Distribution System Analysis for Smart Grid Roger C. Dugan Sr. Technical Executive, EPRI Webcast Feb 8, 2011
EPRI Power Systems Modeling/Analysis Group Resource group -- systems modeling, simulation, analysis Consulting services from generation to end-use Resource support for R&D collaborative efforts Transmission planning Operations Distribution planning and operations Substation design Power quality 2
The Smart Grid SG is different things to different people Communications and control Typically not represented in DSA (at present) Distributed Resources Generation, Storage, Demand Response Test Feeders WG has done large induction machines Monitoring Protection Energy Efficiency 3
Smart Grid Features Distributed Resources Generation Renewable Generation Variable sources Energy Storage Demand Response 4
Smart Grid Features, cont d Communications and Control AMI deployed throughout the system High-speed communications to Metering and Controls State Estimation 5
Impact of SG on Distribution System Analysis? What DSA framework is needed to support all features of the SG? Will there be a need for DSA if everything is monitored thoroughly? What could we do if we know more about the system? How will merging of planning, monitoring and DSE change DSA tools? 6
Role of Distribution System Analysis Distribution state estimation Emergency reconfiguration Account for missing data, failed comm EPRI vision Planning and DMS will converge into one set of tools (Real time and planning will merge) Continued need for DSA tools Different form and more capabilities 7
Advanced Simulation Platform -- OpenDSS Open source of EPRI s Distribution System Simulator (DSS) developed in 1997 open sourced in 2008 to collaborate with other research projects OpenDSS designed to capture Time-specific benefits and Location-specific benefits Needed for analysis of DG/renewables energy efficiency PHEV/EV non-typical loadshapes Differentiating features full multiphase model numerous solution modes dynamic power flow system controls flexible load models Download for free from http://sourceforge.net/projects/electricdss 8
Jan 21 Apr Jul 17 Oct 13 9 5 1 Load, MW Computing Annual Losses Peak load losses are not necessarily indicative of annual losses 70 60 50 Year 5 Losses: total 2413 MWh 25000 20000 40 30 15000 10000 kwh 20 5000 10 Hour 0 0 Month 9
Overall Model Concept Inf. Bus (Voltage, Angle) Power Delivery System Comm Msg Queue 1 Control Center Comm Msg Queue 2 Power Conversion Element ("Black Box") Control 10
Supporting Renewables
MW Difference, MW Solar PV Simulation 1-hr Intervals What is the Capacity Gain? 5 5 Without PV With PV 4 Peak is not reduced Difference 4 3 3 2 2 1 1 0 0-1 2 Weeks 12-1
MW Difference, MW Can DMS Enable Increased Capacity? 5 5 4 Shorter Duration Of Peak Without PV Difference With PV 4 3 3 2 2 1 1 0 0-1 2 Weeks 13-1
Per Unit of Maximum Cloud Transients: 1-sec Interval 1-Sec Solar PV Output Shape with Cloud Transients 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 500 1000 1500 2000 2500 3000 Time,s Impact on Feeder Voltage 14
Solar Ramping Basic Solar Ramp Function Voltage Pushed over limit on recovery Regulators compensate for drop 15
Regulator Response for Series of Cloud Transients Regulator Operations 16
Headroom for PV Voltage Profile for 40% Load Voltage Profile for 100% Load Use DMS to Regulate Lower to Allow More Headroom for DG More efficient, too?? 17
per-unit voltage per-unit voltage Steady-State Voltage Y Device Locations / Voltage Maximum change in voltage PV at increased penetration until limit exceeded 189000.00 Substation PV Use of volt/var control accommodates added PV before violations occur 188000.00 187000.00 1.075 0.02 1.05 0.015 1.025 0.01 1 0.005 0.975 Primary 20% Bus PV 10% 15% 20% Voltages with PV- VVC Base Case Voltage Change 0.95 0 0 0 0.5 1 1 1.5 1.5 2 2 2.5 2.5 3 3 3.5 3.5 4 4 4.5 4.5 5 distance from substation (mi) distance from substation (mi) 186000.00 V(1) V(2) deltav(1) V(3) deltav(2) deltav(3) 185000.00 Analysis results from other feeders indicate 25%-100% more PV can be accommodated using VVC 18 Baseline No PV 10% PV 15% PV 638000 640000 642000 X 20% PV 20% PV and VVC V
V (pu) Primary Voltage Response with Volt/Var Control 12 kv Voltage 1.05 20% PV 1.025 1 20% PV w/ volt-var control 0.975 0.95 0.925 Baseline No PV 0.9 0 4 8 12 16 20 Hour 19
Storage
Generic Storage Element Model (OpenDSS Model) Idle Discharge Charge kw, kvar Idling Losses % Eff. Charge/Discharge kwh STORED Other Key Properties % Reserve kwhrated kwhstored %Stored kwrated etc. 21
Controlling Storage from DMS Discharge Mode Charge Mode kw Target Discharge Time Total Fleet kw Capacity Total Fleet kwh et. al. Time + Discharge rate Peak Shaving Load Following Loadshape Substation controller/dms V, I Comm Link Substation Storage Fleet 22
kw Simple Substation Peak Shaving Load Shapes With and Without Storage Mode=Peak Shave, Target=8000 kw, Storage=75 kwh Charge=2:00 @ 30% 10000 80 9000 8000 7000 6000 5000 4000 3000 2000 1000 70 60 50 40 30 20 10 Base kw Net kw kwh Stored 0 0 0 50 100 150 200 250 300 Hours 23
kw Attempting Peak Shave Every Day Load Shapes With and Without Storage Mode=Load Follow, Time=14:00, Storage=25 kwh Charge=2:00 @ 30% 10000 About Right 30 9000 8000 25 7000 6000 5000 4000 3000 20 15 10 Base kw Net kw kwh Stored 2000 1000 Too Early 5 0 0 0 50 100 150 200 250 300 Hours 24
kw Accounting for Storage Losses Load Shapes With and Without Storage Mode=Time + fixed rate, Time=14:00 @ 25% Storage=25 kwh Charge=2:00 @ 30% 10000 30 9000 8000 25 7000 20 6000 5000 4000 15 Base kw Net kw kwh Stored 3000 10 2000 1000 DIscharging Charging 0 0 200 210 220 230 240 250 Hours Charging energy > Discharging energy (compare areas) 25 5
Key Distribution Modeling Capabilities for Smart Grid Distributed generation modeling Time series simulations Efficiency studies Meshed networks Large systems Parallel computing Distribution state estimation Protective relay simulation AMI Load data Modeling controllers Modeling comm Work flow integration 26
Key Challenges Merging Planning and Real-Time Analysis Very Large System Models Systems Communications Simulations Large Volume of AMI Data AMI-based Decision Making Time Series Simulations Distribution State Estimation 27
Key Challenges, Cont d Detailed LV Modeling Including multiple feeders, transmission DG Integration and Protection Generator and Inverter Models Meshed (Looped) Network Systems Regulatory Time Pressures 28
Reference R.C. Dugan, R. F. Arritt, T. E. McDermott, S. M. Brahma, K. P. Schneider, Distribution System Analysis To Support the Smart Grid, presented at 2010 IEEE PES General Meeting, Minneapolis, MN 29
Together Shaping the Future of Electricity 30