A toolbox for (re)designing the Urban Water Cycle

Transcription

1 A toolbox for (re)designing the Urban Water Cycle Dr Christos Makropoulos National Technical University of Athens, Greece

2 2 Context Strategic planning for the water industry: rethinking the Urban Water Cycle Strategic planning is about formulating a long term vision and getting it delivered. It can be incremental (asset replacement) or more forward-looking (adopting more distributed solutions or moving into circular economy) But it is always subject to (significant) uncertainty. DO WE SEE IT COMING? CAN WE GUARANTEE OUR ASSETS WILL HOLD?

3 3 The extent and rate of change is significant and we cant afford to do nothing CLIMATE IS CHANGING A LOT AND FASTER THAN WE WOULD HAVE LIKED DEMOGRAPHICS ARE ALSO SHIFTING THROUGH (ALSO) MIGRATION AND THE ECONOMY IS STILL IN TOO MUCH FLUX Supply patterns are affected (both in terms of quantity and quality) Mostly slow change in age distributions but even this is now affected by geopolitical shifts The water sector is always affected by energy prices and related policy and these can change drastically.

4 There are promising interventions (and technologies) that can help Blue green options/nature based solutions/water sensitive city options: integrated solutions for complex problems Integrate blue and green infrastructure to minimize flooding and increase ecosystem services Exploit groundwater not only as source but also as water and energy reservoir Turn waste into resources

5 Do we have the tools to design our systems in this more integrated way? The UWM toolbox The UW Optioneering Tool (UWOT) The UW Agent Based Modelling Platform (UWABM) The UW System Dynamic Environment (UWSDE)

6 Enter UWOT A cloud to sea model of the complete urban water cycle Advantages Use a single model to simulate supply (reservoirs, aqueducts), distribution and consumption (end users), as well sewerage, drainage wwtp and recipient waters. This allows options of reuse, synergies between flows as well as synergies with other infrastructure (energy) to be explored. The focus on demand signals is closely related to social behaviour, hence easier integration with social simulation models. It can therefore act as a major part of a full socio-technical model of the whole system at multiple scales Training, Education and Awareness Scenarios, Futures and Decisions Data and Information Tools and Models Options, Interventio ns and Infrastructu re UWOT IS DESIGNED TO UNDERPIN MUCH OF THIS KNOWLEDGE SUPPORT CYCLE

7 The model starts from the very local Generation of demand signals at the appliance level UWOT component / element Push signal Demand signal to drain used water A connection between two components carries two signals: the quantitative (L/dt) and the qualitative (mg/l). Pull signal Demand signal to bring potable water

8 Household level simulation: potable water Potable water demand signal

9 Household level simulation: green water Green water demand signal Green water tank

10 Household level simulation: stormwater Garden, urban green, pervious areas Runoff signal

11 Household level simulation: wastewater Wastewater signal

12 Moving up: Neighbourhood scale simulation UWOT has a number of components that can help it upscale MULTIPLICATION ELEMENT SERVICE RESERVOIR Multiple identical households can be simulated with this component (the neighborhood should be sufficiently homogenous and the timestep large enough depending on number of houses) It can handle multiply incoming signal with a constant number (or time series) of the representative household. The service reservoir is a signal buffer between water distribution and trunk mains. Receives an irregular demand signal and emits a constant topup demand signal

13 Moving even further up: city simulation Checking the status (quantity, quality) of water storage components The quantity splitter diverges demand signals to resources according to the availability of water. Splitters The quality splitter diverges demand signals to secondary resource if primary recourse quality is below a threshold.

14 City scale simulation (quantity example) As timeseries or simulated by UWOT (see household) Splitter: Demand diversion to service reservoir in case of failure of treatment plant or surface water resource

15 City scale simulation (quality example) Splitter: Demand diversion to boreholes in case of unacceptable water quality in surface water resources Borehole or group of boreholes

16 Water supply management Parametric reservoir operation rule: Optimised for best trade-off between cost & reliability (energy consumption / production)

17 Water supply management: GIS interface Reservoirs Aqueduct s Boreholes Small hydroelectric plants WWTPs Water Treatment Plants

18 Water Demand Management: the case of a new urban development in Athens In the case of Eleonas (Athens) we investigated three alternative urban designs (from a very green to a more traditional dense) and looked at water management options for each one.

19 Assessing the effects of Water Demand Management Reduction of potable water demand through reuse and rainwater harvesting Reduction of peak runoff volume through rainwater harvesting and SUDS

20 Wastewater Reuse: multiple scales Phase 1 Phase 2 Used UWOT to develop a roadmap for wastewater reuse for the City of Athens Looking at different intervention scales (central to local) and different time horizons (in 4 phases) Phase 3 & 4 Psyttalia WWTP Thriassio WWTP Sewer mining

21 Blue-Green Infrastructure: Heat Island Effect Assess the urban heat island effect and help in finding a viable solution to mitigate adverse effects of urbanization and climate change. In this case: sewer mining for irrigation and MAR

22 Blue-Green Infrastructure: Heat Island Effect Evaporative cooling is simulated. In this configuration we use treated wastewater for irrigation (from a sewer mining unit)

23 Backup up by a technology library For each component (from a shower to a treatment plant) there is a table with specs in the tech library (including capital, operational and energy costs)

24 Catchment Management Simulate water flows and combined water uses (irrigation, water supply, hydropower) of Thessaly Estimate optimum allocation of available water. Explore impact of Climate Change Scenarios

25 Energy hydropower Simulation of hydropower generation based on efficiency curves (expressed parametrically).

26 Energy renewable Simulate pump-storage schemes. Pumps powered by wind turbines driven by historic or synthetic wind timeseries.

27 Comparing alternative urban water systems at the city scale under different scenarios Configuration 1: Centralized systems at medium scale (BAU) Configuration 2: Innovative and mixed scale

28 We are able to compare their performance against (rather sophisticated!) scenarios 28 DEMAND SCENARIOS SUPPLY SCENARIOS

29 Explore design approaches (here redundancy) and efficiency (energy consumption) 29

30 Also link to Urban Development modelling (scenarios) House typologies Potable water demand 2200 m 3 /d Urban Growth (Cellular Automata) Model Conventional Innovative Years But also ww generation, drainage etc

31 UWOT is modular and extensible UWOT engine in C (4k lines) and interface in Python 2.7 /Qt (10k lines). Engine has been implemented in: dll (called from UWOT python interface or other applications) Mex (run from within MATLAB) OpenMI Django New components can be easily added (couple of hours of coding). Through OpenMI it can be coupled in run time to other (legacy) software

32 But what about the actors and users? Agent Based Models Missing link between technical and social subsystems: modelling the complete socio-technical water system Modelling individual autonomous actors interacting with the physical system (e.g. water sources) through water infrastructure (reservoirs, aqueducts, WDNs). How are infrastructure decisions made (and what governs them?). How do different actors and users interact in the design and use of the hydro system? How do different policies affect the main actors and their decisions? % of households Subsidy-RH(40%) - Subsidy-WR(0%) Subsidy-RH(30%) - Subsidy-WR(0%) Subsidy-RH(30%) - Subsidy-WR(10%) Subsidy-RH(0%) - Subsidy-WR(40%) Subsidy-RH(10%)-Subsidy-WR(40%) Baseline Scenario (Subsidy-RH(0%) - Subsidy-WR(0%) RWH Time (months) 10 year simulation of the diffusion of alternative water demand technologies to a hypothetical urban population with different subsidizing policies

33 UW system dynamic model Looking one scale further up The SD uses available online hydroclimatic information to assesses whether there is a need to trigger a WDM policy A risk index counts the times within a year when the availability indicator falls below a certain threshold Investigate the correct threshold for policy triggering But also pricing and EPI effects

34 Scenarios, Futures and Decisions Using the tools to develop best options MO Optimisation (on a budget if heavy) But what is optimal? New Robust or Real Options optimisation approaches Ultimate goal: Improve the system s resilience (and identify the way to get there: roadmapping) Woodward et al., 2011

35 Strategic roadmaps Resilience and how to get there! Identify, model, optimise series of phased interventions optimise for robustness/no-regrets against high order socio-economic uncertainties Consider costs & benefits distribution (optimisation rarely does) Keep your options open. Prediction is uncertain especially about the future discharge (m³) 3.10E E E E E E E E E+07 Demand of potable water steps

36 UWOT can provide quantification of impacts for different interventions to the urban water cycle New Data: Distributed Sensors Data Analytics & Mining MO Optimisation (Evolutionary Optimisation on a Budget) Roadmapping (Real Options Algorithms) The Agent Based model can be fed social environment variables and provide it with user behaviors (e.g. to different policies or technologies) System Dynamic Environment (blocks, processes, interactions, cause effects, graph theory) The CA urban growth model to more accurately account for spatial patterns in city changes.

37 Take away message: The UW Toolbox can be used to improve the City s BlueGreenprint Technologies and Design Centralised or decentralised options? What is the impact of Demand Management on Reliability? What is the impact of combined BG on demand? (are there regional answers to this?) What are the optimal operating rules for my system? Would increasing redundancy in the system make a difference and when? Soft Measures What will be the impacts of alternative awareness raising measures to demand management (with ABM) When to trigger a campaign and what its link to other policies will be? What would be the impact of different subsidies to technology adoption rates? (with ABM) Under Scenarios, Policy Mixes and also looking at Energy How would my system behave under different world scenarios? How does it behave under pressure? What are is an optimal mix of hard and soft engineering interventions for sustainable UWM What are the energy/cost/risk implications of decentralised systems? How Resilient can a given mix make my city under a range of scenarios (climatic and socio-economic)

38 A toolbox for (re)designing the Urban Water Cycle Dr Christos Makropoulos National Technical University of Athens, Greece