close

Вход

Забыли?

вход по аккаунту

?

ESTS.2017.8069298

код для вставкиСкачать
Power Hardware-in-the-Loop Simulation Testing of a
Flywheel Energy Storage System for Shipboard
Applications
James Langston∗ , Michael Steurer∗ , Karl Schoder∗ , Joseph Borraccini† ,
Don Dalessandro† , Tim Rumney‡ , Tom Fikse§
∗ Center
for Advanced Power Systems, Florida State University, Tallahassee, FL, USA
Surface Warfare Center Philadelphia Division, Philadelphia, PA, USA
‡ GKN Hybrid Power, Abingdon, Oxfordshire, UK
§ Syntek Technologies, Arlington, VA, USA
† Naval
Abstract—Energy storage is anticipated to play an important
role on future surface combatants as specialized weapons and
sensors systems are introduced, which may require high power
and fast ramp rates over short periods of time. The energy
magazine (EM) concept has been proposed as one approach
for incorporating these types of systems into future, as well as
existing, platforms. The EM is envisioned as a power converter
with inherent energy storage capability, providing buffering
between connected loads and the shipboard power generation
system. This paper presents results from power hardware-in-theloop (PHIL) simulation experiments in which an actual flywheel
energy storage system (FESS) is tested as part of an emulated
energy magazine, supplying a pulsating load.
^ŚŝƉ
^ĞƌǀŝĐĞƵƐ
ŶĞƌŐLJDĂŐĂnjŝŶĞ
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
/ŶƉƵƚ
WŽǁĞƌ
ŽŶǀĞƌƚĞƌ
&
^
DĞĚŝĂ
^ŚŝƉ
^ĞƌǀŝĐĞƵƐ
>ŽĂĚƐ
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
>ŽĂĚƐ
ŶĞƌŐLJDĂŐĂnjŝŶĞ
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
/ŶƉƵƚ
WŽǁĞƌ
ŽŶǀĞƌƚĞƌ
I NTRODUCTION
WŽǁĞƌ
ŽŶǀĞƌƚĞƌ
^
Energy storage is expected to play an increasingly important role in surface combatants of the future as the shipboard
power systems accommodate loads of increasing magnitude
and with more demanding characteristics [1]. These loads,
including new weapons and sensory systems, may present
short-term power consumption that may exceed the power delivery capabilities of the plant, or the loads may require ramps
in power that would exceed the limits specified by present
military standards (e.g. MIL-STD 1399-300 [2], MIL-STD
1399-680 [3], etc.). Ships employing integrated power systems
(IPS) including electric propulsion may inherently possess
sufficient generation capacity to handle many of these loads.
However, for non-IPS systems, the incorporation of energy
storage may be necessary to accommodate these demanding
loads envisioned in the future.
^
DĞĚŝĂ
>ŽĂĚƐ
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
>ŽĂĚƐ
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
>ŽĂĚƐ
(b) Converter Coupled ES Media
Fig. 1: General Topologies for Energy Magazine
other media, such as batteries, could also be used in this
configuration. In general, it is not necessary that each output
port makes use of an output power converter. If the internal
DC bus is tightly regulated, or if the load does not require
a tightly regulated supply, some form of circuit protection
(e.g. solid state circuit breakers) can be substituted for the
output converter. The EM simply provides a means whereby
sensitive, demanding, or high priority loads can be supplied.
The EM can serve to provide buffering to the ship generation
system for loads requiring high ramp rates, and can provide
uninterrupted power to critical mission loads in the event of
loss of ship service power. Additionally, the EM provides a
means to supply high power, short-duration loads, for which
the short-term power demands may exceed the capabilities of
the power generation plant.
One of the concepts proposed for integrating energy storage
into existing and future ships is that of the energy magazine
(EM) [1]. With this approach, energy storage is integrated
into power conversion equipment, as illustrated by Fig. 1.
Fig. 1a shows one possible manifestation of an EM, consisting
of an input power converter (typically an active front end
(AFE)) with one or more output converters supplying loads.
One or more energy storage media (e.g. batteries, capacitors,
etc.) are directly connected to the common DC bus to which
the input and output converters are connected. An alternative
configuration is illustrated by Fig. 1b, in which the energy
storage media are connected to the common DC bus through
a power converter. This configuration allows energy storage
media such as flywheels, for example, to be employed, but
978-1-5090-4944-8/17/$31.00 ©2017 IEEE
KƵƚƉƵƚ
ŽŶǀĞƌƚĞƌ
ŽƌŝƌĐƵŝƚ
WƌŽƚĞĐƚŝŽŶ
(a) Directly Coupled ES Media
&
I.
>ŽĂĚƐ
With the introduction of these types of loads and the energy
storage systems to accommodate them, it is important to understand the potential impacts of these systems on the behavior
and performance of the power plant. As described in [4], the
naval community has identified a number of risks relating to
305
the introduction of pulsating loads and energy storage systems,
including a lack of standards, a lack of industrial base, and a
limited test capability and infrastructure. Other identified risks
included controls for energy and power management, power
continuity, and the lack of fidelity in models used in analyses.
Efforts under the Advanced Electric Power and Propulsion
Systems Development Project (AEP3) agreement between the
U.S. Navy and the Royal Navy are underway to attempt to
more closely examine a number of the identified risks through
simulation studies and power hardware-in-the-loop (PHIL)
simulation [5] experiments. This paper describes one of the
efforts that is currently underway, involving PHIL simulation
testing of a 320 kW, 3.5 MJ flywheel energy storage system
(FESS) within the context of an emulated EM. At the time of
writing, the PHIL experiment was still in the commissioning
phases, but the paper describes the approach and presents
preliminary results from commissioning experiments. The test
setup is described in Section II, including the FESS, the
PHIL test bed, and the real-time simulator. Commissioning
activities and results are described in Section III, including
initial results from PHIL tests. This includes a description of
the emulated surrounding system and the energy management
controls. Additional details on planned PHIL demonstrations
and experiments, along with concluding remarks, are given in
Section IV.
II.
Zd^
sŽůƚĂŐĞĂŶĚ
ƵƌƌĞŶƚ
DĞĂƐƵƌĞŵĞŶƚƐ
sƌ
>Ɛ
ZĨ
ŵƉ
Ĩ
Ɖ
/ƌ
/ĚĐ ͕/ĚĐϮ
sĚĐ ͕sĚĐϮ
&^^
Fig. 2: Test Setup
B. DC Amplifier
The DC amplifier is a commercial product [6], which
employs a modular multilevel converter (MMC) topology. The
amplifier consists of four bi-directional converters, each of
which can operate at up to 6 kV and up to 210 A, which can be
configured in any parallel or series configuration. This allows
operation at up to 24 kV or operation at up to 840 A. For these
experiments, three MMC units will be operated in parallel,
although a single MMC is used for initial commissioning
tests at reduced power levels. The converters are operated
at the lower end of the voltage range to match the interface
voltage of the DUT. The converters can operate with no DC
link capacitance, in which case a switching transient of 1 kV
is observed, occurring with a frequency of 12 kHz. A shunt
filter (Rf and Cf in Fig. 2) consisting of a 5 Ω resistor and
a 10 μF capacitor is introduced to substantially reduce the
magnitude of the switching component in the voltage. An
additional parallel capacitor (Cp ) of 20 μF has been introduced
for redundancy. A series inductor (Ls ) of 50 μH has also been
introduced to attenuate high frequency currents into the FESS.
For these tests, the amplifier is operated in voltage control
mode, accepting a voltage reference from the DRTS.
T EST S ETUP
The test setup in illustrated by Fig. 2. The DC terminals
of the FESS are connected through a filter to the DC amplifier
used for the PHIL interface. A digital real-time simulator
(DRTS) provides a current reference to the FESS (Ir ) and
a voltage reference to the DC amplifier (Vr ), making use
of the measured voltage (Vdc and Vdc2 ) and current (Idc
and Idc2 ) at the terminals of the FESS. Redundant voltage
and current measurements were used in order to provide a
means for detecting sensor failures, in order to trigger the
experiment to revert to open-loop operation in the event of
a sensor failure. The DRTS serves the roles of simulating
the surrounding system and implementing the algorithms and
controls for the PHIL interface between the FESS and the
emulated surrounding system. Additionally, the DRTS is used
for monitoring, data recording, and implementing protection
for the setup. The components of the test setup are described
in more detail in the following subsections.
C. Digital Real-Time Simulator
A. Flywheel Energy Storage System
The DRTS is a commercial product [7], executing electromagnetic transient simulations in real time using time-step
sizes on the order of 50 μs, with the capability to simulate
converters in small subsystems with time-step sizes on the
order of 2 μs. These capabilities are used for simulation of
the surrounding system, as well as simulation of the entire
PHIL setup for off-line testing. The simulator also supports
implementation of analysis and controls systems, which are
used for such functions as implementing HIL interface algorithms, system monitoring (e.g. calculation of filtered quantities, power, harmonic distortion, etc.), and protection. The
simulator provides substantial I/O capabilities for interfacing
with the rest of the test setup, including accepting feedback
from instrumentation and providing references to the power
amplifier and FESS. The simulator also provides data capture
The device-under-test (DUT) is a flywheel energy storage
system developed by GKN Hybrid Power Limited for the
UK Defence Science and Technology Laboratory. The unit
supports bi-directional energy storage through a 650 V - 850 V
DC interface, using a modular flywheel system for energy
storage. The unit being tested is composed of three flywheels.
The energy storage capacity for the unit as configured with
three flywheels is rated for 3.5 MJ (useable energy), and the
interface is designed for charging and discharging at up to
320 kW (tested up to 180 kW). The unit has been configured
to accept an external current reference through an analog input,
allowing the charging and discharging to be controlled from
an external source (in this case the DRTS). For these tests, the
negative DC rail of the FESS was grounded.
306
250
facilities, with the ability to capture at sampling rates up to
the time-step size used for the simulation.
C OMMISSIONING T ESTS
150
Current (A)
III.
200
The plan for commissioning of the FESS within the PHIL
test bed included a sequence of tests at increasing voltage and
power levels to verify appropriate and safe operation of the
device. This sequence of tests began with operation at reduced
voltage and power to verify operation of the FESS and to verify
that all protection systems functioned correctly. This included
applying short-circuit faults at the terminals of the FESS at
reduced voltage. These were followed by tests at full voltage
(750 V) and reduced power (160 kW) to verify operation and
control from the DRTS. During this phase of testing, two issues
arose that required measures for resolution before proceeding
with the commissioning. One of the issues was an oscillation
(of approximately 800 Hz) that was induced when connecting the DC amplifier to the FESS. Investigations with the
amplifier and with a controller hardware-in-the-loop (CHIL)
simulation of the amplifier revealed that the oscillations were
triggered simply by connecting the DC amplifier to a large
(approximately 2 mF) capacitor, equivalent to the impedance
presented by the FESS. This resulted in oscillations in current
of up to 50 A in magnitude. Retuning the controls in the CHIL
simulation resulted in essentially eliminating the oscillations,
and this was verified with the amplifier in a hardware test, and
finally in testing with the FESS.
100
Measured
Reference
50
0
-50
0
0.5
1
1.5
2
2.5
Time (s)
3
3.5
4
3
3.5
4
(a) Voltage
0.51
0.5
State-of-Charge
0.49
0.48
0.47
0.46
0.45
0.44
0.43
0.42
A second issue that arose during this phase of commissioning was spurious trips by the FESS when operating at
750 V. It was believed that the environment noise, largely
from the switching of the DC amplifier, was causing the false
trips. Diagnostic tools were employed to trace the source of
the trip to logic in the FESS controls intended to identify
flywheel speed measurement errors. The logic was based on
observation of multiple position sensor readings that should
occur in a specific sequence. Noise from the test environment
was causing sporadic errors in the sensor readings, resulting in
trips. The logic was adjusted to require multiple occurrences
of the errors within a specified time period to trigger a trip,
increasing the immunity of the controls to the environment
noise.
0
0.5
1
1.5
2
2.5
Time (s)
(b) Current
Fig. 3: Results from Current Ramp Commissioning Test
most prominent aspects from the perspective of the FESS.
While the actual PHIL simulation experiments are intended to
include portions of the generation plant, ship service loads, and
multiple specialized loads supplied from an EM, the simplified
system used for commissioning, illustrated by Fig. 4, was
composed of an emulated EM with a simple, constant-power
pulsating load. In this configuration, the EM consists of an
active front end (AFE), DC bus capacitor (C1 ), and energy
management controls, with the FESS hardware serving as the
energy storage medium, virtually connected to the common
DC bus. A periodic pulsating DC load (Load 1) is supplied
directly from the common DC bus (omitting circuit protection
for this case). For these tests, the ramp rate of Load 1 was
unlimited, so that the load could transition from zero to full
load within a single time-step of the simulation, representing a
load with no local energy storage. The capacitor is needed to
supply the fastest dynamics of the DC load, while the AFE and
the FESS ramp up their respective contributions. The capacitor
was sized for 2 F for these tests, based primarily on the rating
of Load 1, the delay and ramp characteristics of the FESS, and
the desire to maintain the DC bus voltage within 100 V of the
750 V nominal set point. The voltage-type ideal transformer
model (ITM) interface algorithm [8] was employed to provide
the virtual coupling between the FESS DC terminals and the
common DC bus. Due to the large capacitance on the DC bus
(relative to the capacitance presented by the FESS), stability
was not an issue with this interface algorithm.
A. Characterization Tests
Resolution of the issues noted above allowed the planned
commissioning tests to be executed, verifying fundamental
operation, while also collecting valuable data for characterization of the FESS and for model validation. Results from
one of the tests are shown in Fig. 3. This shows a ramp of
the current from zero to 200 A (discharging) in Fig. 3a, along
with the associated state of charge of the FESS in Fig. 3b.
From these types of tests, characteristics such as the delay in
response and the maximum ramp rates of the FESS could be
gleaned for modeling and model validation. These provided
key information needed for formulating and preparing for the
PHIL simulation tests.
B. PHIL Tests
For commissioning in the context of the PHIL simulations,
a simplified surrounding system was developed to include the
307
ŶĞƌŐLJ
DĂŶĂŐĞŵĞŶƚ
ŽŶƚƌŽůƐ
/ƌͲ^
^K
Voltage (V)
780
&^^
WŚLJƐŝĐĂů
,ĂƌĚǁĂƌĞ
ŵƉ
>ŽĂĚϭ
>ŽĂĚ
sĚĐ͕/>
,/>/ŶƚĞƌĨĂĐĞ
Hardware
Sim
800
760
740
720
/ƌͲ&
700
ϭ
/ŶƉƵƚ
ŽŶǀĞƌƚĞƌ
&
0
0.5
1
1.5
Time (min)
2
2.5
3
(a) Voltage
Fig. 4: Surrounding System for PHIL Commissioning Tests
AFE
FESS (Sim)
FESS (HW)
Load
150
100
Current (A)
The role of the energy management controls is to ensure regulation of the common DC bus voltage, while also
maintaining a target state-of-charge for the energy storage
medium, by providing current references to the input converter
and the actual FESS. An actual energy management control
system may include other considerations and may be quite
complicated. However, for the purposes of the commissioning
and PHIL demonstrations, a rudimentary energy management
control was implemented, as illustrated in Fig. 5. The control
primarily consists of three branches. The load current (IL ) is
used as an input to a feed-forward branch used as the basis for
the current references when the load is drawing current. A time
delay (Delay 1) is included to make the branch more realistic.
The sensed load current is then multiplied by a constant, ksh ,
intended to specify the portion of the load current nominally
intended to come from the AFE. The current reference for the
AFE is then passed through the hard limits and rate limits for
the AFE, representing the current limitations of the AFE (this
signal is passed through these limits a second time to generate
the actual reference Ir−AF E , as will be further discussed
below). The difference between the sensed load current and the
current contribution from the AFE is then sent as a reference to
the FESS (Ir−ES ), after passing through the hard limits and
rate limits representing the current limitations of the FESS.
This basic structure is augmented with a DC voltage control
branch. This branch represents a traditional feedback control
using a PI controller, and the output of the branch is added to
the AFE current reference. It is necessary that this component
of the AFE current reference does not get included in the
reference to the FESS, as this would consume the current
needed to affect the bus voltage. The AFE limits are applied
again to the total AFE current reference before generating the
actual reference to the AFE, Ir−AF E . The final branch, labeled
ES SoC Control, is used to maintain a target state-of-charge
(SOCr ) for the FESS. A simple proportional control is used,
with the resulting current reference passing through limits,
and then adding to the AFE current reference. A comparator
(Comparator 1) is employed to zero out the additional current
request from this branch if the load current is above a specified
threshold (Ith ). This gives priority to maintaining the bus
voltage and supplying the load when the load is drawing
current, allowing the FESS to recharge during periods when the
50
0
-50
0
0.5
1
1.5
Time (min)
2
2.5
3
2
2.5
3
(b) Current
0.75
0.7
SOC
0.65
0.6
0.55
0.5
0.45
0
0.5
1
1.5
Time (min)
(c) FESS State-of-Charge
Fig. 7: Results from PHIL Commissioning Test
load is not drawing current. The current limits for this branch,
IU L−SoC and ILL−SoC , are adjusted dynamically, based on the
DC bus voltage, as illustrated in Fig. 6. This allows the current
demand from this branch to be gradually reduced as the voltage
drifts farther away from nominal. This formulation of the
energy management controls allows the salient characteristics
of the components and controls to be easily adjusted in order to
explore the design space, with minimal adjustments to control
gains.
Example results from one of the PHIL commissioning tests
are shown in Fig. 7, illustrating results for two cycles of the
308
ŶĞƌŐLJDĂŶĂŐĞŵĞŶƚŽŶƚƌŽů
sĚĐ
sŽůƚĂŐĞŽŶƚƌŽů
^^Ž ŽŶƚƌŽů
'ĂŝŶϭ
^Kƌ
^K
н
/h>Ͳ^Ž
Ͳ
/>>Ͳ^Ž
ф/ƚŚ
/>
&,ĂƌĚ>ŝŵŝƚƐ
ŽŵƉĂƌĂƚŽƌϭ
ĞůĂLJϭ
dĚ
ŬƐŚ
н
sĚĐͲƌ
ŬƉͲϭ
W/
&,ĂƌĚ>ŝŵŝƚƐ
&ZĂƚĞ>ŝŵŝƚƐ
/h>Ͳ&
н
н
W/Ϯ
Ͳ
н
/ƌĂƚĞͲ&
/h>Ͳ&
н
/ƌͲ&
/ƌĂƚĞͲ&
/>>Ͳ&
>ŽĂĚ&ĞĞĚ
&ŽƌǁĂƌĚŽŶƚƌŽů
&ZĂƚĞ>ŝŵŝƚƐ
/>>Ͳ&
^,ĂƌĚ>ŝŵŝƚƐ
/h>Ͳ^
Ͳ
н
^ZĂƚĞ>ŝŵŝƚƐ
/ƌͲ^
/ƌĂƚĞͲ^
/>>Ͳ^
Fig. 5: Energy Management Controls
ϭ
sĚĐͲƌ
sĚĐ
ϭ
н
ŬŶŽƌŵ
Ͳ
н
Ͳ
Ϭ
/h>Ͳ&
/h>Ͳ^Ž
/>>Ͳ&
/>>Ͳ^Ž
ϭ
ͲŬŶŽƌŵ
ϭ
Ϭ
Ͳ
н
Fig. 6: Dynamic Limits for SoC Branch of Energy Management Controls
from the simulated system, including the current drawn by
the load (Load), the current injected into the DC bus by the
AFE (AFE), and the current injected into the DC bus through
the PHIL interface (FESS (Sim)), representing the current
contribution from the FESS. The actual current measured out
of the terminals of the FESS is also shown (FESS (HW)).
Again, the current injected into the simulated capacitor closely
matches the measured current from the FESS. Fig. 7c shows
the state-of-charge of the FESS.
Load
AFE
FESS
Capacitor
100
Power (kW)
50
0
-50
The first of the load pulses is applied at time t = 0, prior to
which the system is at steady state with the common DC bus
voltage held close to 750 V and the state-of-charge of the FESS
held close to 0.75. At t = 0, the first load pulse is applied,
and the current for the load transitions to approximately 175 A
within a single time-step of the simulation. In the time interval
prior to reaction of the energy management controls and the
AFE, the current is initially drawn from the capacitor, resulting
in a decrease in the DC bus voltage. The AFE is the first to
respond to the request by the energy management controls,
quickly ramping to inject 50 A (its maximum current) into
the DC bus. The FESS subsequently ramps up its current
to approximately 150 A, at which point the DC bus voltage
begins to recover, causing the current draw from the constantpower pulse load to slightly decrease. This also results in a
reduction in current from the AFE. The contribution of the
FESS in supplying the load power is evident as the state-of-
-100
-5
0
5
10
15
Time (s)
20
25
Fig. 8: Power Contributions of Components in PHIL Test
pulsating load. For this test, the parameter values given in
Tables I and II were used. Fig. 7a shows the simulated voltage
at the common DC bus (Sim), along with the voltage actually
applied to the FESS (Hardware). Although the actual voltage
at the FESS contains harmonics due to the switching of the
DC amplifier, the voltage presented to the FESS generally
tracks the simulated voltage well. Fig. 7b shows quantities
309
TABLE I: Load Parameters for PHIL Commissioning Test
Parameter
Description
fpulse
Dpulse
Ppulse
The
The
The
The
Imax−pulse
frequency of the load pulses.
duty cycle of the load pulses.
power magnitude of the load pulse.
maximum current draw of the load.
TABLE II: Energy Management Control Parameters for PHIL
Commissioning Test
Value
0.01 Hz
0.15
130 kW
200 A
Parameter
IU L−AF E
ILL−AF E
Irate−AF E
IU L−ES
ILL−ES
Irate−ES
Td
ksh
charge decreases from 0.75 to below 0.45 while the load is
active.
At t = 0.25 min, the load becomes inactive, with the
load current dropping to 0 A within a single time-step. This
results in a sharp increase in the bus voltage, until the energy
management control can react to the load change, and until
the AFE and FESS can react to the rapidly changing current
references from the energy management control. Once the
load is inactive, the ES SoC control branch of the energy
management controls becomes active, and the FESS begins
to charge. The current from the AFE is again ramped up to
its maximum level of 50 A to supply the charging current for
the FESS. As the FESS state-of-charge begins to approach
the target value of 0.75, the charging current demand from
the FESS begins to decrease according to the proportional
gain of the control. By t = 1.5 min, the system has again
reached steady state, with the bus voltage and FESS stateof-charge close to nominal, and the AFE only providing
approximately 5 A to maintain the state-of-charge of the FESS.
At t = 1.67 min, the second load pulse is applied, and the cycle
repeats itself.
Vdc−r
kp−2
kI−2
SOCr
kp−1
Ith
knorm
Description
Value
AFE current limit.
AFE current limit.
AFE current ramp rate limit.
Energy storage medium current limit.
Energy storage medium current limit.
Energy storage medium current ramp rate limit.
Time delay for load feed forward control branch.
Gain in load feed forward control branch designating the proportion of the load current to be supplied
by the AFE.
Reference voltage for DC voltage control branch.
Proportional gain for DC voltage control branch PI
controller (PI 2).
Integral gain for DC voltage control branch PI
controller (PI 2).
Reference state-of-charge for ES SoC control
branch.
Proportional gain (Gain 1) for ES SoC control
branch.
Threshold load current for Comparaotr 1 above
which the ES SoC control branch becomes active.
A gain for normalization of the deviation of voltage
from nominal for the dynamic limits for the ES
SoC control branch.
IV.
50 A
−50 A
250 A/s
180 A
−180 A
200 A
10 ms
0.2
750 V
0.5 A/V
0
0.75
1000 A
20 A
0.01 V−1
C ONCLUSION
As discussed herein, PHIL simulation experiments are
currently underway, interfacing a 3.5 MJ, 320 kW FESS to a
simulated shipboard environment, in which the FESS forms the
primary energy storage component of an energy magazine. At
the time of writing, the testing was still in progress. However, a
number of commissioning tests have been completed, including a set of characterization tests and PHIL commissioning
tests, conducted at a reduced power level of 160 kW. The PHIL
commissioning tests involved a simplified system, consisting
of relevant portions of an energy magazine and a constantpower load with periodic pulsations. A rudimentary energy
management control was implemented to handle regulation of
the EM DC bus voltage and maintaining the state-of-charge
of the FESS, while meeting load demand. The FESS was
tested for a scenario in which the AFE of the EM could
not support the magnitude of the load, and the FESS was
needed to accommodate interfacing the load to the system. The
commissioning test successfully exercised the FESS through
charging and discharging cycles, while also exposing it to the
voltage fluctuations that occurred on the simulated DC bus. It
should be noted that the performance shown for the FESS
reflects its performance at the time of commissioning, but
improvements to the transient response times of the control
system are expected.
The relative contributions of each of the components of
the EM are better illustrated in Fig. 8. This shows the power
drawn by the load (Load), as well as the power contributions
from the AFE (AFE), FESS (FESS), and the DC bus capacitor
(Capacitor). Fig. 8 shows that the load power remains constant
at 130 kW over the demand period, indicating that the load
demand was successfully met. During the initial second of
the load pulse, the majority of the power is drawn from the
capacitor, as the current contributions from the AFE and FESS
are ramped up. Over the remainder of the time for which
the load is active, the power from the AFE and the FESS
is generally shared in the proportion specified by the energy
management control (20% and 80%, respectively). Extra power
is provided by the AFE and the FESS in the first five seconds
as the capacitor is recharged to bring the DC bus voltage
close to nominal. When the load abruptly becomes inactive
(t =15 s), the capacitor absorbs energy as the voltage increases,
but this energy is subsequently removed over the next five
to ten seconds as the DC bus voltage is brought back to
nominal. Thus, this demonstrates the substantial contribution
of the FESS in delivering energy to the load for a powerconstrained scenario in which the AFE alone cannot supply the
required load power. This test also serves for commissioning
of the PHIL simulation capabilities with the FESS, exercising
the FESS, the PHIL interface, and the energy management
controls and models of the simulated surrounding system for
a typical scenario involving cycles of charging and discharging
and the associated voltage fluctuations.
Additional commissioning tests are planned, which will
test the FESS up to the full rated capability of 320 kW.
These commissioning tests are planned to be followed by a
set of PHIL demonstrations, involving different configurations
of energy magazine systems within different shipboard power
system architectures. The goals of the experiments are to
demonstrate the use of the FESS as a component of an energy
magazine system and to gain insights into the relative merits
of this approach, as well as to understand possible issues and
310
shortcomings. The outcomes of this work should generally help
to provide guidance for future efforts related to integration of
energy storage within shipboard power systems and should
contribute to addressing some of the risks identified within
the naval community.
MIL-STD-1399(Navy): Department of Defense Interface Standard, Section 300B: Electric Power, Alternating Current, U.S. Department of
Defense Std. MIL-STD-1399-300B, 2008.
[3] MIL-STD-1399 Section 680: Department of Defense Interface Standard,
Section 680: High Voltage Electric Power, Alternating Current, U.S.
Department of Defense Std. MIL-STD-1399-680, 2008.
[4] J. Langston, K. Schoder, M. Steurer, F. Bogdan, J. Hauer, D. Dalessandro,
and T. Fikse, “Evaluating energy storage applications for naval platforms
using hardware-in-the-loop testing,” in Advanced Machinery Technology
Symposium, 2016 ASNE. ASNE, 2016.
[5] J. Langston, M. Sloderbeck, M. Steurer, D. Dalessandro, and T. Fikse,
“Role of hardware-in-the-loop simulation testing in transitioning new
technology to the ship,” in Electric Ship Technologies Symposium,
Arlington, VA, April 22-24. IEEE, 2013.
[6] M. M. Steurer, K. Schoder, O. Faruque, D. Soto, M. Bosworth,
M. Sloderbeck, F. Bogdan, J. Hauer, M. Winkelnkemper, L. Schwager,
and P. Blaszczyk, “Multifunctional megawatt-scale medium voltage dc
test bed based on modular multilevel converter technology,” IEEE
Transactions on Transportation Electrification, vol. 2, no. 4, pp. 597–
606, Dec 2016.
[7] R. Kuffel, J. Giesbrecht, T. Maguire, R. Wierckx, and P. McLaren,
“RTDS-a fully digital power system simulator operating in real time,” in
WESCANEX 95. Communications, Power, and Computing. Conference
Proceedings. IEEE, vol. 2. IEEE, 1995, pp. 300–305.
[8] W. Ren, M. Steurer, and T. Baldwin, “Improve the stability and the
accuracy of power hardware-in-the-loop simulation by selecting appropriate interface algorithms,” Industry Applications, IEEE Transactions
on, vol. 44, no. 4, pp. 1286–1294, July 2008.
[2]
ACKNOWLEDGMENT
The authors would like to thank other team members at
FSU CAPS contributing to this work including John Hauer,
Ferenc Bogdan, Michael Coleman, Mark Stanovich, Michael
Sloderbeck, Dionne Soto, Matthew Bosworth, Isaac Leonard,
and CAPS director Roger McGinnis. The authors would also
like to acknowledge the contributions from the engineering
teams at GKN, project managed by Tim Rumney. The authors
would also like to thank Brian Lounsberry of Cardinal Engineering, along with Stephen Markle and Maria McLaughlin of
the US Navy Electric Ships Office (PMS320). The authors
would also like to thank Andy Tate and Chris Broadbent
of the UK Defence Science and Technology Laboratory and
Peter Deverill and Simon Lewinton of the UK MOD Defence
Equipment and Support Technology Office. This cooperative
work was sponsored through the OSD Coalition Warfare
Program and made possible by the US Navy Electric Ships
Office (PMS320) as part of the joint Advanced Electric Power
and Propulsion Systems Development Project (AEP3) Project
Arrangement (PA) with the U.K.
R EFERENCES
[1]
J. Kuseian, “Naval power and energy systems technology development
roadmap,” Electric Ships Office, PMS 320, Tech. Rep., 2015.
311
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 257 Кб
Теги
2017, ests, 8069298
1/--страниц
Пожаловаться на содержимое документа