SANITAS Sustainable and Integrated Urban Water System Management - - PowerPoint PPT Presentation

Sustainable and Integrated Urban Water System Management

SANITAS

Sustainable and Integrated Urban Water System Management

Qualitative Modelling for Urban Water System Decision Support

4th SANITAS e-Seminar Jose Porro – Universitat de Girona

Sustainable and Integrated Urban Water System Management

Agenda

Laboratory of Chemical and Environmental Engineering

SANITAS Overview Intro to Qualitative Modelling Qualitative Risk Models for WWTP Control Benchmarking Qualitative / Mathematical Modelling Framework for Assessing Integrated UWS GHG Emissions Eindhoven Case Concluding Remarks

Sustainable and Integrated Urban Water System Management

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Focus

• n

extending knowledge and filling gaps for practical implementation of sustainable and integrated approaches to UWS managment and policy-making decision support in meeting emerging challenges

• Energy
• Climate Change
• Stricter Water Quality Stds
• Emerging Contaminants
• Resource Recovery
• Efficiency
• Minimizing environmental impact while meeting water objectives
• Holistic / Integrated solutions

SANITAS Research Objectives

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Economic Social Political Sustainability

GHG emissions Water Quality Life Cycle Assessment Noise Employee Health Job Creation Public Perception Life Cycle Cost Capital Costs Rate Increases Return on Investment

5

Environmental

New Criteria Traditional Criteria

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

SANITAS Projects and Decision Pathways

Castro (ESR10) Modelling GHG granular sludge anammox Rehman (ESR6) Modelling GHG CFD Ricken (ESR5a) Micropollutants Snip (ESR9) Modelling GHG, Micropollutants Porro (ER1) Qualitative Modelling Batista (ESR5b) Controlling Sulfide,GHG Sewer Saagi (ESR7) BSM System-wide Hadjimichael (ESR1) EDSS Arnaldos (ER2) IMS Modelling Stefani (ESR2) IMS Energy Meng (ESR4) Catchment- Based/Real-time Consenting Vallet (ER3) Modelling CSO WQ Paulo (ESR3) Biogas formation Solon (ESR8) BSM Plant-wide River Water Supply & Treatment Garcia (ESR7) Integrated Master Planning Decision-Making Decision-Making

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Qualitative Modelling

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

• When phenomena of biological nature cannot be predicted adequately

by general and validated deterministic models due to lack of sufficient mechanistic understanding of the underlying kinetics and population dynamics (Comas et al., 2008)

• When understanding of complex mechanism is not needed to answer

practical questions or provide decision support

Why use Qualitative Modelling?

Sustainable and Integrated Urban Water System Management

Artificial Intelligence (AI) or Knowledge-based systems mimic human perception, learning and reasoning to solve complex problems (Chen et al., 2008)

Rule-based reasoning Case-based reasoning Black Box Statistical / Data Mining Machine learning / Data mining Agent technology Neural networks

Qualitative Modelling Techniques

Sustainable and Integrated Urban Water System Management

ARTIFICIAL INTELLIGENCE

Symbolic processes Numerical processes Complex processes “Normal” processes Mathematical resolution Approximate solutions “exact” solutions Approximate information Exact information

Differences of AI with Numerical or Deterministic Methods

NUMERICAL METHODS

Heuristic resolution

QUALITATIVE QUANTITATIVE

Sustainable and Integrated Urban Water System Management

Rule-based System Development

Knowledge Acquisition Knowledge Representation in a graphical way Codification of the branches by means of production rules: IF <conditions> THEN <conclusions> Knowledge Base

Sustainable and Integrated Urban Water System Management

Rule-based System: knowledge acquisition

Knowledge acquisition : Sources and methods

EXPERT experience DATA BASE data SPECIFIC KNOWLEDGE

DATA MINING

LITERATURE theory GENERAL KNOWLEDGE

KB

PROCESS interviews

time

REVIEW

EXPERT experience experience DATA BASE data SPECIFIC KNOWLEDGE LITERATURE theory GENERAL KNOWLEDGE

KB

PROCESS interviews

time

(Comas, 2012)

Sustainable and Integrated Urban Water System Management

Environmental Decision Support Systems (EDSS)

 Single AI techniques could not succeeded

• advantages but limitations
• all knowledge could not be

captured in

• ne

reliable model  Link control algorithms and mathematical models to AI techniques  Environmental Decision Support Systems, which integrate

“a new tool INTEGRATING different reasoning models (mahemtical, AI, GIS, et.) complementing each other and thus increasign the overall potentialities. This tool helps to reduce the time in which decisions are made, and improves the consistency and quality of those decisions”

 Complex management of environmental systems

• numerical control
• mathematical modelling,
• heuristic knowledge (literature, experts),
• experiences

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

• Poch et al., 2004 – KBS EDSS for WWTP Supervision and Technology

Selection

• Garrido et al. 2012 – KBS EDSS for WWTP Technology Selection
• NOVEDAR_EDSS
• Rodriguez-Roda et al., 2002 – Hybrid KBS/CBS DSS for WWTP microbial
• perational problems
• Comas et al., 2008 – Hybrid numerical/KBS for assessing risk of WWTP

settling problems of microbial origins

Qualitative Modelling Examples

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Integrating Mathematical and Qualitative Modelling

UWS Deterministic Modelling tools have done well to address complexity and dynamics

• Sewer models
• ASM models
• BSM models
• River water quality models

Build on previous success and extend capabilities and decision support by leveraging Deterministic model output data for Qualitative Assessment

• Microbial operational problems
• GHG emissions
• Integrated models and benchmarking

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Qualitative WWTP Risk Assessment Models

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Qualitative AS Risk Model for Assessing Risk of Solids Separation AS Risk Model (Comas et al., 2008)

• Knowledge Acquisition from vast heuristic knowledge from experts and

literature

• Knowledge formalized in Decision Trees
• Implemented in fuzzy-logic Rule-Based system

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

AS Risk Model Development

AS Bulking Knowledge Representation thru Decision Trees

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering BSM2 plant configuration (Nopens et al. 2010)

AS / AD Risk Model Implementation in Benchmark Simulation Platforms

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

AS Risk Model BSM1 Results

Open-loop Dry Weather Integrated Overall Risk (Comas et al., 2008)

Sustainable and Integrated Urban Water System Management

AS Risk Model BSM1 Results

Comparison of Three Control Strategies (Comas et al., 2008)

Highlights importance of considering operational problem dimension to typical WWTP control strategy benchmarking. Only considering typical cost (OCI) and WQ (EQI), one would be led to high risk conditions of sludge bulking and foaming.

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Extending Qualitative AS Risk Model and Concept

• AS Risk Model needs validation based on full-scale data
• Using digitized microscope photographs
• Modelling foaming events
• Extend AS Risk Model for other configurations and extend concept for
• ther complex UWS problems (eg. assessing UWS GHG Risk)

Porro (UdG) SANITAS ER1 – Qualitative Modelling in UWS

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Applying AS Risk Model to other Configurations

Terrassa WWTP (Terrassa, Spain) approx 7000m3/d – MBR train

• Measurement campaign to characterize MBR process performance
• Model development and calibration
• Simulation of various optimization strategies
• Adapt and apply AS Risk Model to confirm optimization is not

increasing risk of AS bulking / foaming and MBR operational problems

• First step in developing generic AS Risk Model for any

configuration

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Terrassa MBR Model

Sustainable and Integrated Urban Water System Management

Terrassa MBR Model-Based Optimization

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5

N-NO3 (mg·L-1)

5 10 15 20 25 No3- 1.2DO-4Q No3- 0.8DO-2Q

Current Operation

• Promising WQ and Energy Results
• Risk of Microbial Operational Problems?

Days

3600 3700 3800 3900 4000 4100 4200 4300 0,5 1 1,5 2 2,5 3 3,5 TSS (mg·L-1) 1.2 DO - 4Q 0.8 DO - 2Q

Days

Sustainable and Integrated Urban Water System Management

Terrassa MBR Model-Based Optimization AS Risk Model Results

• Need to determine proper sampling points for rbCOD, DO, sensitivity to

different sampling points, calculation of SRT and F/M for other configurations besides BSM1 (Comas et al., 2010)

• Integrating Mathematical / Qualitative Models adds greater Decision Support
• When permanent improvements are required, evaluating trade-off between,

capital costs, energy savings, and risk of operational problems becomes critical

temps DO SS1 SRT FtoM_2 vec FtoM_1 vec BOD5to N BOD5to P Ss/Xs SsOut Reac5 SNOOu tReac5 SoOut Reac5 XBHOu tReac5 XBHbotto nclarifier Slugevolum eclarifier Qrflow 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 (d) (g/m3) (g/m3) (g/d) (g DBO/ (g X·d)) (g DQO/ (g X·d)) (g/m3 DBO/gm 3 N) (g/m3 DBO/gm 3 P) (g/m3) / (g/m3) (g/m3) (g/m3) (g/m3) (g/m3) (g/m3) g (g/m3)

AS Risk Model Inputs

Sustainable and Integrated Urban Water System Management

Extending Qualitative Risk Modelling concept for Assessing Risk of N2O emissions in Full-Scale WWTPs

N2O Risk Knowledge Parallels AS Risk Knowledge

• Complex microbial phenomena
• No widely validated models
• Significant amount of knowledge from literature
• Clear
• perational

parameters already associated with ¨risk¨ (Kampschruer et al., 2009; Foley et al., 2010; Ahn et al., 2010; Chandran et al.; GWRC, 2011)

(From GWRC, 2011)

First attempt at Synthesizing Knowledge of N2O Risk

Sustainable and Integrated Urban Water System Management

NH2OH

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

AOB Pathway N2O Models

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Heterotrophic Denitrificaiton N2O Models

Hiatt and Grady, 2008 Houweling et al., 2011

4 step Denitrification 4 step Denitrification

Sustainable and Integrated Urban Water System Management

Status of N2O Models

• Due to importance of N2O (300 x CO2 GWP) a lot of focus has

been placed in modelling N2O production in Full-scale WWTPs

• ver the last few years
• For Heterotrophic Pathway Hiatt and Grady (2008) ASMN

model continues as common base

• From the various AOB pathways models, Ni et al. (in press) is

first successful full-scale validation

• Not yet rigourously validated – only two configurations
• Consensus on model, dominant pathways, how to implement

and calibrate all pathways yet to be reached

• Need to determine if pathways selected will affect mitigation

modelling despite successful validation (i.e. Good for reproducing but not good for hypothetical)

• Still a lot of work for AOB pathway models!

Sustainable and Integrated Urban Water System Management

Castro (UGent) SANITAS ESR10 – Granular Sludge Anammox A B C D E F Lab-scale Model Benchmarking A B C D E F

• Determining / Adjustment of N2O Model

Parameters

• Sensitivity Analyses
• Identify measurements / experiments for better

determining default/site-specific parameter values Unified Model(s) Full-scale Model Benchmarking

Task Group on GHG

N2O Model Benchmarking Framework

Snip (DTU) SANITAS ESR9 – N2O ASM Castro (UGent) SANITAS ESR10 – Granular Sludge Anammox

Sustainable and Integrated Urban Water System Management

Qualitative N2O Risk Model Development

Table 1. Operational parameters considered in AS N2O Risk Model Process/Condition Parameter Signal Potential Mechanisms References for Mechanism/Parameter Nitrification Internal Recycle Anoxic/Oxic transitions Delta DO between reactors

• High DO non-limiting

OUR increased NO2 NO2→N20 High NH4 non-limiting low DO AOB switch to NO2 NO2 OUR DO Internal Recycle Rate XQ Excessive NH4 Loading High NH4 NH4 AOB nitrification Chandran et al., 2011 AOB nitrification / denitirification Yu et al., 2010; Chandran et al., 2011 Foley et al., 2010 Kampschreur et al. 2009; Foley et al., 2010 Ahn et al., 2011 AOB denitrification AOB nitrification Ahn et al., 2010 AOB switch to NO2 for electron acceptor Kampschreur et al. 2009 AOB nitrification increased NO2 in anoxic Denitrification Kampschreur et al. 2009; Foley et al., 2010; Ahn et al., 2010; Kampschreur et al. 2009; Foley et al., 2010; Ahn et al., 2010; Kampschreur et al. 2009 NO2 increased NO2 low COD/N high DO influent COD/N DO AOB denitrification - incomplete denitrification limited carbon substrate, incomplete Hetero denitrification inhibition Rapid Process Changes Spikes in NH4, flow, swings in COD:N Delta AOB nitrification / denitirification Kampschreur et al. 2009; Foley et al., 2010

Work in progress

Sustainable and Integrated Urban Water System Management

Qualitative N2O Risk Model Development

1 2 3 4 5 6 7 8 10 9 11 Assign Individual Parameter and Overall N2O Risk Denitrificaiton Nitrification Process Conditions Conceptual N2O Decision Tree Structure

Sustainable and Integrated Urban Water System Management

Next Steps in N2O Risk Model Development

• Assign values to parameters and implement fuzzy logic rule-

based system

• Implement in BSM2 to compare results with mechanistic

representation of N2O production by (Guo et al., 2013; Flores- Alsina et al., accepted)

• Validate with full-scale data
• Investigate pathway identification

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Qualitative / Mathematical Modelling Framework for Assessing Integrated UWS GHG Emissions

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

UWS GHG Sources

CH4, N2O, CO2 CH 4 River Water Supply & Treatment N2O, CH4

Sustainable and Integrated Urban Water System Management

Integrated UWS Modelling

UWS Deterministic Modelling tools have done well to address complexity and dynamics

• Buschets et al., 2010 IUWS Numerical/Qualitative Modelling for

WQ

• Murla et al., 2013 IUWS Numerical/Qualitative Modelling for WQ
• Benedetti et al., 2013 IUWS Numerical Modelling for WQ
• Guo et al., 2013 Sewer / WWTP Numerical Modelling for GHG
• No IUWS (sewer, WWTP, river) GHG models

Objective: Extend IUWS modelling capabilities and fill gaps to include system-wide qualitative GHG assessment for added decision support

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

UWS GHG Modelling Gaps

Sewers

• SewEx Model robust for H2S and CH4 modelling but only for rising

main (pressure) sewers

• Big gap in gravity sewers
• De Graaff et al., 2012 found significant CH4 emissions in

Amsterdam gravity sewers

• No valid models for GHG but significant opportunity to

leverage knowledge from hydraulic models and previous sewer WQ modelling efforts

• Gudjonsson et al. 2002, measured and modelled DO in

Danish gravity sewers and found anaerobic conditions for significant periods of time – Diurnal and temperature effects

• CH4 potential exists for gravity sewers based on

literature

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

UWS GHG Modelling Gaps

Rivers

• No GHG model, but ample opportunity to leverage knowledge from

river WQ models

• Beaulieu et al. (2011) found nitrification and denitrificaiton contributing

to N2O emissions three times greater than IPCC river N2O emission factor

• N2O potential exists based upon literature

Sustainable and Integrated Urban Water System Management

Source: Benedetti et al., 2013

UWS Modelling for Assessing Risk of GHG Emissions from Integrated Control Strategies - Eindhoven Case

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Proposed GHG Measurements in De Dommel River to assess effects of WWTP on GHG emissions

GHG Measurements WWTP Discharge

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Proposed Measurements in the Eindhoven Sewer to Characterize CH4 and H2S emissions

Diagnostic Sewer Measurements WWTP

Sustainable and Integrated Urban Water System Management

Next Steps in IUWS GHG Risk Model Development

• GHG Knowledge Acquisition in river and sewer
• Sewer – DO, CH4 liquid and gas, H2S online, flow, sediment depth
• River – DO, N2O lqiuid and gas, CH4 liquid and gas, NH4, NO2
• Collaborate with ongoing ULaval river WQ monitoring efforts
• Develop CBS/KBS risk assessment models for river and sewer

linking field and model data to measured emissions

• Develop more comprehensive sewer measurement campaign

leveraging knowledge acquisition from Diagnostic Campaign

• Implement N2O Risk Model for WWTP and validate with

measurements and mechanistic model by Guo et al., 2013

• Develop mathematical / RB model approach to assess overall

UWS GHG risk from integrated strategies using IUWS model developed from Kallisto Project

Sustainable and Integrated Urban Water System Management

Concluding Remarks

• Qualitative modelling has clear role in UWS Decision Support
• AS Risk Model adds value to control strategy benchmarking as

seen in BSM1 results and potential in Terrassa WWTP and

• ther WWTPs
• Qualitative Risk Assessment Model Concept can be extended

for assessing UWS GHG Risk and additional Decision Support for sustainable and integrated UWS management

• Qualitative efforts and knowledge acquisition will be

leveraged for continuing and complimenting work in describing mechanistic pathways and quantification of GHG from emissions

Sustainable and Integrated Urban Water System Management

Laboratory of Chemical and Environmental Engineering

Thank You