eqn. 15. With non-steady state tests q is again zero. It should be noted that the respiration rate should also be constant with desorption measurements. If the respiration rate of the mixed liquor falls during the test then a too high value for αkLa is calculated; with increasing respiration a smaller value for αkLa is found. 2.3.3
Evaluation of Absorption and Desorption Measurements
With the measurement of absorption and desorption one obtains value pairs Ct and t. Previously it was usual to plot the saturation deficit (CS - Ct) or C* - Ct against time on semi-log paper and to calculate kLa or αkLa from the slope of the compensating straight line. If the value pairs (Cs - Ct) against t showed a bent path, CS was altered until a linear path was found [11].
June 1996 12
ATV M 209E Today the parameters C0, CS or C* and kLa or αkLa are determined by non-linear regression from the given value pairs Ct and t. A computer programme from BROWN and FISETTE [12] is most frequently used. In the ASCE Clean Water Standard [3], programmes are printed in BASIC and FORTRAN; the standard provides a disk in Turbopascal for PCs. There is also a disk provided with this Advisory Leaflet (see Appendix 4). In addition to the parameter C0, the DO at the selected at time t0, the oxygen transfer coefficient kLa (or αkLa) and the saturation value CS (or C*) and the residues, i.e. the differences between calculated and measured DO values are also issued. The distribution of the residues over time says a great deal about the quality of the test data. If residues show a random distribution (Fig. 4) the result is, in general, all right. If, however, residues show a curve, then usually the initial values of DO are falsified by the lagging sulphite oxidation or by still not yet stabilised mixing conditions. In individual cases the curved path of the residues can also be caused by incorrect values at the end of the adsorption or desorption curve. Then a new calculation is necessary, leaving out several initial values (postponement of to) or several final values. This is demonstrated in an example (see Appendix 4).
Fig. 4:
Plot of residues of DO from a good test (above) and a disturbed test (below).
June 1996 13
ATV M 209E 2.3.4
Off-Gas Measurements
Off-gas measurements are possible only in activated sludge tanks with diffused air aeration. For a series of measurements the air flow is set to a constant value. As aeration tanks, as a rule, do not have a gas proof covering, the exhaust air (off-gas) is collected using a floating collection hood, AA [m2]. The following values are measured: the off-gas flow rate, QL,e [m3/h], the volumetric fraction of oxygen of the off-gas, Xe, as well as of the ambient air Xi, the temperature in the off-gas (TL) and the DO (C) and the temperature (T) of the mixed liquor below the hood. The DO of the mixed liquor should be constant (C = C*) otherwise dC/dt must be taken additionally into account. The measured value of the off-gas method is the volumetric oxygen uptake rate, OVR* [g/m3·h] calculated from the differences of the mass of oxygen in the input air and the offgas. With the air flows QL,i and QL,e in m3N/h the following applies: OVR * = 1000 ⋅ ρ O 2 ⋅
Q L,i ⋅ X i − Q L,e ⋅ X e
(16)
A A ⋅ hW
From the oxygen uptake rate the oxygen transfer coefficient results as: αk L a 20 = OVR * ⋅
1 ⋅ 1.024 (20 − T ) β ⋅ C S,T − C
(17)
As QL,i is not measurable, one applies some sleight of hand: the off-gas to be measured is dried and freed of CO2, i.e. it contains only O2 and N2. Then the molar ratio MV [mol O2/mol N2] can be formed: MVi =
Xi 1− X i
(18)
MVe =
Xe 1− Xe
(19)
Under the assumption that no N2 from denitrification processes is stripped, the number of moles of N2 is the same in the inflow air and the off-gas. Then the actual oxygen transfer efficiency, αOA* [%], becomes: αOA* =
MVi − MVe ⋅ 100 MVi
(20)
The oxygen transfer efficiency for standard conditions is obtained as follows: ß ⋅ CS,T C MVi 100 ( 20 − T ) αOAh = 1 − . ⋅ ⋅ SS,20 ⋅ 1024 ⋅ MVi he ß ⋅ CS,T − C CSS,T
(21)
The saturation value CS,T can be taken from earlier tests with clean water in the same tank. When it has to be calculated according to Eqn. 11, the atmospheric pressure at the time of the test has to be taken into account. June 1996 14
ATV M 209E h p C S,T = C SS,T ⋅ 1 + E ⋅ 20.7 1013
(22)
From Eqns. 17 and 21 it is clear that, with small saturation deficits (CS,T - C) due to a false assumption of the saturation value (CS,T) or an erroneous DO probe calibration and, resulting from this, a false DO concentration (C), the error band for the oxygen transfer efficiency can be considerable. Therefore the tests should only be carried out at DO concentrations of C ≤ 0.5 CS,T. With smaller DO concentrations of C ≤ 1.0 mg/L denitrification can take place. Due to the stripping of nitrogen the measured value Xe is falsified. Therefore at low DO concentrations tests should not be conducted. With several measurements with the same air flow in the same tank with evenly distributed diffusers, it will be found that QL,e and αOAh vary in opposition (see Appendix 2). In circular tanks the diffusers are frequently arranged in strings. On the surface of the water there are areas in which bubbles occur and areas without bubbles. This is similar in closed loop tanks in which the diffusers are arranged in fields (Fig. 5). In such tanks several measurements have to be made in each field of bubbles.
Fig. 5:
Tanks with diffusers arranged in strings and in fields
As relevant value for any case the weighted mean of the oxygen transfer efficiency is formed for n measurements with the same aeration setting: n
αOA h =
∑ (αOA
h
⋅ Q L,e )
1
(23)
n
∑Q
L,e
1
The mean volumetric off-gas flow rate based on the tank volume below the hood is, in addition, calculated as reference value: n
qL,e =
∑Q
L,e
1
n ⋅ A A ⋅ hW
(24)
June 1996 15
ATV M 209E This can be used for comparison with the volumetric input air flow rate qL,e [m3N/(m3·h)] of the aeration tank in which the measurements are carried out. qL,R =
QL VBB
(25)
In tanks with evenly distributed diffusers qL,e should lie between 0.75 and 1.25 qL,R. If one wishes to establish a reference between the oxygen transfer efficiency and the specific input air flow rate of the tank, one must calculate QL,i. This requires, additionally, the measurement of the CO2 content of the off-gas, moreover the partial pressure of the water vapour of the inflow air and the off-gas must be taken into account [13] (see Appendix 2). The results, however, are no more precise than the here recommended calculation using the molar ratios. 2.4
Influences of Tank Geometry and Mixing Conditions
Strictly, Equations 14 and 15 are only valid for completely mixed tanks. That is theoretically such tanks in which with aeration, if required supported by mixing facilities, the same concentrations can be found at the same time at all points. Rectangular tanks without separating walls (e.g. tanks with three cone surface aerators) and rectangular tanks with even diffuser density over the floor or along the longitudinal axis of the tank can, with regard to oxygen transfer testing, be considered as completely mixed (Fig. 7). Rectangular tanks with staged diffusers are not completely mixed tanks. Due to the pumping effect of the aeration there is an exchange of water between areas of different diffuser density. Therefore one will find, in the areas of the highest diffuser density a smaller value of kLa value than in an independent, completely mixed tank with the same diffuser density. Conversely, in the area of the smaller diffuser density a higher kLa value is measured.
Fig. 6:
Typical tank shapes with different aeration arrangements
June 1996 16
ATV M 209E Closed loop tanks are considered to be completely mixed. In closed loop tanks with few aerators (e.g. carousel) the rising or decreasing DO curve is stepped. The residues, therefore, are larger than in truly completely mixed tanks but they are randomly distributed. 2.5
Selection of the Procedure for the Measurement of Oxygen Transfer
The following procedures are available: (1) In clean water: (1.1) Absorption measurements following the lowering of the DO concentration using sodium sulphite. (1.2) Absorption measurements following the lowering of the DO concentration by injection of nitrogen gas. (1.3) Desorption measurements following the raising of the DO concentration by the injection of pure oxygen gas. (2) In mixed liquor: (2.1) Absorption measurements following the lowering of the DO concentration by the respiration of the mixed liquor. (2.2) Desorption measurements following the raising of the DO concentration by the addition of hydrogen peroxide or by the injection or admixture of pure oxygen gas. (2.3) Off-gas measurements. In clean water absorption measurements according to (1.1) are considered as standard practice. The stripping of the oxygen with nitrogen (1.2) is only possible in tanks with diffused air aeration; this applies practically also for the raising of the DO concentration with pure oxygen (1.3). Both procedures (1.2) and (1.3) are practical if a large number of tests are to be carried out with the same water. With greater diffuser depths and larger tank volumes desorption tests (1.3) can be more economical than absorption tests according to (1.1) and (1.2). Absorption and desorption measurements in mixed liquor in a non-steady state (2.1 and 2.2) have the greatest reproducibility. For measurements in tanks with diffused air aeration with steady state conditions, off-gas measurements (2.3) are advantageous as, during the short duration of the individual measurements, a change of the respiration rate of the mixed liquor has hardly any negative influence.
June 1996 17
ATV M 209E
3.
Execution of Tests
3.1
Measurement Equipment
3.1.1 DO probes At least 3 DO probes are necessary for absorption and desorption measurements. For tests in large tanks 6 probes are recommended in order that, with failures of probes, the expensive tests do not need to be repeated (comp. 3.3.3).
A very accurate calibration of the DO probes is not necessary for absorption and desorption measurements if the oxygen saturation value in accordance with Eqns. 10 or 11 is to be the basis for the calculation of OC. In tanks with greater diffuser depths the measurement of the supersaturation [9] instead of the accurate calibration of all probes is recommended for clean water tests. With this it is determined by how much higher the DO concentration in the tank at the end of the test (saturation) is compared with the condition in a flat vessel. In order to eliminate errors as far as possible, water with the same salt content and same temperature as in the test tank should be used in the flat vessel (e.g. bucket). In each case a bucket of water is taken out of the aeration tank approximately halfway through the test and aerated intensively. The bucket is placed in a water bath filled with water from the tank (constant temperature). A DO probe is placed in the bucket. By the end of the test a constant DO concentration (saturation) will be measured in the bucket. A possible initially present supersaturation is removed through the aeration. By changing the steepness setting of the DO meter a value of, for example, 60 % saturation is set for the DO concentration in the bucket. At the end of the test the DO concentration in the tank is read with this probe as, for example, 76.4 %. The supersaturation is 76.4 : 60 = 1.27. The standard saturation value (CSS,20) is multiplied by this, one then obtains CS,20. It should be noted that saturation values higher than in accordance with Eqn. 11 have still not been measured. For example, in a tower-biology system (hE = 20 m) supersaturations were measured corresponding to hE/3 [9]; Eqn. 11 is based on hE/2. For absorption and desorption measurements the response of the DO probes is important. The time which a probe requires to indicate a jump in concentration of 90 % DO should be less than 1/20th of the time which is required in the aeration tank to be tested in order to achieve a 90 % DO change [14]. As the DO measurement also follows a saturation function after a jump of the DO concentration a ”probe kLa" can be calculated. This should be 20 times greater than the kLa of the aeration facility to be tested. As the kLa value of the aeration facility is known approximately the following applies: t90%-probe [seconds] ≤ 415/kLa. With kLa = 5 h-1 one obtains the necessary reaction time t90% ≤ 83 seconds. Various DO probes require a relatively high flow rate (turbulence at the membrane), i.e. with reducing turbulence the instruments show a decreasing DO. In a pre-test it is to be tested whether, with the selected aeration facility, the turbulence is sufficient. This is not the case if the instrument shows a varying DO or if it increases when the probe is moved
June 1996 18
ATV M 209E by hand in the water. In cases of doubt precautions are to be taken to increase turbulence at the probes. In general, probes with short response times require a higher turbulence at the membrane. For measurement of the DO concentration in aeration tanks during off-gas measurements, the probes are to be calibrated with the greatest care (comp. 2.3.4.). In order to keep errors as small as possible the temperature at which the probes are calibrated should be as near as possible to the temperature of the mixed liquor in which the measurements are to be conducted. 3.1.2
Recording of DO Concentration
The signals from all DO probes are to be recorded quasi continuously at absorption and desorption measurements. For this either pen recorders of sufficient printing width and chart speed, printers or data loggers or PCs with converters are to be employed. The rising or decreasing DO curves must be recorded, without pre-run, for a period of at least 2·t90% or 280/kLa (in minutes). With, for example, kLa = 5 h-1 the recording time is at least 56 minutes. With the employment of data loggers or PCs with converters the data should be saved at intervals of ∆t = 5/kLa (minutes). With kLa = 5 h-1 there would be a value from each probe for every minute to be saved. With a recording time of 56 minutes there would be, in this case, 56 values of C available for evaluation. The chart speed of pen recorders is to be adjusted accordingly. In any case the calculation of kLa must be based on at least 30 equidistant values of Ct. 3.1.3
Technique of Off-Gas Measurement
For the off-gas measurement a floating gas hood and the measurement installation are required. A measurement arrangement which has proved itself is described below; design modifications are permitted. The gas hood should be as large as possible. For transport reasons the dimensions of ca. 1.25/2.50 m are not, as a rule, exceeded. It is important that the hood is gas tight (Fig. 7). The floats are to be chosen in a way that the hood is submerged in the water by approximately 10 cm. The free air space under the hood should be 20 to 30 cm. There are ropes attached at the corners of the hood by which the hood can be moved and by which it is fixed in the tank. Preferably two DO probes (previously calibrated) hang from the hood into the body of the water. The DO concentration of the mixed liquor should be recorded by a pen recorder. From the centre of the hood a hose leads to the measurement installation. The hose diameter is so to be selected that the pressure loss remains small and that the gas trapped by the hood reaches the gas flow measurement. In general a hose of ∅ 50 mm is sufficient for a collection hood of AA = 2.5 m2. At measurements in cold periods of the year it should be noted that condensed water can collect in the hanging hose. Care is also to be taken that no foam penetrates into the gas hose.
June 1996 19
ATV M 209E
Fig. 7:
Collection hood for off-gas measurements.
The measurement equipment consists of a gas flow rate measurement facility, a threeway valve with inlets for off-gas and reference air and an outlet for air dehumidification which follows CO2 bonding, the sampling gas pump, a gas cooler and the oxygen measurement equipment (Fig. 8).
Fig. 8:
Schematic diagram of off-gas measurement device
The gas flow measurement equipment must be matched to the off-gas flow rate to be collected. Rotameters are not suitable as they have a too high pressure loss. Impeller measurement recorders with appropriate indicator and evaluation instrument have proved themselves. The gas to be measured is taken off after the gas flow measurement system. Hose diameters of 5 mm have proved successful. Using the three-way valve it is possible to switch from measurement gas to reference air for the calibration of the gas oxygen measurement. Reference air is brought in using a hose from untouched ambient air.
June 1996 20
ATV M 209E The gas sampling pump should have a flow rate of ca. 100 L/h. The pump may not ingest any outside air. Bellows gas pumps have proved successfully. The gas pump is located after the dehumidification stage in order not to have problems with condensation water. A glass column (∅ 30 mm, h = 0.3 m, V = 0.2 l) filled with silica gel serves for dehumidification. The CO2 absorber should have the same dimensions; it is filled with soda lime. Paramagnetic oxygen measurement equipment can be used for the measurement of the gas oxygen concentration (partial pressure in the gas). Normal commercial membrane DO probes are cheaper. When using DO probes attention is to be paid that the temperature sensor, under certain circumstances, reacts slower or even faster than the membrane on temperature changes in the gas. Therefore the gas should be fed through a metal snake which lies in a water bath to dampen temperature variations. The probe is also inserted here (Fig. 9). It is not important to obtain absolute temperature constancy but to prevent sudden temperature changes during a measurement period (duration ca. 30 minutes).
Fig. 9:
Tempering of the gas and the DO probe
Independent from the oxygen measurement equipment employed it is recommended that, for the ambient air, a measurement value, MVi, of 100 (e.g. %) instead of the volumetric fraction of oxygen in CO2 free, dry ambient air of Xi = 0.21, is set. One then measures MWe in the off-gas of the hood. For small changes of MWi during measurements a calibration is not necessary. The following applies with Xi = 0.21: June 1996 21
ATV M 209E X e = 0.21⋅
MWe MWi
(26)
If, for example, MWi = 102.3 and MWe = 85.7 then X e = 0.21⋅
3.1.4
85.7 = 0.176 102.3
Power Input Measurement
The power input (gross effective power input, P [kW]) of the aeration installation, including the mixing facility or bridge drive located in the aeration tank, must be measured during every test. It is to be agreed beforehand, or already with the design of the electrical installations how the power input is to be measured (meter, clip-on instrument, power measurement equipment). The use of calibrated power meters with appropriate converters is preferred. It is always to be checked whether only the aeration installation is really recorded by the power meter. The revolutions of the rotating disk are counted for some 2 - 5 minutes. The power input is obtained with the aid of the meter constant. With every test the power input should be measured at least twice, at the start and at the end of the test. With surface aerators, the power consumption can vary; then more frequent measurements or the employment of a power recorder is necessary. The individual readings are to be recorded in the test protocol. In each case the mean value is used for the calculation of the aeration efficiency. The measurement of the power input is problematic in the cases where one or more blowers are not clearly allocated to the tank being tested or where the lowest blower air output is higher than the tank to be tested can handle. In this case the only solution is to measure the air fed to the tank accurately and carry out a conversion using the power input and the quantity of air delivered by the blower. This must, in any case, be agreed beforehand. 3.1.5
Air Flow Rate Measurement
The majority of air flow measurement facilities installed in wastewater treatment plants serve to check the distribution of air to various tanks. The absolutely correct air flow rate is, however, not recorded. Installations in accordance with DIN 1952 should be employed for precise measurements of the air flow rate. For small blowers one may measure the intake air flow rate of the blower using turbine meters. If individual blowers can be allocated to the aeration tank to be tested then one can determine the air flow from the blower graphs. Here the number of revolutions per minute of the blower and the conditions (pressure and temperature) of the intake air, and these immediately in front of the blower, i.e. behind the air filters, as well as the resultant backpressure are to be measured. The techniques for measuring of the air flow should be agreed beforehand.
June 1996 22
ATV M 209E 3.1.6
Temperature
The temperature of the water is to be measured accurately to 0.5°C using a laboratory thermometer at the start and end of the test. If the temperature varies by more than 2° C during the test, the test should be repeated. The readings are recorded, the mean value is used for the calculation. Tests should not be carried out with water temperatures below 4° C. In special cases specific arrangements are to be made. 3.2
Chemicals
3.2.1
Cobalt Catalyst
Cobalt in the form of a cobalt salt, e.g. CoCl2·6H20, is added as catalyst to accelerate the oxidation of sodium sulphite. Normally a cobalt concentration in water of 0.5 g/m3 is recommended, in many cases a tenth of this concentration suffices. It is therefore recommended that the cobalt requirement for the water used is determined previously by tests [15]. For this a clean container, e.g. a 10 l bucket, is filled with water and this is aerated to saturation using a small blower. One starts with an addition corresponding to 0.05 mg/L Co. With the aeration switched off, the content of the bucket is stirred thoroughly using a laboratory mixer. A DO probe, connected to a pen recorder with a high chart speed, is suspended in the water. Just enough dissolved sodium sulphite is added so that the DO concentration does not sink to zero but rather to 2 - 3 mg/L. Following the addition, the DO concentration must decrease practically as rapidly as with a sudden transfer of the DO probe from saturated water into such with a lesser DO concentration. If this is not the case then the cobalt dose is raised by steps in further tests. If no drinking water or water of a similar quality shall be used then it is to be checked whether the cobalt does not become complexed and thus become ineffective. To test that the water is aerated following the last test for the determination of the cobalt requirement and allowed to stand for several hours. Then without a renewed addition of cobalt, sodium sulphite is added. If the reduction rate of the DO concentration is now lower than previously this is a sign that the cobalt is complexed. Such water is unusable for the tests. Even raising the cobalt dose is no remedy. If several tests are to be carried out in an aeration tank with the same water cobalt salt is to be added once only. The required quantity of cobalt, dissolved in warm water, is to be added sufficiently early so that it is distributed evenly over the complete tank at the time of the addition of the sodium sulphite. For closed loop tanks the notes for the dosing of chemicals are to be observed (comp. 3.3.4). 3.2.2
Sodium Sulphite
Water-free, commercial sodium sulphite (Na2SO3) is employed which is oxidised to sodium sulphate (Na2SO4) by the oxygen dissolved in the water. 8 kg of Na2SO3 are necessary for the bonding of 1 kg oxygen; through the addition of 1 kg of Na2SO3 the salt mass increases by some 1.13 kg. The sodium sulphite should be dissolved beforehand. It can also be scattered dry. This must be previously agreed.
June 1996 23
ATV M 209E The solubility of Na2SO3 at T = 10° C is approx. 195 kg/m3. To be on the safe side one should use 1.0 m3 water for the solution of 100 kg Na2SO3. The requirement of sodium sulphite for one test results from the requirement for the bonding of the oxygen dissolved in water and the requirement for a suitable lag time (comp. 3.3.3) for admixture and, if required, stabilisation of the flow. For example, a tank with V = 1000 m3 is to be tested which has a DO concentration of C = 11 mg/L. The oxygen transfer capacity is, according to the given details, OC20 = 120 kg/h. With a lag time (application, admixture and flow stabilisation) of 15 min the following is calculated: Na2SO3 for DO bonding: Na2SO3 for lag time: Total requirement
1000 · 0.011 · 8 = 88 kg 120 · (15/60) · 8 = 240 kg 328 kg
Very often, in addition to the sodium sulphite required per 1000 m3 for the bonding of the DO, 300 kg of sodium sulphite is added in place of the calculated 240 kg. One recognises that, for the lag time, far more sodium sulphite is necessary than for the bonding of the DO present at the start of the test. With this example the salt content is made up by 328 · 1.13/1000 = 370 mg/L. In order not to exceed the salt content of 2000 mg/L one can carry out 5 tests with the same water. Strictly speaking, the required time for the mixing of the tank is related to the mixing intensity which, on its part, is characterised by the volumetric oxygen transfer rate OCR [g/(m³·h)]. According to investigations by the Bayerisches Landesamt für Wasserwirtschaft (Baverian State Office for Water Management) the necessary lag time for admixture (calculated from the end of the application of the sulphite) can be estimated as t (min) = 90/√OCR. 3.2.3
Hydrogen Peroxide
Normally 35 percentile hydrogen peroxide is delivered in 65 kg balloons. 1 kg H2O2 (35 %) releases 0.165 kg O2 on dissociation. Thus a 65 kg balloon contains some 10 kg O2. In order to use a range of 10 mg/L DO for the calculations, it has proved useful to add peroxide to raise the DO by approx. 15 mg/L. For a 1000 m3 tank the following would be required: 1000 · 0.015 : 0.165 = 90 kg (35 %) H2O2 Note: Hydrogen peroxide is a hazardous chemical. It causes caustic burns on skin. When handling hydrogen peroxide always use protective gloves and protective glasses and as far as possible wear protective clothing!
As far as possible the peroxide should be dosed from the 65 kg balloons straight into the aeration tank. Electric barrel pumps have proved themselves for the dosing. The dosing hose (3/4") is weighted so that the peroxide comes into contact with the mixed liquor near the bottom of the tank, as otherwise the spontaneously formed oxygen bubbles escape unused. June 1996 24
ATV M 209E If a complete balloon is not dosed each time care must be taken of further peroxide flow as a result of the siphon effect with the pump switched off. One must therefore, on switching off the pump with the balloon above the water level, lift it up briefly in order to break the siphon. If the balloon is alongside the tank below the water level a return flow check valve must be installed to prevent the reverse siphoning of the mixed liquor into the peroxide balloon. 3.2.4 Pure Oxygen or Nitrogen Gas Either a bottle set or a liquid gas tank with evaporator and gas flow meter are rented for the tests. The gas is fed into the air pipeline as closely as possible to the tank to be investigated. An appropriate connection piece should be welded on to the pipeline already at the time of assembling the air header. It is recommended that contact is made as early as possible with a gas delivery firm. The involvement of specialist firms for the laying of the pipelines for oxygen gas is essential in particular in view of the explosion hazard of gases with increased oxygen concentration. In covered tanks the admixture of pure oxygen may be practically impossible due to the explosion hazard.
Attention is drawn to Appendix 3 for the estimation of the oxygen requirement. 3.3
Absorption and Desorption Measurements
3.3.1 Determination of the Test Procedure The following two variants are differentiated:
•
Test with continuous operation of the aeration installation. Sodium sulphite, peroxide or pure oxygen are added with the aeration operating at the setting to be tested.
•
Test from standstill. Sodium sulphite (or, if required, peroxide or pure oxygen) is added to the mechanically mixed but not aerated tank, or is added to slightly aerated tanks. Aeration is switched on only after an appropriate mixing period of 15 - 30 minutes.
With clean water tests from standstill one requires less sodium sulphite than for tests with continuous operation of the aeration. Tests from standstill are only possible if it is guaranteed that the sodium sulphite has actually distributed itself evenly in the tank within the mixing period. Even if a DO concentration of zero is measured everywhere in the tank this does not indicate that the sulphite is evenly distributed in the tank. Tests from standstill are, in addition, only possible if the flow in the tank is stabilised within a few minutes after switching on the aeration. Tests from standstill are therefore not possible in closed loop tanks with mammoth rotors or cone surface aerators (carrousels) as it takes a long time for the flow to stabilise and as, in addition, the oxygen transfer is influenced by the flow velocity. Once it has been decided how the tests are to be carried out, with clean water tests and sulphite addition it is to be calculated how many tests with one water filling can be carried out in order to stay below the permitted salt concentration of 2000 mg/L (comp. 3.2.2).
June 1996 25
ATV M 209E In order to obtain an accurate answer for the oxygen transfer capacity at least two tests with the same aeration setting are necessary. For guarantee tests the following is to be determined between the customer and the contractor: (1) For which aeration setting(s) tests are to be carried out? (2) How many tests are to be carried out with each aeration setting? In addition to the total required number of tests it is also necessary to include an initial test. With this one gains i.e. information on the chemical dosing, the mixing conditions and the correct positioning of the DO probes. If the result of the initial test shows itself to be acceptable then it can be used to calculate the mean value of OC. Immediately following each test the calculation of kLa or αkLa is to be carried out. The distribution of the residues says much more about the quality of the test than the progression of the DO concentration.
If, using the same water for clean water tests, there are so many tests to be carried out that the upper limit of the salt concentration will be reached, then the tests with the most important aeration settings should be carried out first. 3.3.2
Test Preparation
Before testing the volume of the tank is to be determined by measurement. A height mark is to be placed above the water level from which the immersion depth of the surface aerators can be measured. Immersion depth or zero freeboard is to be determined on filling the tank. If there are several surface aerators present in one tank the water level is at the zero immersion depth if half the number of aerators just touch the surface of the water. In addition, for clean water tests the following is required: (1) Testing of the water quality if not drinking water is used, comp. 2.1.1. (2) Determination of the cobalt requirement, comp. 3.2.1. (3) Testing of the cleanliness of the tank. Residual water containing oil or algae changes the water quality. Cobalt can complex, comp. 3.2.1. (4) Testing of the sealing of the tank and the shut-off installations. In particular with surface aerators it must be guaranteed that the water level does not sink by more than 2 cm during the test. (5) Laying down the type of measurement of the power input, comp. 3.1.4. (6) Laying down of the type of recording of the air flow rate, comp. 3.1.5. (1) and (2) are dispensed for tests with mixed liquor; here one must consider how it is to be ensured that the respiration of the mixed liquor does not alter during the test.
June 1996 26
ATV M 209E 3.3.3
Arrangement of DO probes
In a completely mixed tank at least three DO probes are to be installed, that is one 0.5 m or hw/6 above the tank floor, one at half water depth and one 0.5 m or hw/6 below the water surface. All three probes mounted on a bar or fixed to a weighted rope has proved successfully. It is also possible, for example, to arrange the probes at different heights and locations. The probes should be located at least 1 m from the tank wall if rotating bridges do not force the placement of the probes to be nearer to the wall. Three probes also suffice in closed loop tanks. They should not be within the spraying range of surface aerators but at a position where the DO concentration is approximately the same across the water depth. In longitudinal flow tanks with diffused air aeration the DO probes should be distributed along the length of the tank. In large aeration tanks there should be, for security's sake, six or more probes, comp. 3.1.1. If in longitudinal flow tanks the diffusers are stepped then for every area of even diffuser density at least one, better two probes are necessary to be placed in the centre. Attention is again drawn to the independence of the measurement signal of the probes from the turbulence, comp. 3.1.1. 3.3.4
Warming up of the Aeration Installation
In most cases, for clean water tests, the aeration system will not have been operated for some time. In order to wash free the pores of fine bubble diffusers, and also in order to blow out any water that has penetrated or condensed in the pipes the aeration is to be operated at full power for at least one day, if possible longer, before testing. Surface aerators should run at least half an hour before the test in order to bring the gears (oil) to the operating temperature. 3.3.5
Addition of Chemicals
With clean water tests the quantity of cobalt determined is to be distributed evenly over the complete tank volume as far as possible in advance of the tests. The aeration is then switched on for at least one hour for the even distribution of the cobalt throughout the water mass. In completely mixed tanks it is irrelevant at what point the chemicals (Na2SO3, H2O2 ) are added. In all tanks with water circulation (closed loop tanks) there is the danger that a circulating "cloud" with a high sulphite or DO concentration (due to the addition of peroxide) superimposes a fluctuation of the DO corresponding with the circulation period on the DO increase or decrease curve. As a rule this can already be detected on the DO concentration curve and, in particular, on the heavily fluctuating residues. Such tests are, as a rule, not usable. It is pointed out that, in closed loop tanks with few aerators, the
June 1996 27
ATV M 209E oxygen increase curve can have a stepped progression, comp. 2.5. However, this always has, as opposed to fluctuations due to chemical clouds, an increasing tendency. Two techniques for the application of chemicals in closed loop tanks have proved successful: (1) Addition at one point for the duration of one, two or three water circulation cycles. The duration of a water circulation cycle can be ascertained in that one adds a small amount of sulphite or peroxide at one point and measures the time to achieve the lowest DO (with addition of sulphite) or the highest DO (with the addition of peroxide) at the same point. Now one must adjust the pumping rate for the chemicals in such a way that the necessary quantity is added in the selected number of cycles. For peroxide dosing the end point of the input can be controlled using the DO concentration at the input point. With sulphite dosing this is not possible because due to overdosing the DO concentration can, under certain circumstances, already be zero after the second cycle. Whether the dosing was correct or not will only be detected from the progression of the DO concentration curve and the residues. (2) Addition of equal quantities of chemicals at several points at the same time. This is very expensive at large closed loop tanks with a long pathway. Here the addition at distances of 20 to 30 m, i.e. from the bridges over the mammoth rotors, has proved successfully. A constant and over the tank even electrical conductivity is the best indicator for the even distribution of the sodium sulphite. In longitudinal flow tanks with diffused air aeration the addition of equal quantities at the same time at several equidistant points is necessary. If, in such tanks, the aeration is stepped, then unusable test results can be achieved with heavy overdosing of sulphite at clean water tests as, in regions with lower diffuser density, more time (less OC, comp. 3.2.2) is required for the consumption of the overdose of sulphite. The start of the increase of the DO concentration takes place deferred in line with the decreasing diffuser density. Due to the exchange of water between the regions of different diffuser density, kLa can again be reduced in the region of the highest diffuser density, comp. 2.4. 3.3.6
Measurement and Recording of Data During a Test
In addition to the DO concentration, which is continuously or quasi-continuously recorded, the following are to be measured and recorded during the test: (1) The water temperature at the start and the end of the test, comp. 3.1.6. Further calculation is to be based on the mean value. The difference should be smaller than 2° C. (2) The power input at the start and end of the test, comp. 3.1.4. The mean value is formed for each test. (3) If required, the air flow rate at the same time as the power input was measured, comp. 3.1.5. The mean value is formed for each test.
June 1996 28
ATV M 209E (4) The still water level without aeration operating before and after the test. The difference should be smaller than 2 cm. (5) With clean water tests: the electrical conductivity at the end of the test. (Only when sodium sulphite has been employed). (6) With measurements in mixed liquor: Measurement of oxygen uptake rate of mixed liquor at the start, in the middle and at the end of the test. (7) With measurements in mixed liquor: Sampling at the middle of the test for the determination of MLSS and sludge volume as well as for the determination of COD and NH4-N (immediately filter sample) and, if required, other parameters. 3.4
Off-Gas Measurements
Off-gas measurements always take place in a tank with a specified air flow. This should be measured as precisely as possible, comp. 3.1.5. The gastightness of the complete measurement system, in particular the gas collection hood, is to be checked before the test. For this the hood is placed in an unaerated tank, e.g. a secondary sedimentation tank, and connected to the measurement system. All outlets of the measurement system are closed. A pressure of 5 cm water column is then applied. The facility is considered leakproof if the pressure drop in 15 minutes is not greater than 1 cm water column. The number of measurements which are necessary in order to obtain informative results is dependent on many factors. Independent of tank size and shape at least 10 measurements with the same air flow rate must be carried out at various points in a tank at DO concentrations of 1.0 mg/L ≤ C ≤ 0.5·CS,T.
If, with the air flow rate to be tested, the DO concentration is always greater than 0.5·CS,T, more wastewater can be fed to the tank in order to increase the oxygen uptake rate. The reverse process is to be applied if the DO concentration lies below 1.0 mg/L. In rectangular tanks with stepped diffuser density the number of measurements is to be increased dependent on the tank size (length). At least two measurements are necessary per area of equal diffuser density. In circular tanks with separate mixing the hood is to be placed according to the bubble pattern. With this, both regions with high air outlet as well as regions with low air outlet are to be recorded. One test with recording of data and evaluation proceeds as follows: (1) Position the gas collection hood. Observe that no foam penetrates the off-gas hose and that no condensation water collects in the hose. (2) Connect DO probes which are in the water and in the gas measurement to the recorder or to the data logger (PC) with monitor in order to be able to detect the variations. The DO concentration in mixed liquor should be as constant as possible! If possible also connect the signal from the air flow measurement to the recorder or data logger.
June 1996 29
ATV M 209E (3) Set the three-way valve to reference air. Following stabilisation of the signal of the gas measurement probe, calibrate this to 100 (%), MWi. (4) Set the three-way valve to off-gas. The measurement takes 10 to 15 minutes following stabilisation of the signal of the gas measurement probe (MWe). If the offgas flow is not continuously recorded several readings shall be taken. (5) Set three-way valve to reference air. Following stabilisation of the signal of the gas measurement probe the previously set value of MWi = 100 (%) should show. With deviations of more than 2 % the measurement is to be repeated. Otherwise the mean of the initial and the end value is to be formed and one obtains MWi . (6) The mean value of the DO concentration in the water over the duration of the measurement is calculated for each probe. Should the mean values of DO of both probes deviate from each other by more than 10 % then the calibration of the probes has to be checked and the measurement has to be repeated. Otherwise the mean value C is to be calculated. If the DO concentration in the water is higher than 0.5 CS,T the result is uncertain! (Comp. 2.3.4). (7) Mean values for MWe and QL,e are calculated over the duration of the test. (8) Calculation of Xe with Eqn. 26: X e = 0,21⋅
MWe MWi
(9) Calculation of the molar ratio with Eqns. 18 and 19: MVi =
0.21 = 0.266 1 − 0.21
MVe =
Xe 1− Xe
(10) Calculation of αOAh with Eqn. 21: ß ⋅ CS,T MVe 100 9.09 ( 20 − T ) αOAh = 1 − ⋅ ⋅ ⋅ 1024 . ⋅ 0.266 hE ß ⋅ CS,T − C CSS,T
(27)
Completely covered tanks are ideal for off-gas measurements. But off-gas measurement is only applicable in completely mixed tanks, i.e. in tanks in which the DO concentration at every point is the same. Thus covered rectangular longitudinal flow tanks are in every case excluded since the DO concentration increases as a rule in the direction of flow along the tank. The procedure otherwise is the same as for hood measurement. Several precisely calibrated DO probes are installed in the water. The air flow is set to a constant value. A possible air extraction must be shut down in order to avoid sucking in outside air. Caution: No noticeable overpressure shall form in the airspace of the tank.
June 1996 30
ATV M 209E The measurement gas hose is put into the airspace of the tank at a suitable location. With constant DO concentration of 1 mg/L ≤ C ≤ 0.5·CS,T the volumetric fraction of oxygen (Xe) in the off-gas is measured, as with hood measurement, over 10 minutes and, as the values of the probes in the tank, recorded. The measurement is repeated at least five times with the same air flow. The oxygen transfer efficiency is calculated after every measurement with Eqn. 27. If individual values for OAh with the same air flow deviate from the mean value by more than 10 % then the number of measurements is to be increased to 10.
4.
Test Evaluation and Report
4.1
Adsorption and Desorption Measurements
For each DO probe the parameters C0, CS and kLa (or C* and αkLa), the residues and OC (for T = 20° C) are given as output from the computer. With curved, non-randomly distributed residues t0 is to be displaced, and if required the DO values near to CS or C* are also not to be taken into account, comp. 2.3.3. With absorption tests in clean water C0 should not be larger than 0.25·CS,T. With desorption measurements |CSS,T – C0| should be ≥ 6 mg/L. The calculation of kLa must be based on at least 30 equidistant values of Ct which should cover a period of 1.5 t90% or 210/kLa [minutes]. The kLa or OC values of each test are determined. Should the kLa values of individual probes deviate by more than ± 5 % of the mean value it is to be checked, using the residues, whether irregularities exist. If required, a new mean is to be formed following the removal of the most deviant kLa value. In tanks with stepped diffuser density the mean value of kLa or OC is to be determined under weighting using the tank volume associated with each measurement point. With measurements in mixed liquor one records αkLa and αOC. If required, ß is also to be taken into account. The aeration efficiency is calculated for each test with the mean power uptake with Eqns. 3 and 4. The oxygen transfer efficiency (OAh) or (OCL,h) is calculated for every test using the mean air flow rates with Eqns. 5 and 6. If several tests are carried out with the same aeration setting then, finally, the mean values for OC, OP and, if necessary, OAh and OCL,h are to be formed. At measurements with mixed liquor the results are corresponding to αOC, αOP etc. 4.2
Off-Gas Measurements
For the minimum of 10 measurements required for the same air flow the mean oxygen transfer efficiency is calculated with Eqn. 23. The mean volumetric air flow rate qL,e is
June 1996 31
ATV M 209E calculated with Eqn. 24. Attention is drawn to the fact that qL,e should lie between 0.75 and 1.25 qL,R, comp. 2.3.4. If the air flow rate (QL) fed to the tank or a part of the tank is known, αOC can be calculated with the aid of Eqn. 6.
4.3
Test Report
The following are to be entered in the tests report: (1) Purpose of the test. (2) Details on deviations from this Advisory Leaflet and agreements reached. (3) Description of the tank and the aeration installation illustrated with drawings. Shown therein the arrangement of the measurement point(s) and the point(s) for chemical dosing, and the positioning of the off-gas collection hood. (4) Overview, as far as possible tabular, of the various test settings and the operated aerators or blowers with name plate details such as rpm, power uptake, etc. (5) Description of how the power input was measured. (6) Description of how the air flow rate was determined. (7) For clean water tests: with the employment of Na2SO3, description of the water quality with details of the salt content at the beginning and end of each test. If required, details of further analysis values of the water. (8) For clean water tests: description of the determination of the cobalt requirement. (Dispensed at desorption tests with pure oxygen). (9) For clean water tests: description of the dissolving, dosing and admixture of sodium sulphite and the mass of sodium sulphite added for each test or corresponding details on stripping the oxygen using nitrogen or the building-up of the DO concentration with pure oxygen. (10) With mixed liquor: concentration of MLSS and COD of the filtrate, ammonia content at the beginning and end of each test and, if required, further the water quality characterising parameters. (11) With mixed liquor: description of the dosing and admixture of the hydrogen peroxide; details of the mass of peroxide added for each test or description of the admixture of pure oxygen as well as details on the consumption of oxygen. (12) Water and air temperature as well as ambient air pressure at the beginning and the end of each test. (13) Still water level at the start and the end of every test. With surface aerators details on the mean immersion depth or the freeboard for each test. (14) For absorption and desorption tests: table of the kLa or αkLa values for each DO probe and mean value for each measurement point for every test. Calculation of the mean value for all measurement points of each test, if required, weighted with the associated volume. The customer can require tables or graphical representation of the DO curves as well as the DO residues separately for each DO probe. The original data shall be saved for two years following completion of the tests.
June 1996 32
ATV M 209E (15) For off-gas measurements: Tables with the values of T , C , MWi , MWe , Xe, MVe, αOAh, qL,e and the associated air flows QL and qL,R. The original data shall be saved for two years following completion of the tests. (16) OC, OP and, if required, OAh or OCL,h for each test. (17) Mean values for the above mentioned values with tests with the same aeration setting. (18) Particular observations with the progression of the tests such as foams, changes of colour of the water, etc.
5.
Measurement Tolerances and Guarantees
5.1
Measurement Tolerances
In this Advisory Leaflet measurement tolerances (%) are understood to be the variation band width of the values of repeated measurements about the mean value. Each measurement value has its own tolerance. In clean water tests one can determine kLa with a tolerance of ± 3 to ± 5 %. With measurements in mixed liquor without wastewater and return sludge flow one can apportion αkLa a tolerance of ± 15 %. With wastewater and return sludge flow α can vary considerably depending on the loading rate and the composition of the wastewater. If the saturation value (CS) is measured with clean water tests, then this value is to be apportioned a tolerance of ± 3 %. If table values in accordance with Eqns. 10 and 11 are used, then the tolerance is dispensed with. The power input (P) can, even using calibrated counters or special measurement equipment, have tolerances of ± 3 % due to reading inaccuracies. The diffuser depth is to be precisely measured. The air flow rate can deviate by more than ± 5 % from the true value, independent of whether one measures the value or calculates it from the blower characteristic curve. Table 1: Possible measurement tolerances at oxygen transfer tests
Parameter
In clean water CS measured
CSS from tables
Without flow
With flow
± 7% ± 10 %
±5% ±8%
± 15 % ± 18 %
± 20 % ± 23 %
OC OP
5.2
In mixed liquor
Guarantees
Fundamentally it is to be agreed beforehand for which parameters (OC, OP) with which deviation of a parameter below a guaranteed value it is to count as met. If nothing different has been agreed the following is to apply for clean water tests:
June 1996 33
ATV M 209E "The guarantee for the oxygen transfer in clean water (OC) is met if the mean value from the tests plus the measurement tolerance of 5 % is at least equal to the guaranteed value." "The guarantee for the aeration efficiency in clean water (OP) is met if the mean value from the tests plus a measurement tolerance of 8 % is at least equal to the guaranteed value". For the guarantee of the oxygen transfer in mixed liquor the measurement conditions are to be laid down precisely beforehand. In particular, the number of measurements and the measurement period are to be laid down for the calculation of the mean value in order to be able to balance deviations which occur as a result of the composition of the wastewater.
June 1996 34
ATV M 209E
6.
Symbols (see note Page 4)
AA C Ct
m2 mg/L mg/L
C0
mg/L
CS
mg/L
C*
mg/L
CS,T
mg/L
CSS,T
mg/L
hE
m
hW kLaT αkLaT MVi
m h-1 h-1 -
MVe MWi
-
MWe OC
kg/h
αOC
kg/h
(α)OCR
g/(m3·h)
αOCR,H
g/(m3·h)
OAh
%/m
αOA*
%
OCL,H
g/(m3N·m)
Area of the collector hood for off-gas measurements Concentration of dissolved oxygen(DO) in the water Concentration of DO in the water at time t at absorption and desorption measurements Concentration of DO in the water at time t = 0 at absorption and desorption measurements DO saturation concentration from the evaluation of a clean water test Calculated apparent DO saturation concentration from the evaluation of an absorption or desorption test with mixed liquor Measured DO saturation concentration or calculated DO saturation concentration at T°C at clean water tests Standard DO saturation concentration (water vapour saturated air, p = 1013 hPa) at T° C Diffuser depth (height of water above apertures of diffusers with aeration switched off) Depth of water in the tank (still water level) Oxygen transfer coefficient in clean water at T°C Oxygen transfer coefficient in mixed liquor at T°C Molar ratio, mol O2/mol N2, for dry, CO2-free reference air, MVi = 0.266 Molar ratio, mol O2/mol N2, for dry, CO2-free off-gas Calibration of an instrument for the measurement of the volumetric fraction of oxygen in the reference air, e.g. MWi = 100 Measured volumetric fraction of oxygen in the off-gas Oxygen transfer capacity in clean water under standard conditions (T = 20° C, p0 = 1013 hPa, C = 0) Oxygen transfer capacity in mixed liquor under standard conditions (T =20° C, p0 = 1013 hPa, C = 0) Volumetric oxygen transfer rate under standard conditions. OCR = OC/V, αOCR = αOC/V Volumetric oxygen transfer rate under the projection surface of the off-gas collection hood Oxygen transfer efficiency in clean water under standard conditions (T = 20° C, po = 1013 hPa, C = 0) referred to the diffuser depth hE Oxygen transfer efficiency in mixed liquor calculated from an offgas measurement Oxygen transfer efficiency in clean water under standard conditions (T = 20° C, po = 1013 hPa, C = 0). OC referred to the June 1996 35
ATV M 209E 3
αOCL,H
g/(m N·m)
OP
kg/kWh
αOP
kg/kWh
OVR*
g/(m3·h)
P p QL
kW hPa m3N/h
QL,i
m3N/h
QL,e qL,e
m3N/h m3N/(m3·h)
qL,R
m3N/(m3·h)
qL,H
m3N/(m3·h)
q
m3/(m3·h)
T TL t90%
°C °C min
TDS Xi Xe V VBB
mg/L m3 m³ kg/m3N
α ß ρO2
air flow QL and the diffuser depth hE Oxygen transfer efficiency in mixed liquor under standard conditions (T = 20° C, po = 1013 hPa, C = 0). αOC referred to the air flow QL and the diffuser depth hE Aeration efficiency in clean water under standard conditions (T = 20° C, po = 1013 hPa, C = 0). OC referred to the power consumption P Aeration efficiency in mixed liquor under standard conditions (T = 20° C, po = 1013 hPa, C = 0). αOC referred to the power consumption P Volumetric oxygen uptake rate of the mixed liquor, calculated from off-gas measurements Power input of an aeration installation Atmospheric pressure, p0 = 1013 hPa Air flow rate referred to standard conditions (T = 20° C, p0 = 1013 hPa, dry) Air flow rate to the volume under the projection area of the hood, determined by off-gas measurements, corrected for standard conditions Flow rate of off-gas, corrected for standard conditions Volumetric flow rate of off-gas (qL,e = QL,e/(AA · hW), corrected for standard conditions Volumetric air flow rate to the aeration tank, corrected for standard conditions Volumetric air flow rate, QL,i referred to the volume under the projection area of the hood Volumetric water flow rate. Sum of all water and sludge flows referred to the tank volume Water temperature Air temperature Time for the decrease of the saturation deficit (CS - C) by 90 %. t90% = 60 . 2.303/kLa; kLa in h-1 Content of dissolved solids, vaporisation residue of the filtrate Volumetric fraction of oxygen in dry, CO2-free injected air Volumetric fraction of oxygen in dry, CO2-free off-gas Volume Aeration tank volume Interfacial factor Salinity factor Density of oxygen (1.43 kg/m3N at 0° C and 1013 hPa). The density of oxygen in dry air correspondingly is 0.299 kg/m3N
June 1996 36
ATV M 209E
7.
Bibliography
[1]
ATV: Arbeitsanleitung für der Bestimmung der Sauerstoffzufuhr von Belüftungssystemen in Reinwasser. [German Association for the Water Environment: Instruction for the Determination of the Oxygen Transfer by Aeration Systems into Clean water]. Korrespondenz Abwasser 26 (1979) p. 416
[2]
ÖNORM M 5888: Abwasser-Kläranlagen, Sauerstoffzufuhr-Leistung von Belüftungseinrichtungen in Reinwasser, 1978 [Austrian Standard Specification M 5888: Wastewater Treatment Plants, Oxygen Transfer in Clean Water by Aeration Installations]
[3]
ASCE Standard: Measurement of Oxygen Transfer in Clean Water. Second Edition (1992). Am. Soc. of Civil Engineers, 345 East 47th Street, New York, N.Y. 10017-2399, USA
[4]
ASCE Standard: Guidelines for In-Process Oxygen Transfer Testing (1993). Am. Soc. of Civil Engineers, 345 East 47th Street, New York, N.Y. 10017-2399, USA
[5]
Verordnung über Trinkwasser und über Wasser für Lebensmittelbetriebe (Trinkwasser Verordnung TrinkwV). [Ordinance on Drinking Water and on Water for Foodstuff Concerns (Drinking Water Ordinance TrinkwV)]. BGBL 1 (1991) p. 2612
[6]
Lister, A. R. and Boon, A. G.: Aeration in Deep Tanks. An Evaluation of a Fine Bubble Diffused Air System. Water Pollution Control, 72 (1973), p. 590
[7]
Boon, A. G.: Oxygen Transfer in The Activated Sludge Process. Proceedings: Workshop Toward an Oxygen Transfer Standard. Asilomar Conference, April 1978, EPA (USA), 1979
[8]
Ziokarnik, M.: Koaleszenzphänomene im System gasförmig/flüssig und deren Einfluß auf den O2Eintrag bei der biologischen Abwasserreinigung [Coalescence Phenomena in the Gaseous/Fluid System and Their Influence on the Specific Oxygen Transfer with Biological Wastewater Treatment], Korrespondenz Abwasser 127 (1980), p.728 - 734
[9]
Kayser, R.: Möglichkeiten und Grenzen zur Bestimmung der Sauerstoffzufuhr in Reinwasser und unter Betriebsbedingungen [Possibilities and Limitations on the Determination of Oxygen Transfer in Clean Water and under Process Conditions], Wiener Mitteilungen Wasser, Abwasser, Gewässer, Vol 64 (1986), p.1
[10] Kayser, R. und Dernbach, H.: Weiterentwicklung der Methoden zur Messung der Sauerstoffzufuhr unter Betriebsbedingung. Berichte aus Wassergütewirtschaft und Gesundheitsingenieurwesen [Further Development of Methods for the Measurement of the Oxygen Transfer under Process Conditions. Reports from Water Quality Management and Health Engineers], TU München (1980), Vol 28, p. 29 [11] V. d. Emde, Kayser, W. und R.: Beitrag zur Praxis von Sauerstoffzufuhrversuchen [Contribution to the Practice of Oxygen Transfer Tests], GWF Wasser-Abwasser 106 (1965), p. 1337 [12] Brown, L. C. and Fisette, G. R.: Non-Linear Estimation for Unsteady State Oxygen Transfer (Basic and Fortran Programme) Manuscript, ASCE Oxygen Transfer Workshop, San Diego, 1979 [13] Wagner, M. und Pöpel, H. J.: Erste Erfahrungen in Deutschland mit der Abluftmethode zur Messung des Sauerstoffeintrages unter Betriebsbedingungen [First Experiences in Germany with the Off-Gas Method for the Measurement of the Oxygen Transfer under Process Conditions], Schriftenreihe WAR, Vol. 71, Darmstadt 1993, p. 267 [14] Philichi, T. L. and Stenstrom, M. K.: Effects of Dissolved Oxygen Probe Lag on Oxygen Transfer Parameter Estimation. Journal WPCF 61 (1989), p. 83
June 1996 37
ATV M 209E [15] Terry, D. W. and Thiem, L. T.: Potential Interferences in Catalysis of Unsteady State Reaeration Technique. Journal WPCF 61 (1989), p. 1464
June 1996 38
ATV M 209E
Appendix 1: Table of the oxygen concentration [mg/L] of air saturated water in balance with water vapour saturated air with an atmospheric pressure of 1013 hPa. (Standard oxygen saturation value, CSS,T) Values from EN 25814 (1992) or DIN 38408, Part 23.
T [° C]
+ 0.0 °C
+ 0.2° C
+ 0.4° C
+ 0.6° C
+ 0.8° C
0 1 2 3 4
14.62 14.22 13.83 13.46 13.11
15.54 14.14 13.75 13.39 13.04
14.46 14.06 13.68 13.32 12.97
14.38 13.98 13.61 13.25 12.90
14.30 13.91 13.53 13.18 12.84
5 6 7 8 9
12.77 12.45 12.14 11.84 11.56
12.70 12.38 12.08 11.78 11.50
12.64 12.32 12.02 11.73 11.45
12.57 12.26 11.96 11.67 11.39
12.51 12.20 11.90 11.61 11.34
10 11 12 13 14 15
11.29 11.03 10.78 10.54 10.31 10.08
11.23 10.98 10.73 10.49 10.26 10.04
11.18 10.93 10.68 10.44 10.22 10.00
11.13 10.88 10.63 10.40 10.17 9.95
11.08 10.83 10.58 10.35 10.13 9.91
16 17 18 19 20
9.87 9.66 9.47 9.28 9.09
9.83 9.62 9.43 9.24 9.06
9.79 9.58 9.39 9.20 9.02
9.75 9.55 9.35 9.16 8.98
9.71 9.51 9.31 9.13 8.95
21 22 23 24 25
8.91 8.74 8.58 8.42 8.26
8.88 8.71 8.55 8.39 8.23
8.85 8.68 8.51 8.36 8.20
8.81 8.64 8.48 8.32 8.17
8.78 8.61 8.45 8.29 8.14
26 27 28 29 30
8.11 7.97 7.83 7.69 7.56
8.08 7.94 7.80 7.66 7.53
8.05 7.91 7.77 7.64 7.51
8.03 7.88 7.74 7.61 7.48
8.00 7.85 7.72 7.58 7.46
June 1996 39
ATV M 209E
Appendix 2: Derivation of essential equations 1. Absorption and Desorption Measurements
With absorption and desorption measurements the DO concentration is lowered or raised in the aeration tank, which must be considered as being completely mixed. Through the aeration the DO concentration rises (Figs. 1 and 2) or falls (Fig. 3). The oxygen transfer coefficient kLa or αkLa is calculated from the change of the DO concentration. Using a mass balance (comp. Fig. A2-1) one can derive the equations for calculation. The following applies for the general case with mixed liquor:
Fig. A2-1:
Relationships in a tank with surface aeration
dt ⋅ [∑ Q ⋅ C i + αk L a ⋅ (β ⋅ C S − C)] = V ⋅ dC + dt ⋅ ( V ⋅ OVR + ∑ Q ⋅ C)
(A2-1)
On the left is the oxygen transfer, on the right the oxygen consumption. Transposed one obtains: dC = α ⋅ k L a ⋅ (β ⋅ C S − C) − OVR − q ⋅ (C − C i ) dt
(A2-2)
If OVR, q, Ci and T, and naturally αkLa also, are constant, then after a long period of aeration there is a constant DO concentration, the apparent saturation value, C*, (C = C*); then dC/dt = 0. From Eqn. A2-2 one obtains: α ⋅ k L a ⋅ (β ⋅ C S − C*) = OVR + q ⋅ (C * −C i )
(A2-3)
As one can measure the values OVR, q, C* and Ci and can assume the value for CS (Eqns. 10 and 11), αkLa can (theoretically) be calculated: OVR + q ⋅ (C * − C i ) α ⋅ kLa = β ⋅ CS − C
(A2-4)
Due to the uncertainty of the measurement of OVR, αkLa shall not be calculated according to Eqn. A2-4. An exception is the off-gas measurement; this is clear from the similarity between Eqns. 17 and A2-4. OVR* in Eqns. 16 and 17 contains the DO consumed for the water throughflow q · (C* - Ci). June 1996 40
ATV M 209E Eqn. A2-3 can be used in Eqn. A2-2 for the elimination of OVR. For this Eqn. A2-3 is solved for OVR and applied to Eqn. A2-2: dC = α ⋅ k L a ⋅ (β ⋅ C S − C) − α ⋅ k L a ⋅ (β ⋅ C S − C*) + q ⋅ (C * −C i ) − q ⋅ (C − C i ) dt dC = α ⋅ k L a ⋅ (C * −C) + q ⋅ (C * −C) dt dC = (α ⋅ k L a + q) ⋅ (C * −C) dt
(A2-5)
In clean water is OVR = 0, q = 0, α = 1 and ß = 1. Eqn. A2-2 simplifies to: dC = k L a ⋅ (C S − C) dt
(A2-6)
One obtains Eqn. 15 through integration. 2. Off-gas measurements
With a certain aeration installation the air flow depends on (comp. Fig. A2-2): • the geographical elevation of the wastewater treatment plant, • the ambient atmospheric pressure, p, • the pressure losses up to the blower (fittings, air filter, exhaust silencer), p - p1, • the back-pressure at the blower outlet, p2 (diffuser depth, fittings, pipes).
The nominal performance of a blower is always given as intake air flow rate, QL (m3N/h), dry air at T = 0° C and p = 1013 hPa for a certain back-pressure p2 (hPa). The intake air flow can be determined from the blower characteristic curves for given values of p1 and p2. Installed air flow instruments can be checked in the same way.
Fig. A2-2:
Relationships in a tank with diffused air aeration
June 1996 41
ATV M 209E The partial pressures of the inlet air and off-gas are made up from the partial pressures of nitrogen, argon, oxygen, carbon dioxide and water vapour. p = p N2 + p Ar + p O 2 + p CO 2 + p H2O
(A2-7)
Written as partial pressure ratio resp. volumetric fractions the equation becomes: 1 = X N2 + X Ar + X O 2 + X CO 2 + X H2O
(A2-8)
In the inlet air pCO2 + pH2O are, as a rule, negligibly small. On the other hand the off-gas is water saturated and contains CO2 in concentrations which are not negligible. As, in addition, the volume of the oxygen used is not the same as the volume of resultant carbon dioxide, it therefore is QL,i ≠ QL,e. Strictly speaking, a balancing of the oxygen is only possible if QL,i and QL,e as well as the fractions of oxygen in the inlet air and the off-gas are measured. Under the assumption that the flow rate of the inert gases (nitrogen and noble gases) does not alter, one can theoretically calculate the inlet air flow rate (QL,i) with the aid of the off-gas flow rate and the partial pressure distribution in the off-gas (Eqn. A2-8) by means of a balancing of the inert gas. The following applies: Q L,i ⋅ (1 − 0.21) = Q L,e ⋅ [1 − ( X O 2,e + X CO 2,e + X H2O,e )] Q L,i = Q L,e ⋅
1 − ( X O 2,e + X CO 2,e + X H2O,e ) 1 − 0.21
(A2-9)
In the intake air the carbon dioxide and water vapour were ignored. At off-gas measurements the volumetric air flow rate, fed to the water volume under the hood, can be calculated as follows: qL,H =
Q L,i
(A2-10)
A A ⋅ hW
It is also possible, for every measurement, to calculate the volumetric oxygen transfer rate, αOCR,H [g/(m3·h)] for the mixed liquor volume under the hood, using αOAh (see Eqn. 21) and Eqn. 6: αOC R,H = αOA h ⋅ qL,H ⋅ hE ⋅
0.299 100
(A2-11)
If, for a series of measurements with the same blower setting, one plots the value pairs αOCR,H and q L,H, then these should form a straight line (Fig. A2-2).
June 1996 42
ATV M 209E
Fig. A2-3:
Possibility for evaluating off-gas measurements
With the air flow fed to the tank (QL) one knows the volumetric air flow rate: qL,R =
QL VBB
(A2-12)
The volumetric oxygen transfer rate for the tank, αOCR [g/(m3·h)], then can be read off, see Fig A2-3. Due to the pretentious measurement technique such investigations should be carried out only by institutions with appropriate equipment and experience.
June 1996 43
ATV M 209E
Appendix 3: Desorption tests with the application of pure oxygen gas 1. Fundamentals
As a rule, if pure oxygen is to be employed to increase the DO concentration for desorption tests, a tank with liquid oxygen including vaporiser and flow measurement equipment is hired. Gas flow rates up to 2000 m3N/h, and with special equipment, up to 4000 m3N/h are possible. The oxygen gas is fed into the main air header behind the blowers. If such tests are planned an appropriate connection piece with flange connection, dimensioned for v= 20 m/s, should be welded on when assembling the air pipe. Special regulations are to be observed when handling oxygen due to the explosion hazard.
It is assumed that an increase of the DO concentration by at least ∆C = 10 mg/L, better 15 mg/L, beyond the air saturation value is sufficient if the aeration is continuously operated with the setting to be tested. If one is to work with a smaller air flow during the admixture of oxygen, it can be practical to raise the DO concentration by more than 15 mg/L, as some time passes before the air flow rate is adjusted. Fig. A3-1 shows the progression of DO at a desorption test in clean water. For comparison the build-up curve of DO at an absorption test is presented. The DO concentration here increases from C = 0 to CS. During the admixture of the oxygen for the duration t02 the DO concentration increases by ∆C. The build-up curve here is steeper as the flow of air plus oxygen is greater than that for air alone. The oxygen transfer coefficient has the value kLa´. Should gassing with the mixture be continued, the oxygen concentration would increase to CS´. After switching off of the oxygen the DO concentration falls again to CS, this range is used for the evaluation of the desorption test.
Fig. A3-1:
Relationships with the desorption test in clean water
June 1996 44
ATV M 209E A high supersaturation CS'/(CS + ∆C) leads to a short admixture period but requires a high flow rate of pure oxygen. At a large tank volume the flow rate of oxygen may be limited through the evaporation capacity. This requires clarification following a first rough estimate. It is further assumed that the kLa-value increases proportional to the flow of the air, QL, plus the flow of oxygen gas, QO2, (kLa is independent of the DO concentration of the gas!). 2. Calculation Equations
The air flow rate QL [m3/h] to be tested, the oxygen transfer capacity, OC [kg/h] to be determined and the tank volume V [m3] are known. It is assumed that the oxygen transfer capacity will also be confirmed in the test. Then the following applies: OC V
(A3-1)
OC R CS
(A3-2)
OC R =
kLa =
Following the selection of ∆C the desired value for the saturation value for the oxygen/air mixture is to be selected. It should be: ′ 1.05 ⋅ (C S + ∆C) ≤ C S ≤ 1.2 ⋅ (C S + ∆C)
(A3-3)
The fraction of oxygen in the air/oxygen mixture must then be: X O2
′ CS = 21 ⋅ CS
(A3-4)
The air flow is designated as QL' during the admixture of the oxygen. It is recommended that one works with QL' = QL. Thus the following ratio of the flows results: QO2 X − 21 = O2 ′ 100 − X O 2 QL
(A3-5)
The oxygen transfer coefficient kLa' for the air/oxygen mixture is: ′ QL + Q O2 k L a′ = k L a ⋅ QL
(A3-6)
The theoretical, necessary duration for the admixture of the oxygen tO2 [min] is obtained from the converted Eqn. 14: t O2 =
C ′ − ∆C − C − 60 S ⋅ ln S ′ k L a′ CS − CS
(A3-7)
June 1996 45
ATV M 209E 3. Calculation Example Given:
Tank volume 5000 m3 Water depth 5.50 m Diffuser depth 5.20 m Temperature of water 11° C Oxygen transfer OC to be investigated 550 kg/h Air flow rate 7000 m3N/h Calculations:
CSS = 11.03 mg/L
(Appendix 1, T = 11° C)
5.20 C S = 11.03 ⋅ 1 + = 13.8 mg / L 20.7 OVR = 550 / 5000 = 110 g /(m 3 ⋅ h) kLa =
110 = 7.97 h −1 13.8
(Eqn. 11) (Eqn. A3-1) (Eqn. A3-2)
Further calculations are summarised in Table A3-1, the associated progressions of the DO concentration are shown in Fig. A3-2. Variables are ∆C (Line 2), the supersaturation ratio (Line 3) and the air flow rate QL' during the admixture of the oxygen (Line 1).
Fig. A3-2:
Courses of the DO concentrations of the examples
The requirement for oxygen increases with ∆C and the supersaturation ratio. The smallest flow rate is obtained if one works with a smaller quantity of air during admixture. In addition oxygen is also necessary for the filling of the pipelines. Following the switching off of the oxygen it must be "washed out" from the pipelines. This requires time. Through this the transition from the rising curve to the falling curve of the DO concentration is indeed not a sharp bend as shown in the figures but a curve.
June 1996 46
ATV M 209E Table A3-1: Calculation of the oxygen requirement for desorption measurements
1 2 3 4 5 6 7 8 9 10 11
Q L' ∆C Cs'/(Cs + ∆C) Cs' XO2 QO2/QL QO2 kLa' tO2 VO2 MO2
m3N/h mg/L -mg/L % O2 -3 m N/h h-1 min m 3N kg
given given given Eqn. A3-3 Eqn. A3-4 Eqn. A3-5 L1 · L6 Eqn. As-6 Eqn. A3-7 L6 · L9/60 1.41 · L10
1
2
3
7000 12 1.2 31.0 47.1 0.49 3.456 11.9 6.05 349 491
7000 15 1.05 30.2 46.0 0.46 3.244 11.7 12.5 677 955
2000 15 1.1 31.6 48.2 0.53 1.050 3.5 31.5 552 778
In comparison to absorption measurements, by which the DO is bonded with sodium sulphite, there are advantages for desorption measurements using pure oxygen for large tank volumes and in particular for large diffuser depths. With the same OVR the requirement of pure oxygen is proportional to the tank bottom area while the sulphite requirement is proportional to the tank volume. Due to the coalescence inhibiting effect of the sodium sulphate, the oxygen transfer capacity from desorption measurements (without sodium sulphite) can be comparatively somewhat smaller than from absorption measurements with sulphite.
June 1996 47
ATV M 209E
Appendix 4: Evaluation of absorption and desorption measurements using a computer 1. Brief Description and Installation
This Advisory Leaflet has a Disk with the programme OCA for the evaluation of the adsorption measurements and the desorption measurements (comp. 2.3.3). Prerequisites for the operation of this programme are: • • • • •
an IBM compatible PC, a MS-DOS compatible operating system, (MS-DOS 3.3 or higher), at least 640 KB main memory, an EGA or VGA card, a 3.5" 1.44 MB Disk drive.
The following are, in addition, an advantage: • • • •
a numerical coprocessor, a harddisk, a mouse, a printer plotter.
The essential characteristics of the programme are: • • • • • •
calculation in accordance with this Advisory Leaflet and US Standard [3] respectively (exponential regression according to Brown and Fisette), up to 50 DO concentration values, operating surface in accordance with SAA Standard, formula oriented input data, printout of test results in form of tables and diagrams, comprehensive context and subject related on-line assistance.
Installation
All necessary files in standard form can be found on the original disk. You can therefore use the programme without further preparation direct from the disk. However, you should first make a safeguard copy of the disk. In order to install the programme on the harddisk it is practical to create a new directory. Then copy all files from the disk into this directory. Start
To start the programme select the drive with the programme disk or change to the established directory and type: OCA [ENTER]
June 1996 48
ATV M 209E 2. Brief Instructions
The following instructions are designed to give you, in five successive steps, an overview on handling of the programme and the most important functions. Further more detailed information can be found in the on-line help, which you can call up at any time using the F1 key. Following programme start-up you will see Fig. A4-1 on the screen. The upper screen line is the menu line with the selection pints • • • •
File Window Setting Help
The lowest line is designated as status line; depending on the situation it shows various key sequences with which you can call up certain functions without using the menu. The screen area between the menu and the status lines is the so-called view port.
Fig. A4-1:
Screen format
Step 1: Creation and opening of data input formulas
With the mouse click on to the menu point "File" (if you use the keyboard press Alt + D together). Following this the file sub-menu opens with the functions as shown in Fig. A4-2. Three dots following the menu function indicate that, on calling up this function, a socalled dialogue window appears with whose help the function reference can be specified. Designations such as "F3" or "Alt + X" represent key functions which can be used instead of function commands using the keyboard. Now click on to "New" (or move the selector bar to "New" using the cursor key and press ENTER). A window with input formula appears on the screen. You can now enter all data and indicators which are required for an evaluation and documentation of an oxygen transfer measurement (see Step 2).
June 1996 49
ATV M 209E New Open...
F3
F3
Import... Save
F2
Save under...
Change directory...
Analyse...
DOS activate Terminate Fig. A4-2:
Alt+X File sub-menu
Excel files can be transferred, for example, using "Import...". This is available if your data logger has stored the data in Excel files. For information on this see "Help". As described above reselect the menu point "File" and the function "Open" in the file submenu. The window "Open File" then appears on the screen. It contains i.a. a list of files which have already been saved - with the first programme start only the initial files "DEMO1" and "DEMO2". Over the file list you see an input field for a file name. To the right you find so-called switches (Open, Abort). Double-click on the name "DEMO1" in the list (or press the tabulator key, move the selector bar if required with the aid of the cursor key to the name "DEMO1" and then press [Enter]). You have now created a second input form on the view port. As opposed to the first form this already contains numbers and signs. As the windows are "one on top of another" (superimposed) at first you see only the one created last. You have i.e. the following possibilities of arranging the forms on the view port that all are, at least in part, visible. (1) Select the menu function "Window" and in the sub-menu "Superimposed". The forms are then presented portioned. (2) Using the mouse click on the title row of a window and keep the mouse key depressed while moving the mouse. In this manner you can displace the window and make the one below visible. Step 2: Input data
If the first created, empty input form is not already lying "on top" click on this form window using the mouse on any visible point in order to bring it to the foreground. The input marking is located, if you have not altered it in the meanwhile, on the first input field "Label". If this is not the case move the marking using the cursor key to this field. Type in any desired series of symbols to describe a measurement (e.g. "Sewage treatment plant XYZ, Test 1, Measurement Point 2"). Then press the key [Down] or [Tabulator] or [Enter]. The input marking is now on the field "Date". Enter the date in the form DD.MM.YY or DD.MM.YYYY and note that all other keys than the space bar, fullstop, the number keys 0....9 and various control keys are blocked.
June 1996 50
ATV M 209E Carry on in the same manner in order to enter numbers and symbols in the other fields. As decimal point please always use the fullstop. Should the significance of the contents of an input field not be clear press F1 in order to obtain the appropriate help. The "Help Window" can be closed, if required, using the [Esc]. The saturation value (CS,20) is calculated and applied for surface aerators automatically with CS,20 = 9.09 mg/L and for diffused air aeration also automatically in accordance with Eqn. 11. Only if you have measured the saturation value or have determined it in any other way you do set a value in the input form. Step 3: Carry out evaluation
Bring the input form "DEMO1" into the foreground and select the function "Evaluate" from the data menu. Dialogue window as shown in Fig. A4-3 then appears. In the top left-hand corner you see below "Default settings" information on the scope of the actual oxygen measurement series. Below there are two input fields in which you can enter which values are to be taken into account with the next evaluation. Thus it is possible to cut out possible false initial and end values.
Fig. A4-3: Evaluation menu
Now click on the switch "Compute" or press the key combination Alt+R. The most important characteristic values of an oxygen transfer measurement then appear in the box "Results": •
the oxygen transfer coefficient (α)kLa for the test temperature,
•
the standard deviation of the measured DO values of the calculated curve (E.E),
•
the oxygen transfer capacity under standard conditions (α)OC,
•
the ratio ta/t90%,which should be greater than 150 %, comp. 4.1.
In order to obtain a graphic representation actuate the switch "Graphic". With this two diagrams are produced: the upper diagram shows the individual DO concentration values and the calculated curve, the lower one the deviations between measured and calculated DO values (residues). The presentation of the residues is particular suited for the assessment of the "quality" of a measurement. The DO value series of the Demo file has been deliberately set so that a "wave form" path for the residues results. This can, for example, be an indication of an insufficient mixing of the chemicals. In such cases you should repeat the evaluation with a different range of values (comp.3. Example, Fig. A47).
June 1996 51
ATV M 209E Step 4: Print
The programme offers the possibility of printing out the measured data in the form of a list and as a diagram. Before you call up a function you should select and adjust your printer. For this select the menu function "Settings" and in the sub-menu "Printer". This is presented as shown in Fig. A4-4.
Fig: A4-4: Menu for printer setting
In the list "Model/Type" there are various well-known printers. Using "Text" a simple, nongraphic capable printer is indicated. If you select this you can print a list but not a diagram. All other printers on the other hand can produce graphics. On the right-hand side of the dialogue window you can select the printer connection; the standard setting is "LPT1". In the case that your printer is connected to a series interface (COM1 and COM2), please note that you must set the transfer parameters (baud rate, word length, etc.) before calling up the programme. In order to divert the printer output into a file, select "File". In the lower part of the window you can input how many character spaces in each row should be left free as margin. An approximately 2 cm margin corresponds to 8 characters. Once you have carried out all necessary settings, close the dialogue window. Call up the function "Evaluate" and actuate the switch "Print list". The printer must then print out the measurement protocol in the desired form. (In the case that you have selected an input on file it is naturally not printed. Instead you are asked for a name for the output file). A print-out for "DEMO1" as an example can be found in Fig. A4-5. At the top of the table the plant-specific values and the default settings are summarised on the left, on the right are the results of the evaluation. Next try out a print of a diagram by actuating the switch "Print graphic". If there are problems with this then the printer is probably not compatible with the selected model. Check the compatibility using the printer handbook.
June 1996 52
ATV M 209E OXYGEN TRANSFER TEST DEMO1, Desorption, Pressure Aeration Test No.: 1 Date: 09.03.95 Sonde: red Tank volume........: 1000.00 m3 Water temperature..: 12.00 ºC Satur.value CS,20...: 11.07 mg/L Power input..: 33.00 kW Diffuser depth hE..: 4.50 m Air flow QL.: 2000.00 m3N/h
C0...........: cs(c*).......: (a)kLa.......: (a)OC........: (a)OCL,h.....: (a)OP
Delta t............: 2.00 min Evaluation range...: 1 - 35
E. E. .......: 0.15 ta/t90.......:
16.04 mg/L 3.12 mg/L 4.42 1/h 59.15 kg/h 6.57 g/m3N*m 1.79 kg/kWh 218
%
I
TIME Min.
C meas. mg/L
C calc. mg/L
Dev. mg/L
I
TIME Min.
C meas. mg/L
C calc. mg/L
Dev. mg/L
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00
16.00 14.80 12.70 11.00 10.00 9.15 8.40 7.75 7.15 6.60 6.15 5.80 5.40 5.12 4.85 4.62 4.40 4.25 4.10 3.95 3.85 3.75 3.65 3.56 3.50
16.04 14.27 12.74 11.42 10.29 9.30 8.46 7.73 7.10 6.55 6.08 5.68 5.33 5.03 4.77 4.54 4.35 4.18 4.03 3.91 3.80 3.71 3.63 3.46 3.50
-0.04 -0.53 -0.04 -0.42 -0.29 -0.15 -0.06 0.02 0.05 0.05 0.07 0.12 0.07 0.09 0.08 0.08 0.05 0.07 0.07 0.04 0.05 0.04 0.02 -0.00 -0.00
26 27 28 29 30 31 32 33 34 35
50.00 52.00 54.00 56.00 58.00 60.00 62.00 64.00 66.00 68.00
3.43 3.38 3.33 3.30 3.25 3.22 3.20 3.17 3.15 3.13
3.45 3.40 3.37 3.33 3.30 3.28 3.26 3.24 3.22 3.21
-0.02 -0.02 -0.04 -0.03 -0.05 -0.06 -0.06 -0.07 -0.07 -0.08
Fig. A4-5: Printout of the results of an evaluation Step 5: Save data
Bring the form "DEMO1" into the foreground and then call up the function "Save" in the file menu. You can determine that the data is then stored on a disk or harddisk without further request. Now repeat the action with the form which was created in the first step. A dialogue window "Save Data" now appears as request to give a file name (Fig. A4-6). June 1996 53
ATV M 209E The reason for this different reaction is that the new form still has no valid file name and is designated provisionally as "NONAME...". You can recognise this in the title row of the form window. If you wish to save a form with a valid file name under another name use the function "Save under" instead of the function "Save". Through this you obtain the named dialogue window and can give it a new file name.
Fig. A4-6:
Menu to save data
3. Example
Call up "DEMO1". Select "File" and "Evaluate". Now go to "Graphic". You now see Fig. A4-7. The wave form of the residues is apparent. Read off kLa = 4.42 1/h and the standard deviation of the residues E.E. = 0.15 as characteristic values. Now change to "Evaluate 2 (instead of 1) from 35" and go to "Graphic". The figure is unchanged. If you now evaluate 3 from 35 the E.E sinks to 0.08 and kLa 60 4.16 1/h. The progression of the residues is still wave form. It is only when you evaluate 4 from 35 that the residues are more randomly distributed, E.E. sinks to 0.02 and kLa to 3.98 1/h. If you then evaluate 5 from 35 you will recognise that practically nothing alters (Fig. A4-8). If one had based the calculation on all 35 measured values then, in this example, one would have overestimated kLa and thus also OC by ca. 10 %.
June 1996 54
ATV M 209E
Fig. A4-7:
Diagram of the test results applying all measured values (1 to 35)
Fig. A4-8:
Diagram of the test results after leaving out the first four measured values (1 to 35)
June 1996 55
ATV M 209E 4. Help
The programme has been developed by AQUADATA GmbH. AQUADATA also takes on programme maintenance on behalf of ATV. If you have problems with the programme you can apply to AQUADATA. The most competent specialist there is Mr. Gero Fröse. AQUADATA GmbH Große Straße 5 D-38116 Braunschweig Germany Tel: (0049) 531-501452, Fax: (0049) 531-500907.
June 1996 56
