Use of integrator systems (operational data and model management platforms) to enhance model performance and value.
Presented at the CIWEM Urban Drainage Group Annual Conference 2018
Richard Dannatt - Intertek
Steve Flood - DHI
TechTAC® CFD Report Summary: A Comparison of Two Types of Tubing Anchor Catchers
Models Done Better... - UDG2018 - Intertek and DHI
1. USE OF INTEGRATOR SYSTEMS
(OPERATIONAL DATA AND MODEL
MANAGEMENT PLATFORMS) TO
ENHANCE MODEL PERFORMANCE AND
VALUE
(MODELS DONE BETTER…)
RichardDannatt– Intertek Energy andWater Consultancy Services
Steve Flood - DHI
UDG Annual Conference, 2018
2. AIMS – FOR
THE NEXT 20
MINUTES
2
Describe the need and value in water quality modelling
Describe methods and drawbacks in applications to date
Describe the value in integrating tools and data more effectively,
and how this can be useful
Describe the process for developing planning asnd design
models and data into operational models with increased value
Show where we are in this process, and how we plan to develop
it
3. BACKGROUND
3
Water quality underpins a great deal of the legislation that we
design industry assets to.
Water quality in many ways equates to health –the need behind
much of our water engineering
Modern water quality studies integrate different models, field
data, monitoring data and meteorological data to provide
detailed understanding of asset impact, and the justification and
scope of investment and improvement of assets
The data generated by these studies also provides significant
future value, including future capital planning, operational
forecasting and warning, contingency, climate change planning,
etc
But we need to make it available and useful
9. WHAT DO WE
NEED?
DATA
9
‘ENVIRONMENT’
•Where are the (non-asset) sources
•What are the (non-asset) sources
•How do they behave?
•Rainfall
•Flows
•Tides
‘ASSETS’
•All water company assets – continuous and intermittent
•Dynamics of their operation – ten year series
•Any interaction with environment
11. WHAT HOLDS
US BACK?
11
Difficulty in models ‘talking’ to one another
Errors –units, outputs, scale, scenarios –there are
plenty
Deliberate partitioning –by license, format, code,
or just learning differences
Legacy of models –maintenance, different
software, etc
Agreed protocols for WQ modelling
16. CENTRALISED
DATA AND
DIGITAL
WORKFLOWS
16
There is a need to ensure that our numerical models
are more than just BIM compliant or compatible
(common data formats, naming conventions, etc).
We must embrace the fundamental principles of
BIM; collaborative working between all (multi-
interdisciplinary) stakeholders of a specific project.
There is a real need for simplified, scalable
workflows to facilitate such collaborative working (to
share data and knowledge).
We need intelligent, model-based processes that
integrate all relevant information required for
planning, design, construction and operation
17. CENTRALISED
DATA AND
DIGITAL
WORKFLOWS
17
Planning & Design
Operation
We need intelligent, model-based processes that
integrate all relevant information required for
planning, design, construction and operation
19. CENTRALISED
DATA AND
DIGITAL
WORKFLOWS
19
What’s needed?
• All relevant data and processing in one system
• Closer integration of data and models
• Combinations of models from different
sources/vendors
• Openness and flexibility are key
Elevating data to Information:
24. CHANGING
THE WAY WE
INTERACT
WITH MODELS
24
User Interfaces:
OUTPUT
REPORTING
DESIGN DATA
APPLICATION
PROGRAMMING
INTERFACE (API)
FORECASTING
REAL-TIME
CONTROL
OPTIMISATION
Application Programming Interfaces:
25. EXAMPLE:
CENTRALISED
DATA, MODEL
MANAGEMENT
& ANALYTICS
PLATFORM FOR
UNITED
UTILITIES
25
Vision
To establish a centralised, scalable data and model management
system for collaborative, future proof water quality assessments
Where are we today?
Proof of concept (prototype – groundwork for sequential
coupling)
Where are we going next?
Apply the prototype to a full suite of cross domain models
(network, river and coastal waters)
Where can we go?
Real-time analysis (spills), optimisation and control (MPC)
26. What’s needed?
• Existing models
• ICMExchange
• MIKE OPERATIONS
EXAMPLE:
CENTRALISED
DATA, MODEL
MANAGEMENT
& ANALYTICS
PLATFORM FOR
UNITED
UTILITIES
26
Prototype workflow:
GET NEW RAINFALL OR HYDROGRAPH
BOUNDARY CONDITIONS
PROCESS DATA
RUN INFOWORKS ICM MODEL WITH NEW
BOUNDARY CONDITIONS
EXTRACT RESULTS FROM INFOWORKS ICM
AND PROCESS
VISUALISE RESULTS IN MIKE OPERATIONS
RUN MIKE 21/3 FM WITH EXTRACTED
AND PROCESSED RESULTS
32. CONCLUSIONS
32
There is a wealth of tools and data to be used
Many good applications at present, but lack
efficiency and flexibility
Methods to integrate diverse tools, models and
data would significantly enhance current practice
Test cases are currently being developed, in order
to test and demonstrate potential
We hope to report back in 2019 on the
development of a working system