Improved Network Operations Through Computer Modeling and Support System

Transcription

1 Improved Network Operations Through Computer Modeling and Support System Lindell Ormsbee, P.E., Ph.D., Sebastian Bryson, P.E. Ph.D. Scott Yost, P.E., Ph.D University of Kentucky Abdoul A. Oubeidillah, Ph.D., Andrew N.S. Ernest, Ph.D., P.E., Joseph L. Gutenson University of Alabama Jim Uber, P.E., Ph.D, Dominic Boccelli, Ph.D University of Cincinnati Srini Lingireddy, Ph.D, KYPIPE, LLC. 1

2 Workshop Agenda 1. Project Overview (UK) 15 minutes 2. Overview of Project Website (UK) 15 minutes 3. Overview of Visualization Models (UK) 30 minutes Graphical Flow Model Pipe Break Model 4. Off-line network modeling (UK) 10 minutes 5. Overview of SCADA (UK) 25 minutes 6. Overview of Sensor Placement (UK) 25 minutes Sensor Placement Model 7. On-line network modeling (UC) 60 minutes Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit (UA) 30 minutes 9. Wrap up discussion 30 minutes 2

3 Workshop Agenda 1. Project Overview/Objectives 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off-line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 3

4 Project Goal To assist water utilities in improving the operation of their water distribution systems through a better understanding of the impact of water distribution system hydraulics and flow dynamics on operational decision making: Normal operations Emergency operations Natural events Man made events Research Knowledge Tools

5 The potential exists for real time on-line network models to produce a more sophisticated event detection filter SCADA On-Line Network Model Hyd. Model WQ Model Measured Chlorine + Estimated Chlorine Event Detection filter Prediction Error Hydraulic Sensor (P, Q, Pump Status, ) Quality Sensor (Cl, Sp. Cond., TOC, ) 5

6 Real Time Operations On-Line Water Quality Model Water Distribution System Operations Hierarchy On-Line Hydraulic Model Supervisory Control and Data Acquisition (SCADA) Off-Line Water Quality Model Off-Line Hydraulic Model Spatial Visualization Model SCADA Database Telemetry/Communication Systems Water Quality Sensors Hydraulic Sensors

7 Project Objectives ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS Develop knowledge and tools to support water distribution system operations Develop a decision support system Operational guidance Operational toolkit 7

8 WDS Operational Objectives Maintain Adequate Pressures Minimize Water Quality Problems Minimize Operational Cost Schedule Maintenance Emergency Response 8

9 Operational Questions How long will it take to fill or drain my tanks under: Normal conditions? Emergency conditions? When and how long should I run my pumps to minimize cost? Can I improve my water quality by changing my operations? How will my system perform if I have to close several pipes? Which valves do I need to close to isolate a pipe break? What are the impacts on pressures and flows? 9

10 Operational Support Supervisory Control and Data Acquisition System Computer Models 10

11 Need for SCADA Why do we need a monitoring/ control system? Normal Operations Pipe Breaks/Leaks Pump/Tank Failures Contamination Events Maintain Energy Efficiency Minimize Operational Costs 11

12 Need for Models Graphical representation of system Network schematic Background map Infrastructure database Customer database Computer analyses 12

13 1. Project Overview Workshop Agenda 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off-line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 13

14 Water Distribution System Operations Website PURPOSE: To assist water utilities with designing a monitoring/control system. Operational Applications Normal Operations Emergency Response Management Water Quality Management Energy Management Event Detection 14

15 WDS Operational Decision Support Tool User Defined Decisional Process User Assisted Decisional Process 15

16 User Defined Decisional Process (Guidance) User Defined Decisional Process User Assisted Decisional Process 16

17 17

18 User Assisted Decisional Process (toolkit) User Defined Decisional Process User Assisted Decisional Process 18

19 WDS Toolkit

20 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 20

21 Spatial Visualization ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS 21

22 SPATIAL VISUALIZATION OF NETWORK COMPONENTS 22

23 SPATIAL VISUALIZATION OF NETWORK COMPONENTS 23

24 Graphical Flow Model User Interface 24

25 File and Program Options Map Setting Options Map Function Menu Map Window Pipe and Node Data Menu

26 Program Command Bar

27 Map Function Menu Map Layout Mode Fixed Mode 1 Cannot move or add nodes Select Multiple Nodes and Pipes Attach a note to the map View data tables Add north arrow to map Undo last map change Map Text Mode Fixed Mode 2 Cannot move/can add nodes Select Nodes/Pipes using a polygon Clear selections Refresh map Views Redo last map change Zoom and pan functions Zoom and pan functions Shift map window functions Shift map window functions 27

28 Network Data Management 28

29 29

30 30

31 You can also use GFM to manage facilities data If you click on the Table icon, the program will take you to a table that contains information about your system 31

32 These are the parameter values that have been read in From the KIA database. However, you can change them if you want. 32

33 These results can also be viewed from the Map menu. In order to turn on a particular set of data, first go to the Map Settings Tab and then click on the Labels tab. Now you can use the drop down menu to select a particular type of data. Here we have selected diameter data. To show this data on your map, make sure to click on the box next to the selected parameter (e.g. Diameter as shown) Now, click on the Map tab to take you back to the map viewing area. 33

34 Here is the map view with all the pipe diameters (in inches) shown. 34

35 Hydraulic Model 35

36 Model Results Display Flowrates Display Pressures Display Pressures Contours Print Out Report 36

37

38 Generating a Flow Analysis 38

39 Analyze Error Check Analysis Inventory/Costs Quick Results 39

40 Flow analysis options are provided in a menu after the run completes 40

41 Pipe flowrates displayed by yellow boxes. Nodal system inputs/outputs displayed by blue boxes. 41

42 Pressures at System Nodes Displayed by Labels

43 Pressure Contours (Hatch) 43

44 Result of Screen Capture 44

45 Elements of a Flow Analysis Report 1. File name of the network 2. Regulatory valve data and properties 3. Pipeline data and properties 4. Node data and properties 5. Regulatory valve flow analysis results (upstream and downstream pressure and through flowrate) 6. Pipeline Flow analysis (flowrates) 7. Node flow analysis results (external demand, hydraulic grade, pressure head and node pressure in psi) 8. Summary of system inflows and outflows 45

46

47

48 Pipe Break/Contamination Extent Visualization 48

49 Pipe Break/ Contamination Feature If the system needs to be isolated, using what valve(s)? What is the associated volume of contaminated water? Number of customers? Total demand? What is the associated volume of isolated water? Number of customers? Total demand? What critical pieces of infrastructure are impacted? If the system should be flushed, can the upstream part of the system be used or does the system need to be isolated and flushing water pumped through a hydrant? If so, which hydrant? What will be the final disposition of the flushed water? Where will the water flow? 49

50 Generate a contamination report by a point 1. Click on Facilities Management on the top menu bar 2. Select Contamination (point) from the menu 3. Click on a pipe in the network map in the location of the contamination 4. Click on a valve to expand the contamination beyond the valve 5. Go back to the Facilities Management menu and select Contamination Report 50

51 1 2

52 3 52

53 4. Single-click this valve to see the contamination move north and west in the system. (Note: clicking again on this valve will close it and the contamination will update to the new boundary condition.)

54 Pipes isolated from the system are highlighted green. Volume Summary Box 54

55 5 55

56 Elements of a Contamination Report (Point Selection) 1. File name of the network 2. Date that the analysis was performed 3. Name of the pipe that contains the source of the contaminant (this is defined by where the user clicked to insert the point intrusion) 4. Volume of water contained in the contaminated pipes 5. Volume of water contained in pipes that are not contaminated but are isolated from a source 6. List of valve that must be turned off in order to isolate the contamination 7. List of hydrants that are in the contaminated region 8. Name and elevation of the lowest hydrant in the contaminated area (for use in flushing) 9. List of the lengths of all of the types of pipes within the 56 contaminated region; categorized by material, rating, and diameter

57 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 57

58 OFF-LINE COMPUTER MODELS ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS 58

59 Computer Models Computer models for network modeling have been around since the late 1950s. 59

60 Potential Model Uses Static hydraulic analyses Flows and pressures in the system Fire flow tests Valve closure Dynamic hydraulic analyses Tank turnover Pump operations Emergency response Water Quality analyses Age analysis Chlorine residual analysis Tracer analysis Real Time Analysis Real Time Operations Emergency Response 60

61 61

62 OFF-LINE COMPUTER MODELS 62

63 OFF-LINE COMPUTER MODEL GUIDANCE Network Analysis Model Development Model Calibration Model Calibration Literature Laboratory Model Calibration Case Study Actual System Calibration Case Studies Model Application Examples of Model Applications Model Selection EPANET KYPIPE

64 64

65 65

66 66

67 OFF-LINE COMPUTER MODEL GUIDANCE Network Analysis Model Development Model Calibration Model Calibration Literature Laboratory Model Calibration Case Study Actual System Calibration Case Studies Model Application Examples of Model Applications Model Selection EPANET KYPIPE

68 68

69 69

70 70

71 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off-line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 71

72 SCADA ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS 72

73 SCADA 73

74 SCADA SYSTEMS GUIDANCE Potential SCADA Uses SCADA Functions SCADA Survey SCADA Components Supervisory Control Schemes SCADA Implementation Process SCADA Sensor Location SCADA Sensor Placement Decision- Making Sequence SCADA CWS Sensor Placement Optimization Program Inputs SCADA Sensor Placement Guidance SCADA Sensor Placement Software

75 SCADA Functions Data Acquisition (Collection) Data Communication (Monitoring) Data Presentation Equipment Control 75

76 SCADA Components Sensors and Controllers SCADA Interface Units Communications Network SCADA Master 76

77 Hydraulic Sensors Types of Hydraulic Sensors Pressure Sensors Flow Sensors 77

78 Hydraulic Sensors Sources of Hydraulic Sensors: ABB, abb.com Ashcroft, ashcroft.com Holykell, holykell.com Honeywell, honeywell.com Keyence, keyence.com Truck, truck-usa.com 78

79 Water Quality Sensors Types of Water Quality Sensors Chlorine Residual Sensor TOC Sensor Turbidity Sensor Conductivity Sensor ph Sensor ORP Sensor - 79

80 Water Quality Sensors Sources of Water Quality Sensors ABB, abb.com GE, ge.com Hach, hach.com Siemens, siemens.com Emerson, emersonprocess.com Yokogawa, yokogawa.com/us 80

81 Picture of Water Quality Control Station 81

82 Control Equipment (Pumps) 82

83 Control Equipment (PRVs) 83

84 Control Equipment (Control Valves) 84

85 SCADA Interface Units Remote Telemetry Units (RTU1s) Remote Terminal Units (RTU2s) Programmable Logic Controllers (PLCs) 85

86 RTU: Remote Telemetry Unit Field interface unit compatible with the SCADA system language Convert electronic signals received from field sensors (i.e. pressure sensors, tank level sensors, etc.) into protocol and transmit data to the SCADA Master Typically consist of a box which contains a microprocessor and a database 86

87 RTU: Remote Terminal Unit Field interface unit compatible with the SCADA system language Convert electronic signals received from field sensors (i.e. pressure sensors, tank level sensors, etc.) into protocol and transmit data to the SCADA Master Typically consist of a box which contains a microprocessor and a database 87

88 PLC: Programmable Logic Controller Basic alternative to RTU; higher cost Typical components include CPU Memory Control Software Power Supply Input/Output Modules A digital computer used to monitor and control certain aspects of equipment such as motor speed, valve actuation, and other functions 88

89 SCADA Supervisory Control and Data Acquisition 2 Supervisory Control Schemes Control Hierarchical Control Distributed Control RTU PLC Control Control Monitor Instruct Control Monitor Instruct PLC Control Control RTU Monitor Instruct PLC Control RTU 89

90 Control Schemes: Advantages/Disadvantages Hierarchical Control ADVANTAGES: Low cost (RTUs vs. PLCs) Distributed Control ADVANTAGES: Normal operations maintained despite communications failure Potential differences are minimized for scan and transmission rates DISADVANTAGES: Inability to operate in case of communication failure Potential problems from data transmission rates and computer scan rates DISADVANTAGES: High cost (PLCs vs. RTUs) 90

91 Why do we need a monitoring/ control system? Normal Operations Pipe Breaks/Leaks Pump/Tank Failures Need for SCADA Contamination Events Maintain Energy Efficiency Minimize Operational Costs 91

92 EPA Water Security Initiative The Water Security (WS) initiative is a U.S. Environmental Protection Agency (EPA) program that addresses the risk of contamination of drinking water systems. EPA established this initiative in response to Homeland Security Presidential Directive 9, under which the Agency must develop robust, comprehensive, and fully coordinated surveillance and monitoring systems, including international information, for water quality that provides early detection and awareness of disease, pest or poisonous agents. 92

93

94 EPA Water Security Operational Phases

95 EPA Water Security Initiative 95

96 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 96

97 Water Quality Sensor Placement General guidelines Single sensor placement Graphical method Simplified graphical method Placement Software TEVA SPOT KYPIPE 97

98 98

99 99

100 Sensor Location Tool INPUT Number of sensors to be placed in system Mass injection rate of contaminant (1000 mg/min) Duration of injection (1 hr) OUTPUT Optimal sensor placement locations within system Contamination Report * Be sure you are using an Extended Period Simulation (EPS)

101 Sensor Location Tool

102 Sensor Location Tool

103 Sensor Location Tool Press SHIFT Shift + F7

104 Sensor Location Tool

105 Sensor Location Tool

106 Sensor Location Tool

107 Sensor Location Tool 2 sensors placed at optimal locations

108 Sensor Location Tool

109 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 109

110 ON-LINE COMPUTER MODELS ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS 110

111 ON-LINE COMPUTER MODELS 111

112 ON-LINE COMPUTER MODEL GUIDANCE Model / SCADA Integration / Calibration Implementation Strategies and Barriers for Real-Time Modeling Potential Applications of Real-Time Simulation Actual Applications of Real-Time Simulation

113 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 113

114 Operational Toolkit for a Water Distribution System WDS Wizard A Water Wizard Module Abdoul A. Oubeidillah, Ph.D. Andrew N.S. Ernest, Ph.D.,P.E., BCEE, D.WRE Joseph L. Gutenson The Environmental Institute The University of Alabama Project Workshop May 27 th 2014 Northern Kentucky Water District

115 User Assisted Decisional Process (toolkit) User Defined Decisional Process User Assisted Decisional Process 115

116 Decision Support Categories ON-LINE COMPUTER MODELS SUPERVISORY CONTROL AND DATA ACQUISITION (SCADA) OFF-LINE COMPUTER MODELS SPATIAL VISUALIZATION OF NETWORK COMPONENTS 116

117 Wizard Knowledgebases Rule-Bases A set of rules that are stored separately that fire the questions using backwards chaining logic, match patterns and guide the user through different decision trees using the PyKE inference engine. The rules can also be retrieved from Water Wizard online Subsequent fired question are dependent on previous question answers Fact-Base A set of established knowledge information stored separately on a text file or dynamically retrieved from Water Wizard online Question-Base An alternate fact-bases containing placeholders for in information required of the user during a run-time session Outputs Outputs are displayed on the GUI screen and also captured on an output file which the user can print out.

118 Semantic Knowledge Development SCADA Graphic Models SCADA Off-Line Modeling S Real Time Modeling If.Then. Rules

119 Knowledge Representation (intended-usage (intended "System Improvements") (type Design) (simulation steady-state)) (intended-usage (intended "Master Planning") (type planning) (simulation steady-state)) (intended-usage (intended "Sizing Storage Tanks") (type Design) (simulation extendedperiod)) (calibration-criteria (type Planning) (detail low) (reading flow) (nbr-of-readings 1) (accuracy 10) (unit gpm)) (calibration-criteria (type Design) (detail moderate) (reading pressure) (nbr-of-readings 2) (accuracy 2) (unit psi) (accuracy-nbr 90)) (calibration-criteria (type Operation) (detail low) (reading pressure) (nbr-of-readings 2) (accuracy 2) (unit psi) (accuracy-nbr 90)) (sensor-kb (type flow) (kb A flow sensor is a device for sensing the rate of fluid flow. Typically a flow sensor is the sensing element used in a flow meter, or flow logger, to record the flow of fluids. As is true for all sensors, absolute accuracy of a measurement requires a functionality for calibration.)) (sensor (name "Pressure Sensor 2") (function hydraulic)(type pressure) (accuracy 0.98)) (sensor (name "Flow Sensor 1") (function hydraulic) (type flow) (accuracy 0.95)) (sensor (name "Flow Sensor 2") (function hydraulic) (type flow) (accuracy 0.925)) (sensor (name "Depth Sensor 1") (function hydraulic) (type depth) (accuracy 0.95)) (question (ident model-calibrated) (type multi) (valid yes no) (text "Is your model calibrated?"))

120 Operational Toolkit Backward-Chaining (defrule MAIN::display-sensor-info (declare (auto-focus TRUE)) (answer (ident purpose-hyd) (text?hindri)) (sensor (name?dzina) (type?usedfor)) (test (eq?usedfor?hindri)) => (display-sensor?dzina?usedfor)) Backward-chaining engine Pattern matching Display matches Rule needs answer Ask question and submit answer Needed-answer purpose-hyd Identify Question

121 Sensor Information

122 Operational Toolkit Explicit Decisional Response + Backward-Chaining (defrule MAIN::hydraulic-monitoring (declare (auto-focus TRUE)) (model available) (answer (ident sensor) (text no)) (answer (ident monitoring) (text yes)) (answer (ident monitoring-type) (text hydraulic)) => (assert (give-sensor-info hydraulic)) (recommend-action "Please use the Calibration Report for the Hydraulic sensor calibration and choose your sensor type below.")) Model Available Sensors Not Installed Monitoring Needed - Yes Type? Hydraulics BRANCH OUT TO OTHER NODES RECOMMENDATIONS

123 Water Wizard

124 SCADA & Spatial Vizualization

125 Off-Line & On-Line Models

126 Cloud Deployment

127 Workshop Agenda 1. Project Overview 2. Overview of Project Website 3. Overview of Visualization Models Graphical Flow Model Pipe Break Model 4. Off line network modeling 5. Overview of SCADA 6. Overview of Sensor Placement Sensor Placement Model 7. On-line network modeling Real-time predictive analytics: process and benefits Real-time network modeling: NKWD field study 8. Overview of Toolkit 9. Wrap up discussion 127

128 Survey On a scale of 1-9, how important are the following knowledge databases to your operations? Information on sensors and telemetry. [ ] Information on SCADA. [ ] Information on network visualization. [ ] Information on sensor placement. [ ] Information on off-line modeling. [ ] Information on model calibration. [ ] Information on real-time modeling. [ ] What additional content would you like to see added? 128

129 Survey On a scale of 1-9, how likely would you be to use the following models? Graphical flow model [ ] Decon/pipe break model [ ] Sensor placement tool [ ] Real-time operations [ ] 129

130 Survey How useful was the project presentation? What additional information would you like to see? What information, if any, do you feel was not necessary? What changes could be made to make the presentation more informative or useful to your operations? 130

131 Questions? 131