modelling network visualisation
play

Modelling & Network Visualisation Rebecca Harrison Anglian - PowerPoint PPT Presentation

Near Real Time Modelling & Network Visualisation Rebecca Harrison Anglian Water All About Us Who? Where? Anglian Water is the largest water and Our huge region stretches from water recycling company in England and Humber to Thames


  1. Near Real Time Modelling & Network Visualisation Rebecca Harrison Anglian Water

  2. All About Us Who? Where? Anglian Water is the largest water and Our huge region stretches from water recycling company in England and Humber to Thames estuaries, from Wales by geographic area. We employ Buckinghamshire to Lowestoft more than 5,000 people and supply water and water recycling services When? to more than six million customers in the East of England and Hartlepool. Our company has been delivering for What? customers and for the environment for more than 30 years. We operate and maintain Why? 78,000km of sewer mains We supply 4.3 million customers with Because we every drop high quality drinking water, and collect used water from over 6 million customers across our region. And we’re passionate about our people, local businesses and the communities we serve – we support the health and wellbeing of our people, encourage growth and prosperity in our region and we help to future-proof it against the challenges of climate change and a growing population.

  3. The challenges we face Especially acute in the East of England Common to the whole water industry Population and Planning for the Climate change Affordability and economic growth long term customer expectations Environmental Markets, structure and Cyber security risk protection financing of the industry

  4. Project Overview Purpose Objectives Outcomes Improved service • Development of live models To deliver enhanced Reduced internal flooding and migration to the cloud modeling, monitoring and Reduced pollutions • Deployment of level and visualisation capabilities in Reduced Opex cost velocity monitors at critical which will improve our Improved EPA locations Improved benefits realisation ability to effectively • Development of a SaaS capability visualization platform manage performance and Improved capital allocation risk within the network Keep pace with digital and data capabilities

  5. Near Real Time Modelling - Road Map Data Models Surfacing Engagement Identify Needs Validate PoC Outputs 10 models in system What are the benefits the system can achieve Agree the Vision PoC – NRTM on Cloud IUD models NRT modelling tested on Implementation and testing Import new models into Cloud & Data integration. system Benefits Flooding/Pollution Mitigation. Innovate Define Outcomes NRTM Production Strategies Operational support provided. Investigate potential of What are the benefits the Production starts. Cloud Align with ICM Live system could achieve becomes BAU Modelling Strategy Sept – Oct 2019 2019 2017 2018 2020 Monitor Data Data Monitor Data Strategies 55 active locations Integration exercise to add 90+ monitoring locations Align with Weather, Technolog and IRIS data Data Strategy Monitor and Review Telemetry into OTP. Define alerts and thresholds. 70+ EDM Locations Weather Data Monitor Data Notifications Visualisation PoC Integration Strategies Data received via FTP. Data received via API. Framework. Emails to OMC and Ops Web based system Test visualisation options for ICM Align with Live at production stage and No process. Best endeavours (WWNet TM ) to share data across Visualisation Strategy deploy to wider business business units. Telemetry Weather Data Data received via csv exports. No Data received via API. Fully process. Data issues. automated. Framework. Strategies Align with Communication Strategy

  6. Development Approach Multi-skilled project team • Domain Experts • End-users (operations and modellers) • Technical experts • IS Project Management • Client Side Project Management • Supply chain collaboration

  7. Near real time modelling “A hydraulic model is a digital representation of our sewerage network where we can simulate flows moving around our systems and try replicate any problems we may face” Predictive…….Analysis…….Alerting ✓ Our customers should not be used as a sensor to tell us when things go wrong.

  8. Hydraulic Model Hydraulic modelling used to forecast issues both Capacity and Other causes (flooding/pollution) therefore enabling actions to either mitigate or prevent incidents. Hind cast Forecast They are run every 3 hours in Dry weather and up to 1 hour in wet weather. They are run with a 6 hour hind cast and 6 hour forecast period (typical). Each model run will generate outputs as well as alerts which are then emailed to Tactical Ops and Collection to enable prioritisation of resources. Review/Deploy Model Run Create/ Alert if Asset to under forecast triggered Monitor Assign an Alert Rainfall

  9. Output Overload The final effluent conduits struggles to discharge into the long sea outfall at times of high tides and rainfall. When this occurs flows can escapes from a manhole on the Esplanade which the pollutes the beach (EA aware). This also creates a potential risk to MOPS due to location of manhole (middle of roundabout). Running the model to predict when the flows in the final effluent will be large enough as well as high tide. Final Effluent Pollution Risk This results in an email alert sent to the Tactical Ops team who then dispatches PSS on site to make the area safe.

  10. Output Rising Main burst This is a temporary example on how we can support either emergency or temporary works as part of capital delivery. The burst main is quite large (1000l/s + & 1meter diameter). While the repair work is on-going, the system is checking for rainfall and flows expected arriving at the TPS. This will allow for site teams to plan for temporary measures in order to cope with the increased flows. Rising Main Burst.

  11. Data - Rainfall All datasets used as part of the model have been uploaded directly into our OTP (Operational Technology Platform) within Azure. Rainfall data (Observed and Forecast) to predict where hydraulic flooding issues are likely to happen. • Data acquired on the OTP and stored in folders • Gridded data files (No DB) • Timing

  12. Data - Monitoring All datasets used as part of the model have been uploaded directly into our OTP (Operational Technology Platform) within Azure. Level & Flow monitoring data – Localised monitoring to prevent blockages that may lead to flooding and pollutions. • Stored in SQL Database accessed directly by ICM Live

  13. Data - Telemetry All datasets used as part of the model have been uploaded directly into our OTP (Operational Technology Platform) within Azure. Telemetry data (IRIS) – Pumps, Wet wells, Overflows, Alarms, Flow… to improve the understanding of the system and help with incident reporting. • Data acquired on the OTP and post processed to suit ICM Live (2min resolution) • Pumps/WW/Alarms/EDM • Data from IRIS is primarily UTC but can also be BST. • Monitor Maintenance/Data quality

  14. Visualisation into AMP 7 https://z2cujf.axshare.com/#g=1

  15. Thank you for listening

Recommend


More recommend