© 2009 IBM Corporation
Advance Water Management
Smart Water
December 2009
Pnina Vortman
IBM Haifa Research Lab
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Organizations will manage their environmental impactby addressing three critical resources:
CO2 emissions contribute to climate change, which impacts water systems.
Energy production results in CO2
emissions.
Water is needed to generate energy
and energy is need to provide water.
WATERWATER
CARBONCARBONENERGYENERGY
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Something profound is happening….
INSTRUMENTED
We now have the ability to measure, sense and see the exact condition of practically everything.
INTERCONNECTED
People, systems and objects can communicate
and interact with each other in entirely new
ways.
INTELLIGENT
We can respond to changes quickly & accurately, getting
better results by predicting and optimizing for future events.
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Water Management – what do we mean ?
When you say ‘water’ there are many issues involved: – Sources: traditional, alternative
• Traditional: Lakes, Rivers, Aquifer, Wells• Alternative: Desalination, Recycled, Types
– Impact on the environment• Must monitor to be able to measure and know• Must be able to control in real-time
– Quality of water• Various quality based on usage : drinking, agriculture, industrial
– Recycling• Water purification for reuse in agriculture and industrial• Water purification to protect the underground water and rivers• Water reuse for residential use
– Wastewater• Protecting the environment• Monitor to prevent contamination
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Water Sources
Rivers and lakes
Underground water
Desalination
Water sensors and meters
Integrated View of Intelligent Water Management
Co
lle
cti
on
Asset Management with QoS Extension
Real-time Analytics
Historical Analytics
Water Network Design
Optimized Leakage ReductionOptimized Electrical PowerOptimal Water BlendingProactive Water Maintenance
QoS Specification
Purification plant
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Alternative Sources of Water – Desalination Plants
Must monitor to understand the impact of desalination on the environment. How do we get rid of the brine ?
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Fighting Floods
We need to be able to predict the weather within small areas. How can we help ? Require data collection.
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Fighting Draughts
We need to save water, find alternatives, control usage, predict the weather within small areas and conserve rain water. Optimization, simulation, and forecasting.
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Strategic Water Information Management Platform
Provides automated sensing of physical environments
Provides water quality, quantity, integration, storage, analysis, modeling, and a management dashboard on a local or regional basis
Provides records management and reporting capabilities for regulatory and compliance requirements.
Provides energy usage information to pinpoint energy efficiency opportunities related to water
Provides advanced asset management and advanced meter management capabilities
Provide intelligence to proactively manage risk related to extreme weather events
10
10
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
By measuring, monitoring and managing environmental systems, decision makers will have more knowledge of how their decisions will influence the environment & business.
Managing & Monitoring Water Systems
Paraguay-Paraná River Basin, BrazilParaguay-Paraná River Basin, Brazil
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Smart Water Management system in residential areas
Machine Learning based water management to detect anomalies, analyze trends, understand behavioral models, and provide usage predictions and inputs to optimization.
Monitoring and integrate information from sewer and water systems in real-time
Collect and analyze information from water metering in real-time
Machine Learning based analysis
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Water Analytics for generating Usage Patterns, Detect Anomalies, Predict and Forecast Water Demand
Goals– Understand usage patterns to detect anomalies and predict demand – Distribute energy consumption to minimize cost based on the usage patterns– Adapt water pressure based on usage patterns
Technology solution– Machine Learning and analytics which analyze water usage data from smart
metering and smart valves/pumps generating reports and generating alerts when anomaly is detected
Results– Analytical reports, alerts, and dashboard based on smart metering data to
understand usage patterns, applied clustering analysis and provided prediction– Using GIS, map usage patterns over the geographic map based on household
Scenarios– Water utility command center receives alert, monitors and observe an anomaly in
usage. Message is sent to customer– Customer checks problem and fix it. High customer satisfaction.
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Example: Before and after leakage fixing
Time-of-day analysis shows the leak drowned the actual demand
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Water Leakage Reduction Through Dynamic Pressure Management – with company in Israel
Problem Overview:– Significant amounts of water are lost due to leakage in water networks– The leakage can be reduced by lowering the pressure of the supplied water– However, sufficient pressure must be maintained to provide the required time varying
demand for water flow Solution Approach:
– Minimize leakage by dynamic control of valves so as to reduce pressure while providing required flow• Utilizes Combination of forecasting, hydraulic simulation and optimization
models/techniques– Includes real time monitoring and control infrastructure to constantly adapt based on actual
consumer water flow requirements Benefit: Significant reduction in water leakage
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Pressure Setting Results
6
8
10
12
14
16
18
20
22
24
26
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Non Regulated Pressure
Maximum Regulated Pressure
Regulater Pressure
IBM Confidential
IBM Haifa Research Lab
© Copyright IBM Corporation 2009
Quality of Service in Water Systems
Need to agree on pressure and quality levels of water for residential, commercial, and industrial
Asset management systems manage and store information about every element in the water network systems such as pumps, valves, pools, meters
Asset management systems can be expanded to “remember” incidents and location of every asset in the water network
Can introduce maintenance optimization based on location and state
Can optimize power cost for pumping water Leakage detection combined with GIS system
can assist in minimize network closure Integrated with usage analytics, can control
pressure and adapt dynamically Integrated with quality sensors, can minimize
the contamination expansion
17
17
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
The Water for Tomorrow Project
Combine science-driven conservation with I/T expertise and computing power
Build a novel software modeling framework to simulate the behavior of river basins around the world
Inform policy and management decisions that conserve the natural environment and benefit the people who rely on these resources
Sample questions:- What impact will development have on water availability for
wetlands and farming downstream? - Will clear-cutting a forest in the upper part of a river's
watershed imperil fish stocks local people depend on for food? - Management tradeoffs including crop production, water
balance, water quality, carbon storage, biodiversity and economic metrics.
Data: climate, rainfall, land cover, soil moisture, land use, etc.
Software:- Scenario forecasting tools- Integrative modeling framework- 3-D visualization
18
18
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
The Beacon Institute
Mission: “To create a global center for interdisciplinary research, policy-making and education regarding rivers, estuaries and their connection with society.”
–The goal: turn all 315 miles of the Hudson river into a distributed network of sensors that will collect and analyze biological, physical, and chemical information–Create a river and estuary observatory through a collaborative program of technology innovation and development aimed at advancing science, management, and education–Link models to the river ecosystem to understand phenomena across a range of spatial and temporal scales
19
19
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
Smart Bay—Galway Bay, Ireland
Collaboration with the Marine Institute, Ireland- A research, test and demonstration infrastructure for monitoring and managing aquatic
environmental data such as tidal flow, wave heights, temperature, and phytoplankton via an integrated network of sensors, robotics and computational technology distributed throughout the Bay.
20
20
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
Intelligent Sensor Networks - Ijkdijk
• Dutch “Smart Levee” Test Site
• Collaboration with Ijkdijk Foundation, Deltares & TNO
•6 meters high (coastal levee category)•100 meters long•30 meters wide
Active monitoring via sensor network to gain insight on levee internal stability
21
21
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
Advanced Water Management - Utility Scale
4) Wireless network links meters and sensors to central console
6) Web console acts as a control center for integrating all water data and enabling drill down on any parameter related to water availability, quality or associated energy costs. Alerts generated if necessary. Multiple locations can be rolled up.
3) Data are also drawn from external sources such as historical databases or remote sensing / geographical surveys
1) Data collection may begin upstream from the treatment plant
2) Sensors are deployed on all water inlets and/or outlets to detect key contaminants. Discharge outlets are also metered. Data are also drawn from SCADA systems. (Manual inputs may also be used)
5) Weather, climate and hydrological models provide additional insight into water quality or availability impacts
7) Advanced analytics provides additional predictive capability
All assets of the same type in this area highlighted in blue
Search for all assets of the Type ‘natural channel’ with condition = 3 in this area. Results are highlighted in blue
22
22
IBM Haifa Research Lab
© Copyright IBM Corporation 2008
IBM 200 mm and 300 mm wafer fabs – Water management
Challenge Reduce water consumption (and associated need for
energy, chemicals, maintenance and labor) to reduce operating cost and minimize environmental impacts
Leverage end-to-end data acquisition, storage and visualization techniques to monitor water usage and improve efficiency
Approach Implemented data collection and storage
infrastructure: sensors, IT network and servers Statistical process control techniques used to
continually analyze vast amounts of operational data and present information in efficient, concise interface
IBM’s Green Sigma™ methodology breaks down water usage by process:
– Dashboards convey key process indicators– Identifies process improvements that reduce water
consumption and provide other benefits, such as reduced electrical power consumption, heat recovery, cooling load reduction, process efficiency etc.
ResultsIBM has achieved over $3M in annual savings by reducing water usage by 27% and increasing manufacturing production over 30%