Simulation and Visualization of Power Grid Operations with High Renewable Penetration Thomas J Overbye Fox Family Professor of Elect. & Comp. Eng. University of Illinois at Urbana-Champaign overbye@illinois.edu GCEP Seventh Annual Research Symposium October 5, 2011 The authors gratefully acknowledge the financial support provided by the GCEP program
The Present: US Energy Sources About 80% Fossil Fuels 2009 Energy Consumption was down about 5% from 2008 CO2 emissions fell about 7% About 40% of our energy is consumed in the form of electricity, a percentage that is gradually increasing. The vast majority on the nonfossil fuel energy is electric! In 2009 we got about 0.75% of our energy from wind and 0.04% from solar (PV and solar thermal) 2
3 The Present: US Generation Sources Difference between capacity (MWs) and Energy (MWh). Some generation sources have a high capacity factors (nuclear, >90%), while others have relatively low values (oil, <20%) Wind and solar fall in-between, at around 40% for wind The price of natural gas is a key driver of electricity prices 3
The Present: US Generation Sources 2009 US Data Left and bottom figure from US EIA According to EIA, wind and solar provided 2.3% and 0.02% respectively of US electricity in 2010 and 3.2% and 0.04% in 1 st half of 2011 4
The Present: US Transmission Grid Voltages up to 765 kv; Highly interconnected but with some what limited long distance power transfer capabilities 5
The Current Power Grid: Not Quite as Dumb as Some Think GE s If I Only Had a Brain SmartGrid 2009 Superbowl ad; www.youtube.com ISO New England Control Center 6
An Important Consideration as We Move Forward With its user-friendly, plug-and-play design the humble outlet has made the electric grid easily accessible to billions. Yet it is really a simple gateway to the world s most complex machine. As we move forward with the Smart Grid it is important to not lose this simplicity. 7
What Makes the Grid Unique Fast system propagation of disturbances throughout an interconnect. There is no mechanism to efficiently store electric energy: generation must equal load only several seconds of kinetic energy stored no equivalent of busy signal, or holding pattern With few exceptions, there is mechanism to directly control power flow in grid flow is dictated by impedance of lines; loop flow is a significant problem on some systems 8
Renewable Generation Implications Some renewable generation functions as conventional thermal generation Solar thermal, geothermal, biofuels Focus is primarily on wind and solar PV Since the incremental fuel cost is essentially free, these resources are usually operated at maximum power output for the given wind/ solar conditions 9
Power System Time Frames Lightning Propagation Switching Surges Stator Transients and Subsynchronous Resonance Transient Stability Governor and Load Frequency Control Boiler and Long-Term Dynamics; power flow 10-7 10-5 10-3 0.1 10 10 3 10 5 Time (Seconds) 10
Power Grid Disturbance Example Figures show the frequency change as a result of the suddent loss of a large amount of generation in the Southern WECC Green is bus quite close to location of generator trip while blue and red are Alberta buses. Black is BPA. Time in Seconds Frequency Contour 11
Frequency Response for Generation Loss In response to rapid loss of generation, in the initial seconds the system frequency will decrease as energy stored in the rotating masses is transformed into electric energy Solar PV has no inertia, and for most new wind turbines the inertia is not seen by the system Within seconds governors respond, increasing power output of controllable generation Solar PV and wind are usually operated at maximum power so they have no reserves to contribute 12
Governor Response: Thermal Versus Hydro Thermal units respond quickly, hydro ramps slowly (and goes down initially), wind and solar usually do not respond. And many units are set to not respond! Normalized output Time in Seconds 13
Frequency Response for Generation Loss In the minutes after a generation loss changes to generator dispatch schedules; i.e., more permanent changes to make up for the lost generation; restoring the frequency to 60 Hz. Wind and solar PV would not be able to participate in this rescheduling. Load and/or storage could be used Actual System Disturbance 14
Impact of Generator Inertia and Governor Response for Small Case Figure shows inertia determines initial frequency drop rate, and governor speed the recovery The least frequency deviation occurs with high inertia and fast governors 15
Renewable Generation Implications The integration of large amounts (e.g., 50%) of wind and solar PV requires enhanced controls to handle the potential for larger frequency excursions More dispersed renewable resources are less likely to suddenly fail, but can be subject to more prolonged, correlated changes Cloud bank moving in over a region that contains lots of solar PV Rapid decrease in wind over large region 16
Control Implications Possible solutions include Operating renewable generation at values below maximum power output to provide reserves; this helps with governor response but not inertia More controllable load; if response is fast (less than about 2 seconds) this can help with inertia response Modified wind controls to mimic inertia Frequency provides a useful control signal Universally available; because of propagation delays communication based control may be faster 17
Fast Load Controls Any fast control (on the order of seconds) probably requires local or at most distribution level control Local frequency and voltage magnitude are easy to measure but their meanings are quite different Frequency is global to the system and relatively fixed; control response is straightforward (e.g., f < threshold decrease load) Voltage is very local (feeder specific with LTC control); control response is difficult to determine. 18
Pluggable Hybrid Electric Vehicles (PHEVs) The real driver for widespread implementation of controllable electric load could well be PHEVs. Recharging PHEVs when their drives return home at 5pm would be a really bad idea, so some type of Toyota Prius PHEV; image from motortrend.com load control is a must. With V2G car could provide large amounts of power to grid in emergency 19
An Example: Community Energy Storage AEP is investigating local, 240v storage located at distribution transformers, shared among several houses Devices along feeders could be networked. Cost forecast over the next five years should come down to $500/kWh (25K for 50 kwh of storage) Source: www.aeptechcentral.com/ces/cespresentations.htm 20
Wide Area Visualization As power grid operation becomes more complex with the integration of less controllable and more variable renewable resources, coupled with more controllable load, wide area power grid visualization will become even more important Two of the four causes of the 8/14/03 blackout were due to lack of situational awareness San Diego Union Tribune reported on 9/23/11 that Grid Operators didn t share info during blackout. 21
August 14 2003 Wide Area View 22
San Diego Blackout, 9/8/11 7890 MW Interrupted SONG S Encin a Path 44 Tallega Old Town Mission San Diego Co. MEF Palomar Escondido Centra l Imperial Co. Substation, 230 kv and 500 kv Power Plants 230 kv lines El Centro Outage started here Arizo na SWPL Miguel Otay Mesa Larkspur Imperial Valley North Gila Tijuana La Rosita Mexico Image Source: Slides from CPUC Public Agenda 3281, 9/22/11 *Map not to scale 23
Questions? 24