Scientific and Technical Report No. 21 Uncertainty in Wastewater Treatment Design and Operation aims to facilitate the transition of the wastewater profession to the probabilistic use of simulators with the associated benefits of being better able to take advantage of opportunities and manage risk. There is a paradigm shift taking place in the design and operation of treatment plants in the water industry. The market is currently in transition to use modelling and simulation while still using conventional heuristic guidelines (safety factors). Key reasons for transition include: wastewater treatment simulation software advancements; stricter effluent requirements that cannot be designed for using traditional approaches, and increased pressure for more efficient designs (including energy efficiency, green house gas emission control). There is increasing consensus among wastewater professionals that the performance of plants and the predictive power of their models (degree of uncertainty) is a critical component of plant design and operation. However, models and simulators used by designers and operators do not incorporate methods for the evaluation of uncertainty associated with each design. Thus, engineers often combine safety factors with simulation results in an arbitrary way based on designer ‘experience’. Furthermore, there is not an accepted methodology (outside modelling) that translates uncertainty to assumed opportunity or risk and how it is distributed among consultants/contractors and owners. Uncertainty in Wastewater Treatment Design and Operation documents how uncertainty, opportunity and risk are currently handled in the wastewater treatment practice by consultants, utilities and regulators. The book provides a useful set of terms and definitions relating to uncertainty and promotes an understanding of the issues and terms involved. It identifies the sources of uncertainty in different project phases and presents a critical review of the available methods. Real-world examples are selected to illustrate where and when sources of uncertainty are introduced and how models are implemented and used in design projects and in operational optimisation. Uncertainty in Wastewater Treatment Design and Operation defines the developments required to provide improved procedures and tools to implement uncertainty and risk evaluations in projects. It is a vital reference for utilities, regulators, consultants, and trained management dealing with certainty, opportunity and risk in wastewater treatment.
Author(s): Evangelia Belia, Lorenzo Benedetti, Bruce Johnson, Sudhir Murthy, Marc Neumann
Series: Scientific and Technical Report, 21
Publisher: IWA Publishing
Year: 2021
Language: English
Pages: 232
City: London
Cover
Copyright
Contents
List of Contributors
Preface
Acknowledgements
Introduction to the Scientific and Technical Report
Chapter 1: Key concepts of the STR
1.1 Introduction
1.2 Risk
1.3 Uncertainty
1.3.1 Classification of uncertainty
1.3.1.1 Nature of uncertainty
1.3.1.2 Level of uncertainty
1.3.1.3 Location of uncertainty
1.3.2 Separating variability and uncertainty
1.3.3 Sources of variability and uncertainty
1.3.4 Uncertainty analysis approaches
1.4 Incorporating Variability and Uncertainty Analysis in Models
1.4.1 Variability and uncertainty in model steps
1.4.2 Sources of variability and uncertainty in models
1.4.2.1 Model input variability
1.4.2.2 Model input uncertainty
1.4.2.3 Model structure uncertainty
1.4.2.4 Model parameter uncertainty
1.4.2.5 Numerical uncertainty
1.4.2.6 Model output uncertainty
1.4.3 Evaluation methods
1.5 Summary
References
Chapter 2: Uncertainty in wastewater treatment – current practice
2.1 Introduction
2.2 General Approaches for Addressing Uncertainty in Wastewater Treatment
2.2.1 Design guidelines
2.2.1.1 Overview
2.2.1.2 Design criteria
2.2.1.3 Safety factors
2.2.1.4 Reliability and redundancy standards
2.2.1.5 Development of tight contract documents
2.2.1.6 Staffing and monitoring
2.2.2 Statistical methodologies
2.2.3 Scenario analysis
2.2.4 Mathematical modelling
2.3 Addressing Specific Sources of Uncertainty and Variability in Current Design Practice
2.3.1 Addressing sources of variability and uncertainty in flow and load determination
2.3.1.1 Use of historical information to develop design flows and loads
2.3.1.2 Use of per capita flows and loads
2.3.1.3 Screening of influent wastewater data
2.3.1.4 Wastewater characteristics when data are not available
2.3.2 Addressing sources of uncertainty in unit process design
2.3.2.1 Selection of design aerobic solids retention time
2.3.2.2 Selection of design sludge volume index
2.3.2.3 Selection of design denitrification rates
2.3.2.4 Selection of dissolved oxygen concentration in bioreactors
2.3.2.5 Selection of design oxygen transfer efficiency
2.3.3 Addressing uncertainty via effluent permit selection
2.3.3.1 Effluent limits
2.3.3.2 Selection of effluent design criteria
2.3.4 Summary of uncertainty analysis methods in current practice
2.4 Implications of Current Practice on Degrees of Freedom in Engineering Decisions
2.5 Summary
References
Chapter 3: Incorporating uncertainty analysis into model-based decision making – opportunities and challenges
3.1 Introduction
3.2 Incorporation of Safety in Current Model-Assisted Design
3.3 Opportunities of Explicitly Considering Uncertainty and Variability
3.4 Scope and Limitations of Models
3.4.1 Evolution of wastewater treatment modelling
3.4.2 Desirability criteria for models
3.4.3 Example of wastewater treatment plant model limitations
3.5 What Don’t We Know about Dealing with Uncertainty
3.5.1 How conservative are we with the safety factor approach?
3.5.2 How to move from guidelines with the safety factor approach to probabilistic model-assisted design?
3.5.3 Determination of prior uncertainty ranges
3.5.4 Parameter (uncertainty) estimation in systems with poor identifiability
3.5.5 How to adequately deal with biokinetic model structure uncertainty?
3.5.6 Full-fledged probabilistic model-based design
3.6 How Can We Currently Account for Variability and Uncertainty
3.6.1 Accounting for variability
3.6.1.1 Temporal variability
3.6.1.2 Spatial variability
3.6.2 Accounting for uncertainty
3.6.2.1 Uncertainty related to design scenarios
3.6.2.2 Uncertainty related to data
3.6.2.3 Uncertainty related to process modelling
3.6.3 Sensitivity analysis
3.7 Opportunities of Combining Models with Uncertainty – Example
3.8 Summary
References
Chapter 4: Available methods for uncertainty analysis in model-based projects – critical review
4.1 Introduction
4.2 Methods and Literature Review Results Summary
4.3 Assessment of Input and Parameter Uncertainty
4.3.1 Input uncertainty (measurement errors)
4.3.1.1 Overview of statistical techniques used in measurement error detection
4.3.1.2 Error propagation
4.3.1.3 Examples of measurement error detection
4.3.1.4 Multivariate statistical methods
4.3.1.5 Statistical process control and fault detection methods
4.3.1.6 Data reconciliation
4.3.2 Parameter uncertainty
4.3.2.1 Inference vs. confidence regions
4.3.2.2 Application to wastewater treatment models
4.3.2.3 More sophisticated methods
4.4 Assessment of Model Structure Uncertainty
4.4.1 Macroscopic vs. microscopic mixing scales
4.4.2 Unquantified model structure uncertainty
4.4.3 Mathematical methods for quantification of model structure uncertainty
4.4.3.1 Monod growth model
4.4.3.2 Non-linear dynamical and chaotic behaviour
4.5 Propagation of Uncertainty for Model-Based Decisions
4.5.1 Review of uncertainty propagation methods
4.5.2 Discussion
4.5.2.1 Model calibration
4.5.2.2 Sensitivity analysis
4.5.2.3 Design optimisation
4.5.2.4 Computational demand
4.5.2.5 Method accuracy
4.6 Summary
4.6.1 Input and parameter uncertainty assessment
4.6.2 Model structure uncertainty assessment
4.6.3 Propagation of uncertainty in model-based decision making
References
Chapter 5: The DOUT uncertainty analysis methodology – combining models, statistics and design guidelines
5.1 Introduction
5.2 The Inclusion of Uncertainty Analysis in a Model-Based Project
5.2.1 General tasks
5.2.2 Linking process modelling steps and uncertainty methodology tasks
5.3 Bridging Design Guidelines and Steady-State Design with Dynamic Stochastic Modelling
5.3.1 Define project objectives
5.3.2 Select configurations to be evaluated
5.3.2.1 Generation of a set of pre-designs with different levels of safety
5.3.2.2 Screening of pre-designs
5.3.2.3 Preliminary evaluation of pre-designs with dynamic data
5.3.3 Identify sources of variability and uncertainty to be evaluated
5.3.3.1 Input variability and uncertainty
5.3.3.2 Model structure and parametric uncertainty
5.3.3.3 Model numerical uncertainty
5.3.4 Prioritise and reduce sources of uncertainty
5.3.5 Describe sources of variability and uncertainty explicitly
5.3.5.1 Influent variability and generation of input time series
5.3.5.2 Parameter uncertainty
5.3.6 Model set-up and model structure uncertainty
5.3.7 Propagation of uncertainty and variability using Monte Carlo simulation
5.3.7.1 Monte Carlo simulations
5.3.7.2 One-dimensional Monte Carlo simulation
5.3.7.3 Two-dimensional Monte Carlo simulation
5.3.7.4 Pragmatic Monte Carlo method
5.3.7.5 Effluent constituents cumulative distribution generation
5.3.8 Synthesise evaluation metrics (output analysis)
5.3.8.1 Calculation of PONC
5.3.8.2 Calculation of total cost
5.3.9 Communicate results
5.4 Summary
References
Chapter 6: Case studies
6.1 Introduction
6.2 Steady-State Uncertainty Analysis Example: Operation of the Durham WRRF
6.2.1 Project objectives
6.2.2 Conventional design approach using safety factors
6.2.3 Probabilistic design approach
6.2.4 Results and discussion
6.3 Dynamic Uncertainty Analysis Example: Design Upgrade for the Eindhoven WRRF
6.3.1 Project objectives
6.3.2 Generation and screening of steady-state pre-designs
6.3.3 Variability and uncertainty propagation
6.3.3.1 Influent variability
6.3.3.2 Model parameter uncertainty
6.3.4 Quantification of probability of non-compliance (PONC)
6.3.5 Total cost estimates
6.4 Summary
References
Chapter 7: The bigger picture
7.1 Introduction
7.2 Engineering Project Phases
7.2.1 Overview
7.2.2 Regulatory phase
7.2.3 Planning phase
7.2.4 Preliminary (conceptual) design
7.2.5 Detailed design, construction, and start-up
7.2.6 Operations
7.3 Stakeholders
7.3.1 Overview
7.3.2 Regulators
7.3.3 Utilities – owners and operators
7.3.4 Engineers
7.3.5 Public
7.4 Contract Delivery Methods
7.4.1 Overview
7.4.2 Examples of delivery methods
7.4.3 Stakeholder involvement as a function of contract type
7.5 Summary
References
Chapter 8: Perspectives
8.1 Introduction
8.2 Socioeconomics and Applied Mathematics
8.2.1 Socioeconomics
8.2.2 Applied mathematics and statistics
8.3 Accounting for Uncertainty in Projects
8.3.1 Regulatory phase
8.3.2 Planning phase
8.3.3 Preliminary design
8.3.4 Detailed design
8.3.5 Operation
8.4 Alternative Ways of Handling Uncertainty
8.5 Outlook
References
Appendix A: Terms and definitions – application and discussion
A.1 Introduction
A.2 Modelling
A.3 Statistics
A.4 Uncertainty
A.5 Discussion of Terms Often Confounded with Uncertainty
A.5.1 Precision and variability
A.5.1.1 Quantification of precision and variability
A.5.2 Accuracy and uncertainty
A.5.2.1 Quantification of accuracy and uncertainty
A.5.3 Error and residual
A.5.4 Trueness and bias
A.5.4.1 Quantification of trueness and bias
A.5.5 Note on true values
A.5.6 Note on repetitions
A.5.7 Bias, variability and uncertainty: a graphical example
A.5.8 Link between measurement, modelling and prediction
A.5.9 Qualitative model performance criteria
A.5.9.1 Identifiability
A.5.9.2 Generalisation and domain of validity
A.5.9.3 Optimality
A.5.10 Reliability and redundancy
A.5.11 Robustness and resiliency
References
Appendix B: Methods for uncertainty analysis
B.1 Uncertainty Frameworks
B.1.1 Frequentist
B.1.2 Bayesian
B.2 Monte Carlo Simulation
B.2.1 Random sampling and LHS
B.2.2 Introducing correlations between parameters
References
Appendix C: Existing methods for uncertainty analysis in WWT model-based projects – Complete literature search results
C.1 Introduction
C.2 Assessment of Input and Parameter Uncertainty
C.3 Assessment of Model Structure Uncertainty
C.4 Propagation of Uncertainty for Model-based Decisions
C.5 Uncertainty in Wastewater Treatment Plant Operational Control Data and Methods of Addressing in Online Control
C.6 Uncertainty in the Fate of Pollutants in the Environment and Resulting in Regulatory (WWTP Effluent Standards) Issues
C.7 Updated Literature 2011–2019
Appendix D: Application of uncertainty analysis methods – knowledge from other fields
D.1 Introduction
D.2 Review of Uncertainty Analysis Methods in Chemical Engineering
D.2.1 Comparison of chemical engineering with wastewater treatment
D.2.1.1 Background
D.2.1.2 Similarities
D.2.1.3 Differences
D.2.1.4 Comparison in the context of model-based projects
D.2.2 Uncertainty methods used in chemical engineering
D.2.2.1 Methods overview
D.2.2.2 Error propagation analysis
D.2.2.3 Explicit modelling of input and output disturbances
D.2.2.4 Sampling-based methods
D.2.2.5 Stochastic simulation algorithm (SSA)
D.2.2.6 Bounded uncertainty
D.2.2.7 Polynomial chaos expansion
D.2.3 Applicability to WWT
D.3 Review of Uncertainty Analysis Methods in Hydrogeological (Groundwater) Engineering
D.3.1 Comparison of hydrogeological engineering with WWT
D.3.1.1 Background
D.3.1.2 Comparison in the context of model-based projects
D.3.2 Uncertainty methods used in hydrogeological engineering
D.3.2.1 Model structure uncertainty
D.3.2.2 Model identifiability (equifinality)
D.3.2.3 Conceptual model uncertainty
D.3.2.4 Model conditioning
D.3.3 Applicability to wastewater treatment
References
Appendix E: Current practices in different countries
E.2 Current Practice in North America
E.2.1 Planning phase
E.2.2 Design-bid-build contracts
E.2.2.1 Preliminary design
E.2.2.2 Detailed design and construction
E.2.2.3 Operation
E.2.3 Design-build contracts
E.2.3.1 Preliminary design
E.2.3.2 Detailed design and construction
E.2.3.3 Operation
E.2.4 Design-build-operate contracts
E.2.4.1 Preliminary design
E.2.4.2 Detailed design and construction
E.2.4.3 Operation
E.3 Current Practice in Other Countries
E.3.1 Questionnaire
E.3.2 United Kingdom
E.3.3 The Netherlands
E.3.4 Switzerland
E.3.5 Czech Republic
E.3.6 South Korea
E.3.7 South America
References
Index
