Experiences on Instrumentation and Control of ... - Science Direct

0 downloads 0 Views 2MB Size Report
Specifically, process analysis and control methods, experi- ences on measurement ... Microprocessor; Digital control. .... util ized to minimize the need for manual.
EXPERIENCES ON INSTRUMENTATION AND CONTROL OF ACTIVATED SLUDGE PLANTSA MICROPROCESSOR APPLICATION R. Aarinen*, J. Tirkkonen* and A. Halme** -Project Department, Dy Labko Ab, Tampere, Finland - -Division of Control Engineering, University of Dulu, Duiu, Finland Abstract. The paper presents experiences and results from a research and development project, where feasibility and advantages of instrumentation and automatic control of activated sludge wastewater treatment processes were investigated. Specifically, process analysis and control methods, experiences on measurement techniques and instrumentation and the developed digital, microprocessor based monitoring and control system are discussed. Results pertaining especially to dissolved oxygen control are given. Keywords . Water pollution; Wastewater treatment, Activated sludge process ; Microprocessor; Digital control . I NTRODUCT I ON

While quite a many studies on the same problems have been reported in the 1 iterature, many of them are quite theoretical and deal only with some aspects of the problems. For reI iable, high performance and economical operation, however, all these aspects should be considered. In this paper a unified approach to the problems is taken.

Wastewater treatment processes are becoming more important subjects for instrumentation and automation as their significance increases with increasing amounts of wastewaters and more stringent norms in environmental prote c tion. More attention is now being paid to the functioning and economy of the treatment processes. Users' knowledge and requirements for automation have been increased from the almost zero level where it was some years ago. At the same time sensor technology has made rapid progress and can now offer many solutions with moderate reliabil ity for the hard conditions typically existing in wastewater treatment plants. The processes themselves are, because of large load variations and many simple bas i c operations, quite suitable subjects for automation. In this paper instrumentation and automatic contr o l of the activated sludge process used for urban wastewater treatment are considered. Results and ideas presented are from a threeyear research project, where an instrumentation and control system for activated sludge plants was developed by the authors. The project was part of a larger "Communities'Water and Environment Project" financed by SITRA (The Finnish National Fund for Research and Development) . The project gave much practical experi e nce both on the process operation and the problems related to the instrumentation and automatic control. Special attention was paid to the control and supervision of the activated sludge process, which determines the functioning quality and economy of the whole plant . Specifically , sludge pumping policies, chemical feeding and pH-control strategies and dissolved oxygen control were considered .

First the activated sludge process operation characteristics, and control objectives are reviewed. Then process control policies and their associated measurements are discussed. Specifically, return and excess sludge pumping, chemical feeding, pH adjustment and dissolved oxygen control are considered. Then the rather unique flow, pH and dissolved oxygen measurement techniques are discussed . Finally the digital microprocessor based data acquisition, reporting and control system is presented, and operation experiences given. EspeCially, the system performance in dissolved oxygen control is studied in more detail. ACTIVATED SLUDGE PROCESS The most harmful constituents of wastewaters include organic materials (biochemical oxygen demand), nutrients (nitrogen and phosphorus), solids, pathogens and other hazardous components (phenols, cyanides , heavy metals etc.). Biochemical oxygen demand (BOO) and phosphorus may be removed very effectively by combined chemical and biological treatment, sol id materials by mechanical means and pathogens through disinfection. The most widely used biological purification method is the activated sludge process with its variants. Activated sludge wastewater treatment processes may be considered as technical appli-

255

R. Aarinen, J. Tirkkonen and A. Halme

256

cations of the biological stabilization (oxidation) process of dis50lved and colloidal organic waste in the receiving streams. By sludge (biomass) recirculation and powerful aeration the reaction rates are, however, increased to a level far exceeding the capability of the environment. Treatment processes typically operate under heavy variations of loading in both qual ity and quantity. In properly designed and controlled activated sludge processes reduction of dissolved organics is, however, more than 90 % and by combined chemical treatment (e . g . simultaneous precipitation) the phosphorus removal efficiency 70 ... 90 %. Operation From the process and control engineering point of view the activated sludge processes are simple continuous flow stirred tank biological reactors. The main functional units are the aeration basin and the final clarifier as shown in Fig . 1. The biological degradation of organic materials occurs principally in the aeration basin, whi le mechanical separation of the purified water and the biomass and the biomass thickening are functions in the final clarifier. Incoming normally mechanically pretreated wastewater is fed to the aeration basin, where aerobic micro-organisms attack the dissolved and colloidal organic materials in the wastewater and metabol ize them producing stabilized endproducts, mainly carbon dioxide and water. At the same time the biomass is growing. To increase the phosphorus removal, ferrous salts (FeSO"7H 2 0) are often added to the aeration basin (~imuT­ taneous precipitation).

F\..xw. XR

F~. x~

:-----R€TuRN-sl"uDGE---l- -EXCESS- SLUDGE-" , I

I I

( alternative 1)

~F~X~~H~~__~__A-~~~~~~'~

,

INA.UE

or / --"

~M . V...x.C . rx

:~------\S

I

:

I 'b

I I I

F...,.x..rX

~---5C~SSL@~--~

AIR

Stabilization of organics, biomass respiration and auto-oxidation, nitrification and precipitant (FeS04' 7H20) oxidation consume dissolved oxygen in the aeration basin . This consumption is replaced by powerful aeration of the basin contents. At the same time sufficient mixing for keeping the biomass in suspension is achieved. In the final clarifier the biologically purified water and the biomass are separated from each other by gravitation. The former is discharged through disinfection to the receiving stream and the latter is recycled (return sludge) to the aeration basin. The net growth of the biomass (excess sludge) is fed to the sludge handl ing processes. The excess sludge is removed from the final clarifier (alternative 1) of preferably directly from the aeration basin (alternative 2). Aeration profiles The dissolved oxygen profile along the aeration basin should be adjusted as equal as possible. Oxygen demand in the aeration basin depends primarily on the used process variant, Fig . 2. Air diffuser clogging or chemical oxidation in simultaneous precipitation may disturb the dissolved oxygen profi le. In a conventional plug flow proces s , Fig. La, the biomass is growing at maximum rate at the beginning of the aeration basin consuming a large amount of oxygen. At the end of the basin the unreacted substrate concentration decreases and oxygen is consumed only by endogenous respiration . The oxygen demand decreases and the biomass begins to flocculate. An equal dissolved oxygen profile in the basin may be adjusted by step aeration (cig. 2b), tapered aeration (Fig. 2c) or by complete mix process (Fig. 2d). In step aeration and complete mix processes the organic load is equally distributed to the basin contents thus improving the effective use of the whole aeration volume. Better circumstances for sludge flocculation and thickening are usually, however, obtained in conventional plug flow and tapered aeration processes.

( alte r native 2)

AERATION BASIN

FINAL CLARI FIER

Process control objectives influent and effluent f l ow, re urn and e xcess (waste ) s ludge flow s. III 3Id

influent, effluent. aer
'A ' Vs

aeration basin and f i nal c larifier volumes . m3

H, .1M/ AT = rx VA

sludge

fl\a SS

and it s ne t growth in the systE!l'l.

kg and kg/ d sludge A9f! . d

yield coeff icient, kg / kg '0

sludge decay coefficient, 1 / d food-to-micro- organ lsm rati o . lId sludge recyc le and wa s ting r a tio, sludge

Fig. 1.

concentrat i on fa c t or, 1

Activated sludge process schematics and operational parameters.

1

The control of activated sludge processes may well be compared to that of similar production processes in chemical and biotechnical industry, where automation has been appl ied for many years successfully. Automation has improved process performance, productivity, product qual i ty, process reI iabi I i ty and stabil ity and at the same time reduced operational and maintenance costs . General and special purpose digital computers are currently an integral part of such measurement and control systems. In activated sludge process the raw materials are the incoming wastewater and the chemicals used in the treatment. Products are the purified water and the sludge. The removal

Experiences on Instrumentation and Control

~ 'onk~h

_

Olr supply (JlygL

= a L'>T

NH

+ bM + c

L'> 3 -xr-

where OR = oxygen consumption, a = 0.4 ... 0,65 kg02/kgBOD5' b = 0.1 . . . 0.15 kg02/kgSSd and c = 4.5 ... 4.6 kg02/kgNH3' Because the oxygen content in the aeration basin is very small (corresponds to oxygen consumption of only a few minutes), the consumption must be immediately balanced, i . e. the oxygenation capac i ty OC

= CXKLa(C~

- C) VA

= the dissolved oxygen saturation

where C*

concent~ation in the aeration basin and

a = the total mass transfer coefficient ofLoxygen, must be equal to OR, or

~K

OC

= OR

As the mass balance of dissolved oxygen is difficult to formulate quantitatively exactly enough, a first order lag plus dead time model T

~ dt

T

e K u AIR

+ C

= KU(t - e)

where

time I