Final Report Odoco1 TUB MPA Barjenbruch080411 KWB 080707 Final Klein
Final Report Odoco1 TUB MPA Barjenbruch080411 KWB 080707 Final Klein
Final Report Odoco1 TUB MPA Barjenbruch080411 KWB 080707 Final Klein
24
D-10709 Berlin
Germany
Tel +49 (0)30 536 53 800
Fax +49 (0)30 536 53 888
www.kompetenz-wasser.de
REPORT
Gustav-Meyer-Allee 25
Berlin-Brandenburg
13355 Berlin
Voltastrae 5
Germany
13355 Berlin
Germany
for
Kompetenzzentrum Wasser Berlin gGmbH
Berlin, Germany
2008
Copyright 2008 by the KompetenzZentrum Wasser Berlin gGmbH. All rights including translation into other languages,
reserved under the Universal Copyright Convention, the Berne Convention or the Protection of Literacy and Artistic
Works, and the International and Pan American Copyright Conventions.
ii
Abstract
Project name:
Duration:
4/2007 to 12/2007
Veolia Water
KWB contact:
Odours emerging from sewage networks are unpleasant, can cause health impacts on
sewer workers and impair public perception of the operator companies. Corrosion is one
of the causes for the cost of repairs for damages to wastewater systems in the public
sewage network, which are rising extremely [DWA, 2004]. Both phenomena can have
their origin in biogenic acid corrosion that is illustrated in this report. The Kompetenzzentrum Wasser Berlin (KWB) commissioned the Technical University Berlin and the Material Testing Institute of Berlin-Brandenburg to give a report on:
State of the art on control systems for odours and corrosion problems in sewer
networks,
State of the art on sensor technologies for water, gas and corrosion parameters
to follow corrosion and H2S production,
Control systems
The report contains recently acquired knowledge about processes and process parameters that lead to odour and corrosion. Biochemical and physical mechanisms as well as
their enhancing and hindering frame conditions are explained in detail. Literature study
iii
and own investigations show that retention time, temperature and biological activity are
relevant in all parts of the system, that is liquid and gaseous phase as well as biofilm below and above water level. Furthermore, redox potential, pH-value and ratio of submerged biofilm to wastewater volume are important quantities in sewage and biofilms.
Type of sewage network, type of pumping and constructive features demonstrate an influence on production of odour and corrosion, too. Attempts were made to quantify the
influence of these parameters. But for all those efforts, neither the critical point of retention time could be determined nor correlation with temperature is generally quantified.
Not even the inclusion of a number of relevant parameters in empirical equations (COD,
sulphur content etc.) of various authors (THISTLETHWAYTE, BOON & LISTER, ATV, US EPA,
POMEROY, HVITVED-JACOBSEN)
ple keeping in a certain H2S-conentration or retention time are not available. So detailed
investigations should be carried out, to get more knowledge about the processes and
their influencing parameters.
Odours and corrosion can be controlled with various countermeasures. Countermeasures that are often used and approved as well as new approaches will be explained.
They utilise methods of enrichment with oxidants, precipitation of sulphide, pHregulation, gas treatment with biofilters, covering systems and constructive solutions.
Feedback control systems permit to ensure the right choice and adjustment of countermeasures. Data on the state of material, online-measured parameters and experienced
patterns trigger the adequate adjustment. Several examples are given for this relatively
new field. It can be concluded, that deeper knowledge about relationships between parameters and odour or corrosion, including knowledge about kinetics of these processes,
would be helpful to adjust countermeasures automatically. Optimized dosage concepts
can achieve economical application of those chemical methods. A reliable prediction
technique based on modeling, could support this.
Measurements
The investigation on existing sensors in relation to the necessary parameters to be
measured, determining the condition of technical wastewater facilities, regarding the
presence of odour and corrosion effective substances, results in the following:
Plastic corrosion (aromatic and halogenated hydrocarbons)
Sensory processes for the continuous monitoring of concentration of halogenated and
aromatic hydrocarbons in sewage do not exist. For detection in the headspace, the IRsensor shows the highest selectivity (detection of CH-binding) and is suitable for environments of very high humidity.
iv
Data processing
The necessity to implement an integrated data processing and storage system is
strongly recommended. Large data sets, documenting sensorially and analytically won
measurements, are emerging in the process of monitoring sewage systems for control of
odour and corrosion. Units of data have to be reasonably and automatically captured and
pre-processed. They form the basis for any calibration and validation of the numerical
models, as well as the foundation to assess the model-quality. Since a comprehensive
study investigating the field of data processing was not targeted, only a short depiction of
the required tasks and the potential approaches are given.
Modelling
Empirical formulas are just capable of giving a very rough estimation of the sulphide development in sewers and they are poorly transferable to general conditions. Urban drainage, wastewater plant and water quality models have some more or less close relations
to the simulation of flow, transport and reactions processes in sewers systems. Among,
the very few models which are capable to simulate the above-mentioned processes, only
the WATS model has been applied to odour and corrosion problems in sewer systems.
The WATS model has its strength in simulating the microbial and chemical reaction
processes of organic matter, sulphur and other components in the water and gas phase
including mass transfer water-gas and corrosion with a focus on dry weather conditions.
Current deficits of WATS are in the flow simulation (no non-uniform and instationary water flow, no turbulence, no gas flow) and the transport simulation (no diffusion / dispersion in water and gas, no sediments and erosion / deposition). Problems dealing for example with rainwater conditions, ventilation, flushing, flow over drops can not be treated
with WATS. Furthermore, the numerics and the software design can be improved. The
WATS model is not public domain, and the sources and the software are not available.
However, the developers are very open for collaborations (personal communication). The
amount of the source code is not very high and all methods are published. The quality of
the WATS model strongly relies on the availability and quality of data, and many data of
the WATS group are published, too.
In future work, a three-phase model will be developed including the phases water, gas
and solid (sewer wall) for prediction odour and corrosion in sewer systems. The model
will solve the coupled equations for flow, transport and reaction in the water and gas
phase including mass-transfer processes (hydrogen sulphide, oxygen). Based on previous work, it will use the Finite-Volume Method and it will be embedded in an objectorientated framework. The major focus of the model will consist of biological and chemivi
cal transformations of carbon, sulphur, and nitrogen. First, a calibration and validation in
a pilot sewer under controlled conditions is required. In further steps, it can applied for
designing countermeasures (ventilation, dosages, flushing, ...) and it can be transferred
to real sewer systems. The three-phase model will be part of a closely linked monitoring
and prediction system.
Outlook
The results of the feasibility study (ODOCO-1) show that it is possible to develop a prediction technology. Further studies shall be carried out at a sewer pilot plant where conditions can be controlled and parameters of odour and corrosion are adjustable. Existing
sensor technologies will be applied and improved for capturing of correlations. After data
processing and storage, this information is available for investigation on biochemical and
transport processes. A three-phase model will be built on the base of this information.
With the help of this model, countermeasures can be adjusted accurately in a feedback
control system. A detailed draft of a sewer pilot plant is given in the proposal for Developing a prediction technology using studies at a pilot plant ODOCO-2 in the appendix.
vii
Zusammenfassung
Projektname:
Dauer:
Volumen:
48.750 Euro
Durchfhrung:
Prof. Dr. Matthias Barjenbruch, Technische Universitt Berlin, Institut fr Bauingenieurwesen, Fachgebiet Siedlungswasserwirtschaft,
Stiftungsprofessur KWB Veolia Wasser
Dr. Bodo Weigert, KWB
Christophe Renner, Anjou Recherche
Sponsor:
Veolia Water
KWB-Kontakt:
Aus Kanalisationsnetzen aufsteigende Gerche sind unangenehm, sie knnen Gesundheitsprobleme bei Kanalarbeitern hervorrufen und den Ruf der Betreiberfirmen schdigen. Korrosion ist eine der Ursachen fr die stark steigenden Kosten, die bei der Beseitigung von Schden an Abwassersystemen in ffentlichen Kanalisationsnetzen anfallen
[DWA, 2004]. Beide Erscheinungen knnen durch die in diesem Bericht betrachtete biogene Surekorrosion verursacht sein. Das Kompetenzzentrum Wasser Berlin (KWB) beauftragte die Technische Universitt Berlin und das Materialprfungsamt BerlinBrandenburg mit einem Bericht ber:
den Stand der Technik bei Kontrollsystemen fr Geruchs- und Korrosionsprobleme in Kanalisationsnetzen,
den Stand der Technik bei Sensortechnologien fr Wasser-, Gas- und Korrosionsparameter zur Beobachtung der Korrosion und der H2S-Erzeugung,
Kontrollsysteme
Der Bericht enthlt neue Erkenntnisse ber Prozesse und Prozessparameter, die Geruch und Korrosion hervorrufen. Biochemische und physikalische Mechanismen sowie
ihre frderlichen und hinderlichen Rahmenbedingungen werden im Detail erklrt. Ein Literaturstudium und eigene Untersuchungen der Autoren zeigen, dass Retentionszeit,
viii
Temperatur und biologische Aktivitt in allen Teilen des Systems relevant sind, also in
der flssigen und der Gasphase ebenso wie in dem Biofilm ber und unter dem Wasserspiegel. Auerdem sind das Redoxpotential, der pH-Wert und das Verhltnis zwischen
dem Biofilm unter dem Wasserspiegel zum Abwasservolumen wichtige Gren im Abwasser und in Biofilmen. Effekte von Geruch und Korrosion werden auch durch die Art
des Kanalisationsnetzes und der Pumpeinrichtungen sowie die Bauweise beeinflusst. Es
wurde bereits vielfach versucht, den Einfluss dieser Parameter zu quantifizieren. Bei all
diesen Bemhungen konnte aber weder der kritische Punkt der Retentionszeit bestimmt
werden, noch wurde die Korrelation mit der Temperatur berhaupt quantifiziert. Selbst
die Bercksichtigung einiger relevanter Parameter in empirischen Gleichungen (CSB,
Schwefelgehalt usw.) verschiedener Autoren (THISTLETHWAYTE, BOON & LISTER, ATV, US
EPA, POMEROY, HVITVED-JACOBSEN)
spielsweise sind keine praktischen Kriterien fr die Einhaltung einer bestimmten H2SKonzentration oder Retentionszeit verfgbar. Es sollten also eingehendere Untersuchungen durchgefhrt werden, um mehr Wissen ber die Prozesse und die sie beeinflussenden Parameter zu erhalten.
Gerche und Korrosion knnen mit verschiedenen Gegenmanahmen kontrolliert werden. Es werden oft angewendete und erprobte Gegenmanahmen und auch neue Verfahren erlutert. Sie nutzen Methoden der Anreicherung mit Oxidantien, die Sulfidausfllung, die pH-Einstellung, die Gasbehandlung mit Biofiltern, Abdecksysteme und bauliche
Lsungen. Rckkopplungsregelungssysteme geben die Mglichkeit, die richtige Auswahl
und Anpassung der Gegenmanahmen zu gewhrleisten. Daten ber den Materialzustand, online gemessene Parameter und erfahrungsgeme Muster dienen der adquaten Anpassung. Fr dieses verhltnismig neue Gebiet werden mehrere Beispiele gegeben. Man kann daraus schlieen, dass vertieftes Wissen ber Beziehungen zwischen
Parametern und Geruch und Korrosion, einschlielich von Wissen ber die Kinetik dieser
Prozesse, hilfreich bei einer automatischen Anpassung der Gegenmanahmen wre.
Optimierte Dosierungskonzepte knnen fr die wirtschaftliche Anwendung dieser chemischen Verfahren sorgen. Ein auf Modellierung beruhendes zuverlssiges Vorhersageverfahren knnte das untersttzen.
Messungen
Die Untersuchung vorhandener Sensoren in Bezug auf die zu messenden Parameter fr
die Bestimmung des Zustands technischer Abwasseranlagen unter Bercksichtigung der
Anwesenheit geruchs- und korrosionswirksamer Stoffe kam zu den folgenden Ergebnissen:
ix
Die Verfahren zur Bestimmung der biologischen Aktivitt von Biofilmen sind noch im
Entwicklungsstadium. Es ist unklar, ob die Messung der Biofilmaktivitt spezifisch genug
ist, um relevante Informationen zu den Problemen von Geruch und Korrosion zu liefern.
Andererseits ist die Messung physikalischer Parameter (relative Feuchtigkeit, Temperatur, Luftdruck und Durchsatz) durchaus bewhrt.
Datenverarbeitung
Die Anwendung eines integrierten Datenverarbeitungs- und Speichersystems ist dringend zu empfehlen. Bei der berwachung von Abwassersystemen zur Kontrolle von Geruch und Korrosion fallen groe Datenmengen an, die durch Sensoren oder Analyse
gewonnene Messungen dokumentieren. Die Dateneinheiten mssen in vernnftiger
Weise und automatisch erfasst und vorverarbeitet werden. Sie bilden die Basis fr jede
Eichung und Validierung der numerischen Modelle und sind die Grundlage fr die Bewertung der Modellqualitt. Da eine umfassende Untersuchung in Bezug auf das Gebiet
der Datenverarbeitung nicht beabsichtigt war, wird nur eine kurze Beschreibung der erforderlichen Aufgaben und der potentiellen Verfahren gegeben.
Modellierung
Empirische Formeln knnen nur eine sehr grobe Schtzung der Sulfidentwicklung in
Abwssern liefern und sind schlecht auf allgemeine Bedingungen bertragbar. Stadtentwsserungs-, Klrwerks- und Wasserqualittsmodelle haben einige mehr oder weniger
enge Beziehungen zur Simulation von Strmungs-, Transport- und Reaktionsprozessen
in Kanalisationssystemen. Von den sehr wenigen Modellen, die die oben genannten
Prozesse simulieren knnen, ist nur das WATS-Modell auf Geruchs- und Korrosionsprobleme in Abwassersystemen angewendet worden.
Die Strke des WATS-Modells liegt in der Simulierung der mikrobiellen und chemischen
Reaktionsprozesse von organischer Substanz, Schwefel und anderen Bestandteilen in
der wssrigen und der Gasphase, einschlielich des Stoffaustauschs Wasser-Gas und
der Korrosion, mit einem Schwerpunkt auf den Bedingungen bei trockenem Wetter. Die
derzeitigen Defizite des WATS-Modells liegen in der Strmungssimulation (keine ungleichmige und instationre Wasserstrmung, keine Turbulenz, kein Gasstrom) und
der Transportsimulation (keine Diffusion/Dispersion in Wasser/Gas, keine Sedimente
und Erosion/Ablagerung). Beispielsweise knnen Probleme, die mit Regenwasserbedingungen, Belftung, Splung oder berlufen verbunden sind, nicht mit WATS behandelt
werden. Auerdem sind die Numerik und die Software-Gestaltung verbesserungsfhig.
Das WATS-Modell ist nicht lizenzfrei, und die Quellen und die Software sind nicht verfgbar. Die Entwickler sind jedoch sehr offen fr Zusammenarbeit (persnliche Mittei-
xi
lung). Der Umfang des Quellcodes ist nicht sehr gro und alle Methoden sind verffentlicht. Die Qualitt des WATS-Modells hngt stark von Verfgbarkeit und Qualitt der Daten ab, viele Daten der WATS-Gruppe sind auch verffentlicht.
In der zuknftigen Arbeit wird ein Dreiphasenmodell zur Vorhersage von Geruch und
Korrosion in Kanalisationssystemen entwickelt werden, das die Phasen Wasser, Gas
und Feststoff (Kanalwand) einbezieht. Das Modell wird die gekoppelten Gleichungen fr
Strmung, Transport und Reaktion im Wasser und in der Gasphase lsen, einschlielich
der Stoffaustauschprozesse (Schwefelwasserstoff, Sauerstoff). Auf frheren Arbeiten beruhend wird die Methode finiter Elemente angewendet. Der Schwerpunkt des Modells
wird die Abbildung von biologischen und chemischen Transformationen von Kohlenstoff,
Schwefel und Stickstoff bilden. Zuerst sind eine Eichung und die Validierung in einer Pilotkanalisation unter kontrollierten Bedingungen erforderlich. In weiteren Stufen kann das
Modell fr das Entwerfen von Gegenmanahmen (Belftung, Dosierungen, Splen usw.)
eingesetzt werden, und es kann dann auf reale Kanalisationssysteme angewendet werden. Das Dreiphasenmodell wird Bestandteil eines eng vernetzten berwachungs- und
Vorhersagesystems sein.
Ausblick
Die Ergebnisse der Machbarkeitsstudie (ODOCO-1) zeigen, dass es mglich ist, eine
Vorhersagetechnologie zu entwickeln. Weitere Untersuchungen sollten an einer Abwasserpilotanlage durchgefhrt werden, in der die Bedingungen kontrolliert werden knnen
und die Parameter von Geruch und Korrosion anpassbar sind. Hier knnen vorhandene
Sensortechnologien angewendet und verbessert werden, um Korrelationen zu erfassen.
Nach Datenverarbeitung und Speicherung stehen diese Informationen fr Untersuchungen von biochemischen und Transportvorgngen zur Verfgung. Beruhend auf diesen
Informationen knnte ein Dreiphasenmodell erstellt werden. Mit Hilfe dieses Modells
knnen Gegenmanahmen in einem regelungstechnischen System genau angepasst
werden. Im Anhang ist ein detaillierter Entwurf einer Abwasserkanalpilotanlage enthalten
mit einem Vorschlag zur Entwicklung einer Vorhersagetechnologie auf Grund von Untersuchungen an einer Pilotanlage ODOCO-2.
xii
Rsum
Nom de projet :
Dure :
04/2007 12/2007
Volume :
48.750 euros
Ralisation :
Parrain :
Veolia Water
Contact KWB :
xiii
Systmes de contrle
Le rapport contient de nouvelles connaissances sur les processus et les paramtres de
processus provoquant les odeurs et la corrosion. Les mcanismes biochimiques et
physiques ainsi que les conditions cadres utiles et embarrassantes sont expliqus en
dtail. Une tude bibliographique et les analyses faites des auteurs montrent que le
temps de rtention, la temprature et lactivit biologique dans toutes les parties du
systme sont intressants, donc aussi bien dans la phase liquide et dans la phase
gazeuse que dans le biofilm au-dessus et sous le niveau deau. En outre, le potentiel
redox, la valeur de pH et le rapport entre le biofilm sous le niveau deau et le volume des
eaux uses sont des donnes importantes pour les eaux uses et les biofilms. Les effets
des odeurs et de la corrosion sont galement influencs par le type du rseau
dassainissement et des installations de pompage ainsi que par le mode de construction.
Il a dj t tent plusieurs reprises de quantifier linfluence de ces paramtres. Toutes
ces tentatives nont cependant pu aboutir ni une dtermination du point critique du
temps de rtention ni une quantification de la corrlation avec la temprature. Mme la
prise en compte de certains paramtres intressants dans des quations empiriques
(DCO, teneur en soufre etc.) de diffrents auteurs (THISTLETHWAYTE, BOON & LISTER, ATV,
US EPA, POMEROY, HVITVED-JACOBSEN)
dispose par exemple daucun critre pratique pour le respect dune certaine
concentration en H2S ou du temps de rtention. Des analyses plus approfondies doivent
donc tre menes afin dobtenir plus dinformations concernant les processus et les
paramtres qui les influencent.
Il est possible de contrler les odeurs et la corrosion avec diffrentes contre-mesures. Le
rapport explique les contre-mesures appliques et prouves mais aussi de nouveaux
procds. Sont utilises des mthodes denrichissement avec des oxydants, la
prcipitation de sulfate, le rglage du pH, le traitement de gaz avec des biofiltres, des
systmes de recouvrement et des solutions de construction. Les systmes de rglage
par rtroaction offrent la possibilit de garantir le bon choix et le bon ajustement des
contre-mesures. Les donnes concernant ltat des matriaux, les paramtres mesurs
en ligne et les modles expriments permettent de raliser un ajustement adquat.
Plusieurs exemples sont donns pour ce domaine relativement nouveau. On peut en
conclure quun savoir approfondi concernant les relations entre paramtres et odeurs et
corrosion, incluant un savoir sur la cintique de ces processus, serait prcieux dans le
cadre dun ajustement automatique des contre-mesures. Des concepts de dosage
optimiss peuvent permettre lapplication rentable de ces procds chimiques. Un
procd de prvision fiable reposant sur une modlisation pourrait servir de support.
xiv
Mesures
Lanalyse des capteurs existants en regard des paramtres mesurer pour la
dtermination de ltat des installations techniques dvacuation des eaux uses, avec
prise en compte de la prsence de matires provoquant odeurs et corrosion, a conduit
aux rsultats suivants :
Corrosion des matires plastiques (hydrocarbures aromatiques et halogniss)
Il nexiste pas de procd de captation pour une surveillance en continu de la
concentration dhydrocarbures halogniss et aromatiques dans les eaux uses. Lors de
la mesure dans lespace de tte, le capteur IR affiche la plus grande slectivit
(affichage de la liaison CH). Il convient aux environnements avec une humidit trs forte.
Corrosion du bton (acide sulfurique)
Il nexiste pas de mthodes prouves pour la mesure par capteur en continu des
valeurs pH et de la concentration en sulfate dans le biofilm. Le dveloppement de
capteurs pour la mesure en continu de valeurs pH dans le biofilm de lespace de tte
serait trs intressant en particulier pour la dmonstration prcoce de dommages
ventuels au niveau dinstallations dvacuation des eaux uses par corrosion dtruisant
le bton.
Il existe un prototype en fonctionnement dun capteur fibre optique pour la
dtermination de la valeur pH dans du bton arm dinstallations dvacuation des eaux
uses, actuellement dvelopp en collaboration avec lOffice dexamen des matriaux
de Berlin-Brandebourg. Le systme est actuellement test dans le cadre dun projet de
recherche.
Odeurs (H2S)
Il existe diffrents procds de mesure de H2S dans lespace de tte. Il nest pas
possible de dterminer avec certitude dans quelle mesure les capteurs sont en mesure
de fournir suffisamment de signaux de mesure stables dans le cadre de la mesure
longue dure en continu de H2S dans lespace de tte dinstallations dvacuation des
eaux uses.
Autres paramtres de modlisation
En
ce
qui
concerne
la
dtermination
dautres
paramtres
importants
pour
xv
Modlisation
Les formules empiriques ne peuvent fournir quune estimation trs grossire de
lvolution de sulfure dans les eaux uses et sont difficilement applicables aux conditions
gnrales. Les modles de drainage urbain, dusines dpuration et de qualit de leau
prsentent quelques rapports plus ou moins troits avec la simulation des processus
dcoulement, de transport et de raction dans les systmes dassainissement. Parmi les
rares modles capables de simuler les processus mentionns ci-dessus, seul le modle
xvi
Le travail futur dveloppera pour la prvision des odeurs et de la corrosion dans les
systmes dassainissement un modle trois phases intgrant les phases eau, gaz et
matire solide (paroi de canal). Le modle solutionnera les quations couples
dcoulement, de transport et de raction dans leau et dans la phase gazeuse, incluant
les processus de transfert de masse (sulfure dhydrogne, oxygne). Reposant sur des
travaux dj accomplis, la mthode emploiera des volumes finis et elle sera incorpore
dans un cadre tourn vers les objets. Les transformations biologiques et chimiques de
carbone, de sulfure et dazote formeront la dominante du modle. Tout dabord, un
jaugeage et une validation dans une canalisation pilote dans des conditions contrles
doivent tre effectus. Le modle peut tre utilis dans dautres tapes pour lbauche
de contre-mesures (ventilation, dosages, rinage, ) et peut ensuite tre appliqu sur
des modles de canalisations rels. Le modle trois phases fera partie intgrante dun
systme de surveillance et de prvision troitement mis en rseau.
Perspectives
Les rsultats de ltude de faisabilit (ODOCO-1) montrent quil est possible de
dvelopper une technologie de prvision. Dautres analyses doivent tre menes sur
une installation pilote dvacuation des eaux uses avec possibilit de contrler les
xvii
xviii
Table of Contents
Chapter 1 Introduction........................................................................................................... 1
1.1
Odour ....................................................................................................................... 1
1.2
Corrosion.................................................................................................................. 2
2.3
2.3.1 Introduction............................................................................................................. 12
2.3.2 Biogenic acid corrosion .......................................................................................... 12
2.4
Introduction............................................................................................................. 22
4.2
4.3
4.4
4.5
xix
Chapter 5 Countermeasures............................................................................................... 69
5.1
5.3
7.2
Urban-drainage models.......................................................................................... 92
7.3
Wastewater-plant models....................................................................................... 95
7.4
7.5
7.6
Flow, transport and reaction models in sewers and the WATS model................. 100
xx
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 1
Introduction
The increasing centralisation of wastewater disposal has led to the extension of sewerage networks causing long detention times. A long detention time is an important precondition for wastewater to ferment. Fermenting of sewage in sewers causes several
undesirable effects:
Emergence of odour, particularly at transfer shafts from pressure into gravity systems
Especially the emergence of odour increasingly causes acceptance problems for the
wastewater disposal companies. For this reason, the amount of chemicals used for septicity prevention has risen immensly [YANG & HOBSON, 2001]. In most cases, corrosion
leads to considerable costs in the long run. E.g. Veolia Eau Germany employs 800 t/a
Nutriox for about 200 000 /a.
First, problems of odour and corrosion will be illustrated. In the second section, causing
processes are presented. Chapter 3 and Chapter 4 present sensor technologies to detect all necessary parameters and show how to process and store the resulting data. The
state of the art on countermeasures and control systems for odour and corrosion problems in sewer networks is described in Chapter 5. Existing model concepts are introduced and their potentials for prediction of odour and corrosion are assessed in Chapter
7. Based on the knowledge of causing processes and effects of countermeasures, approved and possible feedback control systems are presented in Chapter 6. A preproposal for further investigation on An Online-Monitoring and Operating System to prevent
Odour and Corrosion in Sewer Networks Pilot Plant and Prediction Model has been
elaborated by the authors and is a supplement to this report.
1.1 Odour
Odour is the property of a substance that is perceptible by the sense of smell [OXFORD
ENGLISH DICTIONARY].
ters the nasal cavity and excite the osmoreceptors ... Certain parts of an odourous mole-
Report ODOCO-1
_________________________________________________________________________________________________
cule are responsible for the characteristic smell. These are called osmogenes. The sulphuryl group (SH) is one such osmogene [ENVIRONMENT AGENCY UK, 2001].
Raw sewage from households has a characteristic musty smell, resulting from its components. Although it is not pleasant, it rarely gives rise to odour problems [HORAN &
MARA,
Secondary osmogenes need time to develop. Therefore, they emerge from aged sewage
in wastewater treatment systems. Common sewage odour compounds are sulphides
(hydrogen sulphide, methylmercaptan, ethylmercaptan, (di)methyl(di)sulphide), nitrogen
containing compounds (ammonia, (di)methylamine, indole, skatole), acids (acetic, butyric, valeric), aldehydes (formaldehyde) and ketones. The amount of a compound released to the atmosphere depends on various parameters. E.g., the pH-value determines the chemical state of a compound in the liquid phase. Low pH-values favour the
emission of H2S, mercaptans and volatile fatty acids, while high pH-values favour the
emission of ammonia and amines. In general, odours at high pH-values are less intense,
causing rarely odour nuisance. Therefore, it is justified to focus on odours emerging from
neutral to acid wastewater [HORAN & MARA, 2003]. In this range of pH-value, hydrogen
sulphide is acknowledged as a keynote parameter for odour nuisance from household
sewage. As H2S is the dominant odorant here, concentration of H2S in gas phase is an
acceptable surrogate for odour [FRECHEN, 2007]. The buildup of hydrogen sulphide is
described below in sec. 2.1.1.
1.2 Corrosion
The term corrosion describes all reactions on metallic and non-metallic construction
materials with their environment, which lead through chemical, electro-chemical or
microbiological processes to a detraction of the construction material. Inside sewer systems made of concrete or steel three types of corrosion are distinguished. In the submerged parts they consist of
Above water level, dominating process of corrosion is the aerobic biogenic corrosion
of sulphuric acid. Under certain conditions, this is the fastest mechanism of corrosion,
so we focus on this type of corrosion [MATSCH & SARACEVIC, 2007].
Report ODOCO-1
_________________________________________________________________________________________________
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 2
Causing processes
2.1 Aerobic biogenic corrosion of sulphuric acid
Hydrogen sulphide, produced below water level, strips out to the sewer atmosphere. It
dissolves in condensed water of the sewer wall where it is transformed to sulphuric acid
attacking the material. Sulphur species are key reactants in this process. Their transformation reactions are described below.
Report ODOCO-1
_________________________________________________________________________________________________
They utilise products of fermenting bacteria like alcohols, organic acids and protons as
reducing agent [SEWER PROCESSES AND NETWORK GROUP, 2006; MARA & HORAN, 2003].
Sulphate dissimilation takes place in anoxic parts of submerged biofilm, bulk water and
sediments. HVITVED-JACOBSEN (2002) explains, that because of their slow growing rate,
SRB have no chance to establish in flowing sewage, so biofilms and deposits are relevant locations of their activity. Sulphate dissimilation is the main process of H2Sformation:
SO42- S2- + HS- + H2S
Desulphuration
A less important source of sulphide is desulphuration. Fermenting bacteria hydrolyse
sulphur-containing amino acids and produce H2S [LAWA, 2004]:
R-S S2- + HS- + H2S
2006]. At redox potentials within the range of -200 to -300mV they pro-
duce H2S, resulting in a further drop in Eh [MARA & HORAN, 2003; ODOM, 1993; FUCHS,
2007]. According to BITTON (1999), SRB favour a redox potential between -150 and 200mV. HEITZ (1997) measured Eh =-375 to -250 mV on a steel surface below a biofilm
of SRB compared to 50 to 150 mV without biofilm.
Anabolism of SRB is quite variable and flexible between autotrophic and heterotrophic
[FUCHS, 2007]. Some varieties oxidise lactat to acetat, others oxidise it completely
[MATSCHE, 2007]. On account of this, the concentration of hydrocarbons has no direct
influence on H2S-production. The height of concentration of hydrocarbons and their
biodegradability induce the time until dissolved oxygen is depleted. This way it is a
Report ODOCO-1
_________________________________________________________________________________________________
contributing factor to H2S-production. For this reason, COD or ratio of COD/BOD5 is directly related to the amount of sulphide formed [BOON, 1995; HVITVED-JACOBSEN et al.,
1988 in MARA & HORAN, 2003].
Biofilms on the submerged conduits are supplied with sufficient substrate and consume about 700 mg/m/h oxygen at 15C [BOON & LISTER, 1975 cited in MARA & HORAN,
2003, POMEROY, 1991]. Their influence is discussed controversially. KLOSE (1981) states
that the build-up of sulphide happens independently of the thickness of the submerged
biofilm [LAWA. 2004]. By contrast, YANG & HOBSON (2001) show that it takes up the most
oxygen. The ratio of biofilm surface to wastewater volume is higher in pipes of small diameter. Therefore, their relative rate of sulphide-formation is bigger [ATV-DVWK M 154,
2003]. This is of importance especially in rising mains with small diameters. ECKER
(2003) mentions values of 700 mg O2/(mh) as the maximum oxygen uptake rate in
submerged biofilms [LAWA, 2004]. LOHSE (1986) found a high variability of oxygen uptake between 0,2 and 7 g O2/(mh).
Beside activity of microorganisms in the submerged biofilm, activity in the sewage itself
is important. Raw domestic sewage takes up 2 4 mg O2/l/h [ECKER, 2003 IN LAWA,
2004]. During the retention time, bacteria populations establish and the number of microorganisms in sewage increases. Accordingly, the respiration rate and the rate of uptake of dissolved oxygen accelerate. Then, under anaerobic conditions, a growing number of microorganisms break down complex organic compounds [MARA & HORAN, 2003],
producing readily biodegradable material. This is why the rate of uptake of dissolved
oxygen increases with the retention time from 2 mg/l/h to 14 mg/l/h at 15C [BOON, 1995
cited in MARA & HORAN, 2003].
LOHSE
merged biofilm and sewage to predict oxygen required for preventing the development of
sulphide in rising mains:
D z Abw
OC = 0,024 D L z Sh +
Q24 (cO 2 1)
4
OC
D
L
ZSh
=
=
=
=
ZAbw
Q24
CO2
=
=
kg
d
Report ODOCO-1
_________________________________________________________________________________________________
The influence of the retention time is acknowledged. LOHSE (1986 in LAWA, 2004)
proves that the formation of sulphide increases up to retention times of 6 hours. BOON
[1995; HVITVED-JACOBSEN et al., 1988 in MARA & HORAN, 2003] states that the retention
time is proportional to the amount of sulphide formed. This is valuable up to a retention
time of 6 hours. After 6 hours stagnation begins [LAWA, 2004]. MARA & HORAN (2003) focus especially on retaining time under anaerobic conditions as a valuable parameter. A
distinct specification of a critical age does not make sense, because of various parameters [LAWA, 2004].
A small flow velocity (originating from high diameters, improper slope and low water
level) contributes to a higher retention time and causes deposits that enhance anaerobicity [FRECHEN, 2007].
Temperature is a dominant parameter for concentration of H2S in sewer atmosphere
[MATSCH & SARACEVIC, 2007; FRECHEN, 2007]. Its formation increases with temperature
ranging from 15-27C according to KITAGAWA in [LAWA, 2004]. Formation rate is likely to
double for a 10C increase in temperature within the range of about 5-25C [MARA &
HORAN,
2003]. The reasons for this are increased biological activity in general and on the
other hand a smaller solubility of gases like oxygen and H2S in warmer fluids, leading to
a faster depletion of oxygen and a risen part of released H2S (s. sec. 2.1.2) [BOON, 1995;
HVITVED-JACOBSEN
increases in temperature will start to have an adverse effect on growth rate of microorganisms [MARA & HORAN, 2003]. The equation for reaction kinetics of ARRHENIUS describes effects of temperatures on microbiological activity. LOHSE (1986) cites the modified form:
kT = k20 T-20
kT
k20
= temperature [C]
Temperature coefficients vary over a wide range from 1.024 to 1.139 [cited in LOHSE,
1986]. Their vast variety and dependency on temperature range and constitution of
wastewater constrict application of the temperature coefficient [LOHSE, 1986].
Report ODOCO-1
_________________________________________________________________________________________________
pH tolerance of SRB is between 5.5 and 9. An optimal condition for sulphide production
is a pH between 6.5 [MARA & HORAN, 2003] or 7 [LAWA, 2004] and 8. Its main importance is connected to gas release [FRECHEN, 2007] (s. sec. 2.1.2).
As the largest part of H2S production takes place in the submerged biofilm, a high ratio of
submerged surface area to sewage volume increases specific H2S production [LAWA,
2004; MATSCH & SARACEVIC, 2007; BOON, 1995; HVITVED-JACOBSEN et al., 1988 in MARA
& HORAN, 2003].
Mass transfer of H2S from water phase to gas phase is liable to the general equation for
mass transfer:
m& = A k (c z c x )
m&
Report ODOCO-1
_________________________________________________________________________________________________
cz
cx
pi = H i xi = yi ptotal
H
x
ptotal
The Henry coefficient depends on constituents of the system and on the temperature.
For a system consisting of H2S, air and water, the Henry coefficient has a value of
481.38 bar at 20C (536 atm/mol at 25C [HVITVED-JACOBSEN, 2002]). It rises with higher
temperatures. Reaching of equilibrium conditions is accelerated by turbulences. The
more intensive the contact between fluid and gas is the merrier is the mass transfer. It
increases with turbulences, sudden falls and changes of direction. They are dominant
parameters for outstripping of H2S [FRECHEN, 2007]. When H2S got into the sewer atmosphere, it quickly builds up a comparable equilibrium with the water phase that is
condensed at the inner wall of the conduit. This place is simultanously a sink, as described in sec. 2.1.3.
If catchment areas are spacious, meteorological differences of pressure increase aeration. On the other hand, locations of odour exposure and risk of corrosion change with
the meteorological situation [FRECHEN, 2007].
Report ODOCO-1
_________________________________________________________________________________________________
as electron receptor is between -0.22 and +0.8V [FUCHS, 2007]. Sulphide and its chemical oxidation products are oxidised by bacteria like Thiobacilli to sulphuric acid, which reacts with concrete to form gypsum [SAND & BOCK, 1984 cited in ODOM, 1993]. As the water at the surface of the conduit is condensed water, it does not provide sufficient alkalinity to prevent low pH-values. Therefore, pH drops without barriers. The optimal pH for
sulphuricating bacteria is 1-3 [FUCHS, 2007].
Sulphuricating bacteria are mostly fastidious lithotrophic, using H2S or other not-oxidised
sulphur-compounds as electron donators. Their anabolism is autotrophic, sometimes
facultative [FUCHS, 2007], which means that they build up their cells with carbon out of
CO2.
H2S + O2
2 H2O + 2 S
S2- + 2 O2
2 SO42-
S + H2O + 1.5 O2
H2SO4
Table 1:
Factors
Sulphate contents
Temperature
Organic pollution
Redox potential
Biofilm
pH-value
Flow velocity
Flow time
Operation method
Effects
development of sulphide increases with increasing sulphate contents (Monod);
inhibition < 30 mg/l [LOHSE, 2002]
development of sulphide increases with rising temperatures
different results: partly no influence, partly increasing
production of sulphide only if wastewater is in anaerobic
conditions (< 0,1 mg O2/l)
consumes the most oxygen
tolerant in width range of pH; between 5,5 to 9,0
generally no considerable influence; a thicker biofilm is
produced in case of lower flow velocities
in case of long detention periods, wastewater changes
into anaerobic state; critical detention period difficult
(appr. > 2 to 4 hours)
with pumps operating continuously and with flow times of
up to 6 hrs, values are 10-20 % higher than they would be
with discontinuous operation [LOHSE, 2002]
10
Report ODOCO-1
_________________________________________________________________________________________________
Table 2:
Problems
depends on construction and
operation
In view of the fact that frame conditions change in many places (only partly developed
settlement and industrial areas, decreased specific wastewater production), the transferring shafts of rising mains with long retention times are particularly problematic, because osmogenes can be emitted from the wastewater with an anaerobic milieu into the
surrounding air. In the long term, it must be expected that corrosion problems will arise
as well (Figure 3).
11
Report ODOCO-1
_________________________________________________________________________________________________
Osmogenesis
Air out
1) Good design:
low turbulence
top of the pressure pipe
constantly covered
Air in
Corrosion
Turbulence
Corrosion
12
Report ODOCO-1
_________________________________________________________________________________________________
ment (CEM I) is particularly acid-soluble. During the hydration of cement, calcium hydroxide crystallises primarily in those areas that were previously taken up by the mixing
water and on the surface area of the aggregates. This results in the formation of a threedimensionally linked Ca(OH)2-structure through the whole of the concrete.
Figure 4 shows the polarisation microscopic image of a cement matrix. The CSH-phases
and cement clinker are black, the Ca(OH)2 is light-coloured (yellow).
Figure 4: PolMi-image of the three-dimensionally linked calcium hydroxide structure. CSH phases, cement clinker: black; calcium hydroxide: lightcoloured
If the calcium hydroxide is solved by an acid, the dissolution quickly proceeds along
these paths into the centre of the binder matrix. Initially, the resulting damage to the concrete is not visible.
A further reason why concrete is affected so rapidly by an acid is the Grotthuss transport
mechanism for H+ ions (acid ions). Conventional ions, such as sulphate ions, diffuse
through a solution along the concentration gradient. Thus, the ions have to move from
the point of origin to the surface of the substrate during the diffusion process. In contrast,
13
Report ODOCO-1
_________________________________________________________________________________________________
H+ ions attach themselves to existing water molecules and form H3O+- ions. The H3O+
ions themselves do not diffuse along the concentration gradient, but give off their surplus H+ ion to an adjacent water molecule. The result is a transport of H+ ions, during
which the H+ ions jump from one water molecule to the next along the concentration
gradient (bucket chain principle). The Grotthuss transport mechanism is virtually independent of the counter diffusion of solved parts of the matrix and up to 5 times faster
than the conventional diffusion of ions.
Figure 5:
During biogenic corrosion by sulphuric acid, the second step is the diffusion of the sulphuric acids sulphate ions into the concrete along the solved Ca(OH)2 paths, where they
further destroy the concrete in an expanding attack.
The concrete matrix, which is damaged by (biogenic) sulphuric acid (H2SO4), therefore
has two damage fronts. Within the first damage front, seen from the damaged surface,
the concrete matrix is completely destroyed by solving (H+) and expanding (SO42-) attack.
This completely destroyed layer can be removed fairly easily by mechanical means, if it
doesnt fall off of its own accord.
A second damage front, which runs deeper, is invisible to the human eye and can only
be detected microscopically. Within this second layer the acid has damaged the concrete
matrix merely by solving attack. The mechanical durability of this layer has largely been
preserved, as there has been no (or hardly any) expanding attack yet. As the acid hits
unhydrated cement clinker, this begins to hydrate, too. This process leads to a change in
volume which results in the formation of microcracks. Furthermore, the hydration releases alkalines which counteract the lowering of the pH-value by the acid. Thus, this
damage front can neither be detected by a sulphate analysis of drilling powder nor
through the use of phenolphthalein.
Figure 6 shows mortar damaged by sulphuric acid viewed under a scanning electron microscope (backscattered electron image). The two damage fronts are visible.
14
Report ODOCO-1
_________________________________________________________________________________________________
If the concrete is not removed to a sufficient depth in the course of reconstruction measures and if the attack by the acid deep within the concrete is not removed completely, the
corrosion protection system applied to the concrete or the re-profiling will possibly fall off
after a couple of months. The reason for this is that the high pressure water jet, which is
usually used to remove damaged areas of concrete, releases the sulphate from the plaster within the first, completely destroyed damage front, which then reaches the deeper,
damaged areas of the concrete full of cracks. After the re-profiling or the corrosion protection coating is applied, additional plaster is formed, which can push away the reprofiling or the corrosion protection coating in an expanding attack.
Furthermore, the attack deep within the concrete remains invisible for a long time, because the layer of plaster which is formed makes the concrete appear undamaged. Only
15
Report ODOCO-1
_________________________________________________________________________________________________
after approx. 2 to 4 years of continuous damage will the completely destroyed outer layer
fall off, and the first damages become visible. The process begins anew, so that the further progression of corrosion again does not become visible until 2 to 4 years later. During the respective intermediate phases no progression of corrosion is visible.
At the MPA BB a testing method has been developed with which the durability of concrete, or rather the depth of the damage, can be determined exactly.
To assess the durability of concrete comparatively for the development of new sewage
systems, test objects are exposed to sulphuric acid with simultaneous simulation of an
abrasive attack in a custom-built system over the course of 3 months. Durability is analysed via the microscopic measurement of the attack depths and the further attack by
acid deep within the concrete.
The damage of concrete in existing sewage systems is likewise assessed via the microscopic measurement of the attack depths and the further attack by acid on drilling cores.
In both cases the attack depths are determined on microsections using a stereomicroscope. The attack by acid deep within the concrete is determined using a polarisation
microscope or scanning electron microscope (backscattered electron image) on thin sections.
Limiting values of acid and sulphate exposure for concrete
DIN EN 206-1, table 2 states, among other things, the limiting values for SO42
concentration and the pH-value for an external groundwater attack on concrete (see
Table 3). These limiting values apply in case the attacking medium is supplied continuously by the groundwater. According to DIN EN 206-1, Table 3, much higher limiting values apply for affected soil, as the attacking medium is not supplied continuously and for
the most part is used up quickly through being bound (sulphate) or buffered by the concrete (sulphate).
Table 3:
Corrosion by
groundwater
Chemical property
XA1
XA2
XA3
SO42-
200 to 600
6,5 to 5,5
in mg/l
pH-value
The most restrictive value for each chemical property determines the exposition class.
16
Report ODOCO-1
_________________________________________________________________________________________________
In exposition class XA3 (sulphate concentration > 3000 mg/l or pH-value < 4,5) or above
concrete must be protected by anti-corrosion systems.
According to the ATV-M 168 guideline, which is in effect only in Germany, the limiting
values and requirements for concrete shown in Figure 4 apply to internal corrosion by
communal wastewater.
Table 4:
600
< 3000
pH-value
6,5
short-term
posure
The most restrictive value for each chemical property determines the requirements.
According to ATV-M 168, concrete (with HS cement) needs additional protection by corrosion protection systems if the sulphate concentration is > 5000 mg/l or the pH-value is
< 5,5 (long-term exposure).
17
Report ODOCO-1
_________________________________________________________________________________________________
Table 5: Chemicals frequently used in industry and trade according to DIN 28052-1
Organic solvents are typical organic compounds with a wide range of uses. These substances are employed in technical products or processes, but can also be found in many
ordinary household products.
18
Report ODOCO-1
_________________________________________________________________________________________________
Small amounts of organic solvents reach the sewage systems as part of industrial or private wastewater. Large amounts of solvents in sewage systems can be caused by accidents or deliberate illegal disposal.
Depending on the location and the history of the soil, these substances can also affect
the construction material of sewage systems externally. In possibly polluted areas (i. e.
on former sites of the metal-working or chemical industry, or of petrol stations) heavy pollution with the compounds mentioned above, especially aromatic or chlorinated hydrocarbons, can occur. As these are practically water-insoluble, even large amounts of hydrocarbon will lead to only small amounts in the water phase, while the remainder forms
a fluid plume which remains above or below the groundwater, depending on its density.
Within these plumes of harmful substances, the hydrophobic hydrocarbons are thus virtually undiluted. Therefore, the effect of organic solvents on the construction material of
sewage systems has to be examined both from the inside of the pipe as well as from the
outside.
Whereas the attack on the outside of the pipe can be expected primarily from the liquid
phase, an attack from the gas phase must be considered inside the pipe, because all
substances in this group are more or less volatile.
Among the compounds with especially high relevance with regard to the amount of the
substance or the corrosive effect on plastics are most notably:
halogenated hydrocarbons
aromatic hydrocarbons
The MPA Berlin-Brandenburg is involved in the planning work for the materials to be
used in the Emscher canal in North Rhine-Westphalia, which is presently the largest
wastewater construction project in Europe, with a total investment volume of 4,5 Billion
Euros. In the course of the evaluation phase, various studies were conducted to assess
the risk of corrosion in sewage systems.
Wastewater and groundwater analyses indicate that the corrosion of plastics (pipes and
seals, for example) by aromatic and halogenated hydrocarbons must be taken into
account in sewage systems located in industrial and densely populated areas.
During the loading of aromatic or halogenated hydrocarbons to plastics or plastic
systems it could be observed that many of the standard plastics for sewage systems
showed considerable corrosion to the point of complete disintegration after only a few
hours. An example is shown in the following image.
19
Report ODOCO-1
_________________________________________________________________________________________________
An unexpected result was that the rapid disintegration after several hours or days of
storage in aromatic or halogenated hydrocarbons occurred not only in the liquid phase,
but also in the gas phase. Several plastic systems also showed signs of corrosion after
storage in sulphuric acid. These facts indicate that plastic corrosion is possibly relevant
to the durability of sewage systems.
Figure 8: Various polymer concretes after 20 hours in the gas phase of halogenated hydrocarbons
20
Report ODOCO-1
_________________________________________________________________________________________________
21
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 3
Measurement
3.1 Introduction
Online monitoring, within the scope of a pilot plant, should allow for a most comprehensive acquisition of relevant test measurement parameters, associated with formation and
occurrence of the activity of odour and/or of corrosion substances or groups of correlated
substances.
As discussed in sec. 2, depending on the material of technical wastewater facilities,
mainly following substances or groups of substances are active corrosives:
Concrete material / reinforced concrete primarily damaging: sulphuric acid, (nitric acid)
The following Table 6 presents the odour-active and corrosion-active substances for
plastics and concrete as well as the essential ones participating on their formation and
shows where they appear (wastewater, gas, biofilm, pipe material).
In an additional row (desirable to measure) an estimation is made where (wastewater,
gas, biofilm, pipe material) it is meaningful to carry out detection tests for odour- and corrosion-active substances or participating substances in their formation, by sensors (for
modelling).
Table 6:
Corrosive compounds, their causing parameters, precursors and reaction product and importance to measure
Corrosive Compound
Sewage
Gas
Biofilm
Plastics
Organic solvents
Desirable to measure
+
+
++
++
0
-
O2,
org. sulphuric
compounds,
SO42-, HS-, S2+ (O2, SO42-S2-)
H2S
(odour)
acid (H3O+),
sulphate
(SO42-),
bioactivity
++ (acid)
+ (sulphate)
Concrete
Sulphuric acid: (H2SO4)
Desirable to measure
++
22
Material
gypsum,
pH-value
+ (pH-value)
Report ODOCO-1
_________________________________________________________________________________________________
23
Report ODOCO-1
_________________________________________________________________________________________________
to be captured in order to ascertain the current condition of the wastewater systems regarding corrosion and odour.
24
Report ODOCO-1
_________________________________________________________________________________________________
Table 7:
Sewage
x
Air
x
Temperature
organic load:
S-species:
sulphate
pH
N-species:
hydrogen sulphide
x
x
HNO3
x
x
x
x
biological activity
velocity, flow
organic solvents
concentration of applied chemicals
nitrate, nitrite
Biofilm Pipematerial
x
x
x
x
moisture
3.2.1.1 pH-value
ISFET pH-meter
Functionality
ISFET stands for ion sensitive field-effect transistor. Analogue to the general field-effect
transistor (FET) the measurement principle is based on the change of the field effect
(training of a space-charge zone) between SOURCE and DRAIN. The function mode of
FET is not described in the following. It is expected to be familiar. In place of the electri-
25
Report ODOCO-1
_________________________________________________________________________________________________
cal contact at the GATE a pH-sensitive layer with the ISFET (ex. Ta2O5) applied, which is
brought directly in contact with the measured liquid. Over a bias voltage, which is applied
over a reference electrode, which likewise is in the solution, the operating point of the
ISFET can be specified analogue to the establishment of the operating point of a FET.
Dependent on the concentration of the ions in the solution which can be examined, an
additional surface barrier at the contact between liquid and ion sensitive layer (Nernst
equation) is formed. This potential adds itself on the constantly applied bias voltage and
affects thus the space charge zone between SOURCE and DRAIN. This leads to a
change of the SOURCE drain stream, which can be measured. The changes are directly
proportional therefore to the change of the analyte concentration. Over a calibration, can
be reckoned back, with help of the measured river on the analyte concentration. The ISFET is electrically regarded thereby as a trans-impedance transducer. An electrical potential is measured potentiometrically, without considerable current, and it is converted
into a measurable SOURCE-DRAIN-current. This is in first approximation independent of
following circuits and thus as high impedance power source [GRNDLER, 2003].
Practicability
There are sensors available on the market for measurements in the wastewater system
which are compiled in the following table:
Measurement range
Accuracy
Speed of response
Materials process tangent
Allowed process temperature
Allowed process force
Minimal allowed process
conductibility
Knick,
pH-Sensor SE 545
0-12 pH
PEEK, Elastomeric
EPDM
0 to 80C @ 6 bar
0 to 6 bar @ 80C
Specific literature:
Emerson Process Management / Rosemount Analytical: PDS49-TFS396 Rev. C
http://euedocs.emersonprocess.co.uk/groups/public/documents/markcom/pds_ph_tfs396
_rev_c.pdf
Knick, Produktkatalog 2007, http://www.knick.de/
26
Report ODOCO-1
_________________________________________________________________________________________________
Galvanometric pH-meter
The pH meter is a standard device for measuring pH-values of samples. It is mainly used
in laboratories for discrete measurements. There are pH-meters that can be applied continually.
Functionality
A glass diaphragm ball filled with a potassium chloride solution is immersed into the liquid that is measured. By the inclination of the hydrogen ions to deposit itself in thin layer
at the glass surface a galvanic voltage develops itself inside the ball. A galvanic cell is
developed, whose electric motor force is relative to a reference electrode independent of
the hydrogen ions one measures [SCHWABE, 1976; BATES, 1954; HLL, 2002; DEV, 2003].
Practicability
For continuous application, special electrodes are available, which are characterised by
increased robustness. The following table arranges the characteristic values of a selection of such sensors:
27
Report ODOCO-1
_________________________________________________________________________________________________
Driesen&Kern,
pH-Log530
Measurement range
Accuracy
Speed of response
Material tangent
1...14 pH
0,02 pH
Hach Lange,
pH-/RedoxKombiElektrode
8350
014 pH
0,05 pH
Casing V4A
PTFEcompound,
safety glas,
PPS case
Allowed process
temperature
Allowed process
force
0,2C...+80C
max 110C
Comments
Data logger
with integrated
power source
designed for
field use in river
monitoring
Pt-temperature
measurement
with automatic
compensation
10 bar @
80C
Temperature
compensation
Brkert,
pH-Sensor
8201
010 pH
0.1 pH
WTW,
ph-Einstabmesskette Sensolyt SEA
212 pH
Enamelled
steel pipe,
Ground diaphragm ceramic enamelled, Process
connection
1.4404,
Electrode
head PVDF,
Seal EPDM
0 to +140C
-16 bar rel.
< 10 bar @
20C,
< 1bar @
60C
Mounting in
special
Sensolytmounting
Specific literature:
{ Driesen & Kern, pH-Log530, http://www.driesen-kern.de/downloads/phlog530.pdf }
{ Hach Lange, ph-8350, Angebotskatalog 2007, Angebot 194460, S. 53-56 }
{ Brkert, pH-Sensor 8201, http://www.buerkert.com/products_data/datasheets/DS8201standard-EU-EN.pdf [2007-09-07] }
{ WTW, pH SEA, Katalolg Online Messtechnik 2006, S.18}
Report ODOCO-1
_________________________________________________________________________________________________
Functionality
The thick film pH sensor consists of a sturdy mother board of steel which is isolated. On
the isolation layer in the thick-film procedure on one side the test electrode, on the other
side the reference electrode is attached. The sensor works like a pH meter, gets along
however without liquid system components. It is pressure insensitive and compatible to
commercial pH measuring instruments.
Practicability
The ranges of application which are named for the sensor are the chemical industry,
food technology and medical technology. Its potential strength is shown in the economical fabrication as well as in its break resistance and the adaption possibility of the sensor
layout because of the requirements of application. An application in wastewater is not
described and must be exactly examined. A commercial manufacturer is not known.
Measurement range
Accuracy
speed of response
Allowed process temperature
Comments
Specific literature:
{Fraunhofer-Institut fr Keramische Technologien und Systeme, Informationsblatt Thick
Film pH-Sensor, Messestand Sensor+Test, Nrnberg 2007}
3.2.1.2 O2-Concentration
Oxygen is present both in the gaseous phase in gas form and in the wastewater in dissolved form. The respective concentrations are determined by sensory measurement.
Measurement principles are described below. For the description of the practicability with
all presented measurement principles the same characteristics were selected. If to one
of the characteristics no indication is given, no information is present for this.
Clark-Oxygen Electrode
The main application range of Clark oxygen electrodes lies in the regulation of oxygen
concentration in liquids. Lesser represented, although also feasible, is the measurement
if oxygen concentration is in gas mixtures.
29
Report ODOCO-1
_________________________________________________________________________________________________
Functionality
The Clark-electrode consists usually of a platinum cathode and a silver anode as reference electrode, which stands over an electrolytic solution connection. They are separated by an oxygen-permeable diaphragm from the sample. On the platinum cathode a
bias of -0,8 V rests against the silver anode.
If one brings the solution which can be examined to the measuring chamber now to the
diaphragm, then the CO2 partial pressure difference between diaphragm outside and inside to a CO2 diffusion leads by the diaphragm foil into the measuring chamber. The oxygen is reduced at the cathode, takes up thus electrons of the cathode, whereby hydroxyl
ions (OH) are developed. At the anode to Silver chloride one oxidizes to silver, therefore an Ag/AgCl electrode is also usually used.
Cathode:
Anode:
The result is thus a current, which is directly proportional to the partial pressure p of the
oxygen CO2. This current is measured and becomes back-closed on the concentration of
the oxygen.
Since the oxygen solubility and the permeability of the diaphragm are temperaturedependent, compensation is necessary [HAMANN & VIELSTICH, 1985; GRNDLER, 2003;
RUMP, 2003; HLL, 2002; DEV, 2003].
Practicability
The application in wastewater, sensors were developed which are especially protected
against contamination. The following table arranges some sensors. No data mean that
no information is present.
30
Report ODOCO-1
_________________________________________________________________________________________________
Measurement range
Accuracy
Detection limit
Response time at 25
C (Air > N2)
Signal in Air
Residual signal in
O2-free medium
Flow dependency
Pressure range
Temperature range
Material process
tangent
Knick,
SauerstoffSensor SE 703
pO2 < 1200 mbar
WTW,
TriOxmatic
700
060 mg/l
1 % + 30 ppb
30 ppb
98 % of final
value < 90 s
40 110 nA
0,3 % of signal
in air
5%
0,5 2 bar absolute
0 60 C
Silicone and FKM
(Viton), Membrane:
PTFE/Silicone/
PTFE (Steel-net
reinforced), PPS
0,1 mg/l
Hamilton,
Oxysenes
Ahlborn,
FYA640O2
40ppb saturation
0,040mg/l
< 1%
t98 < 60 s
< 10 bar
5%
04 bar
0...50C
060C
-5..50C
Sensitivity shift
< 10% every 2
months with
25C water
and stable
conditions
Durability
Comments
Operating time
with single
Electrolyte filling: several
months,
Total: several
years
ReferenceElectrode increasing signal stability
and selfmonitoring of
electrolyte replacement
Specific literature:
{Co Knick Berlin, Sauerstoff-Sensor SE 703, http://www.knick.de/hps/ client/knick/public/
index.hbs }
{Co WTW, WTW Welt der online Messtechnik, http://www.wtw.com}
{Co Hamilton, http://www.hamiltoncompany.com/eu-sensors/oxysens.asp}
{Co Ahlborn, Gesamtkatalog Ausgabe 2007/2008, S. 15.08}
31
Report ODOCO-1
_________________________________________________________________________________________________
Fibre-optic Sensor
In the last decade a large number of measuring procedures have emerged from the
range of fibre optics. Thereby, the absorption and fluorescence characteristics of materials are used in dependence of the light wave length as well as the change of refraction
characteristics of the light conductor with accumulation of certain materials within ranges
of the conductor.
For the measurement of oxygen, this principle of measurement can be applied both in
the gas phase and for the measurement of liquids.
Functionality
The fibre-optic oxygen sensors are coated probes that use fluorescence quenching to
measure the partial pressure of dissolved or gaseous oxygen. Different sensing formulas
are available (ruthenium or Pt-porphyrin formulation). The system works like this:
An LED sends excitation light to one leg of a bifurcated optical fibre assembly. The fibre
carries the light to the oxygen probe, which is polished to a 45 angle. The distal end of
the probe tip consists of sensor formulation trapped in a sol-gel matrix, immobilized and
protected from the sample. The light from the LED excites the ruthenium or pophyrin
sensor formulation at the probe tip. The excited complexes fluoresce, emitting energy at
~600 nm and 650nm, respectively.
If the excited complex at the probe tip encounters an oxygen molecule, the excess energy is transferred to the oxygen molecule in a non-radiative transfer, decreasing or
quenching the fluorescence signal. The degree of quenching correlates toe the partial
pressure of oxygen in the sol-gel, which is in dynamic equilibrium with oxygen in the
sample.
The fluorescence is collected by the probe and carried to the Spectrometer/MFP
Fluorometer via the second leg of the bifurcated optical fibre assembly.
The following assemblies are available:
Oxygen sensor with spectrofluorometer:
Oxygen is sensed by measuring the decrease in fluorescence intensity of a fluorophore
bound to the tip of an optical fibre. The sensor responds to the partial pressure of oxygen.
Oxygen Sensor with multi-frequency Phase Fluorometer(MFPF):
Oxygen is sensed by measuring the phase shift of fluorescence of a fluorophore bound
to the tip of an optical fibre. The sensor responds to the partial pressure of oxygen.
{cp http://oceanoptics.com/Products/catalogsensors.pdf [2007-09-07]}
32
Report ODOCO-1
_________________________________________________________________________________________________
Practicability
In the following table characteristic values are arranged for a fibre-optic measuring system. The characteristic values refer to the assigned indicator matrix. The advantages of
the procedure are independent from changes in the pH-value, amount of salt and ionic
strength. There is no influence by humidity, CO2 or methane.
Measurement
range
Accuracy
Detection limit
Response
time at 25 C
(Air > N2)
Allowed process temperature
Frequency of
necessary recalibration
Durability
Comments
Ocean Optics,
FOXY Sensor Formulation
0-40 ppm (dissolved),
0-100% (mole percent)
0.02 ppm, 0.05%
Ocean Optics,
HIOXY Sensor Formulation
0-40 ppm (dissolved),
0-100% (mole percent)
0.02 ppm, 0.05%
-5080 C
-5080 C
Probe: Recondition
once per year
Benign environment,
aqueous liquids and
vapours
Probe: Recondition
once per year
designed for monitoring in non-aqueous
fluids: Hydrocarbonbased liquids and vapours, fuels
Hach-Lange,
Lange LDO- Sensor
LXV416.00.00001
0.05 to 20 mg/l
0.05 to 20 ppm
0.01 mg/l, 0.01 ppm,
0.01 % saturation 0.1
C
Not necessary
Air or inline calibration
possible
One year warranty
No interference from
the following: H2S,
pH, K+, Na+, Mg2+,
Ca2+, NH4+, Al3+, Pb2+,
Cd2+, Zn2+, Crges.,
Fe2+,Fe3+, Mn2+, Cu2+,
Ni2+, Co2+, CN-, NO-,
SO42-, S2-, PO43-, Cl-,
Cl2, anion active tensides, cruide oils
Specific literature:
{Co http://www.hach-lange.de :: DOC053.52.03203.Apr03}
{Ocean Optics Ltd., http://oceanoptics.com/Products/catalogsensors.pdf [2007-09-07]}
33
Report ODOCO-1
_________________________________________________________________________________________________
Practicability
Sensors, that are similar to each other, are available from various manufacturers. Two
sensors are described exemplary:
Electrode type
Measurement range
Accuracy
Response time
Material process tangent
Allowed process temperature
Allowed process pressure
WTW,
SensoLyt PtA
Gel-Polymer solid electrolyte, double-fold perforated
diaphragm
+/- 2000 mV
Hach Lange,
1200-S sc Redox
Ag/AgCl Polymer, perforated diaphragm
+/- 1500 mV
t90 < 15s
Stainless steel, PPS,
glass/Platin
-560C
max 2 bar overpressure
Specific literature:
{ WTW , pH PtA, Katalolg Online Messtechnik 2006, S.18 }
{ Hach Lange, 1200-S sc, Angebots-Katalog, S.46/76 }
{cp http://www.wtw.com/media/LaborundUmweltKatalog2006_I.pdf [2007-09-06] S.22}
34
Report ODOCO-1
_________________________________________________________________________________________________
3.2.1.4 COD
The sum value of chemical oxygen demand (COD) is a measurement for all in the water
existing and under certain conditions oxidizable materials. It indicates the quantity of
oxygen (in mg/L) which was needed for its oxidation, if oxygen were the oxidizing agent.
The determination of the COD is usually to be accomplished wet-chemically in the laboratory on the basis of a water test. It is standardized. This process is very complex. For
the determination without the complicated chemical process cycle cuvette high-speed
tests were developed. These measure the dichromate consumption of the sample of water in the test liquid of a prefabricated cuvette according to a photometric principle.
For the quasi-continuous measurement instruments with a sampling mechanism and
automatic analysis are available. A continuous detection is reached using spectrometer
probes. With this procedure it will furthermore be dealt with [RUMP, 2003; HLL, 2002;
DEV, 2003].
Spectrometer Probe
A wide-band spectral measurement is made and reckoned back over correlation functions on the COD content.
Functionality
The probes measure the spectrum from ultraviolet to the long-wave visible light. From
the high information content the measured values are determined. The foundation is a
long term built up database and the whereupon developed algorithms. The wide-band
measurement permits the compensation of the disturbing influence of the turbidity.
Practicability
Sensors are used in sewage engineering and are characterised by the following advantages in relation to other procedures:
direct measurement in the process medium
on-line-processing
precise measurement due to the spectral analysis of the scanned UV/VIS range in
place of a punctual measurement
high durability by automatic compressed air cleaning before each measurement
One disadvantage of the procedure is the indirect measurement and the necessity to fall
back to an existing database and algorithms. The selection of used data and methods for
correlation has substantial influence on the result and can, with strongly varying wastewater compositions, lead to bad results. Likewise unfavourable is the necessary compressed air connection (3-7 bar) which effects the installation. Such a sensor is offered to
35
Report ODOCO-1
_________________________________________________________________________________________________
WTW under the designation CarboVis, the most important characteristics are arranged in
the following table. A study about process stability for the application of UV-VIS-Systems
for sewage and surface water is ongoing at Kompetenzzentrum Wasser Berlin and Institute of Civil Engineering Department Urban Water Management.
Principle of Measurement
Measurement range
CSB
Measurement range
TOC
Accuracy
Measurement range
solids
Materials
Allowed process
pressure
Allowed process
temperature
Flow velocity
pH-range
Salinity of mediums
WTW,
CarboVis
Spectral measurement in UV/VIS
range: 200 - 750nm
0.1800.0 mg/l
s::can,
carbo::lyser
UV-Vis spectrum between
200 and 750 nm.
5.0...1500 mg/l 5.0
1500 mg/l
+/-3 %
03000 mg/l (optional)
Al Mg Si 1, anodized
<= 1 bar
stainless steel
045C
0 - 45 C
<= 3m/s
49 pH
< 5000 mg/l (Chloride)
Specific literature:
{WTW, CarboVis, http://www.wtw.com/media/OnlineMessKatalog2006_I.pdf [2007-0907] S.50ff}
{s::can, Product description http://www.s-can.at/index.php?id=16 }
3.2.1.5 BOD
The laboratory determines the BOD after sampling. There are different methods, which
are all based on the biological processes at constant temperature of the sample within
the entire time. Subsequently, consummation of oxygen can be calculated in different
ways (sometimes collateral). For on-line measurement, these procedures are inadequate, indirect measurements are used [RUMP, 2003; HLL, 2002, DEV, 2003].
Spectrometer Probe
A wide-band spectral measurement is made and reckoned back over correlation functions on the BOD value.
36
Report ODOCO-1
_________________________________________________________________________________________________
Functionality
This is the same principle that also describes the determination of the CSB. The difference in the procedure is only a different underlying data basis and changed correlation
methods.
Practicability
The same statements are valid as under sec. 3.2.1.4, since the same sensor for the determination of the BOD5-value is used. A known sensor is CarboVisTM of WTW.
Specific literature:
{WTW, http://www.wtw.com/media/OnlineMessKatalog2006_I.pdf [2007-09-07] S.50ff}
3.2.1.6 Sulphate
Sulphates are salts or esters of the sulphuric acid. The salts contain as anion the sulphate ion [SO42-]. They are chemically proven, by adding hydrochloric acid and barium
chloride or barium hydroxide solution. A heavy, soluble precipitation of white barium sulphate is developed [JANDER & BLASIUS, 1980; DEV, 2003; RUMP, 2003].
There is commercial sensor for the detection of sulphates in aqueous solutions available.
Research exists, which concentrates on the development of bio sensors. These are,
however, still far away from application in the wastewater range. The work is outlined
briefly.
37
Report ODOCO-1
_________________________________________________________________________________________________
3.2.1.7 Sulphide
Sulphides are salts or esters of hydrogen sulphide (H2S). The sulphide anion (S2-) is
shortly named sulphide. The concentration of the anions is to be determined in the
aqueous phase.
38
Report ODOCO-1
_________________________________________________________________________________________________
tics are the exemplary set up of two exemplary products in the following table. No statements are known about the application in wastewater.
Radiometer analytical,
ISE25S
5 x 10-7 100
0.02 30000
0 60C
12 mm
11 14 (direct meas.)
Solid-state
Equally sensitive to Ag+,
Precipitated cations forming
complexes with sulphides
Recommended reference
electrodes
WTW,
ISE Typ Ag/S 800
10-7...1 mol/l
0,003...32000 mg/l
2-12
Solid-state
Specific literature:
{Radiometer analytical, ISE25S, MeterLab Guide to Reliable pH, Ion and Conductivity
Measurements (Electrodes) Data Sheet (lit 4416) ,
http://www.hach.com/fmmimghach?/CODE%3AL4416_01-0712747%7C1}
{WTW, ISE Ag/S, http://www.wtw.com/media/LaborundUmweltKatalog2006_I.pdf [200709-06]}
{http://www.wtw.com/media/LaborundUmweltKatalog2006_I.pdf [2007-09-06]}
39
Report ODOCO-1
_________________________________________________________________________________________________
Functionality
The principle works potentiometrically with a measuring and a reference electrode. The
function mode of the ion-selective measurement is identical to section 7.1 which has
been described.
Practicability
Electrodes are available on the market for the measurement of nitrate. The characteristics of some products are summarized in the following table.
Since the ion-selective electrodes are affected by interfering, it makes sense to compensate disturbances. The firm WTW has developed the product "VARiON sensor that the
installation of various electrodes and thus allows compensation for the nitrate measurement, the chloride ions. This sensor was constructed especially for measurement in
wastewater.
Radiometer analytical,
ISE25NO3
3 x 10-6 100
0.2 60000
Temperature range
Diameter
pH range
Sensor type
Interfering ions
Durability
Comments
0 50C
12 mm
3 10
PVC membrane
Cl, Br, NO2
4-8 Months
integrated temperature sensor
Specific literature:
{Radiometer analytical, ISE25S, MeterLab Guide to Reliable pH, Ion and Conductivity
Measurements (Electrodes) Data Sheet (lit 4416), http://www.hach.com/fmmimghach?
/CODE%3AL4416_01-0712747%7C1}
{WTW, http://www.wtw.com/media/OnlineMessKatalog2006_I.pdf [2007-09-07] p.38}
40
Report ODOCO-1
_________________________________________________________________________________________________
Practicability
Particularly for the application within the wastewater range, equipment from the company
WTW is available. Its characteristic values are arranged briefly.
WTW,
TresCon ON 510
0,4090,0 mol/l
0,0051,200 mg/l
0,1 mol/l [0,001 mg/l]
Photometric,
year maintenance, calibration 24h,
The Device is analyzing samples taken in 10/15/20 min intervals. Cleaning-, calibration- and reagent-solutions have to be
obtained.
Specific literature:
{WTW, http://www.wtw.com/media/OnlineMessKatalog2006_I.pdf [2007-09-07] p.45}
3.2.1.10
Conductivity
The electrical conductivity is a physical quantity, which indicates the ability of a material
to conduct electric current. In aqueous solutions the conductivity as a sum-parameter is
a measure for the ion concentration. The more salts, acids or also bases are dissociated,
the higher the conductivity. The conductivity of wastewater can be easily measured to
get information about possible dilution of wastewater with stormwater run-off.
Conductometric Analysis
Functionality
The measurement of conductivity is based on the principle of the electro-chemical resistance test. A measuring cell is used which consists of two similar electrodes. An alternating voltage is applied to the electrodes. From the current caused by the ions of the
measured media, the conductivity value is computed. Cell geometry and the temperature
influence the measured current and are considered throughout the computation
[GRNDLER, 2003; HAMANN & VIELSTICH, 1985; DEV, 2003; RUMP, 2003].
Practicability
Conductivity sensors for the employment in wastewater are available from different manufacturers on the market. The most important characteristics are listed on two products.
41
Report ODOCO-1
_________________________________________________________________________________________________
Sensor
Measurement range
Resolution
Accuracy
Allowed process pressure
Temperature
Measurement range
Resolution
Accuracy
Casing
Type of protection
Driesen&Kern,
S-Log540
0,05100ms
Ranges: 0,05 0,4 1,0
2,0 3,8 7,9 100 mS
better than 0,1% upper limit
of respective range
better than 0,1% upper limit
of respective range
Pt1000-Sensing resistor
0,2C...+50C
0,1 C
0,2 C
V4A
IP68
WTW,
TetraCon 700
10S/cm 1000mS/cm
Ranges: 0,0 0,2 2
20 200mS
0-10 bar
NTC
0 50C
+/- 0,2K
PVC, V4A
IP68
Specific literature:
{http://www.driesen-kern.de/produkte/wasserseiten/loggerfuerdenunterwassereinsatz/
01b6f9989a069a33b.html [2007-09-06]}
{WTW, http://www.wtw.com/media/OnlineMessKatalog2006_I.pdf [2007-09-07] S.26}
Practicability
By zirconium dioxide sensors oxygen partial pressure within the ppm-range can be
measured without problems. Preferential applications are flue gas measuring probes and
the Lambda probes used in the motor vehicles.
42
Report ODOCO-1
_________________________________________________________________________________________________
Special designed devices can be used in the gas phase of sewers for the continuous
measurement. The gas mixture must be sucked in by way of a hose and supplied to the
measuring instrument. The characteristics of a measuring device are listed below. Other
measuring devices are not known.
Measurement range
Accuracy
Detection limit
Response time at 25 C (Air > N2)
Allowed process temperature
Comments
Zirox,
Module of oxygen detection ZR5
1 Vol.-ppm100 Vol.-%
relative measuring error < 5 %
1 ppm, made-to-order item up to 10-15 ppm
t90 1 s
050 C, max. 80 % rel. humidity
Ready to operate: < 15 min
flow rate: 8 l/h 2 l/h, controlled by internal pump
Specific literature:
{Co ZIROX GmbH, http://zirox.de/produkte/oem_loesungen/sauerstoffmessmodul_zr5.
html}
43
Report ODOCO-1
_________________________________________________________________________________________________
2Pb + 4OH- 2PbO + 2H2O + 4eAs the two processes above take place, a current is generated which can be measured
externally by passing it through a known resistance and measuring the potential drop
across it. Since the current produced is proportional to the rate at which these reactions
occur, its measurement allows accurate determination of the oxygen concentration.
As the electro-chemical reaction results in the oxidation of the lead anode these sensors
have a limited life. Once all the available lead has been oxidised they no longer work.
Typically oxygen sensors have 12 year life time, however this can be lengthened by increasing the size of the anode or restricting the amount of oxygen that gets to the anode.
{cp http://citytech.com/technology/02-sensors.asp [2007-09-06]}
Practicability
Electro-chemical gas sensors are susceptible against condensation. This requires, owing
to circumstances, a complex preliminary drying of the two measured gas mixtures, which
apparently react negatively on application in the wastewater pipe.
Measurement range
Accuracy
Detection limit
Response time at 25 C
(Air > N2)
Signal in Air
Residual signal in O2free medium
Allowed process pressure
Allowed process temperature
Frequency of necessary re-calibration
Durability
Comments
CitiCel 2FO
0-25% Oxygen, max. 30%
t95 < 10 s
t90 < 15
300 400 A
40 A (in N2)
Atmospheric 10%,
Pressure coefficient: <0.02%
signal/mBar
-20C to +45C
Temperature Coefficient:
0.2% signal/C
Specific literature:
{ City Technology Ltd, http://www.de.citytech.com/PDF-Datasheets/2fo.pdf }
Fibre-optic Sensor
In the last decade a large number of measuring procedures have emerged from the
range of fibre optics. Thereby, the absorption and fluorescence characteristics of materi-
44
Report ODOCO-1
_________________________________________________________________________________________________
als are used in dependence of the light wave length as well as the change of refraction
characteristics of the light conductor with accumulation of certain materials within ranges
of the conductor.
For the measurement of oxygen, this principle of measurement can be applied both in
the gas phase and for the measurement of liquids. For explanation of functionality see
sec. 3.2.1.2. The practicability for measurements in the atmosphere of sewage pipes is
not described.
(0.8)
SR
(0.2)
Specific literature:
{Surface Chrominated Thick Films of Ba(0.8)Sr(0.2)TiO3 for H2S Gas Sensing, Authors: G.H. Jain , L.A. Patil , P.P. Patil and U.P. Mulik, in sensors ISSN 1424-8220
2007 by MDPI www.mdpi.org/sensors}
45
Report ODOCO-1
_________________________________________________________________________________________________
Infrared sensor
The absorption behaviour is used by radiation of a material for the determination of the
concentration.
Functionality
The basic principles of operation of infrared gas sensors rely on the exploitation of the
following facts:
A wide range of materials absorb infrared radiation (due to intramolecular vibrations). For
any one material, the strength of absorption (absorbance) varies with wavelength (its absorption spectrum).Different materials have different absorption spectra.
This causes that certain basic components are common to all infrared gas sensors: an
infrared source (e.g. incandescent lamp), a detector (e.g. thermopiles, pyroelectric detectors), a means to select appropriate wavelengths (e.g. band pass interference filter) and
a sample cell. Radiation from the source passes through the sample cell and wavelength
selector. The choice of wavelength has a large bearing on the relative selectivity of the
sensor. The radiation NOT absorbed by the sample is then detected and the ratio of this
to the incident provides a measure of the concentration of target gas in the sample. A
second detector (or channel) tuned to a different wavelength that is not attenuated by
any species likely to be present in the sample is normally used to provide this reference
measurement.
A further component that enhances the performance of IR gas sensors is a temperature
sensor. All these components have temperature dependencies which must be compensated to provide an accurate measure of gas concentration. This temperature sensor
(typically a thermistor) should be sited within, or in very close proximity to, the detector(s).
Infrared sensors effectively give a measure of the number of target gas molecules in the
light path between source and detector. Consequently, the output signal not only varies
with concentration but also barometric pressure i.e. they are partial pressure devices.
For very high measurement accuracy, compensation for barometric pressure is, therefore, required. This dependency also infers that sensors with longer optical path lengths
(i.e. distance travelled by radiation between source and detector(s)) will have increased
sensitivity and tend to have a lower dynamic range but increased resolution.
In a single target gas, fixed optical light path device under constant barometric pressure,
the signal output (and signal/noise ratio) approximately exponentially decays with increasing concentration i.e. infrared gas sensors are inherently non-linear. The measurement accuracy decreases with increasing concentration.
46
Report ODOCO-1
_________________________________________________________________________________________________
The components described above form a typical infrared gas sensor. However, some
supporting electronics is required in any practical system. The more common detector
technologies provide very small analogue signal outputs that require amplification. Basic
analogue filtering of the amplified output signal can then enhance measurement accuracy.
The source also requires a driver circuit. It is usual practice to modulate the source output by pulsing (although some older design used fixed illumination and mechanical
choppers). This creates periodic variations in the emitted intensity and so allows the use
of synchronous detection techniques.
To carry out the temperature and barometric compensations, it is common practice to
use computational algorithms inside a microprocessor. This first requires the analogue
signals to be converted into digital signals. The compensated data is then transmitted to
the user in some form.
{cp http://citytech.com/technology/irsensors.asp [2007-09-06]}
Practicability
There is no information available, whether this principle can be applied to measure in the
gas phase of sewer networks.
Pellistor
Functionality
Pellistors employ catalytic combustion to measure combustible gases or vapours in air
up to the Lower Explosive Limit (LEL) of the gas.
The standard sensor consists of a matched pair of elements, typically referred to as a
detector and compensator (reference element). The detector comprises a platinum wire
coil embedded within a bead of catalytic material. The compensator is similar except that
the bead does not contain catalytic material and as a consequence is inert.
Both elements are normally operated in a Wheatstone bridge circuit that will produce an
output only if the resistance of the detector differs from that of the compensator.
The bridge is supplied with a constant dc voltage that heats the elements to 500-550C.
Combustible gases are oxidised only on the detector element, where the heat generated
increases its resistance, producing a signal proportional to the concentration of combustible gas. The compensator helps to compensate for changes in ambient temperature,
pressure, and humidity, which affect both elements equally.
Most pellistors have the pairs of elements housed in separate metal cans. In a complete
gas detector (to be used in a potentially explosive atmosphere) the cans will normally be
47
Report ODOCO-1
_________________________________________________________________________________________________
mounted inside a flameproof enclosure consisting of a metal sinter and housing. This
enclosure allows gas to reach the sensor whilst ensuring that the hot sensor elements
cannot ignite an explosive gas mixture. As an alternative complete detectors are available with both elements mounted inside a flameproof enclosure approved to the latest
European (ATEX) and North American (CSA & UL) standards.
Detection of explosive atmospheres relies on the accurate measurement of combustible
gases below the LEL concentration. Safety applications, therefore, are not generally
concerned with measuring the volume concentration of gas. Measurements are more
usually expressed as a percentage of the LEL concentration of the gas (%LEL).
Most combustible gas detection techniques are designed to detect a wide range of
gases. Ideally the output of a sensor will be independent of the gas being measured. In
reality, however, the variation in physical properties affects the output. Catalytic oxidation
sensors are no exception, so the response a pellistor gives to the same volume concentration of different gases will vary. However when exposed to the same %LEL concentration of different gases, the variation in output is fairly small compared to other detection
techniques. As safety applications are interested only in %LEL measurements this is a
major advantage.
The variation in output for the same %LEL concentration of different gases is termed
relative sensitivity.
Practicability
There are effects which limit the application framework of these sensors. The two most
important are outlined briefly:
Poisoning
Some compounds will decompose on the catalyst and form a solid barrier over the catalyst surface. This action is cumulative and prolonged exposure will result in an irreversible decrease in sensitivity. Typical poisons are organic lead and silicon compounds.
Inhibition
Certain other compounds, especially halogenated hydrocarbons, are absorbed or form
compounds that are absorbed by the catalyst. This absorption is so strong that reaction
sites in the catalyst can become blocked and normal reactions are inhibited. The resultant loss of sensitivity is temporary and in most cases a sensor will recover after a period
of operation in clean air.
Pellistors are available for the measurement of H2S on the market. The behaviour of the
sensors depends nonlinearly on concentration, the sensor and the circuit, in which it is
48
Report ODOCO-1
_________________________________________________________________________________________________
operated. On their behalf, important characteristics for a specific sensor head are arranged below.
Operation
Stabilisation time
Power supply
Sensor current consumption
Output in clean air
Response in 10ppm hydrogen sulphide
Deviation of response at 5ppm
Deviation of response at 25ppm
Life time
Specific literature:
{ e2v technologies, http://www.e2v.com/files/sensors/vq101.pdf }
{cp http://citytech.com/technology/pellistors.asp [2007-09-06]}
Electro-chemical
The general electro-chemical measurement principle was already described for the oxygen measurement [sec. 3.2.2.1]. It is dealt with more deeply here but with only with the
characteristics for the measurement of H2S.
Functionality
Electrochemical toxic gas sensors are micro fuel cells. They have a direct response to
volume concentration of gas rather than partial pressure.
The simplest form of electrochemical toxic sensor comprises two electrodes: sensing
and counter, separated by a thin layer of electrolyte. This is enclosed in a plastic housing
that has a small capillary to allow gas entry to the sensing electrode and includes pins
which are electrically attached to both electrodes and allow easy external interface.
These pins may be connected to a simple resistor circuit that allows the voltage drop resulting from any current flow to be measured. Gas diffusing into the sensor is either oxidised or reduced at the sensing electrode and, coupled with a corresponding (but converse) counter reaction at the other electrode, a current is generated through the external circuit. Since the rate of gas entry into the sensor is controlled by the capillary diffusion barrier, the current generated is proportional to the concentration of gas present
outside the sensor and gives a direct measure of the toxic gas present.
The central feature of the design is the gaseous diffusion barrier, which limits the flow of
gas to the sensing electrode. The electrode is therefore able to react with all target gas
as it reaches its surface, and still has electrochemical activity in reserve.
49
Report ODOCO-1
_________________________________________________________________________________________________
From the reaction at the counter electrode, it is evident that oxygen is required for the
current generation process to take place. This is usually provided in the sample stream
by air diffusing to the front of the sensor, or by diffusion through the sides of the sensor
(a few thousand ppm is normally sufficient). However, continuous exposure to an anaerobic sample gas may result in signal drift, despite the oxygen access paths and so it
is recommend that toxic sensors are never potted with resin or completely immersed in
an anaerobic gas mixture.
For certain demanding applications where the sensors are frequently exposed to very
high concentrations of the analyte, for example in flue gas analysis, it may be necessary
to ensure there is an additional source of oxygen access to the counter electrode.
2-electrode sensors are the simplest form of toxic gas sensors. However they have limited measuring range due to polarisation of the counter electrode. This polarisation effect
can be eliminated by using a third, reference, electrode with a stable potential in the sensor design. In these sensors the sensing electrode is held at a fixed potential relative to
the reference electrode (from which no current is drawn) so both maintain a constant potential. The counter electrode is still free to polarise, but has no effect on the sensing
electrode and does not limit the sensor in any way.
3-electrode sensors are the most widely used design of electrochemical sensors for detecting toxic gases. Despite this there are some applications where the 3-electrode design proves inadequate. For example cross-interfering gases or zero-offset changes with
temperature can compromise their overall performance. By introducing a fourth auxiliary
sensor accurate sensor performance can be maintained while also allowing the simultaneous measurement of two gases.
A 4th auxiliary electrode can assist in overcoming cross interference from other gases.
The main sensor is responding to the desired gas while the auxiliary electrode is responding to the interfering gas alone. Once the ratio of the responses on each electrode
is known, a compensated signal can be obtained by subtracting the auxiliary signal from
the sensing electrode signal with an analogue circuit or using a microprocessor with appropriate software.
The baseline signal of most electrochemical sensors tends to increase exponentially with
temperature, approximately doubling for every 10C rise in temperature. For the majority
of applications this does not normally present problems but for applications involving
very low concentrations of gases any baseline shift with temperature could seriously affect the ability to measure these gases accurately. The signals from both the sensing
electrode and auxiliary electrode will both show similar responses to changes in temperature but because the auxiliary electrode is not exposed to reactive gas, its signal can
50
Report ODOCO-1
_________________________________________________________________________________________________
simply be subtracted from that of sensing electrode. This is a useful method of compensating for any baseline shifts that would normally occur as a result of changes in temperature, but is not ideal.
{cp http://citytech.com/technology/toxic-sensors.asp [2007-09-06]}
Practicability
Electro-chemically working sensor measuring heads and sensors are available on the
market. The characteristics for a sensor head are indicated below.
Operation
Life time
Measuring Range
Sensitivity Range
Zero Current at 20C
Resolution at 20C
Bias Potential
Linearity
Response Time at 20C
Long Term Sensitivity Drift
Humidity Range
Specific literature:
{City Technology Ltd., http://citytech.com/PDF-Datasheets/h2s3e100s.pdf [2007-09-06]}
51
Report ODOCO-1
_________________________________________________________________________________________________
Practicability
The most frequently applied detector in the gas chromatography is the FID, since it connects robustness with high sensitivity. An FID is more sensitive up to 1000x than the
heat conductivity detector. The maintenance is extensive because of the supply of pure
detonating gas (H2), which has to be re-filled every few days.
Different manufacturers offer devices with FID-principle for measurement. Different examples with characteristic values of equipment are mentioned below.
Ranges
Accuracy
Response time
Zero drift
Span drift
Temperature of heated block
Operating temperatures
Autonomy
Scan frequency
Photoionization Detector
Functionality
A photoionization detector or PID uses an ultraviolet (UV) light source to break molecules to positively charged ions that can easily be counted with a detector. Ionization occurs when a molecule absorbs high energy UV light, which excites the molecule, and re-
52
Report ODOCO-1
_________________________________________________________________________________________________
sults in temporary loss of a negatively charged electron and the formation of positively
charged ion. The gas becomes electrically charged.
In the PID, the charged particles produce a current that is amplified and measured. This
current correlates to the amount of particles in the analyte.
{cp http://en.wikipedia.org/wiki/Photo_Ionization_Detector }
Practicability
Different manufacturers offer devices of this principle. The PID as well as the FID just
gives a response according to the sum of VOCs. For identification and quantification of
distinct solvents, a separation of the gas mixture using gas chromatography is necessary.
Specific literature:
{Ama Instruments Gmbh, Produktinformation GC 5000 Process, http://www.amainstruments.com
Infrared sensor
The absorption behaviour is used by radiation of a material for the determination of the
concentration.
Functionality
See description above within the sec. 3.2.2.2.
Practicability
There is no information available, whether this principle can be applied for measure in
the gas phase of sewer networks.
53
Report ODOCO-1
_________________________________________________________________________________________________
the pattern information is projected onto a two-dimensional plane, on which similar patterns overlap the same regions.
A number of different technical sensors are used in electronic noses. The main ones are:
Sensors, evaluating mass effects, with the two groups of swinging quartz sensors
(QMB/QCM sensors) and the surface wave sensors (SAW sensors).
Each of these sensor types has its own instrumental characteristic. In particular, the
chemical range measuring of gases differ. MOS sensors preferably measure lowmolecular oxidizable gases. With conducting polymers polar gas components are well
measured, mass-sensitive sensors measure high-molecular materials preferentially. In
technical application however, in particular the measuring stability of the sensors is a
crucial parameter during longer periods, since they must hold the calibration information.
Depending upon the principle of operation of the sensors and their specific application,
hence, measures must be planned for the protection and verification of the sensors.
The gas sensors used in electronic noses are often constructed in the form of specially
grouped gas sensors partially mounted on a microchip. One speaks here of sensor arrays or chemo sensor arrays.
[FRECHEN, 2007; {cp http://de.wikipedia.org/wiki/Elektronische_Nase}]
54
Report ODOCO-1
_________________________________________________________________________________________________
Si/SiO2 or Al2O3
SnO2 or WO3
SiO2
38 or 16
9x10 mm (38 Seg.) or 3x4 mm (16 Seg.)
2 heating elements: 4
200 - 400C
ca. 1 Watt (16 Seg., hanging in ceramic card)
ca. 6 Watt (38 Seg., on bearings in PGAsocket
55
Report ODOCO-1
_________________________________________________________________________________________________
Practicability
A system manufactured by the company AltraSens (OdourVector) is used in practice for
the monitoring of different processes. It consists of 6 sensors, which prefer highmolecular substances with offensive smells.
According to the manufacturer, odour emissions from the sewer system were measured.
Likewise, a biological clarification plant in a long-term monitoring was accompanied.
Specific literature: {Co Altrasens, http://www.altrasens.de}
3.2.3.1 pH-value
Principle of measurement and functionality are the same like in wastewater (see sec.
3.2.1.1).
Galvanometric pH-Meter
Practicability
For the pH measurement in semi-solid materials, for example in the area of the food industry (measurement in meat, sausage and cheese) are puncture-pH-probes for measurements available. These matrices contain bio molecules, as to be expected in biofilms
and in addition triglycerides (fats) and higher protein levels that make pH-value measurements difficult. Therefore, application appears to be possible for pH-determination in
biofilm with these sensors. However, no experience exists for continuous use and application in sewage area with these electrodes.
Measurement area
Accuracy
Allowed process temperature
Comments
56
Report ODOCO-1
_________________________________________________________________________________________________
Measurement range
Accuracy
Speed of response
Allowed process temperature
Comments
Specific literature:
{Fraunhofer Institut Keramische Technologien und Systeme, Informationsblatt Thick Film
pH Sensor, Messestand Sensor+Test, Nrnberg 2007}
Optical Sensor
A sensor has been developed which examines the characteristics of the biofilm with
photometric methods.
Functionality
Transmission, dispersion and fluorescence are measured continuously by the sensor.
The measuring procedure is accomplished in a flashing operational mode of the sensor.
The dark signal is measured in the dark phase between two pulses. Parallel to it a reference signal is measured for the characterization of the linked radiation intensities. Each
measured value is corrected concerning dark signal and standardized on the reference
signal. The transmission, dispersion and fluorescence are measured with 3 different UVwavelengths and a NIR-wavelength. The measured optical values are correlated with the
57
Report ODOCO-1
_________________________________________________________________________________________________
biofilm thickness and the number of cells. The characteristics of the biofilm are thus determined
Practicability
The dynamic behaviour of the different optical parameters in view of the processes in the
biofilm has not yet been sufficiently examined. Precise mathematical-empirical methods
and algorithms for statements about biofilm need to be developed. It is expected that by
the sensor detailed process characteristics and biochemical parameters can be measured.
Specific literature:
{KWB Band 4, Mittenzwey, Szewzyk, Dworak & Schulze Online-Sensor zur Erfassung
von Biofilmen, Berlin 2006}
Ultrasonic Scan
There are studies investigating the application of the procedure of high frequency ultrasound measurement to biofilms. The advantage of this procedure is the production of
three-dimensional in situ pictures of a biofilm sector, which hopefully allows for an exact
analysis of the structure and the development of the biofilm.
The research in this area is still in a very early stage and an application in the laboratory
for the biofilm of a species was examined.
Specific literature:
{http://www.ncbi.nlm.nih.gov/sites/entrez?tmpl=NoSidebarfile&db=PubMed&cmd=Retriev
e&list_uids=17604896&dopt=AbstractPlus}
Electro-chemical
Functionality
Two metallic electrodes which are isolated from each other are provided over an energy
source with a potential difference from 50 to 500 mVs. Due to the potential induced depolarization-effects a flowing direct current through the biofilm is measured. Likewise, the
current conduction, which flows after switching the energy source off, is measured, as
well as the electrical resistance between the electrodes which characterizes mainly the
biofilm.
Practicability
The application of the sensor in wastewater is not clarified. A manufacturer was not
found.
Specific literature: {Source: US Patent 5,356,521 - 18th October 1994}
58
Report ODOCO-1
_________________________________________________________________________________________________
Functionality
Optical coherence tomography (OCT) is an interferometric, non-invasive optical tomographic imaging technique offering millimetre penetration (approximately 2-3 mm in tissue) with micrometer-scale axial and lateral resolution.
OCT is based on low-coherence interferometry. In conventional interferometry with long
coherence length (laser interferometry), interference of light occurs over a distance of
meters. In OCT, this interference is shortened to a distance of micrometres, thanks to the
use of broadband light sources (sources that can emit light over a broad range of frequencies). Light with broad bandwidths can be generated by using superluminescent diodes (superbright LEDs) or lasers with extremely short pulses (femtosecond lasers).
White light is also a broadband source with lower powers.
Light in an OCT system is broken into two arms, a sample arm (containing the item of interest) and a reference arm (usually a mirror). The combination of reflected light from the
sample arm and reference light from the reference arm gives rise to an interference pattern, but only if light from both arms have travelled the "same" optical distance ("same"
meaning a difference of less than a coherence length). By scanning the mirror in the reference arm, a reflectivity profile of the sample can be obtained (this is time domain
OCT). Areas of the sample that reflect back a lot of light will create greater interference
than areas that don't. Any light that is outside the short coherence length will not interfere. This reflectivity profile, called an A-scan contains information about the spatial dimensions and location of structures within the item of interest. A cross-sectional tomograph (B-scan) may be achieved by laterally combining a series of these axial depth
scans (A-scan). En face imaging (C-scan) at an acquired depth is possible depending on
the imaging engine used.
Practicability
The procedure of the OCT was applied in the laboratory at TU-Munich for none-invasive
measurement of biofilms in the pipe. Growth and also peel off processes could be recorded on-line. It was possible to record the processes in a temporal resolution within the
range of seconds to minutes. The authors express hope that the procedure will be established as a standard technique for the monitoring of biofilms.
Specific literature:
Meier, Optical Coherence Tomography, Technische Information, Berner Fachhochschule
Technik und Informatik
{cp http://en.wikipedia.org/wiki/Optical_coherence_tomography [2007-09-06]}
59
Report ODOCO-1
_________________________________________________________________________________________________
Report ODOCO-1
_________________________________________________________________________________________________
two different wavelengths (e.g. at the intensity maximum points or at the isosbestic point)
was applied. Such a ratio of intensities is not altered by external factors. The measurement principle and some more details are described in DANTAN et al. (2005).
light source
coupler
fiber bundle
sensor head
spectrometer
measuring fiber
fiber bundle
light-guiding fibers
glass window
reflection layer
pH sensitive membrane
Figure 9:
61
Report ODOCO-1
_________________________________________________________________________________________________
Steel tubing
Teflon holder
pH 7
pH 13
pH sensitive membrane
Figure 10:
PHOTO: BAM
The diameter is 8 mm; the sensitive layer is protected but sufficiently sensitive for contact to the concrete matrix. The sensitive membrane must not exceed a definite thickness because the ion diffusion is hindered. In order to ensure a reliably stable thickness
of the sensitive membrane, a special powder compacting tool for manufacturing of pHsensitive membranes was developed.
The measurement resolution of the sensor for pH-values between 9 and 12 is in the
range from 0.1 to 0.6 pH units depending on the pH-value. The highest resolution can be
achieved in the middle of the measurement range (between 9.7 and 11). One particular
condition of use is that the pH sensitive membrane must not dry out. This requirement is
mostly fulfilled in hydraulic engineering and geotechnical applications. In order to prevent
drying out before integration into concrete structures, the pH sensitive membrane must
be protected by a small watertight topcoat.
Figure 11:
62
Report ODOCO-1
_________________________________________________________________________________________________
Figure 12:
Suitability of this sensor has been tested in steel anchors installed in the harbour of
Rostock (North Germany). Fig. 18 shows two of totally 10 pH sensor prototypes that
have been fixed on prefabricated anchor bodies (two in each anchor) before introducing
them into the borehole. The topcoat (Fig. 17) was removed shortly before introducing the
anchor into the borehole and concreting it. This procedure ensured that the membrane
maintained its hydrophilic properties. Since sensor installation in July 2005, seven of 10
installed sensors have provided useful information about the pH-value of the grout. (Two
sensors were damaged already during installing the anchors, one sensor failed after few
months.)
The sensor concept has been prepared for commercial supply. Relating to his, special
attention was put on optimization of the robustness of the sensor components and integration technology. A series of investigations have been carried out to clarify whether the
membranes are reliably capable of recording pH changes in concrete structures. For this
purpose, several sensor heads were embedded in fresh mortar test prisms according to
the standardized testing procedures6. After a one-month curing period in water, the mortar samples with integrated sensor heads have been stored in an acid bath for simulation
of pH decrease. The sensor signal has been monitored periodically and showed expected pH changes of the concrete matrix. Currently, the transfer of the research results
into production will be arranged by the research partners of the German research project: BAM Berlin, Schmidt & Haensch GmbH Berlin, MBF GmbH Berlin (MPA Berlin
Brandenburg) and Dietz Geotechnik Consult Hilden.
63
Report ODOCO-1
_________________________________________________________________________________________________
64
Report ODOCO-1
_________________________________________________________________________________________________
Table 8:
Corrosive compounds, their causing parameters, precursors and reaction product, importance to measure and existing sensors
Corrosive Compound
Plastics
Org. Solvents
Desirable to measure
Existing sensors
Concrete
Sulphuric acid:
(H2SO4)
Sewage
Gas
Biofilm
+
+
-
++
++
+
0
-
H2S
(Odour)
acid (H3O+),
(SO42-), bioactivity
++ (acid)
+ (SO42-)
- (acid),
- (SO42-)
++
+ (O2), - (SO42-),
-+ (S2-)
65
Material
Gypsum,
pH-value
+ (pHvalue)
0 (prototype)
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 4
Data processing
The application of a numerical prediction tool to a real sewage system requires a model
of this sewer system. In such a model, temporal-immutable properties, such as the
course of sewage segments, their length, height or slope, pipe diameter, channel types,
intakes or public sewers have to be included. The utilized materials are assigned to various segments. Furthermore, time-varying quantities in the simulation domain are necessary. They are provided as initial and boundary conditions for the simulation. Timevarying variables include all sensory and analytically measured quantities. The parameters controlling the operation of a sewage network have to be included, as well. The
time-varying boundary conditions for the simulation up to the current moment are conveyed of historic measurement data of each sewer segment. For the forecasted periods
the boundary conditions of historical data will be extrapolated or plausibly assumed.
The task of data processing for a pilot plant is to ensure the full and complete documentation of the long-term experiments for a number of years. For this reason an extensive
data base is arising. Out of this data base the necessary input parameters are compiled,
so that the simulations of the processes in sewage pipes are made possible. For the
planned development of hydrodynamic model concepts, a calibration and validation is
essential. For this purpose, recorded measuring data stored in the data base will be
used. For analysis of various hydrodynamic model variations, it is necessary to compare
and evaluate model results with each other. These comparisons are to be documented
for research progress in the numerical simulation and hence, should be stored in the
data base.
The transfer of available techniques for the present task has not yet taken place. Solutions are available for the individual parts linked with the data management for the
planned pilot plant. The implemented work is therefore geared to the application of available concepts and strategies for the necessary data management. Essential tasks in this
context are described below.
66
Report ODOCO-1
_________________________________________________________________________________________________
The measured data has to be read from data loggers, special bus systems or connections to specially purchased software packages (AgilentVee / LabView).
Data in summary reports of analytical investigations are discontinuously expected. Many
analyzer-instruments and working methods of laboratory professionals prefer an evaluation in tabular form, the data from this area should therefore be exchanged in spreadsheet formats.
The field of plant control lets one expect data of both above procedures.
Report ODOCO-1
_________________________________________________________________________________________________
data base. The hurdle to be overcome is the adjustment of the measured values from the
data base with the discretized model, because generally no coincidence of measurement
points or measurement times to the discretization can be expected. In agreement with
the hydrodynamic model implementation, a constraint generator is to be written, that
produces the required values through interpolation and integration and serves them in
the required form to the simulation.
68
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 5
Countermeasures
If planners think odour and corrosion problems are likely to arise or if there are already
problems, countermeasures have to be taken. Basically, planners can take the following
into account:
treatment of sewage
- to prevent formation of H2S,
- to eliminate H2S or
- to inhibit its transfer to atmosphere
treatment of gas or
69
Report ODOCO-1
_________________________________________________________________________________________________
Table 9: Measures in the planning phase to reduce Odour and corrosion in sewers
[BARJENBRUCH, 2002 expanded]
Measure
Operation mode
Remarks
location
could cause a different problem
Care of conventional
roof ventilation
Pneumatic pumping
station
higher investments
Figure 23), if there is abundant slope. Special options for problematic rising-mains are
blowing-in of pressurised air or technical oxygen, compressed air flushing or a vacuum
drainage systems instead of direct water pumping.
Another procedure is to augment redox potential of wastewater by dosage of chemicals.
There are various chemical compounds increasing redox potential of a solution. First
dosage of technical oxygen was applied, but insufficiently spreading in the water phase
is a considerable disadvantage. Dosage of non-volatile products are easier to handle.
The definite market leading procedure in Germany is the dosage of a Ca(NO3)2solution
called Nutriox. In addition to this, potassium permanganate, ozone, hydrogen peroxide,
sodium chloride and sodim hypochlorite are stronger oxidants in use. Another approach
is to admit building-up of H2S and to eliminate it prior to the point of disturbance. The
Thiox-method consists of the dosing of an oxidating substance (e.g. oxygen) at the last
70
Report ODOCO-1
_________________________________________________________________________________________________
section of a pressure pipe. Sulphide oxidating bacteria dominate under these conditions
and metabolise a relevant part of H2S. Some bacteria species that are involved in removing of odourous sulphur compounds have been identified and proved to enhance odour
reduction [GENG et al., 2004, FRIEDRICH et al., 2004]. All fresh keeping methods have in
common that oxidation of organic compounds decreases the fraction of readily biodegradable compounds, which can be necessary at the treatment plant for biological phosphorous elimination and denitrification.
& HOBSON (2001) propose a calculation for nitrate demand. Underlying the demand
of dissolved oxygen cited from POMEROY (1990) (2 mg/l/h respictively 700 mg/ m/h), they
assume that oxygen-respiring microorganisms are capable to catabolise nitrate, too. For
an equivalent number of electron receptors, 1 mg/l of nitrate-nitrogen can replace approximately 2.86 mg/l of dissolved oxygen. Therefore, the equivalent nitrate consumption
rates for sewage and biofilm are 0.7 mg/l/h of NO3--N and 250 mg/m/h of NO3--N respectively. With these values and estimations for nitrate respiration rates, nitrate demand by a
forced main can be calculated:
M =
M
D
LRS + DLR B = 0.00025DL(0.7 + 1)
4
- - Total nitrate demand [kg NO3--N/h]
RS
RB
200
80
180
70
160
Power failure
60
140
120
100
80
60
50
40
Start of dosage
30
20
40
10
20
0 30.07.01
Monday
0
31.07.01
Tuesday
01.08.01
Wednesday
02.08.01
Thursday
03.08.01
Friday
04.08.01
Saturday
05.08.01
Sunday
14.08.01
06.08.01
Monda
Tuesday
15.08.01
Figure 13:
16.08.01
Wednesday Thursday
17.08.01
Friday
18.08.01
Saturday
19.08.01
Sunday
20.08.01
Monday
Weekday
Weekday
71
21.08.01
Tuesday
Report ODOCO-1
_________________________________________________________________________________________________
Dosed
of product
reduced
nitrate
is
WTP
Shaft
River
Sewer
Figure 14:
5.1.2 Precipitation
Iron salts (s. Table 10) are used as precipitants to bind the already generated sulphide
as iron sulphide. It is an adequate coagulant, because its solubility is lower than other
ferric compounds in water. The reaction depends on pH. Between pH 7 and 8, least dosage of iron is necessary. It is a selective reaction to reduce H2S, therefore other odourcausing substances are not affected. Some combined products like solution of
Na[Al(OH)4] + FeCl3, ferric nitrate (sold as Anaerite 263) use both mechanisms: stabilisation of redox potential and flocculation (see below).
72
Report ODOCO-1
_________________________________________________________________________________________________
200
80
180
70
160
60
140
120
100
80
60
50
40
30
20
40
10
20
0 30.07.01
Monday
0
31.07.01
01.08.01
Tuesday
Wednesday
02.08.01
03.08.01
Thursday
Friday
04.08.01
Saturday
05.08.01
06.08.01
09.07.01
Monday
Monda
Sunday
10.07.01
Tuesday
11.07.01
Wednesday
12.07.01
Thursday
Figure 15:
13.07.01
Friday
14.07.01
Saturday
15.07.01
Sunday
Monday
Weekday
Weekday
2 Fe + 3 H S 2 FeS + 6 H + S
2
2+
2-
per mg S
Qd ( FM ) =
c( S 2 ) M Fe Qd
V
M S Fe FM
VFM
Qd ( FM )
c( S 2 )
Qd
MX
VFe
VFM
FM
73
Report ODOCO-1
_________________________________________________________________________________________________
5.1.3 pH-regulation
The augmentation of pH withdraws hydrogen from H2S (s. sec. 2.1.2). The resulting ions
HS- and S2- are not volatile. Dosage amounts based on the pKa-values underestimate
demand for pH-increasing chemicals. YANG & HOBSON (2001) hypothesise that pH at the
water surface (point of H2S-transfer) is directed by CO2-transfer. This way, the alkaline
boundary water layer prevents stripping of H2S, while an acid pH is measured in the bulk
of the water. pH-regulators in use are mentioned in Table 10.
2007]. Some mechanical means are listed in Table 10. VOIGT (2007) re-
ports a measure especially for separate systems: installations to allow a directed discharge of the stormwater at beginning of sewer or at points with sedimentation.
74
Report ODOCO-1
_________________________________________________________________________________________________
Table 10:
Measure
Oxidants
Operation mode
technical oxygen (only to be used
in closed pressure pipes)
Thiox-method; dosing of an oxidant at the last section of the
pressure pipe
Nutriox in form of Ca(NO3)2
Ozone (toxic, AGW value
0,1 ppm)
hydrogen peroxide (H2O2) (limited
stability and explosion hazard)
potassium permanganate
(KMnO4)
NaCl+NaOCl; production of AOX,
effective for a short period only
sodium hypochlorite (NaOCl,
germicidal, possible production of
AOX
ferric chloride sulphate solution,
ferric sulphate solution, ferrous
chloride solution
Fe(OH)3 as sludge or conditioned
sludge (dosing of solid substances requires individual test)
Remarks
preventing anaerobic processes
consumption of organic carbon
compounds
reliable method
relatively high costs
gas bubbles can block the pipe
oxidation of the reduced sulphur
compounds
increased consumption due to
reaction with other wastewater
compounds;
high costs of purchase
slight sustainability
75
Report ODOCO-1
_________________________________________________________________________________________________
Operation mode
removal of depositions and biofilm
in the channel with shields,
ploughs, high pressure rinsing
surge flushing with automatic flap
gate or others
cleaning of pressure mains with
big balls or small balls of expended clay
Installing a (rotating)
reducing substances with high
contactor
odour load which can be easily
degraded
Adding external water reducing detention time
(drinking water, pump reducing concentrations in
water or surface water) wastewater for a short period and
increasing oxygen concentration
76
Remarks
limited effect
constant repetition necessary
requires a lot of personnel and
equipment
addition of inhibitors
Report ODOCO-1
_________________________________________________________________________________________________
Figure 16:
Figure 17:
Examples for extensive measures are installations of coverings or gel plates. Gel
plates (BIOTHYS) contain essential oils (terpene alcohols, aldehydes and ketones).
Odourous substances react to them and form compounds, which smell less intense [BIOTHYS].
To ensure the reaction, gel plates need sufficient air ventilation. Therefore,
77
Report ODOCO-1
_________________________________________________________________________________________________
ROSENWINKEL
Gel plates have to be replaced every 2 to 4 months [BIOTHYS]. For the DRAUSY-measure
a perforated tube is pulled into a gravity sewer. Afterwards air or another oxidant is distributed to keep aerob conditions [LAWA, 2004].
There are covering systems available that are installed inside the manhole. They seal
the manhole with a plastic to avoid escape of gases. A rest of air circulation is manually
adjustable. The type of covering systems that are installed close to the base of the manhole have the advantage to beware the shaft from corrosion. A third type is an amorphous covering system. It consists of a swimming layer of particles that constrict gas
transfer. They are used in pumping stations.
Formverschlusssystem KFVA
Formverschlusssystem FVA
Formverschlusssystem FVK
Figure 18:
78
Report ODOCO-1
_________________________________________________________________________________________________
Before searching for an adequate method of gas treatment, one should note, that risk for
complaints can be reduced, choosing well directed points for exhausting air. For this
purpose manholes with enlarged ventilation openings should be installed in areas
where odour can not become a nuisance for the population. Flowing wastewater is able
to carry along the airstream above in the sewer. If sewer atmosphere has access to fresh
air, it can provide a certain amount of oxygen. This natural ventilation can be promoted
by installation of manholes with enlarged ventilation openings at adequate points [VOIGT,
2007]. This is one example for constructive enrichment of the O2-concentration.
Contaminated
Clean air
Chimney
Fresh air
Figure 19:
If the location of escaping odours is reduced to few spots, a specific gas treatment can
be applied. This is possible at vent openings of rising mains, at pump stations or at
points where flow changes. There is a variety of scrubbers (STUETZ & FRECHEN, 2001).
Their strategy is to oxidise or to sorb sulphur-compounds.
There are biofilters in use with and without forced air stream. Odourants are bound to
materials such as bark mulch, peat, expanded clay etc. Carrier material has to be kept
wet.
Another adsorbent is activated carbon. Its surface can be modified by impregnation,
catalytic substances, gas-injection or with immobilised sulphide-oxidating bacteria [YAN et
al., 2004; STUETZ & FRECHEN, 2001]. In Sydneys pumping station a catalytic incineration
of odourous gases utilising a platinum catalyst is in operation. One measure uses
photooxidation. Constituents of exhaust air are cracked by radiation with ultraviolet
light. In this process emerging strong oxidants like oxygen- and hydroxyl-radicals result
79
Report ODOCO-1
_________________________________________________________________________________________________
Figure 20:
Table 11:
Measure
Specific stripping
with exhaust air
treatment
Covering systems
Treatment inside
manhole
Compressed air
flushing
Operation mode
stripping at points with change of
flow (drop, vacuum aeration)
selective treatment of waste air
(biological or chemical scrubber)
Remarks
reduced hydraulic efficiency is
possible
space at the surface is needed;
additional operation point, energy
requirement possible
stripping of H2S is prevented
odour nuisance is limited to the
is applied in pump stations and
place where the covering sysmanholes
tems are applied
exhaust air is treated with on-thecorrosion of network cannot be
spot-biofilters or gel plates in the
reduced; maintenance is necesmanhole
sary, odour will be transferred to
other emersion points
difference of head-loss about 4-6
has to be taken into considerabar (BHLER system)
tion when planning the pressure
reduction of the detention time in
pipe
pressure pipe
increased investments for comremoval of the biofilm
pressed air stations
reducing substances with high
requires more maintenance
odour load which can be easily de- clogging and blocking
graded
specific ventilation of the channels generally, sanitary public health
by ventilating the house drain of
engineering standard, that has
each individual buildings
often not been taken into consideration recently
80
Report ODOCO-1
_________________________________________________________________________________________________
On behalf of RWE Energie AG a highly durable concrete with maximum possible acid resistance was developed at Technische Universitt Berlin, in cooperation with MPA Berlin
Brandenburg.
With highly durable concrete built of optimally chosen materials and w/z-values between
0.38 and 0.42 a resistance can be achieved, that is approximately 5-to 20 times higher
compared to "traditional" concretes based on CEM I HS. The lifespan increases directly
in proportion to the resistance, too. Above all, no sub-surface-damage to the binding-
81
Report ODOCO-1
_________________________________________________________________________________________________
agent matrix takes place when highly durable concrete is applied. The sub-surface damage of the highly resistant concrete corresponds to the depth of the visible damage.
Figure 21 shows two examples of concrete slabs after about 6 months of exposure to
sulphuric acid pH 2. The concrete on the left in Figure 7 shows a concrete with approximately 380 kg HS CEM I, w / z-value 0.42. The concrete on the right in Figure 7 is a
concrete with optimized resistance, densest packing of the overall mix, low binding-agent
content and binding-agent with maximum resistance, compared to acid. Whereas conventional" concrete shows degradation, the optimized concrete shows a virtually intact
surface with only a salt-crust deposit.
Optimized concrete
320kg BM: Z+FA+MS, w/b=0,42
Conventional concrete
380 kg cement, w/z=0,42
Figure 21:
In 1999, in Niederaussem near Cologne, there was the highest cooling tower in the world
(200 m) built without further corrosion protection on the inside. See also: HILLEMEIER, B.,
HTTL,
82
Report ODOCO-1
_________________________________________________________________________________________________
(about 12.6 at CEM I, and about 12.4 at CEM III B), this is an approximately ten trillion
fold! acid-concentration (H+ concentration).Under such conditions, concrete must be
supplemented with a corrosion protection system, against a further attack of sulphuric
acid.
83
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 6
Feedback control systems
Regulation and adjustment of countermeasures are generally based on information from
suppliers, who provide partly commercial material. Therefore, this type of information is
doubtful. The particular setting for each application is determined by experience and trial
and error [YANG & HOBSON, 2001]. Complaints about odour nuisance are documented
and evaluated. Subsequently, H2S-concentration of the sewer atmosphere is measured
to estimate the demand on countermeasures. Afterwards an apparently suitable countermeasure is implented. Control measurements are carried out to adjust the countermeasure. A more objective and transferable decision process that is applicable for most
cases and types countermeasures should be prefered. It is introduced below.
If it is foreseeable during the planning phase or if problems with odour and corrosion occur within a certain sewer network, the operator has to decide:
1)
2)
3)
The risk of odour and corrosion depends on parameters that can be regulated by a
feedback control system. This kind of control system is supplied with
data about the kind and state of the material of the network and
These data are combined with information about the relations between parameters, results and costs of treatments, resulting in commands to trigger the measure. Afterwards,
data about impacts of the measure is collected and analysed. This way, feedback information enables improvement of the control system.
84
Report ODOCO-1
_________________________________________________________________________________________________
spreads and gets transformed, depending on weather, gas pressure and ventilation. Not
till then, the odour infiltrates and is smelled. Principally, concentration of H2S decreases
during transmission. Therefore, it proved to be practical to aim at a concentration of less
than 3 ppm in sewer atmosphere to assure to go below the threshold-odour concentration of 0.13 ppm H2S in respirated air [BARJENBRUCH, 2006].
Adjustment of countermeasure
To answer number 3) one has to keep in mind, that especially for measures with dosage
of chemicals, adjusting the right amount of dosing is the critical factor. Excessive dosage
has to be avoided because it is wasteful and unwanted side-effects are not justifiable
anymore [STUETZ & FRECHEN, 2001]. The task to find the right amount for dosing is not
trivial, as generation and sorption of H2S at sewer wall does not occur steadily and there
is e.g. no fixed ratio between concentration of sulphide-species and nitrate needed [YANG
& HOBSON, 2001]. Furthermore it was observed that short peaks of very high H2Sconcentrations cause odour problems, meanwhile more than 90% of the time H2Sconcentrations are close to zero. HVITVED-JACOBSEN (2002) assumes that the reason is
the fast degradation of H2S in the biofilm. Adjustment to these very short periods of
odour nuisance is difficult.
Online-Control:
Client
Client-PC
H2S-Measuerements
Company-
Dosierwerte
http://www.....
Set-point
Setting value
Dosage-
Control-
Server
Serverwith con-
pump
modul
trol function
H2Smeasured
value
Product
via mobile-net
H2S
parameters
H2Sprobe
H2S
Measuring
point
Dosage point
Figure 22:
85
Report ODOCO-1
_________________________________________________________________________________________________
Examples
A conventional approach would be to start with a survey of the status quo. IATROU et al.,
(1997) in Houston/Texas took samples of sewage and shaft atmosphere, analysing sulphide and H2S for decision-taking. In Martinez (USA) SMITH et al. (2004) combined examination methods of physical investigation, H2S-measure in sewer headspace and sulphide measure in sewage. With help of a model [ASCE, 1989], they calculated expected
lifetime of several sections of the sewer network. Some sections demanded immediate
rehabilitation by installation of PVC sheet liner, slipliner pipe and others. Other sections
showed to have more than 0,635 cm of sacrificial concrete. Operators opposed the estimated costs for measures to reduce H2S-concentration to certain levels to savings by
extended lifetime. The District chose the 0.3 mg/l dissolved sulphide control level. Dosage of hydrogen peroxide and caustic slugging were applicated and adjusted by continous measurement of H2S in sewer atmosphere.
The producers of precipitation agents VTA, KRONOS and Sdflock utilise the parameters H2S-concentration and temperature of sewer atmosphere and temperature of sewage after point of dosage, to plan and to adapt required quantity.
There are also some trials to adjust countermeasures online without time gap. FRECHEN
(2007) tests at Emschergenossenschaft a control system, using the parameters odour
emission potential and H2S-concentration of sewage and H2S-concentration in gas phase
after point of dosage. For determination of odour emission potential odours are stripped
out of the liquid phase and measured with sensor arrays of an electronic nose (see sec.
3.2.2.4).
Yara sells a control system that consists of long-nose a control unit and dosage of
Ca(NO3)2. They include parameters of H2S-concentration in sewer-atmosphere and detention time [MATSCH, 2007]. Variation of certain parameters with time is considerable.
Efforts have been taken to integrate certain weekly patterns of parameters in control systems [MATSCH, 2007; STUETZ & FRECHEN, 2001]. They focus on temperature, sewage
flow and composition, as there is knowledge about their time variation. It is imaginable to
add diurnal or yearly patterns.
An example that has find its way into official rules is described below. Generation of turbulences by falls can be controlled by a level-depending switching mechanism for the
pump. Furthermore, if there is data about state of aerobicity of the wastewater (e.g. the
information, that a slaughterhouse is discharging wastewater with a high content of BOD
at the moment), fill-levels can be adjusted (s.
Figure 23).
86
Report ODOCO-1
_________________________________________________________________________________________________
aerobic
anaerobic
sewage
sewage
inlet
inlet
switch-on level
turbulences
switch-off level
switch-off level
Figure 23:
87
Report ODOCO-1
_________________________________________________________________________________________________
Chapter 7
Modelling
First, we start with an overview on empirical formulas which are applied to estimate the
sulphide development in sewers. There is a number of models which have a certain relation to flow, transport and reaction processes in sewers and the associated odour and
corrosion problems. These models stem from the fields of urban drainage, wastewater
plants and water-quality models. We give an overview on these models also commenting
on the different requirements for modelling odour and corrosion in sewer systems. Then,
we focus on the very few models which have a closer relation to odour and corrosion in
sewer systems, we introduce one model which is already comparatively far developed in
this direction and we also point out the models limitations.
All models which are mentioned in the following are written in capital letters and are
listed in the appendix.
Figure 24.
zP =
3
b
t
f = 1,4
u hy
D
t
35,31 Q T I S f
D
88
Q
T
QV
0 , 064
Report ODOCO-1
_________________________________________________________________________________________________
ZP
IS
b
uhy
QT
QV
< 5 000
~ 7 500
~ 10 000
~ 15 000
25 000
Figure 24:
no corrosion
slight corrosion
medium corrosion
severe corrosion
very severe corrosion
ATV
(1992)
= C t/D
US EPA
[mg/l]
(1991)
S2
S2
t2 t1
4/D
= temperature [C]
[mg/l]
POMEROY
d[S]/dt
[mg/(lh)]
CS
[mg/l]
89
Report ODOCO-1
_________________________________________________________________________________________________
BOON
and LISTER
d[S]/dt
[mg/(lh)]
CS
[mg/l]
THISTLETHWAYTE
d[S]/dt
[mg/(lh)]
CS
[mg/l]
CS
SO4
= temperature [C]
4/D
HADJIANGHELOU
et al.
= 0,975 t/D
[mg/l]
[mg/l]
The authors state that the values are at least half and at most double the calculated
value, and that therefore a safety factor of 2 is necessary [HADJIANGHELOU et al., 1984].
NIELSEN, HVITVED-JACOBSEN
rS
and RAUNKJAER
[g
S/(mh)]
rS
COD
T
Following further measuring in pressure mains, Nielsen et al. (1998) develop another
formula for the estimation of sulphide production:
rS
[g S/(mh)]
90
Report ODOCO-1
_________________________________________________________________________________________________
140
ATV
Sulfidentwicklung [mg/l]
120
Pomeroy; US EPA
Hadjianghelou
100
Thistlethwayte
80
60
40
20
0
0
Fliezeit [h]
Figure 25:
All approaches calculate the total amount of sulphide, which does not necessarily have
to be stripped out into the canal atmosphere. A comparison in Figure 25 shows that, under equal boundary conditions (tR = 5 h), the approach according to ATV produces values which are approx. 50% higher, so that the approach used has to be carefully tested.
In the following Figure 26, a comparison of measured and calculated hydrogen sulphide
concentrations is shown for different approaches. For analysing this figure, one has to
take into account that the axes for the measured and calculated hydrogen sulphide concentrations are different. Regression lines have been determined for the measurements
and calculations. The very low correlation coefficients for all approaches indicate their
very poor agreement [LAWA, 2004].
91
Report ODOCO-1
_________________________________________________________________________________________________
200
Linear (Thistlethwayte)
180
Linear (Nielsen)
y = 8,9x + 94
R2 = 0,19
Linear (ATV)
160
Linear (Pomeroy)
Linear (Hadjianghelou)
140
Linear (Boon/Lister)
y = 3,5x + 63
R2 = 0,25
120
100
y = 1,5x + 74
R2 = 0,04
80
y = 2,38x + 36
R2 = 0,15
60
40
y = 0,40x + 39,6
R2 = 0,03
y = 0,75x + 9,1
R2 = 0,10
20
0
0
10
y = -0,07x + 6,6
R2 = 0,04 14
12
There is a number of empirical formulas for the prediction of hydrogen sulphide in both,
gravity sewers and pressure mains. However, the formulas are rather limited to very
specific conditions and they are hardly transferable to general conditions. Therefore, they
are a means to get a very rough estimation, but they are not suitable for a more precise
prediction of odour and corrosion in sewer systems.
92
Report ODOCO-1
_________________________________________________________________________________________________
2003]. Urban-drainage and rainfall-runoff models simulate runoff generation and concentration in a catchment. Several models are also capable of simulating polluting loads.
Thus runoff (and polluting loads) is (are) transported to sewer systems (in urbandrainage models) and to rivers (in rainfall-runoff models). In the following, we concentrate on the urban-drainage models.
In the sewers, we distinguish hydrological and hydraulic model concepts for simulating
the flow and polluting loads. Hydrological models are based on translation or retention
concepts, for example linear storage or storage cascasdes. They are computationally
cheap and therefore, they are a very good means, for example for long-term simulations. As the hydrological model concepts only give integral information on the scale of a
catchment, a sub-catchment or a comparatively coarse grid (e.g. 1x1km), they are not
suitable for our purposes because we require refined information about flow fields and
concentration distributions.
Hydraulic models in sewers are based on the solution of the one-dimensional (stationary
or instationary) continuity and momentum equations including concepts for turbulence. If
we have gravity sewers, i.e. a free-surface in the sewer, the above-mentioned equations
can be simplified to the so-called Saint-Venant or shallow-water equations, see for example the models EXTRAN, HEC-RAS. Sometimes, the Saint-Venant or shallow-water
equations are not fully solved in sewer systems, but in a simplified form. It is mentioned
that the flow in rivers is also based on the shallow-water equations. However, the equations are solved for rivers in one, two or three dimensions depending on the problems.
This means that a hydraulic one-dimensional river model is suitable for modelling pipe
flow without any modifications as soon as we have gravity sewers. For pressure mains,
we do not solve the Saint-Venant, but the (one-dimensional) continuity and momentum
equations, also called (one-dimensional) Navier-Stokes or Reynolds equations, for an incompressible fluid, see the models SIR 3S or EPANET [ROSSMAN, 2000], freely available
from the US EPA (Environmental Protection Agency). If a sewer system consists of gravity sewers and pressure mains, the corresponding model concepts must be combined,
see the models (SWMM 5, freely available from the US EPA), SIMBA-Sewer (see also
sec. 7.5), STORM. For the practical use of a hydraulic model, the consideration of special buildings like weirs, culverts, pumping stations etc. is very important.
The simulation of gas flow in pipes is similar to the simulation of water flow in pressure
mains. However, gas is considered instead of water and generally, the compressibility
must be taken into account. Models come from the field of gas-supply (SIR 3S) or from
the field of ventilation of shaft in coal mines [INWET, POLLACK, 1987]. To the nowledge of
the authors, there is no model which simulates independent water and gas flow in a
pipe or gravity sewer.
93
Report ODOCO-1
_________________________________________________________________________________________________
In the literature, hydraulic models are also called hydrodynamic-numerical or HN or hydro-numerical models, while the models which solve the Navier-Stokes or Reynolds
equations (in two or three dimensions) are also called CFD (Computational Fluid Dynamics) models.
Hydraulic models are based on the numerical solution of the underlying partial differential
equations, generally based on the Finite-Difference (FDM), Finite-Element (FEM) or Finite-Volume method (FVM) or combinations of these methods. Therefore, a mesh is
generated for the computational domain which subdivides the domain into cells (FDM) or
elements (FEM) or control volumes (FVM) where the unknowns, for example the flow velocities or the concentrations, are computed on the corresponding nodes. In recent
years, the FVM has become more and more popular in hydro- and environmental engineering because it guarantees local conservation behaviour, while the FEM is only
globally conservative. Both methods, FVM and FEM, are applicable to unstructured grids
which should be chosen if the computational domain has complex boundaries or complex inner structures. The FDM as the only solution method is losing on importance as it
is not necessarily conservative and restricted to structured grids. An overview on simulation methods for modelling hydro- and environmental systems is given in HINKELMANN
(2005). If we compare hydraulic and hydrological models, the hydraulic ones require
(much) more computational effort. However, they compute detailed information about the
water level distribution and the flow fields depending on the chosen mesh resolution.
The polluting loads in the catchment are simulated with accumulation and removal approaches and thus they reach the sewer systems. Polluting loads in sewers can be modelled in a similar way as the flow with hydrological and hydro-numerical methods. For the
same reasons as already mentioned in the context of the flow simulations, the hydrological concepts are not suitable for our purposes. The hydro-numerical concepts for transport consist of the numerical solution of the transport equation with the methods already
mentioned (FDM, FEM, FVM and combinations). Principal transport mechanisms are advection, diffusion / dispersion and reaction. The transported components include dissolved and particulate ones. Particulate components can be deposed and eroded. Reaction processes could be based on concepts similar to the ones from the ASM family (see
next section). As the reaction processes are poorly understood today, as they require
many data and as they seem to be not relevant for several purposes which are not concerned with odour and corrosion, they are neglected in many models and just conservative transport is simulated.
The STORM model has a special emphasis on polluting loads based on different hydrological concepts. It also includes the relevant special buildings like storage-space channels, passage or storage basins. The SIMBA Sewer model forsees hydrological and hy94
Report ODOCO-1
_________________________________________________________________________________________________
dro-numerical components for polluting loads including sedimentation and erosion. However, reaction processes in sewer systems have not been considered in both models
(STORM, SIMBA Sewer; personal communications with the developers). Reaction processes in sewer systems are only accounted for in very few models. This will be discussed together with the aspect of data in sec. 7.6.
For our purposes of modelling odour and corrosion, we require a hydro-numerical model
which is capable to simulate a detailed resolution of the flow processes (distribution of
water levels and flow velocities) and the transport processes (concentration distributions
of pollutants or other components). The Finite-Volume method is desirable because of its
local conservation behaviour. As we just plan one-dimensional modelling, and computers
have become fast in the last years, we generally do not expect to get problems with CPU
times for most of our investigations.
95
Report ODOCO-1
_________________________________________________________________________________________________
Figure 27:
For the description of the reaction and processes, a matrix notation has been introduced
and it has established for such kind of problems since then (see Figure 28).
96
Report ODOCO-1
_________________________________________________________________________________________________
Figure 28:
A further step in that direction was ASM2 (HENZE et al. 1995) and ASM2d [HENZE et al.
1999] which included the biological and chemical elimination of phosphorus. As phosphorus will not be of importance for our purposes, we do not comment on it further.
97
Report ODOCO-1
_________________________________________________________________________________________________
The SIMBA model is tool for the simulation of wastewater treatment plants which is
widely-spread in research and engineering practice, especially in Germany. It includes
ASM1, ASM2(d) and ASM3 as well as different other ASM variants. It foresees also several modules for the sludge treatment. It is further commented on in section 7.6. The two
models DENIKAplus and DENISIM are a linked package for dimensioning, optimizing
and simulation wastewater treatment plants. DENIKAplus mainly deals with aeration
tanks and clarifiers, while DENISIM enables dynamic simulation concerning water purification based on ASM1 and ASM2d.
In many respects, a wastewater plant and a sewer are very similar concerning the occurring microbial processes. Therefore, the modelling concepts such as ASM can be similar,
too. However, there are a number of aspects which are different concerning the definition
of the system, dominating processes and a number of details. The heterotrophic biomass
concentration in bulk water of a sewer is typically low compared to activated sludge.
Substrate in a sewer may be relatively abundant. Easily biodegradable substrate is in
young wastewater often available for non-limited biomass growth.
(1999) developed a water quality model for rivers also based on a three98
Report ODOCO-1
_________________________________________________________________________________________________
dimensional flow model. She included several reaction processes of the ATV water quality model (ATV-DVWK 2002) and she dealt with the down-scaling of multi-dimensional
water quality models. HUNZE (1996) developed a water quality model for an aeration basin on the basis of ASM1 [HENZE et al., 1987] and she coupled it with a multi-dimensional
flow model.
Figure 29:
Furthermore, it is mentioned that the models from the big software developers for modelling the world of water Danish Hydraulics Institute (DHI; MIKE11, MIKE21, MIKE3),
99
Report ODOCO-1
_________________________________________________________________________________________________
from WL / Delft Hydraulics (SOBEK, DELFT3) and from Sogreah (TELEMAC2D, TELEMAC3D) forsee components for simulating water quality.
Water quality models also require many data which partially can be taken from the literature and partially must be measured for driving, calibrating and validating the models.
Some water-quality models for rivers have a certain relation to problems in sewer systems because they also account for some microbial processes which occur in sewers.
However, they generally neither focus on the sulphur cycle nor simulate the mass transfer of components (as hydrogen sulphide) from water to gas.
(2007). The SIMBA model also belongs to the groups of integrated models as it
combines urban drainage (SIMBA Sewer, see sec. 7.2) and wastewater plants (see sec.
7.3) and as it is principally also capable of integrating water quality in rivers.
The foci of the above-mentioned models is generally on integrated studies which concern the optimization of wastewater management, real time control, rainwater harvesting,
reduction of outflow from combined sewers or other problems. As they do not treat odour
and corrosion in sewers, they are not further commented on, here.
7.6 Flow, transport and reaction models in sewers and the WATS
model
As already mentioned, there are only very few models with relation to odour or / and corrosion in sewer systems.
ALMEIDA
et al. (1999) developed a tool for the predicting of in-sewer changes in wastewa-
ter quality under aerobic conditions focussing on dry weather conditions. They used hydro-numerical models for water flow and advective / dispersive transport, while reactions
are based on a concept similar to ASM1. Processes such as reaeration, heterotrophic
growths and decay, hydrolysis, ammonification and biofilm consumption of substrate and
oxygen are considered. They focussed on sensitivity analyses about the model parameters which indicated that the yield and maximum growth rate of heterotrophs were the
most sensitive ones. However, the sulphur cycle was not considered.
HUISMAN
bined sewers which has been included into the AQUASIM model. Water flow and transport were modelled with a hydro-numerical approach, reaction processes were based on
the ASM3. The main objectives of the work have been biofilm activities and oxygen bal-
100
Report ODOCO-1
_________________________________________________________________________________________________
ances. Sulphur components were included in the model concept, however, their influence was negligible.
Figure 30:
Flow and mass-transfer processes (left) and transformation processes (right) (taken from: www.sewer.dk)
101
Report ODOCO-1
_________________________________________________________________________________________________
As WATS concentrates on dry weather conditions, water and gas flow is just considered
in very simple way. The water flow is stationary and uniform based on Mannings friction
law. Therefore, several flow processes which can be determined from the solution of the
Saint-Venant equations such as non-uniform and instationary flow as well as a turbulence parameterisation are not included, i.e. for example that the simulation of the flow
conditions around drops is not possible. However, WATS can be applied to gravity sewers and pressure mains. The gas flow is also assumed to be stationary, i.e. also that
non-uniform and instationary flow as well as turbulence are not treated. The gas flow is
not computed, it is determined from measurements or estimated to be a certain percentage of the water flow, for example 15%. Therefore, WATS is hardly capable to simulate
countermeasure like different ventilation scenarios. A friction condition between the water and gas phase which is common practice in river hydraulics to account for the wind
influence on the water flow is not included.
Concerning the transport simulation, diffusion / dispersion is neglected in the water
phase. In the gas phase, diffusion / dispersion is also not modelled, but estimated from
measurements. Sediment transport, deposition and erosion are also not addressed.
The major emphasis of the WATS model consists of the concept for the microbial and
chemical processes of organic matter, sulphur and nitrogen in the water and gas phase
as well as the mass transfer processes water-gas and gas-solid (pipe wall) including
variable temperature and ph-conditions. An overview is given in Figure 31. The organic
matter (measured as the total COD) is subdivided into four components (heterotrophic
biomass, readily biodegradable substrate, fast and slow hydrolysable substrates), and
three sulphide fractions (dissolved sulphide, particulate sulphide, gaseous hydrogen sulphide) are considered. The kinetics include aerobic and anaerobic transformations of organic matter, hydrogen sulphide formation, sulphide oxidation in water and biofilm, sulphide oxidation in the gas phase / sewer wall (adsorption, corrosion). The mass transfer
between the water and gas phase accounts for reaeration and sulphide emission depending on Henrys law. Corrosion rates are described by a Monod-type reaction depending on the hydrogen sulphide concentration in the gas phase. A matrix notation of
most of the mentioned reaction processes is shown in Figure 32.
102
Report ODOCO-1
_________________________________________________________________________________________________
Figure 31:
Figure 32:
103
Report ODOCO-1
_________________________________________________________________________________________________
The processes in biofilms are very complex and a field of science for itself. An overview
about biofilm kinetics is found in HARREMOS (1978), HENZE et al. (1995), LOHSE (1986)
and BROCKMANN (2006). For odour and corrosion in sewers, simpler expressions are
required. Ficks second law of diffusion and a Michaelis-Menten (Monod) kinetic are used
to describe the transport and transformation processes in the biofilm. The aerobic biofilm
growth is approximated with a half order kinetic. Biofilms in the gas phase are not (directly) accounted for. Indirectly, the biofilms in the gas phase have an influence on the
adsorption and corrosion rates which depend on the ph and temperature conditions.
Additonally to the book of HVITVED-JACOBSEN (2002), other literature of his group is worth
to be mentioned and contains newer results compared to 2002. Overall, the studies of
this group resulted in more than 150 publications. In HVITVED-JACOBSEN (1998) the process and model concept for microbial wastewater transformations in gravity sewers is explained much briefer compared to the book. HVITVED-JACOBSEN & VOLLERTSEN (2001) introduce the formation of hydrogen sulphide and odourous substances in wastewater also
much briefer compared to the book. NIELSEN et al. (2005) present simulations of sulphide
buildup in wastewater and atmosphere of sewer networks. This paper has a special focus on the sulphur cycle in water, gas and the sewer wall, and it has implemented the
corrosion rates from ESOY et al. (2002). In that context, the corrosion rates determined by
WEISSENBERGER
(2002) are also mentioned (also WEISSENBERGER does not belong to the
Report ODOCO-1
_________________________________________________________________________________________________
highly developed numerics, nevertheless state of the art numerics should be applied.
Furthermore, the software design can be improved, for example using object-orientated
methods.
105
Report ODOCO-1
_________________________________________________________________________________________________
The author Hinkelmann has more than 15 years research experiences in the fields of the
development and application of models for simulating flow, transport and reaction processes in subsurface (groundwater, water-gas flow) and free-surface flow systems (rivers,
estuaries) as well as in the fields of integrated models and hydroinformatics. His experiences are summarized in the book Efficient Numerical Methods and InformationProcessing Techniques for Modeling Hydro- and Environmental Systems published by
Springer in 2005. Since 2004, he is head of the Department of Water Resources Management and Modeling of Hydrosystems at the Technische Universitt Berlin. Current
research projects and publications are found at www.wahyd.tu-berlin.de.
106
Report ODOCO-1
_________________________________________________________________________________________________
Bibliography
Almeida M. C.; Butler D.; Davies J. W. 1999. Modelling In-Sewer Changes in Wastewater
Quality under Aerobic Conditions. Water Science & Technology 39: 63-71.
ATV 2004. Anforderungen an die Niederschlag-Abfluss-Berechnungen in der Siedlungsentwsserung. Merkblatt 165.
ATV-DVWK 1992. Besondere Entwsserungsverfahren: Unterdruckentwsserung Druckentwsserung. Bad Hennef.
ATV-DVWK 2003. Geruchsemissionen aus Entwsserungssystemen -Vermeidung oder
Verminderung- Merkblatt ATV-DVWK-M 154.
ATV-DVWK-Arbeitsgruppe GB-4.2 2002. Handbuch ATV-DVWK-Gewssergtemodell.
Barjenbruch M.; Dohse C. 2004. Bewertung von Manahmen zur Verringerung von Geruchs- und Korrosionserscheinungen im Kanalnetz des lndlichen Raumes. Institut fr
Kulturtechnik und Siedlungswasserwirtschaft, Universitt Rostock.
Barjenbruch M. 2006. Komplexeinsatz Amorphes Abdecksystem - Formverschlusssystem. Abschlussbericht zum Forschungsvorhaben.
Bates R. G. 1954. Electrometric pH-Determination. Wiley, New York.
Bauer R. 2007. Elimination von Viren durch Sand/Bodenpassagen an halbtechnischen
Versuchsanlagen fr die Anwendung der Uferfiltration zur Trinkwasseraufbereitung.
Hannover.
Becker M. 2001. Vergleich verschiedener Manahmen zur Geruchs- und Korrosionsvermeidung. Universitt Rostock.
Belie N.; Monteny J.; Beeldens A.; Vincke E.; Van Gemert D.; Verstraete W. 2004. Experimental research and prediction of the effect of chemical and biogenic sulfuric acid
on differnt types of commercially produced concrete sewer pipes. Cement and Concrete Research 34: 2223-2236.
Boon A. G.; Lister A. R. 1975. Formation of Sulphide in Main Rising Sewers and its Prevention by Injection of Oxygen. Progress in Water Technology 7.2: 289 - 300.
Brockmann D. 2006. Modelling Nitrification and Deammonification in Biofilms. Verffentlichungen des Institutes fr Siedlungswasserwirtschaft und Abfalltechnik 135, Hannover.
Dantan N.; Habel W. R.; Wolfbeis O.S. 2005. Fiber optic pH sensor for early detection of
danger of corrosion in steel-reinforced concrete structures. Proceedings of the International Conference on Smart Structures and Materials. San Diego 5158: 274-284.
DEV 2003. Deutsche Einheitsverfahren zur Wasser-, Abwasser- und Schlamm-
107
Report ODOCO-1
_________________________________________________________________________________________________
Report ODOCO-1
_________________________________________________________________________________________________
Heitz E.; Flemming H.-C.; Sand W. 1996. Microbially Influenced Corrosion of Materials.
Springer, Berlin.
Henze M. 1995. Wastewater Treatment - Biological and Chemical Processes. Springer,
New York, Berlin.
Henze M.; Grady C. P.; Gujer W.; Marais G.; Matsuo T. 1987. Activated Sludge Model
No. 1. IAWPRC Scientific and Technical Report No. 1. IAWPRC, London.
Henze M.; Gujer W.; Mino T.; Wentzel M. C.; Marais G.; 1995. Activated Sludge Model
No. 2. IAWQ Scientific and Technical Report No. 3. IAWQ, London.
Henze M.; Gujer W.; Mino T.; Wentzel M. C.; Marais G.; van Loodsrecht M. C. 1999. Activated Sludge Model No. 2d. Water Science & Technology 39: 165 - 182.
Hinkelmann R. 2005. Efficient Numerical Methods and Information-Processing Techniques for Modeling Hydro- and Environmental Systems. Springer, Berlin, Heidelberg.
Hll K. 2002. Wasser- Nutzung im Kreislauf, Hygiene, Analyse und Bewertung. de Gruyter, Berlin, New York.
Huisman J. L. 2001. Transport and Transformation Processes in Combined Sewers. Dissertation. Swiss Federal Institute of Technology, Zrich.
Hunze M. 1996. Numerische Modellierung reaktiver Strmungen in oberflchenbelfteten Belebungsbecken. Dissertation. Institut fr Strmungsmechanik und Elektronisches Rechnen im Bauwesen Nr. 48, Hannover.
Hvitved-Jacobson T.; Vollertsen J. 2001. Formation of Hydrogen Sulphide and Oderous
Substances in Wastewater. Geruchsemissionen. Institut fr Siedlungswasserwirtschaft Nr. 105, Karlsruhe.
IWA Task Group on Mathematical Modelling for Design and Operation of Biological
Wastewater Treatment 2000. Activated Sludge Models ASM 1, ASM 2, ASM2d,
ASM3. London.
Kngeter J.; Rubbert S. 2004. Untersuchungen zur Gewsserbeschaffenheitsentwicklung der Spree. Abschlussbericht des gleichnamigen Projektverbundes im Auftrag des
Bundesministeriums fr Bildung und Forschung.
Kopmann R. 1999. Mehrdimensionale Effekte in dimensionsreduzierten Gewssergtemodellen. Dissertation. Inst. fr Strmungsmechanik und Elektronisches Rechnen im
Bauwesen Nr. 57, Hannover.
Lnderarbeitsgemeinschaft Wasser 1993. Leitlinien zur Durchfhrung von Kostenvergleichsrechnungen. Gebr. Parcus KG, Mnchen.
Landesamt fr Natur, U. u. V. N. (. 2004. Feststellung und Beurteilung von Geruchsimmissionen (Geruchsimmissions-Richtlinie - GIRL -). in der Fassung vom 21. September 2004 mit Begrndung und Auslegungshinweisen.
LAWA 2004. Abschlussbericht zum LAWA-Forschungsvorhaben im Rahmen des Ln109
Report ODOCO-1
_________________________________________________________________________________________________
Anlagen
September
2007.http://www.cpc-
consultants.de/common/downloads/vortraege/060907_1145_rosenwinkel.pdf.
Rump H. H. 2003. Laborhandbuch fr die Untersuchung von Wasser, Abwasser und Boden. Wiley-VCH, Weinheim.
Schneider S. 1999. Modellierung des Einflusses von Mischwasseremissionen auf die
Gte langsam flieender gestauter Gewsser. Schriftenreihe fr Stadtentwsserung
und Gewsserschutz Nr. 21,.
Schrder K.; et al. 2005. Integrated Simulation of the Berlin Sewage System and Evalua110
Report ODOCO-1
_________________________________________________________________________________________________
Processes
and
Network
Group:
The
WATS
model.
URL
http://www.sewer.dk/WATSdetail.html. - 2007_7_18.
Smith J.; Chesler G.; Schafer P. 2004. Extending pipeline life through corrosion control
optimization with atmospheric hydrogen sulfide monitoring and caustic slug dosing.
Thistlethwayte D. K. 1972. The control of sulphides in sewerage systems. Butterworths.
Vincke E., Verstichel S., Monteny J., Verstraete W.H. 1999. A new test procedure for
biogenic sulphuric acid corrosion of concrete, Biodegradation 10, pp. 421-428.
Vincke E., Monteny J., Beeldens A., Belie N.D., Taerwe L., van Gemert D., Verstraete
W.H. 2000. Recent developments in research on biogenic sulfuric acid attack of concrete. In: Lens P.N., Pol L.H. (eds.), Environmental Technologies to treat Sulfur Pollution Principles and Engineering, IWA Publishing: 515-541.
Voigt J. 2007. Betriebliche Manahmen zur Minimierung von biogener Schwefelsurekorrosion in abwassertechnischen Anlagen. 1. Symposium zur Nachhaltigkeit abwassertechnischer
Anlagen
September
2007.
http://www.cpc-
consultants.de/common/downloads/ vortraege/vortraege.pdf.
Vollertsen J.; Nielsen A. H.; Hvitved-Jacobson T. 2005. Effects of In-Sewer Processes: a
Stochastic Model Approach. Water Science & Technology 52: 171 - 180.
Weissenberger J. 2002. Betonkorrosion - ein Forschungsprojekt aus Norwegen. Schwefelwasserstoff in Abwassersystemen. Fachtagung der Hydro-Care.
Yan R.; Ng Y. L.; Chen X. G.; Geng A. L.; Gould W. D.; Duan H. Q.; Lian D. T.; Koe L. C.
2004. Batch experiment on H2S degradation by bacteria immobilised on activated
carbon. Pages 299-308in IWA, editors.Odours and Volatile Organic Compounds II Measurement, Regulation and Control Techniques. IWA Publishing, London.
Yang G.; Hobson J. 2001. Use of chemicals for septicity and odour prevention in sewer
networks. Pages 269-292in Stuetz R., F. F., editors.Odours in Wastewater Treatment.
IWA Publishing, London.
111
Report ODOCO-1
_________________________________________________________________________________________________
Models
AQUASIM
http://www.aquasim.eawag.ch/
Eawag
CAEDYM
http://www.cwr.uwa.edu.au/services/models/caedym/caedym.html
Hipsey, M.R., Romero, J.R., Antenucci, J.P. and Imberger, J., 2006. The Computational
Aquatic Ecosystem Dynamics Model (CAEDYM): a versatile water quality model for coupling with hydrodynamic drivers. Proceedings of the 7th International Conference on Hydroinformatics, Nice, France.
DELFT3D
http://delftsoftware.wldelft.nl/
WL | Delft Hydraulics
DENISIM / DENIKAplus
http://www.isah.uni-hannover.de/denika/denisim.html
Institut fr Siedlungswasserwirtschaft und Abfalltechnik, Universitt Hannover
EPANET
http://www.epa.gov/nrmrl/wswrd/dw/epanet.html
Rossman, L. A., 2000. EPANET 2 Users Manual, Water Supply and Water Resources
Division, National Risk Management Research Laboratory, Cincinnati.
HEC-RAS
http://www.hec.usace.army.mil/software/hec-ras/
US Army Corps of Engineers - Hydraulic Engineering Center (HEC)
HYSTEM-EXTRAN
http://www.itwh.de/S_extinfo.htm
itwh GmbH, Hannover
INFOWORKS-CS
http://www.wallingfordsoftware.com/products/infoworks_cs/
112
Report ODOCO-1
_________________________________________________________________________________________________
MIKE-URBAN
http://www.dhisoftware.com/mikeurban/
DHI Water & Environment
MIKE SHE
http://www.dhisoftware.com/mikeshe/
Refsgaard, J. C. & Storm, B., 1995. MIKE SHE in V. P. Singh (editor), Computer Models
for Watershed Hydrology, Water Resources Publication, pp. 809-846.
MOUSE
http://www.dhigroup.com/Software/Urban/MOUSE.aspx
DHI Water & Environment
NASIM
http://www2.hydrotec.de/vertrieb/nasim/
Hydrotec Ingenieurgesellschaft fr Wasser und Umwelt mbH
Wolf-Schumann, U. (1994): Das komplexe Wasserbilanzmodell NASIM 2.0 mit seiner
Peripherie. Darmstdter Wasserbaukolloquium. Darmstadt.
QSIM
http://www.bafg.de/servlet/is/7455/
Kirchesch, V. & Schl, A., 1999. Das Gewssergtemodell QSIM - Ein Instrument zur
Simulation und Prognose des Stoffhaushalts und der Plantondynamik von Fliegewssern - Hydrologie und Wasserbewirtschaftung, 43, S. 302 - 308.
QUAL2K
http://www.epa.gov/athens/wwqtsc/html/qual2k.html
SOBEK
http://delftsoftware.wldelft.nl/
WL | Delft Hydraulics
113
Report ODOCO-1
_________________________________________________________________________________________________
SIMBA
http://simba.ifak-md.de/simba/
ifak system GmbH
SIR 3S
http://www.3sconsult.de/docu/3s/software/index_de.html
3S Consult GmbH
STORM
http://www.sieker.de/modules/wfsection/viewarticles.php?category=17
Ingenieurges. Prof. Dr. Sieker mbH, 2003. Handbuch STORM
SWMM 5
http://www.epa.gov/ednnrmrl/models/swmm/index.htm
Rossman, L. A., 2006. Storm Water Management Model - Quality Assurance Report:
Dynamic Wave Flow Routing, Water Supply and Water Resources Division, National
Risk Management Research Laboratory, Cincinnati.
TELEMAC2D / 3D
http://www.telemacsystem.com/
Sogreah Consultants
WASIM-ETH
http://homepage.hispeed.ch/wasim/index.html
Schulla, J., Jasper, K., (2006). Model Description WaSiM-ETH. Internal report, IAC, ETH
Zrich, 174 pp
WATS
http://www.sewer.dk/
Section of Environmental Engineering Department of Life Sciences
Aalborg University
WBALMO
http://www.wasy.de/deutsch/produkte/wbalmo/index.html
WASY Gesellschaft fr wasserwirtschaftliche Planung und Systemforschung mbH
114