Sie sind auf Seite 1von 7

Downloaded from http://iranpaper.

ir

http://plcman.vcp.ir
1

Distribution Power Flow for Smart Grid


Technologies
K. P. Schneider, Senior Member, IEEE, D. Chassin, Senior Member, IEEE, Y. Chen, Member, IEEE
and J. C. Fuller, Student Member, IEEE

AbstractSmart Grid technologies hold the promise of being


able to solve many of the problems currently facing in the electric
power industry. However, the large scale deployment of these
new technologies has been limited due to an inability to
accurately model their effects or to quantify their potential
benefits. GridLAB-D is a new open source power system
modeling and simulation environment developed by the United
States Department of Energy specifically to integrate detailed
power systems and end-use models. In order to effectively model
the vast array of possible smart grid technologies GridLAB-D
was developed as a general simulation environment. This paper
describes the basic design concept, the power flow solutions
implemented, and a detailed example of the type of analysis that
can be performed within the simulation environment in order to
support the evaluation of smart grid technologies.
Index Terms current injection method, distribution system
analysis, forward-backward sweep method, power simulation,
power modeling, Gauss-Seidel, smart grid.

I. INTRODUCTION

ridLAB-D is the first of a new generation of distribution


system simulation technologies developed by the U.S.
Department of Energy (DOE) at the Pacific Northwest
National Laboratory (PNNL) in collaboration with Industry
and Academia. GridLAB-D has been developed as an
enabling tool to facilitate the evaluation of smart grid
technologies. Historically the inability to effectively model
smart grid technologies has been identified as a significant
barrier to their adoption; GridLAB-D is an attempt to address
this problem.
GridLAB-D is a complete simulation
environment which incorporates advanced modeling
techniques, high performance computing capabilities, and
integration tools to other existing software platforms.
GridLAB-D is continually under development, by PNNL as
well as additional collaborators, and versions exist that can
currently be downloaded [1]. The most important new
capabilities slated for GridLAB-D include:
This work was funded by the Office of Electricity Delivery and Energy
Reliability of the U.S. Department of Energy (DOE) under contract DE-AC0576RLO1830.
K. P. Schneider, D. Chassin and Y. Chen are with Pacific Northwest
National Laboratory in Richland, Washington 99352. They may be contacted by
email
at
kevin.schneider@pnl.gov,
david.chassin@pnl.gov
and
Yousu.chen@pnl.gov respectively.
J. Fuller is with Washington State University Tri-Cities in Richland
Washington. He may be contacted at jcfuller@tricity.wsu.edu.

978-1-4244-3811-2/09/$25.00 2009 IEEE

Extended quasi-steady state time-series solutions;


Detailed end-use models including consumer
appliances and distribution utility equipment models,
all implemented with the latest agent-based simulation
methods;
Distributed energy resource models, including
appliance-based load shedding technology and
distributed generator and storage models;
Integration of transmission and distribution system
models including unified solvers;
Retail market modeling tools, including contract
selection, business and operations simulation tools,
models of SCADA controls, and metering
technologies;
The ability to run efficiently on multicore and
multiprocessor machines; and
External links to Matlab, MySQL, Microsoft
Excel and Access, and text-based tools, as well as
being able to convert models from the SynerGEE and
Windmil power distribution modeling systems.

GridLAB-D is capable of studying distribution utility


system behaviors ranging from a few seconds to decades,
simulating the interaction between physical phenomena,
business systems, markets and regional economics, and
consumer behaviors. The results include many power system
statistics, including selected reliability indices from IEEE
1366 [2] (e.g., CAIDI, SAIDI, SAIFI), and business metrics
such a profitability, revenue rates of return, and per customer
or per line-mile cost metrics are planned.
GridLAB-D has been validated with both existing end-use
simulation and standard distribution analysis tools such as the
IEEE Radial Test Feeders [3].
II. THE GRIDLAB-D SYSTEM
A. What is GridLAB-D?
GridLAB-D is a flexible simulation environment that can
be integrated with a variety of third-party data management
and analysis tools. At its core, GridLAB-D has an advanced
algorithm that can determine the simultaneous state of
millions of independent devices, each of which can be
described by multiple differential or difference equations
solved locally for both state and time. The advantages of this
algorithm over traditional finite difference-based simulators

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
2

are: 1) it is much more accurate; 2) it can handle widely


disparate time scales, ranging from sub-second to many years;
and 3) it is very easy to integrate with new modules and thirdparty systems. The advantage over tradition differential
algebraic solvers is that it is not necessary to integrate all the
devices behaviors into a single set of equations that must be
solved. The GridLAB-D system also includes modules to
perform the following system simulation functions:
Power flow calculations and device controls, including
distributed generation and energy storage
End-use appliance technologies, equipment and
controls
Data collection on every property of every object in the
system, and boundary condition management including
weather and electrical boundaries.
Additional planned modules are being developed to provide
additional functionality, including:
Consumer behavior including daily, weekly, and
seasonal demand profiles, price response, and contract
choice
Energy operations such as distribution automation,
load-shedding programs, and emergency operations.
Business operations such as retail rate, billing, and
market-based incentive programs.
B. How does GridLAB-D work?
GridLAB-D includes an extensive suite of tools to build
and manage studies, and analyze results. Existing and
planned tools include:
Agent-based and information-based modeling tools
that allow you to create detailed models of how new
end-use technologies, distributed energy resources,
distribution automation, and retail markets interact and
evolve over time.
Tools to create and validate rate structures, examine
consumer reaction, and verify the interaction and
dependence of programs with other technologies and
wholesale markets.
Interfaces to industry-standard power systems tools
and analysis systems.
Extensive data collection tools to permit a wide variety
of analyses.
GridLAB-D is capable of examining, in detail, the interplay
of every part of a distribution system with every other.
GridLAB-D does not require the use of reduced-order models,
so the danger of erroneous assumption is averted. The ability
to use fully detailed models, with no simplifications, will be
necessary to the effective evaluation of smart grid
technologies. GridLAB-D is therefore specifically suited to
address the many issues pertaining to the integration and
adoption of smart grid technologies.

III. POWER FLOW MODELING


The power flow component of GridLAB-D is separated
into a distribution module and a transmission module. While
distribution systems were the original focus of GridLAB-D,
the transmission module was included so that in the future
interactions between the two modules can be examined.
When the integration of these two modules is complete it will
be possible to examine how smart devices on the
distribution system affect the transmission system.
Traditionally the ability to examine interactions on this level
has been limited by computational power; the use of multiple
processors is how this is addressed in GridLAB-D.
A. Transmission Systems Component models
As previously stated, the transmission system is not the
primary focus of GridLAB-D. The primary purpose of
including a transmission module is to allow for the
interconnection of multiple distribution substations. If a
transmission module was not included each substation, and
the associated feeders, would need to be solved
independently. While distribution feeders can be solved
independently, as is common in current commercial software
packages, GridLAB-D will also have the ability to generate a
power flow solution for multiple distributions feeders at
substations which are interconnected via a transmission or
sub-transmission network.
In the current version of GridLAB-D the power flow
algorithm used for the transmission system is the GaussSeidel (GS) method. There are 2 primary reasons why the GS
method was selected over other algorithms for the initial
transmission module. The first was that GS method is able to
solve a single power flow iteration more efficiently than other
methods such as the Newton Raphson (NR) method. This is
especially important since GridLAB-D will only start from a
flat start at the first time step and all other solutions will use
the solution from the previous time step. This is one of the
key differences in GridLAB-D; it is not designed to solve only
a single power flow problem but instead to generate a series
of power flow solutions that approximate a time varying
condition. The ability to examine time-series issues is central
to many of the questions surrounding smart grid technologies.
The second reason for the use of the GS solver is that it
has many desirable characteristics when implemented in an
object orient environment on multiple processors. For
example, NR solvers do not lend themselves to parallelization
because of the need to invert the Jacobian, which is not an
intrinsically parallel process.
The GS method, while
converging linearly, does so in an intrinsically parallel way
[4]. Because GridLAB-D solves a time-series of quasi-steady
solution, the number of iteration for either solver is generally
very low, and the convergence performance advantage of NR
is lost when compared with the high degree of parallelism
possible with GS. In addition, the GS solver is more robust
when starting from a poor initial guess solution. This
means that when a large disturbance is evaluated, the solver is
more likely to find a solution, even if it requires substantially
more iterations.

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
3

In the current version of GridLAB-D, the transmission


module is capable of solving balanced 3-phase models.
Future modules will be capable of solving unbalanced
conditions allowing integration of the transmission model
with unbalanced distribution models.
B. Forward Backward Sweep
In order to accurately represent the distribution system the
distribution feeders are expressed in terms of their conductor
types, conductor placement on poles, underground conductor
orientation, phasing, and grounding [5]. GridLAB-D does not
simplify the component models of the distribution system.
The existing distribution module of GridLAB-D utilizes the
traditional Forward and Backward Sweep (FBS) method for
solving the power flow problem. This method was selected in
lieu of newer methods such as the Newton-Raphson based
Three-Phase Current Injection Method (TCIM) methods of
[6] for the same reasons that the GS method was selected for
the transmission module; converging in the fewest number of
iterations is not the primary goal.
Just as with the
transmission module the distribution modules will only start
with a flat start at initialization and all subsequent solutions
will be from the previous time step.
Even though the FBS method was chosen as the initial
power flow algorithm, the modular nature of GridLAB-D does
not prevent the creation of a power flow module that
implements other algorithms, as in the case of the
transmission module. Section C will describe a new power
flow module that is currently being developed.
While there is discussion over what constitutes the
forward sweep and what constitutes the backwards sweep, for
the purposes of this paper the conventions of [5] will be used.
Equations for the forward and backward sweep are shown
below:
Node 1

[Iabc]1

Node 2
Series
Component

[Iabc]2
[Vabc]2

[Vabc]1

Forward Sweep

[Vabc ]2 = [A] [Vabc ] 1 [B] [I abc ] 2

(1)

Backward Sweep

[Vabc ]1 = [a ] [Vabc ] 2 + [b] [I abc ] 2

(2)

[I abc ] 1 = [c ] [Vabc ] 2 + [d ] [I abc ] 2

(3)

The [a], [b], [c], [d], [A], and [B] generalized matrices are
developed using characteristics of the individual series
components. The matrices are 3x3 and represent three-phase
components, but can also represent two-phase and singlephase components by filling in the unused rows and columns

with zeroes. At this time the following components are


modeled and available for use:
Overhead and Underground Lines.
Multiple
configurations, bare conductor, concentric neutral
underground cables, and triplex lines.
Transformers. Single-phase or three-phase in most
common configurations, and center-tapped.
Voltage Regulators. Three-phase Load Tap Changer
(LTC) type, 3 - single-phase in Wye and Delta
configurations, support for Line Drop Compensation
(LDC).
Fuses. Simplified over-current model.
Switches. Single-phase and three-phase ganged.
Shunt Capacitor Banks. Modeled similar to a load;
static, switched, and automatic control are supported.
Metering is supported for both single phase center-tapped
customers and three phase customers. Metering is also
available to support distributed resource operations including
distributed generation and energy storage. Support for
reclosers, islanding, and overbuilt lines is anticipated in
coming versions.
The ability to model overbuilt lines will be essential in
simulations where multiple feeders originate from a single
substation. When multiple feeders originate from a single
substation it is common for them to share the same support
structures, resulting in mutual coupling. The mutual coupling
means that the simulation of two feeders, sharing a support
structure, cannot be completely decoupled. This result is the
necessity to be able to model an n-phase system.
C. Gauss Seidel Three Phase Current Injection Method
While the power flow problem can be solved effectively
and efficiently using the FBS method, there are certain
limitations. The most notable of these limitations is the
inability to handle networked distributions systems. While the
vast majority of distribution feeders in the United States are
radial there are a substantial number that are not, especially in
urban centers. For this reason PNNL has begun work on a
new power flow module that is based on the well established
TCIM [6]. Since GridLAB-D was constructed with a modular
design the implementation of a new power flow algorithm will
not require modifications to any other modules. Additionally
all of the existing models for power system equipment within
the power flow module are valid; this is a key design feature
of GridLAB-D.
The method discussed in [6] is an effective method to solve
the power flow problem but it does not readily adapt to the
GridLAB-D environment, so a variation of it is being
implemented. Instead of solving the three phase current
injection based power flow problem with a NR technique, the
GS solution method will be used. The primary justification
for this is that the NR method cannot be inherently performed
on multiple processors, while the GS method can. This is the
same reason that the transmission module utilizes the GS
method. The basic formulation of the GS method, extended
to an unbalanced representation, is shown in (4):

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
4

( ) ]

S mj = Vmj ( y jt )mk Vkt


n

k =1

tp

(4)

where j indicates phase a, b, or c, m indicates the bus number,


n indicates the number of buses, and p is the set of phase a, b,
and c.
Equation 4 can be further expanded to give the GS
formulation for the voltage at any PQ bus, as is shown in (5):

Vmj =

(y )

jj mm

Pmj + Qmj
Vmj

[(y
[(y

n t p

k =1
k m t p
t j

) (V ) ]

*
jt mk

t *
k

t *
k

) (V ) ]

jt mk

(5)

In (5) both the real and reactive powers at a given bus are
known values, but this is not true for a PV bus where the
reactive power can vary based on bus voltage. Equation 6
shows how the reactive power injection at a PV bus is
calculated.

j
m

= Im Vmj

[(y
n


k =1 t p

) (V ) ]
*

jt mk

t *
k

(6)

Equations 5 and 6 form the core equations that allow for


the implementation of a GS solution to the TCIM; a method
that is inherently parallel. This work is currently in progress
and a paper detailing the GS solution to the TCIM is being
written and a new GridLAB-D module implementing this
solution method is in development. At the present all
GridLAB-D calculations use the FBS method.
IV.

EXAMPLE CASE

In order to fully demonstrate the effects of modeling large


distribution systems, their end-use loads, and distributed
resources, an example case has been developed. The example
case presented in this section is a 12.47 kV feeder with 1,400
nodes and is representative of a typical West Coast feeder that
is a mix of residential and commercial loads.
Connected to this feeder are three wind turbine generators
(WTG). To increase the level of model complexity 50 of the
residential loads were replaced with complex load models.
While most of the residential loads were kept as constant
power loads supplied through a triplex cable from a centertapped transformer, 50 loads were expanded to include
models for the various appliances within the residence. These
appliances included hot water heaters, lighting, microwaves,
ranges, dishwashers, clothes washers, refrigerators, and
heating, ventilating and air conditioning (HVAC). The
following subsections explain in detail how the WTGs and the
details residential loads were modeled.

A. Wind Turbine Generator Model


A simple model for a generator can be described as a
constant power or constant power factor source, injecting
power or current into the system solely as a function of
voltage. The WTG models implemented in this example are
far more complex. While the level of detail shown in this
example may not always be necessary or practical, it is
illustrative of the capability to model generators in GridLABD.
Additionally, the evaluation of future smart grid
technologies may require this level of modeling detail.
Terminal voltage and meteorological data, including
temperature, wind speed, and air pressure, are used as inputs
into the WTG model, while the final output is the current
injections at the generator terminals. With the inputs and
outputs defined the WTG can then be modeled by breaking
the system into two main components, the mechanical model
and the electrical model.
The mechanical model determines the amount of power
extracted from the wind, what percentage of that is
transmitted through the drive train, and the final amount of
power that is delivered to the electrical generator. By using
the meteorological data, and design specifications supplied by
WTG manufacturers, the mechanical power extracted, Pm, can
be determined by (7):

Pm =

1
r 2C p (u) gear u 3
2

(7)

where is the density of air, which is a function of air


temperature and pressure, r is the radius of the blade, Cp is the
coefficient of performance, u is the wind velocity, and gear is
the efficiency of the drive train.
The coefficient of performance, Cp, is blade specific to each
turbine type and must be extrapolated from manufacturers
tables when available, or approximated through the methods
described in [7] and [8]. Fig. 1 shows a set of example results
from an approximation of a General Electric 2.5 MW machine
using the methods of [7] and [8]. Through the use of higher
order polynomials the GridLAB-D output can be made to
match the manufactures data to any desired level of accuracy,
at the cost of computational speed.

Fig. 1. GridLAB-D active power output for a General Electric 2.5 MW wind
turbine compared to manufacturers data.

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
5

One the mechanical power available at the output of the


drive train is calculated it is then converted to electrical power
through a synchronous generator. The generators for this
example were modeled using balanced Y-grounded circuit
models, but GridLAB-D has the ability to implement models
with unbalanced generator parameters if the user desires. The
equivalent circuits are solved through an iterative process,
where the inputs are terminal voltage and mechanical power,
and the output is injected current.
GridLAB-D is able to incorporate complex meteorological
data, aerodynamic characteristics, and mechanical parameters
into the power flow solution. It is up to the user to determine
what level of complexity is necessary for their particular
application.
B. Residential House Model
For this example the majority of residential, and all
commercial, loads were modeled as time varying constants
power values, i.e. the load was treated as a constant power at
each time step, but allowed to vary between time steps. To
increase the level of model complexity 50 of the residential
house models were treated as complex time-varying loads, i.e.
the power consumed is not an operator defined input. This
model simulates power demand in a single-family home by
incorporating models for end-use products, such as
dishwashers, refrigerators, water heaters, lights, ranges,
microwaves, plug loads, and HVAC. Each end-use device is
modeled individually, and the power demands are then
aggregated and incorporated into the system through the
parent house model.
An example of a relatively simple individual appliance
model can be seen in the electric ranges, incorporated in each
of the 50 residential house models. Each is modeled as a
purely resistive load, where 100% of the power is converted
to radiant heat, which then contributes to heat gains within the
house that must be compensated for by the HVAC system.
A more complicated individual appliance model can be
seen in the hot water heater. These are modeled as twoelement systems where cold water flows from the bottom
through the larger first element, to the second, smaller
element. The water heater is a multi-state model with states
for full, partial, or empty. Corresponding to the multiple
states there are various power consumption levels which are
based on variables such as tank size, thermal mass of water in
the tank, flow rate, ambient temperature, thermostatic setpoints, control dead bands, and tank shell losses. The
complexity of the residential model can be varied as desired
or based on available computational resources.
C. Analysis of Radial Distribution System
To highlight the level of modeling detail necessary to
perform smart grid evaluations, a 1,400 node radial feeder
was created. The feeder model contains three 1.5 MW wind
turbines, 50 residential house models, each with unique, timevarying load distributions, 50 polyphase and 400 single-phase
constant power loads, and over 600 transformers, all feed
through a 12.47 kV substation. The houses were connected
through center-tapped pole-top transformers and triplex
cabling. The WTGs were connected to three equidistant

nodes along the radial feeder. Wind speed data was imported
from a Typical Meteorological Year 2 data set (TMY2)
produced by the National Renewable Energy Laboratory
(NREL) [9], then semi-randomized at one-minute intervals to
provide gust-like conditions. Voltage measurements were
made at five minute intervals at a residential house
approximately halfway along the feeder. Fig. 2 shows the
results for 2 days at 1 minute time intervals and Fig. 3 shows
the results for 8 days at 5 minute time intervals. In Fig. 2 and
Fig. 3 the regulator at the substation was held constant so that
the voltage effects of the WTG can be seen.

Fig. 2. Nominal end use voltage over a 2 day period

Fig. 3. Nominal end use voltage over an 8 day period

From Fig. 2 and Fig. 3 it can be seen that the outputs of the
WTGs have a pronounced effect on the voltage level at the
house, so much so that the normal diurnal variations are
masked. This is not a completely unexpected result for three
1.5 MW turbines on a 12.47 kV feeder. These WTGs are
much large than those that are generally connected to a 12.47
kV distribution feeder.
The point of interest in Fig. 2 and Fig. 3 is that the detailed
residential models and the detailed WTG models were
combined into a single simulation environment. In this
environment the meteorological effects were used to
determine the power output of the WTGs and the
heating/cooling requirements of the house while at the same
time electrical models were coupled to thermal building
models. All of these effects were combined to determine

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
6

what the actual electrical load of the houses would be and


what would be the resultant voltage seen at the outlets. As
previously mentioned this level of detail may not be necessary
for feeder capacity studies but it critical for the evaluation of
new smart grid technologies.
For the radial example system of 1,400 nodes with 3 WTGs
and 50 detailed residential models the system solved at the
rate of approximately 5 minutes simulation time per second,
i.e. 9.6 minutes for a 2 day simulation. This speed was
achieved on an early model Pentium IV desktop. Newer
desktop machines, especially those with multiple processors
complete these simulations in significantly less time.
V. MODELING OF DISTRIBUTION FEEDERS
Due to the size and complexity of distribution feeders it is
rarely practical to generate new distribution feeder models.
For this reason GridLAB-D has incorporated the ability to
convert files from other software formats as well as providing
a set of distribution feeder models that may be used openly.
A. Importing Models
Most distributions utilities utilize some form of software to
model their distribution feeders. Generally these models are
created by importing information from their Geographic
Information System (GIS). Depending on the utility these
models may be combined with their billing system or other
databases to indicate the type of customer served by each
transformer, e.g. residential, commercial, industrial, or
agricultural. While distribution utilities often only utilize
these models for peak capacity studies, and a limited range of
other studies, the models contain much of the information
necessary to generate the distribution feeder models in
GridLAB-D.
To gain the ability to access the information in these
models scripting has been written to import distribution feeder
models from the SynerGEE and Windmil software
packages. These feeder models can then be combined with
detailed residential and commercial models to accurately
model existing distribution feeders. While scripting exists for
importing models it is not a fully automated process because
of the varying quality of model data. Scripting for importing
distribution feeder models from other software packages are
pending.
B. Feeder Taxonomy
While it is desirable to obtain distribution feeders models
directly from a utility, it is not always possible. For various
well founded reasons distribution utilities generally do not
make copies of their system models available for general use.
Without meaningful distribution feeder models it is not
possible to effectively analyze smart grid technologies. To
address this issue the DOE Modern Grid Initiative (MGI) [10]
undertook the task of developing a set of prototypical
distribution feeders that could be disseminated to the general
public, for use in GridLAB-D. The goal was to generate a set
of distribution feeder models that was representative of those
used in the 48 states of the contiguous United States. At
present 25 radial distribution feeders have been constructed in

GridLAB-D. Each feeder model contains all system elements


from the substation regulator to the end-use meters, the same
level of detail displayed in the example of section IV. Each
feeder was designed to represent a specific voltage level and
load composition within the various climate regions of the
United States. A complete description of this work will be
presented in a future paper which is currently being written.
VI. CONCLUSION
This paper has shown how the traditional power flow
analysis can be enhanced through the use of detailed end use
and generator models, and implemented in the GridLAB-D
simulation environment. It has also discussed the level of
modeling detail that will be necessary to determine the
impacts of smart grid technologies. The ability to model these
effects will be essential in increasing the penetrations of these
new technologies and modernizing the electricity
infrastructure.
VII. REFERENCES
[1]
[2]

www.gridlabd.org/
IEEE 1366: IEEE Guide for Electric Power Distribution Reliability
Indices, available online at: www.ieee.org
[3] IEEE Radial Test Feeders- Available through the IEEE Distribution
System Analysis Subcommittee.:
www.ewh.ieee.org/soc/pes/dsacom/testfeeders.html
[4] D. P. Chassin, P. R. Armstrong, D. G. Chavarria-Miranda, R. T.
Guttromson, Gauss-Seidel accelerated: implementing flow solvers on
field programmable gate arrays, IEEE PES GM 2006,
[5] W. H. Kersting, 2nd ed., Distribution System Modeling and Analysis.
Boca Raton: CRC Press, 2007
[6] P. A. N. Garcia, J. L. R. Pereira, S. Carneiro Jr., V. M. Da Costa, and N.
Martins, Three-Phase Power Flow Calculations using the Current
Injection Method, IEEE Transaction on Power Systems, Vol. 15, Issue 4,
May 2000, pp. 508-514
[7] C.G. Justus, Winds and Wind System Performance, Philadelphia, PA:
Franklin Institute Press, 1978.
[8] B. Malinga, J.E. Sneckenberger, and A. Feliachi, Modeling and Control
of a Wind Turbine as a Distributed Resource, Proc. 35th Southeastern
Symp. System Theory, Atlanta, GA, 2003, pp. 108-112.
[9] TMY2weather
data,
available
online
at:
www.rredc.nrel.gov/solar/old_data/nsrdb/tmy2/
[10] www.netl.doe.gov/moderngrid/

VIII. BIOGRAPHIES
Kevin P. Schneider (M06, SM08) received his B.S. degree
in Physics and his M.S. and Ph.D. degrees in Electrical
Engineering from the University of Washington. His main
areas of research are power system operations and visual
analytics. He is currently a research engineer at the Pacific
Northwest National Laboratory (PNNL), working at the
Battelle Seattle Research Center in Seattle Washington. Dr.
Schneider is also an Adjunct Faculty member at the
Washington State University Tri-Cities campus and a licensed Professional
Engineer in Washington State.
David P. Chassin (M03, SM05) received his BS of
Building Science from Rensselaer Polytechnic Institute in
Troy, New York. He is a staff scientist with the Energy
Science and Technology Division at Pacific Northwest
National Laboratory where he has worked since 1992. He
was Vice-President of Development for Image Systems
Technology from 1987 to 1992, where he pioneered a hybrid
raster/vector computer aided design (CAD) technology called

Downloaded from http://iranpaper.ir

http://plcman.vcp.ir
7

CAD Overlay . He has experience in the development of building energy


simulation and diagnostic systems, leading the development of Softdesk Energy
and DOEs Whole Building Diagnostician. His recent research focuses on
emerging theories of complexity as they relate to high-performance simulation
and modeling in building controls and power systems and is currently
responsible for the design and development of DOEs GridLAB-D simulator.
TM

Yousu Chen (M07) received his B.E. in Electrical Engineering from Sichuan
University, China, his M.S. in Electrical Engineering from Nankai University,
China, and M.S. in Environmental Engineering from Washington State
University. Currently he is a Research Engineer at the Pacific Northwest
National Laboratory in Richland Washington. Mr. Chen is an IEEE member and
the Vice-Chair of the Richland Chapter of the Power Engineering Society.
Jason C. Fuller (S'08) received his B.S. degree in Physics from the University
of Washington in Seattle Washington. Currently he is a graduate student at
Washington State University Tri-Cities Campus in Richland Washington. His
main areas of interest are power system analysis and renewable integration. Mr.
Fuller is a Student Member of the IEEE.

Das könnte Ihnen auch gefallen