Integrated System Models Graph Trace Analysis Distributed Engineering Workstation Robert Broadwater dew@edd-us.com 1
Model Based Intelligence 2
Integrated System Models Merge many existing, models together, relating all measurements in context of ISM Community Model Automated Analysis Attach SCADA, all customer load, weather, outage, and other measurements Re use model from planning to training to real time analysis to real time control Push algorithms to data 3
Distributed Engineering Workstation Solves Integrated System Models Has solved transmission distribution system models with more than 3,000,000 nodes and 3,000 complex, multiphase loops Uses Graph Trace Analysis for algorithms Object oriented approach to topology management Time Series Analysis Loads, weather, generation ( NREL interface, Clean Power Research interface), time varying LMP costs 4
ISM as Living Model 5
GTA: Generic Programming Roots Algorithms that process objects in container, independent of object type Container with Objects Iterators CS Algorithms Generic Programming Ex: Sorting algorithms Algorithms that process edges or components of graph ISM Topology Iterators Engineering Algorithms Graph Trace Analysis Ex: Flow simulation Generic analysis code independent of system type (eg, electric, gas, etc.) 6
Matrix Analysis with Edge Node Graph 2 1 2 3 1 3 Graph Trace Analysis with Edge Edge Graph Topology Iterators 1 3 2 Global View N o d e s 5 4 Transform 4 Edges 1 2 3 4 5 1 1 0 1 0 1 2 1 1 0 0 0 3 0 1 1 1 0 4 0 0 0 1 1 Computer Processing Local View 5 4 Edge knows neighbors Topology continuously maintained Algorithms with topology iterators Object oriented topology 7
GTA Based Power Flow 8
ISM Analysis Architecture Mass Storage Memory ISM edge-edge topology Living Model Customer Loads ISM SCADA Measurements Weather Measurements Interface provided by ISM to applications App 1 App 2 Topology iterators, sharing of results, measurements Push algorithms to data 9
Real World Versus Simplified Models The best equivalent is no equivalent Avoid scenario based solutions use scenarios in testing, not control algorithms 10
Relation of ISM to GIS Used to clean GIS data, especially non physics based GIS modeling 11
ISM Model Management for Distributed Environment Measurements Maintained in memory ISM Model Client: Fault Location ISM Model Server Supports distributed computations Model Queue ISM Model Client: Reconfiguration Analysis Processes 12
Analysis Automation Example: Probabilistic Risk Assessment Assembly line computing 13
Collaborative Solutions 14
Peak Load versus Time Varying Design Time varying designs provide better foundation for automation More capacity for automation to work with Time varying designs provide significant improvements in efficiency For medium sized utilities, $1,000,000s / year For large utilities, $10,000,000s / year 15
DEW Programming Interfaces C++ programming interface Used for over 15 years by graduate students Automatic creation of starter application Programmer can focus on calculations Matlab interface Can be used for modeling controller dynamics, such as PV controls, or for modeling power plant dynamics 16
MODEL CENTRIC DMS 17
Model Centric DMS Architecture 18
DEW Control Agorithms DSR control optimization for transmission Standard capacitor, voltage regulator, and LTC controls, including time delays PV controls, including voltage error feedback, voltvar control, and volt watt control Coordinated control with 3 modes CVR, feeder efficiency, and maximum capacity DG control Automated reconfiguration 19
DEW Real Time Controls SCADA interface with OPC client Used for supervisory control of distributed generation Used for coordinated control of capacitor banks (including multi step, individually phase controlled), LTCs, and voltage regulators Fully automated reconfiguration for restoration with fault re isolation and location Runs on blade computers, automatically distributing calculations 20
Model Centric Smart Grid Phase 1: Develop architecture and analyze costs/benefits of alternatives by using simulation Phase 2: Test design concepts and integration in lab setting, using production systems as much as possible Phase 3: Test design concepts and integration in field pilot Phase 4: Gather measurements from pilot project and compare against predicted analysis values 21
Model Centric Smart Grid Phase 5: Use simulation to plan most cost effective system implementation Phase 6: Implement throughout system Phase 7: Gather measurements from system implementation and compare against predicted analysis values Above phases should be iterated 22
VERIFICATION & VALIDATION 23
ORU ISM Validation Experiences Identified SCADA measurements failed 50% low Helped located failed controllers Prediction of annual system losses calculated from 8760 power flow within 0.4% of measured value
Phase Balancing Validation 25
Verification IEEE Standard Transmission and Distribution models Verification of line impedance calculations Verification of power flow calculations Verification of fault analysis calculations Robustness tests specified by independent consultant Passed 89 tests 26
Validation of Load Research Mount Valley Sub 2977 Res, 349 commercial Entergy no longer sends engineers into the field to capture peak measurement 27
Ameren Reconfiguration Validation Reconfiguration for restoration solution prevented major power outage in 2002 using system model of 50 interconnected feeders and over 2000 sectionalizing devices Since then validated in extensive smart grid lab tests at ORU for real time switching automation 28
CEC Funded NREL Study: Test Circuit 29
CEC Funded NREL Study: Sample Results 30
Fault Location Validation at Detroit Edison 31
Motor Start Validation at Detroit Edison 32
DEW APPLICATIONS 33
DEW Applications Data mapper and circuit builder Customer load attachment Monthly, demand, hourly SCADA data attachment Outage data attachment Weather data attachment Automatic schematic builder All schematics synchronized with GIS model 34
Power flow DEW Applications Solves transmission, radial distribution, lightly meshed distribution, and heavily meshed distribution, and secondary distribution, all in same model Diversified power flow Fault analysis Secondary fault analysis Fault location Validated at Allegheny Power Systems 35
DEW Applications Load research analysis Validated at Entergy and Detroit Edison Customer class profiler Load estimation Takes into account weather conditions Component impedance and admittance Takes into account temperature Planning tool 36
DEW Applications Switching sequence analysis Reliability analysis Reconfiguration for restoration Generic application that runs across critical infrastructure, including electric, gas, and water Monte Carlo analysis DER adoption Cascading power flows Storm response with reconfiguration for restoration 37
DEW Applications Outage display and protective device operation analysis Lightning density and outage analysis Phase balancing design Capacitor design Protection/coordination design Phase prediction Used at Detroit Edison 38
DEW Applications Flicker analysis Secondary equipment optimization Transformer load management Can use either monthly or hourly data 100% accurate at predicting overloaded transformers at ORU Contingency analysis Reconfiguration stressor Contingency analysis with reconfiguration 39
DEW Applications DR control DR fuse checker Checks for partial fuse damage DER adoption analysis DER assessment analysis Coordinated control 3 modes: CVR, feeder efficiency, maximum capacity 40
DEW Applications CES battery scheduling Feeder performance analysis Automates system analysis over time varying load Used to compare benefits of alternative designs Distributed series reactance design for flow control of transmission lines 41
DEW Applications Arc flash analysis Measurement based harmonic power flow Revenue flow 42
DEW SAMPLING OF SCREEN SHOTS 43
Auto Schematic Generation 44
Cascading Failure Analysis 45
Display of Capacity Over Google Earth Legend: >3 Required Capacity 2-3 Required Capacity 1.5-2 Required Capacity 1-1.5 Required Capacity 0.8-1 Required Capacity 0-0.8 Required Capacity 0kVA Remaining Capacity (e.g. Since the user entered 1000kVA, the green lines have more than 3MVA remaining capacity) 46
Load Density Over Google Earth 47
Fault Location Results over Google Earth Crew dispatched directly to fault location, eliminating patrol 48
Residential Heat Pump Load Scaling Factor 49
Revenue Flows 50
Storm Outage Forecasts Cumulative Outage Number 450 400 350 300 250 200 150 100 50 Empirical Model Real Time Storm Data Five Hour Forecast with Observer 0 0 20 40 60 80 Time Elapse since Storm Arrival (Hour) Normal Storm Cumulative Outage Number 400 350 300 250 200 150 100 Empirical Model 50 Real Time Storm Data Five Hour Forecast with Observer 0 0 10 20 30 40 50 60 Time Elapse since Storm Arrival (Hour) Abnormal Storm 51
DER Controller Motion Results 52
DER Fault Assessment Results 53
EV/Battery/Solar Analysis 54
Monte Carlo Storm Analysis Results Evaluation of automated system versus manual system for storm response 55
Overview Window Overview window immediately tracks loss of power throughout system 56
Circuit Coloring by Operating Voltage 57
Circuit Coloring by Voltage Ranges Davis 8960 Blue > 118 volts Green < 118 volts Red < 114 volts 58
Lightning Density Analysis Storm 94 300 Number of Flashes inside the corridor above selected intensity 250 200 150 100 50 >10 >20 >30 >40 >50 0 0 50 100 150 200 250 300 350 400 450 Half Corridor Width (feet) Calculates lightning density within corridor around lines
Load Growth Solution: Distribution DGs versus Transmission Lines $125 million transmission solution versus $23 million distribution solution 60
Integrated System-of-Systems Model 61
Zoom in on System of Systems Compartment Fire Electrical Isolation 62