DE10231908A1 - Self-regulating parametric process for oxygenation of activated sludge in waste water treatment plant based on redox potential and/or pH value and waste water plant load - Google Patents

Abstract

A waste water treatment aeration stage is regulated by a self-parametric process. In the process the activated sludge treatment basin, or individual zones within the basin, are continually aerated in conjunction with previous or simultaneous de-nitrification. A variable target oxygenation rate is especially calculated from the redox potential and/or pH value, in accordance with the waste water plant load. The variable target oxygenation value is derived from the difference between the redox potential value actually measured and/or the actual pH value, and the specific redox and/or pH normal values calculated for the respective waste water treatment plant. In determining the specific redox and/or pH normal values for the respective waste water treatment plant, a long-term mid-value of the two is used over a time lapse which is equal to or greater than the waste water residence period within the treatment plant.

Description

The invention relates to a method for load-dependent, self-parameterizing regulation of the ventilation of sewage plants according to the generic term of claim 1.

Optimal process control for sewage plants with upstream or simultaneous denitrification sets a constant need-based Concentration of dissolved oxygen ahead in the nitrification tank or in the ventilated zones.

The problem with upstream denitrification is that the nitrate and oxygen-containing wastewater / sludge mixture from the nitrification basin and the nitrate-containing return sludge in an internal recirculation to the denitrification tank or fed to the denitrification zone is, where it mixes with the inlet.

With simultaneous denitrification flows the activated sludge containing nitrate and oxygen from the nitrification zone directly to the denitrification zone.

In both cases, denitrification through the carryover of the dissolved oxygen from the nitrification into the denitrification zones significantly hampered. Carbon-consuming substance is consumed, resulting in a Abatement especially for those Denitrification urgently needed Biological oxygen demand (BOD) leads. At the same time it means entrained dissolved oxygen a waste of ventilation energy.

Are prevalent today conventional controller with fixed setpoints. controlled e.g. Ammonium, nitrate and phosphate concentrations, or the organic sum parameters such as total organic carbon (TOC), Chemical oxygen demand (COD), biological oxygen demand (BOD), with the supply of oxygen being the manipulated variable is used.

They are used continuously measuring analyzers for actual value determination. A major disadvantage when using these parameters there is a high purchase price and operating costs of the analyzers. A second functional disadvantage is that the required period between sampling from the activation tank and provision of the measured value for the control, is so great that the determined actual value for the control is not more with the actual Condition conditions in the aeration tank matches. A process-optimal This dead time means that it is not possible to react to the control.

The further variant of the ventilation control works with fixed target values for the oxygen concentration in the nitrification zone or the aeration tank. It is based the setpoint specification (as in the first case) is based on experience with it to a great extent subject to subjective judgment. Accordingly different is the result of the regulation and the distance to the optimum of Regulation.

Even when using complex and complicated control models, you are not able to react optimally to the complex and complicated processes of wastewater treatment, as long as fixed setpoints have to be specified (Correspondence Wastewater No. 4/1994, pages 556–563). Such strategies are associated with considerable problems, since the factors determining the wastewater treatment process not only show different dynamic behavior, but with multivariate control there are sometimes opposite control objectives (e.g. excessive NH _{4 run-} off concentrations due to inhibition of the nitrifiers due to insufficient pH value). For some years now, the use of knowledge-based control systems and adaptive controls based on models for the process control of wastewater treatment plants has been worked on (publication series WAR 109/56. Darmstädter Seminar-Abwassertechnik-1997, pages 81-107). So far, these models have only been able to reflect the complexity of the process to a very limited extent. Fuzzy logic regulations show a simplification of the modeling, since modeling can only be carried out at the level of process knowledge. To do this, you use fuzzy "if-then rules" that are gained from expert experience. With all the advantages of fuzzy technologies, this strategy has the disadvantage under very complex application conditions that the number of input variables and the rules is very high. It is known, for example, that a total of 48 "if / then rules" based on 6 input variables are required to regulate the ventilation of an intermittent sewage treatment plant. This means that the basic requirement for rule-based systems for completeness and freedom from contradiction of the rules can hardly be met. It is also essential that the effectiveness of the fuzzy logic rules is largely determined by the quality of the expert knowledge from which these rules are created and by the specific setting conditions specified by experts (e.g. choice of the form and number of membership functions, the fuzzy operators) and the initial weight calculation). Practical experience has shown that controls based on fuzzy logic in certain circumstances lead to poorer operating results than the controls used conventionally with a fixed setpoint.

Self-learning control systems based on neural networks are only just beginning. So far, they have mainly been used to simulate, classify and predict inflow parameters ( DE 198 27 877 A1 ) and ( DE 199 12 655 A1 ).

Model-based regulations based on process simulations aim at the Ba sis of online measurements to determine the current cleaning capacity and to determine the expected runoff concentrations from the previous and current measurement data. These regulations require constant adjustment of the model parameters to the real conditions. Since, as a rule, not all of the measurement parameters required to calculate the model parameters can be determined continuously, considerable simplifications in the model formation are sometimes necessary (series WAR 108/9. Joint Seminar -Waste Water Technology- Darmstadt 1998 , Pages 109-136). The wastewater treatment process to be controlled can therefore only be optimized to a limited extent.

From the analysis of the state of the art shows that the tasks of a process-optimal ventilation control have so far been insufficiently solved in practice.

It is the object of the invention to change the nitrogen-dependent regulation of the oxygen input into the nitrification tank in such a way that with a high ammonium nitrogen value (NH ₄ -N) and at the same time a low nitrate nitrogen value (NO ₃ -N) in the course of the nitrification tank, the oxygen content increases and at opposite relationships is lowered.

The task is solved with the characterizing part of claim 1. Further training in the subclaims specified.

The regulation of ventilation at Wastewater treatment plants with upstream or simultaneous denitrification should be carried out in this way be that optimal energy use among changing Load conditions possible becomes. This is a need-based regulation of the Oxygenation in the nitrification zone in such a way that Oxygen is provided depending on the cargo. Such regulation of the oxygen input depending on the need from the cargo leads to reduce operating costs by minimizing the required Ventilation energy expenditure.

This is the method according to the invention for load-dependent, self-parameterizing regulation of the ventilation of sewage treatment plants, where the nitrification basin or individual zones of the same continuously ventilated are, preferably sewage treatment plants that according to the process of upstream or simultaneous denitrification work, characterized in that from the parameters redox potential and / or A pH value that is variable depending on the load on the sewage treatment plant for one known oxygen control is calculated, the variable oxygen setpoint from the difference of the measured actual redox potential actual value and / or current pH value and that for the respective sewage treatment plant specific standard of the redox and / or pH value is calculated.

For the determination of the for the respective wastewater treatment plant specific norm of the redox and / or pH value becomes an average from the long-term behavior of the measured values, redox potential and / or pH value used in a development of the invention.

The averaging is advantageously carried out via a Period equal to or greater than the average residence time of the sewage material in the sewage treatment plant.

In another advantageous one Embodiment of the invention is used to determine the mean values Calculation of the redox and / or pH value by forming a Amplitude frequency, where the mean is a selectable Quantile of the amplitude frequency is.

In a further, appropriate further training According to the invention, the target value of the oxygen concentration is determined the sum of the mathematical product of an adjustable constant and from the difference between the redox and / or pH value and a predetermined one Basic oxygen value formed.

By measuring the relevant process variables in the nitrification zone over a long period of time, the normal behavior of the sewage treatment plant is recorded, determined from the difference of this standard with the current value of the current load condition of the sewage treatment plant, and an optimal oxygen setpoint for the oxygen control is calculated from this condition value. The oxygen setpoint is fed to a controller known per se (eg PID controller). According to the invention, the redox potential and / or the pH value are used as relevant process variables. With the procedure for regulating ventilation for intermittent sewage treatment plants ( DE 198 198 75 ) it could be shown that process variables dissolved oxygen, redox potential and pH value in the aeration tank are relevant for a description of the condition in the aeration tank.

The procedure is set out below an example and the related Drawings closer explained will show:

1 a schematic representation of a sewage treatment plant with upstream denitrification,

2 a schematic representation of the control method according to the invention and

3a and b an example of the calculation of the wastewater treatment plant standard by the distribution of the amplitude of the redox potential.

An exemplary construction of a sewage treatment plant with upstream denitrification is shown in 1 shown. The sewage material first flows through a rake into a ventilated sand trap / fat trap 1 , then into a pre-clarifier 2 , The sewage material is fed through a distributor 2a on individual streets 3 . 4 and 5 that in cascades 6 . 7 . 8th . 9 . 10 ; 11 and 12 are divided, distributed. After the last cascades 12 of the streets 3 . 4 and 5 the sewage is in a common confluence 13 merged, from which it was in secondary clarification 14 and 15 flows. An internal recirculation 16 promotes part of the sewage from the last cascades 12 in the first cascades 6 , A return sludge conveyor 17 promotes from both secondary clarifiers 14 and 15 Mud to the distributor 2a back. The cascades 9 . 10 and 11 of the streets 3 . 4 and 5 are aerated for an optimal nitrification process.

The redox potential, the pH value and the dissolved oxygen are in the cascades 12 measured that the ventilated cascades 9 . 10 and 11 connect.

The method according to the invention for regulating the ventilation is in 2 shown schematically.

The redox potential or the pH value are used by corresponding sensors as analog values 18 a preprocessing module 19 fed. Individual measured values are taken from the analog values at intervals of 2 s, from which an average value is calculated every 60 s. The mean values are in the preprocessing module 19 cached as time series. The time series are over a normal transmission 20 then for a defined time horizon, which corresponds to the average residence time of the wastewater in the wastewater treatment plant, to a module for determining the wastewater treatment plant standard 21 transfer. In the module for determining the wastewater treatment plant standard 21 the wastewater treatment plant standard is calculated. For this purpose, time series with limited time are used 31 and 32 statistical amplitude frequency distributions 33 and 34 calculated ( 3 ). These statistical distributions reflect the normal behavior of the clarification process over a longer time interval, which experience has shown is not a normal distribution but a multi-peak amplitude distribution 33 and 34 how out 3 can be seen shows. From the amplitude frequency distributions 33 and 34 a quantile (e.g. 70% of the histogram area) is calculated, which is the average wastewater treatment plant standard 35 is interpreted. This value is called the redox or pH setpoint 22 designated. The redox or pH setpoint 22 is smoothly determined at an adjustable interval (eg every 3 hours) and a controller module 23 fed. At the same time, a current redox or actual pH value 24 to the controller module 23 to hand over.

It becomes a reference value 25 for the dissolved oxygen concentration (e.g. 1 mg / l). The reference value 25 corresponds to the required oxygen concentration when the redoxist value 24 the redox setpoint 22 equivalent.

If the redoxist value deviates 24 to the redox setpoint 22 an oxygen concentration change is calculated (e.g.

+3 mg / l each –100mV redox difference between actual and setpoint.). The determined oxygen value is smoothed and limited at the top and bottom and sent to a PID controller 27 as the oxygen setpoint 26 passed. At the same time, an dissolved acid value becomes 28 the PID controller 27 fed and by the PID controller 27 from the difference from the dissolved oxygen actual value 28 and oxygen setpoint 24 another air slide position setpoint 29 calculated. Through a two-point controller 30 the air slide action is triggered.

3a ) and b ) show diagrams to show the calculation of the average wastewater treatment plant standard 35 , 3a shows the development of the redox potential over a period of 90 hours and two time series as an example 31 and 32 , In 3b are the associated relative amplitude frequency distributions 33 and 34 (Histograms) of the respective time series 31 and 32 shown. To the time series 31 belongs to the amplitude frequency distribution 33 , to the time series 32 the amplitude frequency distribution 34 , The calculated average wastewater treatment plant standard 35 is also shown.

1

Sand trap / grease trap

2

primary clarifier

2a

distributor

3

road

4

road

5

road

6

cascade

7

cascade

8th

cascade

9

cascade

10

cascade

11

cascade

12

cascade

13

Confluence

14

secondary clarifier

15

secondary clarifier

16

inner recirculation

17

Return sludge promotion

18

analog value

19

preprocessing

20

normal transmission

21

module for the Determination of the sewage treatment plant standard

22

redox or pH setpoint

23

Control module

24

redox or actual pH value

25

reference value

26

Oxygen setpoint

27

PID controller

28

Gelöstsauerstoffistwert

29

Air slide position value

30

Two-point controller

31

time series

32

time series

33

Amplitude frequency distribution

34

Amplitude frequency distribution

35

average normal treatment plant

Claims (5)

Process for load-dependent, self-parameterizing regulation of the aeration of sewage treatment plants, in which the aeration tank or individual zones thereof are continuously aerated, preferably sewage treatment plants, which operate according to the processes of upstream or simultaneous denitrification, characterized in that the parameters redox potential and / or pH A value that is variable depending on the load on the sewage treatment plant is calculated for a known oxygen control system, the variable oxygen setpoint being the difference between the measured actual redox potential actual value and / or the current pH value and the standard of the redox and / or specific for the respective sewage treatment plant pH value is calculated. A method according to claim 1, characterized in that for the determination of for the respective wastewater treatment plant specific norm of the redox and / or pH value an average from the long-term behavior of the measured values redox potential and / or pH value is used. A method according to claim 1 or 2, characterized in that the averaging over a period of time equal to or greater than average residence time of the sewage material in the sewage treatment plant he follows. Method according to one of claims 1 to 3, characterized in that that the determination of the mean values for calculating the redox and / or pH value by the formation of an amplitude frequency, the Average is a selectable quantile the amplitude frequency is. Method according to one of claims 1 to 4, characterized in that that the target oxygen concentration is the sum of the mathematical product of an adjustable constant and from the Difference between the redox and / or pH value and a given basic oxygen value is formed.