Jacobs Solutions Inc. · NYSE:J

Jacobs Solutions — Technical Presentation (June 2021)

50 slides from Jacobs Solutions's technical presentation, Jun 10, 2021

  1. Slide 1: State-of-the-Art Water Supply Protection

    State-of-the-Art Water Supply Protection In the kNOW Webinar Series June 10, 2021

  2. Slide 2: Presenters

    Presenters  Kelly Taylor, Clayton County Water Authority  Lauren Chamblin, Clayton County Water Authority  David Austin, Jacobs  Roger Scharf, Jacobs  Adam Hosking, Jacobs

  3. Slide 3: Source Water Control an Integral part of OneWater

    Source Water Control an Integral part of OneWater  Growing pressure on water resources  Impacts on source water quality  Reductions in dissolved oxygen levels  Source water control the first line of treatment  Predictive technologies to enable proactive management  Supports resilient, integrat

  4. Slide 4: Agenda

    Agenda  CCWA background and the development of a resilient raw water supply  Reservoir Management system and results  Proactive monitoring  AI for Predicting Taste and Odor

  5. Slide 5: Clayton County Water Authority Background

    Clayton County Water Authority Background

  6. Slide 6: Who are We?

    Who are We? Clayton County Clayton County Water Authority  Created in 1955 /Governed by Board  Serves 289,000+ people in 6 cities and unincorporated area  ~380 employees across 12 departments  Part of metro Atlanta  Home to world’s busiest airport Water Sewer Stormwater  7 municipalities  295

  7. Slide 7: CCWA Services for our Community

    CCWA Services for our Community Water Sewer Stormwater 5 Raw Water Reservoirs 3 Water Reclamation Provides drainage 3 Treatment Facilities Facilities structures to help protect 38.4 MGD Treatment property from flooding, 42 MGD Capacity Capacity sinkholes 1500 Miles of Pipeline 1400 Miles of Pipeline

  8. Slide 8: The History of Indirect Potable Reuse at CCWA

    The History of Indirect Potable Reuse at CCWA  River Basin Headwaters – limited  1980 – Built Land Application  1989 - Blalock Reservoir water supply System constructed  Flint River water quality drove  More than 2,400 forested acres  2000 - Master Plan addressed innovation  70% of reclaimed

  9. Slide 9: CCWA System is complex and integrated

    CCWA System is complex and integrated 9 ©Jacobs 2021

  10. Slide 10: The benefits are sometimes met with challenges

    The benefits are sometimes met with challenges 10 ©Jacobs 2021

  11. Slide 11: How did we get to a Reservoir Management System?

    How did we get to a Reservoir Management System?  Projects initiated in 2016  Oxygenation system recommended for Blalock/Shamrock  Large taste and odor event accelerated activities  Reservoir improvements online since August 2019 11 ©Jacobs 2021

  12. Slide 12: Reservoir Management Systems

    Reservoir Management Systems

  13. Slide 13: High Nutrient Source Water Issues

    High Nutrient Source Water Issues High algae, Temperature and pH Epilimnion Thermocline Intake Pipe High phosphorus, iron, Hypolimnion Low Oxygen manganese 13

  14. Slide 14: Reservoir Restoration Philosophy

    Reservoir Restoration Philosophy 14 ©Jacobs 2021

  15. Slide 15: Shamrock/Blalock: Linear Diffuser Oxygenation

    Shamrock/Blalock: Linear Diffuser Oxygenation  Diffuse oxygen into reservoir hypolimnion  Meet oxygen sediment and water column oxygen demand  Oxygen fed from liquid oxygen tank 15 ©Jacobs 2021

  16. Slide 16: Shamrock/Blalock: LOX Storage and Feed

    Shamrock/Blalock: LOX Storage and Feed  Liquid oxygen (LOX) storage  9,000-gallon horizontal tank  Ambient air vaporizers – Alternate every 12 hours for de-icing  Gaseous oxygen mass flow control valves – 1 Valve for each reservoir  LOX flow based on operator set point 16 ©Jacobs 2021

  17. Slide 17: Shamrock/Blalock: LOX Diffuser Layout

    Shamrock/Blalock: LOX Diffuser Layout Key Oxygen Diffuser In Lake Supply Trenched Supply Destratification Diffuser 17 ©Jacobs 2021

  18. Slide 18: What Results Do We See?

    What Results Do We See?

  19. Slide 19: Minimize Hypolimnetic Anoxia to Quench Algae Growth

    Minimize Hypolimnetic Anoxia to Quench Algae Growth Blalock Reservoir, Dissolved Oxygen (mg/L) 2019/03/04 thru 2019/04/03 (Past 30 Days) 10 Reservoir systems come online 770 8 Eleveation (feet) 6 765 4 760 2 0 3/8 3/13 3/18 3/23 3/28 4/2 Date

  20. Slide 20: Metals Release from the Bottom of the Reservoirs

    Metals Release from the Bottom of the Reservoirs 20 ©Jacobs 2021

  21. Slide 21: Phosphorus Release from the Bottom of the Reservoirs

    Phosphorus Release from the Bottom of the Reservoirs 21 ©Jacobs 2021

  22. Slide 22: Levels of Anabaena Are now Closer to Pre-Huie Levels

    Levels of Anabaena Are now Closer to Pre-Huie Levels 22 ©Jacobs 2021

  23. Slide 23: Levels of Geosmin Have Decreased Over Time

    Levels of Geosmin Have Decreased Over Time 23 ©Jacobs 2021

  24. Slide 24: Treatability of Blalock Reservoir Raw Water

    Treatability of Blalock Reservoir Raw Water Huie Site A Reservoir comes online systems come online 24 ©Jacobs 2021

  25. Slide 25: Decrease Customer Complaints of Taste, Odor, and Discoloration

    Decrease Customer Complaints of Taste, Odor, and Discoloration in Drinking Water 26 ©Jacobs 2021

  26. Slide 26: Current Monitoring Practices

    Current Monitoring Practices

  27. Slide 27: CCWA's Response to Challenges includes a Diversity of Measures

    CCWA’s Response to Challenges includes a Diversity of Measures 28 ©Jacobs 2021

  28. Slide 28: Monitoring

    Monitoring Continuous Reservoir Monitoring  Vertical profiler (Xylem/YSI System)  Buoys (Sea Bird System) Weekly EXO1 Sonde Stream Monitoring  Total algae/BGA  fDOM sonde 29

  29. Slide 29: Monitoring

    Monitoring Bi-Weekly Algae Sampling  5 Reservoirs  Water quality grab samples from epilimnion Bi-Weekly MIB/Geosmin Sampling  5 Reservoirs  3 Water Production Plants  2 Ground Storage Tanks Bi-Monthly Boat Sampling  Shamrock, Blalock, and Hooper Reservoirs  Samples from hypolimnion and epilim

  30. Slide 30: AI for Predicting Taste and Odor

    AI for Predicting Taste and Odor

  31. Slide 31: The Challenge

    The Challenge  Quantitative prediction of taste & odor (T&O) and harmful algae blooms (HAB) in raw water has been elusive  Both T&O can be a BIG surprise that causes BIG problems at the water treatment plant (WTP)  Model prediction needs: – Fair warning: prepare at WTP – Prevention: Inform reserv

  32. Slide 32: Model Inputs

    Model Inputs  Continuous Data Sources – Vertical Profiler – Buoys – Local Weather Station  Discrete Data Sources – Taste Odor Samples – Lake Grab Samples – Stream Grab Samples – Wetland Outfall Samples 34 ©Jacobs 2021

  33. Slide 33: Data – Taste and Odor

    Data – Taste and Odor  T&O data spans May 2017 – November 2020  542 geosmin samples  221 2-MIB samples  CCWA upgraded T&O testing capacity in 2019 – Key to training models 35 ©Jacobs 2021

  34. Slide 34: Feature extraction

    Feature extraction  Large number of features available  Model parsimony – reduce the number of inputs to the minimum  Use recursive feature elimination  Select the smallest subset of inputs (features) that maximizes the accuracy 36 ©Jacobs 2021

  35. Slide 35: Feature extraction results

    Feature extraction results Top 5 Features for Blalock Top 5 Features for Shamrock Top 5 Features for Hooper geosmin geosmin geosmin  [Rate of change of Dissolved  [OUTFALL.003.TOTAL.FLOW]  Rate of change of DO with Oxygen with time in Blalock at 8 t= - 3weeks Depth at 12 ft in Blalock ft ] t= - 2

  36. Slide 36: Models used in this work

    Models used in this work Artificial Neural Convolutional Neural Network and Network Recurrent Neural Networks 38 ©Jacobs 2021

  37. Slide 37: Modelling results - geosmin

    Modelling results - geosmin  Data split into training and testing sets  80:20 split  Evaluate on accuracy or error of testing set  Limited T&O events during testing set but single event that occurred was successfully predicted 39 ©Jacobs 2021

  38. Slide 38: Geosmin Modelling results

    Geosmin Modelling results Limited T&O events during testing set but no false positives were generated by the model 40 ©Jacobs 2021

  39. Slide 39: Geosmin Modelling results (cont)

    Geosmin Modelling results (cont) Limited T&O events during testing set but no false positives were generated by the model 41 ©Jacobs 2021

  40. Slide 40: 4 Week Forecast Model

    4 Week Forecast Model  4 individual models Lagged Inputs  Each model can take Model output different inputs up to T=1 week time t=0 Lagged Inputs  Forecast by predicting Model output geosmin at time T=2 weeks t=1…4 Lagged Inputs Model output T=3 weeks Lagged Inputs Model output T=4 Past Present w

  41. Slide 41: 4 Week Forecast

    4 Week Forecast Comparison forecast with lab results Date Forecast Lab Forecast Lab Forecast Lab 2019-12-15 20 14 7 13 2 8 2020-01-12 7 7 90 62 7 7 2020-02-02 7 7 8 5 11 13 1 week 2 weeks 4 weeks Each of the models produced useful results for the forecast window 43 ©Jacobs 2021

  42. Slide 42: Fall 2020 Retraining

    Fall 2020 Retraining  Model retrained in Fall of 2020  Data split into training and testing 80:20 as before  Model picked up the November T&O spike 44 ©Jacobs 2021

  43. Slide 43: Web-Based Platform

    Web-Based Platform 45 ©Jacobs 2021

  44. Slide 44: Summary and Next Steps

    Summary and Next Steps  Relatively accurate geosmin forecasts obtained with a small set of parameters  Forecast potentially very useful tool for drinking water utilities  Future extension to cyanotoxin and harmful algae bloom prediction could be evaluated 46 ©Jacobs 2021

  45. Slide 45: Our State-of-the-Art System Promotes OneWater

    Our State-of-the-Art System Promotes OneWater  We are constantly working to anticipate, adapt, and evolve to protect our water resources  Challenges have presented themselves over time  Monitoring and data analysis  We have improved interdepartmental coordination and understanding  Monitoring a

  46. Slide 46: Polling Question #1

    Polling Question #1  What level of treatment are you able perform on your raw water source? 1. No treatment 2. Occasional chemical application 3. Continuous treatment 48 ©Jacobs 2021

  47. Slide 47: Polling Question #2

    Polling Question #2  How much forewarning of taste and odor spikes or harmful algae bloom onset would your utility need to take appropriate management actions? 1. 24 to 72 hours 2. 1 week 3. 2 weeks 4. Not applicable 49 ©Jacobs 2021

  48. Slide 48: Questions

    Jacobs AUTHORITY Questions Jacobs 2021

  49. Slide 49: Thank You

    yvacobs noyfea

  50. Slide 50: Copyright Notice

    Copyright Notice Important The material in this presentation has been prepared by Jacobs®. ©2020 Jacobs Engineering Group Inc. All rights reserved. This presentation is protected by U.S. and International copyright laws. Reproduction and redistribution without written permission is prohibited. Jacob

Jacobs Solutions — Technical Presentation (June 2021) on 6ix