Smart Grid Challenges/Opportunities
Smart Grid
Source: http://antwrp.gsfc.nasa.gov/apod/ap001127.html
SMART GRID
An Overview, Challenges and how (research direction) to address them
Prakash Ranganathan
Talk Outline
� Introduction on SMART GRID
� Why SG?, Characteristics, Background, Transmission, Distribution,
Storage
� Grid architecture: Grid design, stability and control
� Examples p overview:
� Network Optimization,
� Distribution Tree Problem,
� Agent g framework
� Renewable Sources adding to grid
� Security Issues
� Engineering work on SG
� Ckt Breaker, TFM Monitoring stations and Power Quality
� Good Resources on SG
Why the Smart Grid Revolution?
“Running today's digital society through yesterday's grid is
like running the Internet through an old telephone
SSwitchboard” hb d”
* Energy Future Coalition, Reid Detchon, AREVA T & D
Visualizing SMART GRID
Source: US Dept of Energy
http://www.oe.energy.gov/DocumentsandMedia/DOE_SG_Book_Single_P
ages.pdf
Context: Cities with 10 million people
Source: IEEE/Amin, UMN
Energy Consumption/Forecast [GAO]
� Today, fossil fuels (coal, oil, and natural
gas) provide about 86 percent of our
� total l energy consumption, with h the h rest
coming from nonfossil sources such as
nuclear (8 percent) and renewables, such
as hydroelectric y energy gy and
� wind power (6 percent). Overall, the
majority of the nation’s energy
consumption is met by domestic
production. However, imports of some
fuels have risen
Fuel Sources: Otter Tail Power Company
S
Source:
Otter Tail Power Company
In our area, using coal gen. electicity
S
Source:
Otter Tail Power Company
Industry Challenges
� Soaring energy demand
�� Power outages’ outages financial impact
� Green energy takes center stage
�� Electricity prices on the rise
� Aging infrastructure/workforce
Frequency and Impact of major disasters
Historical analysis of US outages (1991-2000)
Summary of August 14 Blackout Statistics
� Reported as affecting 50 million people
�� 60 60-65,000 65 000 MW of load initially interrupted
� Approximately 11% of Eastern Interconnection
400+ Generating units tripped
� Cascading lasted approximately 12 seconds
�� Thousands of discrete events to evaluate
Source: NERC and Joint U.S.-Canada Task Force
Electricity Price on rise
Source: EIA
Electricity cost by state – Jan 2009
Peak hours demand
� While supply and demand is a bedrock
concept in virtually all other industries, it is
one with which the current grid g struggles gg
mightily because, as noted, electricity must
be consumed the moment it’s generated.
Without being able to ascertain demand
precisely, p y, at a g given time, , having g the ‘right’ g
supply available to deal with every
contingency is problematic at best. This is
particularly true during episodes of peak
demand, , those times of ggreatest
need for
electricity during a particular period.
� Source: US DOE
Understanding Electric Power Grid
The electric power grid is a complex
adaptive system under semi-autonomous
distributed control.
It is spatially and temporally complex, nonconvex,
nonlinear and non-stationary with
a lot of uncertainties
uncertainties.
Source: IEEE PES 2009, M. Pipattanasomporn
Growing challenges…
� Increasing demand for energy and electricity
�� Moving towards an environmentally environmentally-friendlier friendlier energy mix
(lower carbon emissions, more renewable energy…)
� Volatile energy gy prices p and critical energy gy losses
� Emerging large regional transmission networks
RAISING THE NETWORK’S COMPLEXITY
Solar Potentials in USA
Source: GE data
Wind Potentials
Small Wind
Bio mass to power
Distributed generation: System Integration
Patterns in Power delivery
Source:DOE [Link]
Building Tx Line is Essential
Adding Renewable increases complexity
� The integration of renewable energy such as wind
farms farms, and plug-in vehicles further adds complexity
and challenges to the various operations, decision
making g and controls at all levels of the power p grid g –
generation, transmission and distribution.
Advanced computational methods are required for
planning, monitoring, control and optimization of
power system operations operations.
NREL’s WinDS (Wind Deployment System
Model)solution )
�� Good for Planning and Placement of renewables
� Does not account for power stability or grid resiliency
� Need Network optimization p approximations
pp
DOE “20% wind energy by 2030" report, DOE/GO-102008-2567.
W. Short, N. Blair, and D. Heimiller, Solar Today, pages 28–29, 2003.
Grid Design
� placement of generations and storage
� intermittent wind as an integrated
capacity i
� accounting for outages and intermittency
in planning p g
Network Optimization problem: Optimal Sparsity
[Chertkov, Smartgrids]
� Grid is represented as graph
�� Nodes in grid are generators, generators storage, storage interconnects or
demand units
� Edges g of the graph, g p , namely y the connections between
nodes, are the transmission lines.
� Seek Optimal Placement of Tx for a given set of nodes.
Net.Opt contd.,
� Load flow approximation with resistive loads
� Ci – effective cost per line
{ cost of building the transmission line over the life time)
� K(y) is the sparse conductivity matrix
� J- current (injection/consumption), y is conductivity of a node
Knowledge of KCL/KVL laws
gamma < 1? For optimal Sparsity
Analysis of Control parameter γ
The cost function models the tradeoff between the cost of the dissipated
power and the cost of the transmission lines.
�� The parameter γ varies from one to zero zero, and models economies of scale scale.
� When γ= 1, the cost is linear: doubling the capacity doubles the price.
� A more realistic model is given by a small γ, where the cost rises abruptly with
the addition of a new line, even if ˜y is very small. Intuitively, we expect that
smaller values of γ lead to sparser networks networks.
� The value of this concrete model is that it enables us to determine the optimal
sparsity quantitatively. The algorithmic approaches are quite different in the two
cases.
�� Wh When γ = 1 1; th the problem bl i is convex and d has h a single, i l easily il computed t d optimum. ti
� When γ < 1, the problem is non-convex and there are possibly many local
minima, which requires the use of simulated annealing type algorithms.
The complexity of the latter model faithfully captures several facets of the
complexity l i of f the h real l grid.
id
Power flow Basics can be applied on a network level [F [Fundamental d t l Electrical El t i l Circuit Ci it concept] t]
CComponent t llevel l
Chertkov Proposal deliverables
T, D, C Domains
� The Transmission Domain carries bulk electricity over power transmission lines over long distances,
connecting the bulk generation to the energy consumption centers of the smart grid. It also contains the
power system substations; the transmission and the distribution substations. It may also connects to energy
storage facilities and alternative distributed energy resources at the transmission level level.
� The Distribution Domain distributes the electricity to and from the end customers. The distribution
network connects the smart meters and all intelligent field devices; manages and controls them through a
two-way y wireless or wireline communications network. It may y also connects to energy gy storage g facilities and
alternative distributed energy resources at the distribution level.
� The Customer Domain is where the end users (home, commercial/building, and industrial) of electricity
are connected to the electric distribution network through the smart meters. The smart meters control
and d manage the h flow fl of f electricity l i i to and d from f the h customers and d provides id energy information i f i about b
energy usage and patterns. Each customer has its own domain comprised of electricity premise and twoway
communications networks. It may also generate, store, and manage the use of energy and the
connectivity with plug-in-vehicles.
� Source: IEEE SMART GRID
According to DOE/Smart Grid
Characteristics
� self-healing,
�� consumer friendly friendly,
� attack resistant,
�� provides power quality for 21st century needs needs,
� able to accommodate all generation and storage options,
�� enables markets and
� optimizes assets and operates efficiently.
Agent Modeling Smart Grid [ieee-PSCE09]
Distribution (Heuristic, Tabu)
� Distributed Tree Problem (DTP)
� DTP takes into account characteristics of the substations and
consumer demand. It also determines the optimal topology of
the network to distribute electrical energy at minimum cost.
� Heuristic Determination of Distribution Trees, 2010 IEEE Transactions on Power Delivery, Victor Parada,
Jacques et.al. IEEE paper
Simulated Annealing Vs Tabu Search
Assumptions made in Dist. Planning
The DTP considered in this paper determines the
primary distribution network using radial topology
for a specific period, assuming that:
� the installation costs and the physical locations of
substations and consumer points are known;
� additional transformers may have to be added to
some power substations;
�� each consumer point has a fixed power demand;
� each consumer point is connected to a substation
along a path that may include other consumer
points;
� there is a set of conductor types available with
known parameters (ampacity (ampacity, resistance resistance, reactance, reactance
and cost per unit length);
� the selected conductors for each path are capable
of transmitting the necessary power flow;
� the sum of the capacities of all substations is
sufficient ffi i t ttosupply l power t to all ll customers; t
� several consumer points may belong to the same
client.
DTP [PARADA et.al]
Security [Cisco] [ Silverline]
� Smart grid technologies better identify and respond to man-made or natural disruptions. Real-time
information enables grid operators to isolate affected areas and redirect power flows around damaged
facilities.
�� O One of f th the most t iimportant t t i issues of f resist i t attack tt k i is th the smart t monitoring it i of f power grids, id which hi h i is th the bbasis i
of control and management of smart grids to avoid or mitigate the system-wide disruptions like blackouts.
The traditional monitoring is based on weighted least square (WLS) which is very weak and prone to fail
when gross errors (including topology errors, measurement errors or parameter errors) are present. New
technology of state monitor is needed to achieve the goals of the smart grids.
Video SVP,/GM SG Solution CISCO
Smart Grid Overview
Engineering work on SMART Grid
Control Strategies
Grid Control
� load balancing
� queuing and scheduling
� optimal power flows
� feeder lines control (In power
engineering, a feeder line is part
of an electric distribution
network, usually a radial circuit of
intermediate voltage) g)
� distribution and switching with
redundancy
Self healing grid
Source: EPRI
Situation Awareness tool (SAT)
Smart Meters
Beginning next month, Houston-based
CenterPoint Energy Inc. is preparing to install
more o e than t a two million o ssmart a t meters ete s ove over five ve
years. During a two-year test of the technology,
consumers were able to call up a Web portal
showing the energy consumption of the
hhome's ' major j appliances. li
Consumers also could calculate energy bills in
different situations: What would be the effect
of keeping the house at 75 degrees in the
summer instead of 65?
What adjustments would be necessary to keep
summer electric bills under $200?
Next, a new generation of smarter appliances
could help consumers curtail energy use and
help utilities reduce pressure on the grid.
Smart Grid Challenges/Opportunities
– Infrastructure for Generation/Transmission/Distribution Systems
– Infrastructure for Smart Customer Interface
– Distribution Automation
– Smart metering improves load models and profiles
– DDevice i monitoring i i and d self‐healing lf h li di diagnostics i
– Communication infrastructure provides opportunities for monitoring
and diagnostics
– Distributed Sensing g and Control
– Alternative Smart Grid Architectures
– Infrastructure Security: Controls, Communications and Cyber Security
– Markets and Policy
– Distributed generation and storage adds complexity
TRANSFORMER
MONITORING STATIONS
� Continuous supervision Exact statements about
operating condition
� Optimized operating and lifetime management
� Early detection of incipient faults
� Avoid failures, outages & collateral damage
� Knowledge of transformer life & lifetime
consumption
� Extended transformer lifetime
� Reduction of life-cycle cost
SMART NETWORK
SUPPORTING DEVICE
IMPROVING POWER QUALITY
� All-in-one Real-time
� Compensation
CIRCUIT BREAKER MONITORING
SYSTEMS (24/7)
� MINIMIZING COSTS AND ENSURING RELIABILITY
Main features
� Monitors, via sensors major components of the circuit breakers
� Analyses information and triggers alarms locally and/or remotely
� Current Harmonics
� Voltage Stabilisation
� Flicker Reduction
� Power Factor Correction
� Load Currents Balancing
� Reaction Time is less than 2ms
� Universal Communication Connectivity
C b li d ll Hi h V l Ci i b k Li T k D d T k G
� Can be applied to all High-Voltage Circuit breakers : Live Tank, Dead Tank, Generator
Circuit Breaker
FUTURE SMART GRID VISION
Today’s Grid Vs SMART GRID
Source: DOE
SMART GRID
Publications/Standards/policies
Top 25 papers
�� http://smartgrid.ieee.org/publications/top-ieee-xplore-
http://smartgrid.ieee.org/publications/top ieee xplore
publications/smartgrid
Standards
� http://smartgrid.ieee.org/standards/approved-ieee-
smartgrid-standards
Policies
� http://smartgrid.ieee.org/public-policy/key-smart-gridlegislation
References
� EPRI, GE, DOE,IEEE
� The Electric Power Grid, Today and Tomorrow, Stringer, Amin
�� http://cdtlnet http://cdtlnet.cdtl.umn.edu/presentations.html
cdtl umn edu/presentations html
� http://www.americanprogress.org/issues/2009/02/grid_101.html
� PHEV, Source of Distributed regulation, Sara
Change language