Abstract - Telecom power systems are connected to grid,

diesel generator and batteries through an automatic transfer switch.

This system uses AC bus as the main bus and is connected to the grid
and diesel generator through automatic transfer switch to the main bus.
A typical telecommunication system has four different types of loads;
AC non-critical loads, AC Critical loads, AC ultra-critical loads and
DC critical loads. These loads need the battery backup time of 2 to 8
hours. An example of this load is switching equipments. In this thesis
a telecom power plant using local distributed micro sources (Fuel cell,
micro turbine, wind turbine and photovoltaic cell) and utility grid as
power input to the system is connected to a main DC bus using a
multiple input DC-DC convertor. Most of the micro sources have a
large response time which is the order of second. Some benefits of the
proposed architecture are eliminating the transfer switch, eliminating
the diesel generator and eliminating the battery bank. Another benefit
is that the critical and non-critical telecommunication loads are all
connected to the main bus. The micro turbine is connected to the DC
bus for improved availability. The proposed design uses a multiple
input positive Buck-Boost (MIPBB) convertor to connect all the micro
sources to the main DC bus. Multiple input convertors can combine
different energy sources with different current and voltage
characteristic. Some of the benefits of the proposed MIPBB convertor
are it uses minimum number of parts and it gives positive output
voltage without requiring any extra transformer. Besides, another
benefit of this converter is that it can accommodate up to n different
micro sources with n different voltage and current characteristics.
This converter can be operated in buck, Boost and Buck-Boost modes.
The proposed architecture reduces the energy storage requirements
and is more reliable than a traditional power plant the system
configuration is highly flexible and cost effective. This thesis
introduces a topology for the multiple input positive Buck-Boost
(MIPBB) convertors and explains some control approache. The
analysis is supported by simulation.
Keywords: Telecom power system, multiple input converters, power
system architecture, fuel cell model, micro-turbine model.
1. INTRODUCTION
Conventional telecommunication power systems are
composed of grid, diesel generator, batteries, and an automatic
transfer switch. These systems use AC bus as the main bus
and connect the grid and diesel generator through transfer
switch to the main bus. Most telecommunication systems use
24, 48, or 60V DC power plants that have several hours
(between 2 to 8 hours) of battery backup systems[1]. In these
systems, inverters or Uninterruptable Power Supply (UPS) are
utilized to supply the AC power from utility grid and diesel
generator. The main goal of utilizing diesel generator is to
provide uninterrupted power during long utility outages.
During the power outage, power is temporarily supplied to the
critical loads through batteries, and the automatic transfer
switch will switch to diesel generator. This conventional
system is relied on the diesel generator and batteries to
provide uninterruptable service during long power outage. A
conventional telecommunication power system is shown in
Fig.1.
A typical telecommunication system has four different types of loads; AC
non-critical loads, AC critical loads, AC ultra-critical loads, and DC critical
loads[1].
The AC non-critical loads can survive during the
power outage and they do not need any backup systems such
as UPS or diesel generator. Typically, these loads are
connected directly to the utility grid. The AC critical loads
need to be backed up by diesel generator; however they do not
need to be fed during short period of outages. These loads do
not need back up systems through UPS or batteries. One
example of these loads is air conditioning system. The AC
ultra-critical loads need redundant uninterruptable power
supply with at least 30 minutes of battery backup for the
power outages of more than 3 ms. Some examples of these
loads are network equipments and computers. The DC critical
loads are connected to a DC bus through redundant rectifiers.
These loads need the battery backup time of 2 to 8 hours. An
example of these loads is switching equipments [1].

Fig. 1. Conventional telecommunication power plant.
Although it is proven that the telecom power system
in Fig.1 is a reliable telecommunication power plant, there are
some drawbacks in this system [1]:
The automatic transfer switch is a single point of
failure in the systems.
While diesel generators are reliable if they equipped
with two starters, they require frequent and costly
maintenance Besides, diesel generators are source of
pollution and noise.
A battery bank with long backup time requires space;
and considering replacement and maintenance, battery banks
are one of the most expensive equipments in
telecommunication systems.
Battery banks typically occupy a large area and they
need to be replaced every 5 years[2]. Considering
maintenance, replacement cost, and the area that battery bank
Grid Connected to a Telecom Power System
1
#
S.Hidayathulla, Student (M.Tech), 2
#
V.Ravi Prasad, Associate professor,
3
#
Syed karimulla,Associate professor,4
#
P.S. Niranjan kumar,Senior,Asst professor.
1
st
Department of Electrical and Electronic Engineering.
E-mail: hidayathulla4u@g-mail.com; E-mail:vrpnaidu@gmail.com,E-mail:hodeee@alhabeebcollage.ac.in,E-mail:
niranjan.merits@gmail.com.
occupies; they are one of the most expensive equipments in a
telecommunication power plant. There are. many different
methods that try to address these issues and replace the diesel
generator and/or reduce or eliminate the battery bank.
II. PROPOSED ARCHITECTURE
The proposed system in this paper uses micro-
sources such as Micro-Turbine (MT) and Fuel Cell (FC). Most
of micro-sources have a large response time which is in the
order of seconds or minutes. In order to compensate for heir
slow response time, some sort of energy storage device with
fast response time such as Ultra-Capacitor (UC) is needed.
The UC needs to store power for only a few minutes. Some
benefits of the UC are their wide range of operating
temperature, long life, and superior performance [3]. Proposed
architecture is shown in Fig. 2.
In this system, not only micro-sources are used as
backup systems during the power outage but also they are
used for normal operation. This method reduces the
investment payback time for micro-sources. Some benefits of
the proposed architecture are eliminating the transfer switch,
eliminating the diesel generator and eliminating the battery
bank. Another benefit is that the critical and non-critical
telecommunication loads are all connected to the main bus.
This system supports all the load types (AC non-critical, AC
critical, AC ultra-critical, and DC critical) with high
redundancy and high reliability such that none of these loads
experience the power outage. This system can also be
connected to grid through an ACIDC inverter for higher
reliability and efficiency in the case that one of the
microsources goes out of service.
II.I The case for Eliminating Batteries:
For about 100 years lead acid batteries have been
the primary energy reserve system in D.C power plants used
in telecommunication applications. While this technology is
relatively successfully in telecom applications, there have
been longstandling issues with it.
There have been continuous to the performance
disputes between user and manufacturers, especially with
valve regulated lead-acid (VRLA) battery systems, and those
are compounded with. In the past, battery disposal was a non-
issue or was not monitored. Today, battery disposal represents
considerable maintenance and administrative cost that grows
in complexity and portends liability. Batteries also require
periodic maintenance and replacement that add to the annual
and life cycle cost of the power plant.
While power plant rectifier weight and volume have
decreased by an order of magnitude or more, battery weight
has changed little and remains the heaviest occupier of floor
space in the power plant. Because of battery weight, a power
plants floor loading is high requiring the plant to be placed on
specific floors of a building or undergo costly building
infrastructure modifications to place the battery on a floor
that does not have the intrinsic weight bearing capacity. In
addition, the battery footprint, somewhat reduced with the
introduction of valve-regulated technology, remains the
largest occupier of floor space in the power plant for most
applications. Lastly, the initial cost of the battery represents a
large percentage of the power plant total cost.
II.II Challenges to Eliminating Batteries:
There have been many attempts from different many
attempts from different approaches to reduce or eliminate
betteries in telecom applications, but each approach has run
into signicant obstacles. One principal obstacle is the use of
the utility grid as the prime source of power. While it is the
most available generating source that exists, it still only
provides 99% availability of electricity suitable for
information technology equipment. Standby diesel engine-
alternators do not help since even the most well maintained
machines fail to start 1% of the time and have a failure rate of
15% for 24-hour missions.
Several alternative energy alternative storage
technologies (i.e. flywheels, etc.) are now available on the
market, but none have proven economically viable for the
amount of reserve time required to compensate for the
unreliability of standby engine-alternators. Several telecom
companies are currently evaluating distributed generation
(DG) in the form of reciprocating engines, natural gas turbines
and fuel cells. However, when considered in a telecom
application for prime power, synchoronization of multiple DG
units and the associated complex switchgear severely limits
overall system availability. Plus, interconnection issues
continue to impede the acceptance of distributed generation in
the serves areas of many utilities.
The proposed architecture for telecommunication
power system in this paper uses DC bus as the main bus. The
reason behind using a DC bus is that most of the microsources
have DC output voltage and they can easily be connected to
this bus through a DC/DC converter. While micro-sources are
connected to this bus through DC/DC converters, grid can also
be connected through an AC/DC rectifier. There are three
different commonly used possible nominal voltages for this
DC bus (48V, 150V, and 300V). In the conventional
telecommunication systems, 48 V is used. For DC bus, 150V
is chosen since it has some advantages such as higher system
efficiency, less conductor losses, lower system cost, lower
rectifier cost, and less required copper over the conventional
48V [5]. Additionally, 150V is already used in some
traditional systems powering toll switches [5]. The 300V is
more suitable if there is a need to bidirectional power flow for
utility grid [5].
The load in telecommunication system varies from 3
kW in remote telecom systems, up to 250 kW and higher in
large telecom centers. The designed power plant in this paper
has the power rating of 22.5 kW. This topology can easily be
applied to other telecom systems by changing the size of fuel
cell and micro-turbine accordingly.
III. MULTIPLE INPUT BUCK - BOOST CONVERTER
The designed topology uses a Multiple Input Positive
Buck-Boost (MIPBB) converter to connect all the
microsources to the main DC bus. Multiple input converters
can combine different energy sources with different current
and voltage characteristic [7] with less number of parts. There
are many different multiple input converters available. Some
of the benefits of the proposed MIPBB converter are low
number of parts and positive output voltage without requiring
any extra transformer. Fig. 3 shows the proposed MIBB
converter. His converter uses one common inductor and one
capacitor for all of the inputs. Besides, another benefit of this
converter is that it can accommodate up to N different inputs
with N different voltage and current characteristics. This
converter can be used as a buck or buck-boost converter.
Here, the buck-boost operation is chosen since voltage of
micro-turbine is higher than the output and voltage of fuel cell
is lower than output. In order to drive this converter in buck-
boost mode 2, ... , N) and Q1 should be switching at the same
time.
Fig. 3. The proposed multi-input positive buck-boost converter.
In this thesis, FC and MT are chosen as two micro
sources. The proposed converter is used as a two input
converter. Generally, if another micro-source is added to the
system, same topology can be applied by adding one more
switch.
There are three different modes of operation for two
input buck-boost converter. Switch Q is ON whenever one or
more of the input switches are ON. Inputs are ordered from
low voltage to high voltage (V1 <V2). In first mode both S1 and
S2 are ON. In this mode the input with higher voltage turns
OFF the other switch. Hence, in mode 1 the input with higher
voltage (V2) is conducting. In second mode, switch S2 turns
OFF therefore in this mode, the lower voltage input (V2 )
isconducting. In mode 3, both S1 and S2 are turned OFF and in
this mode D1 and D2 are turned ON.
With this switching strategy each switch has a duty
cycle and an effective duty cycle. For S2 the duty cycle and
effective duty cycle are the same while for S1 the effective
duty cycle is;
Deff (I)= D1 - D2 (1)

Based on Eq. (I) D1 should be greater than D2 to be
able to transfer power from first input to the output. If the
number of inputs is increased to more than two, the equation
for effective duty cycle becomes;
Deff(i) = Di -
N
i 1 +
Deff( j) (2)
IV MODELING FUEL CELL AND MICRO-
URBINE
A fuel cells offer potentially attractive alternative
backup power solut6ions for telecom applications. Recent
major developments in polymerelectrolyte membrane fuel
cells (PEMFC) have resulted in multiple companies
introducing PEMFC-based products. However, there are
important differences in PEMFC technologies that clearly
differentiates these products and their key attributes (e.g.,
lifetime and cost). UTC fuel cells has recently employed their
proprietary6 PEMFC technology in a prototype 5-kW backup
power unit, which provides a seamless power transition
without the use of batteries.
A. Why Fuel Cells?
Fuel cells combine the best features of engines
and batteries; namely, like an engine they can operate for as
long as fuel is available and like a battery they produce
electricity directly from this fuel without combustion (which
can result in undesirable emissions and noise) and without
intermediate steps (which are inefficient). With all of these
desirable features, fuel cells will undoubtedly be replacing
engines and supplementing batteries in a variety of
applications in the future, as long as they can demonstrate that
they are at least equally reliable and are cost competitive. The
fact that fuel cells can potentially offer an attractively solution
for backup power applications is obvious, and an economic
justification for using hydrogen fuel cells in telecom
applications has been provided by others [1]. However, what
is not widely known outside the fuel-cell community is that
there are many different types of fuel cells and, within each
type, a wide variety of technologies have been developed.
This paper will briefly explain some these differences, as well
as why certain fuel cells are now close to commercial viability
for telecom applications.
B. A Very Brief History of Fuel Cells
Fuel cells were first invented over 160 years ago in
England by Sir William Grove. So, why are fuel cells still
considered a relatively exotic technology? Fundamentally,
there are no issue mountable technical obstacles that prevent
fuel cells from enjoying huge commercial success. The main
obstacle has undoubtedly been the cost of this technology. For
instance, fuel cells have been consistently employed for
manned space flight (e.g., the Gemini, Apollo, and Space
Shuttle missions) where a lightweight and very high
efficiency power plant (to reduce the amount of fuel and
oxidant required) is critical and cost is not an overriding
factor.
C. Major Types of Fuel Cells
There are five major types of fuel cells that have
been developed to varying degrees. They are differentiated
and named according to the electrolyte membrance fuel cells
(PEMFC), phosphoric-acid fuel cells(PAFC), molten-
carbonate fuel cells(MCFC). A higher operating temperature
is preferable if one wishes to use a fuel other than pure
hydrogen (e.g., a hydrogen-rich stream obtained from
reformed natural gas), since these fuel cells are less
susceptible to catalyst poisoning. However, at these
temperatures material degradation rates (e.g., corrosion) can
be a serious issue and cell lifetime is therefore a major
challenge. Nevertheless, UTC Fuel Cells (a subsidiary of
United Technologies Corporation, and henceforth referred to
as simply UTC) developed a stationary 200-kW power plant
that operators on natural gas using a PAFC stack that routinely
runs for greater than 40,000 hours; This product, the Pure Cell
TM
200 (formerly known as the PC-25), has been successfully
deployed all over the world since it was commercially
introduced in 1992. The higher temperature fuel cells, namely
MCFC and SOFC, are still under development, although
power plants employing both technologies have been
demonstrated, the primary barrier to their widespread
introduction continues to be limited life of the fuel-cell stacks.
D. Fuel Cell Model
Fuel cells are known to have higher efficiency than
conventional power plants [8]. Fuel cells are environmentally
friendly (environmentally clean), have extremely low
emission and they produce very low noise [8]. Fuel cells are
composed of fuel processing unit (reformer), the fuel cell
stack, and a power conditioning unit [10]. Hydrogen is
produced in the reformer by processing any hydrocarbon fuel
[8]. Some examples of hydrocarbon fuels are propane,
methane, and methanol. The produced hydrogen is supplied to
the fuel cell stack. Using provided hydrogen, and oxygen from
the air; and through the electrochemical process, fuel cell
stack produces electricity and water.
Due to low operating temperature and rapid start up
time, Proton Exchange Membrane Fuel Cell Power Plants
(PEM FCPPs) are typically the best candidate for
telecommunication power systems. A model for 27 kW PEM
FCPPs) is developed. The fuel cell model that is used in this
paper is developed in MATLAB & Simulink. Then, this
model is implemented in Simpower System as a controlled
voltage source [8] and [9].
The relationship between the modular flows of any
gas through the valve is proportional to its partial pressure
inside the channel [11]. For hydrogen, this relationship can be
expressed as follows

2
2
2
2
2
H
K
H
M
H
K
H
p
H
q

(3)
The molar flow of hydrogen that reacts can be found
from the basic electrochemical relationship between hydrogen
flow and the fuel cell system current [9]
FC
I
r
K
F
FC
I
stack
N N
H
r
q ' 2
2
'
0
2
(4)
The hydrogen partial pressure can be obtained by applying
Laplace transform on (3) and (4) [9],
PH2 =

,
_

+
FC
I
kr
H
in
q
H
H
K
' 2
2
2
1
2
1

(5)
Where
RT
H
K
an
V
H
2
2

(6)

Similar operation can be done for water and oxygen partial
pressure. The Nernsts instantaneous voltage can be expressed
as eq. (7) [12].

1
1
]
1

1
1
]
1

+
O
H
P
o
P
H
P
F
RT
E
stack
N N E
2
2 2
log
2
0
0
(7)
The reformer generates hydrogen for stack operation
of fuel cell. The fuel cell system consumes hydrogen based on
the power demand. Mathematical equation of the reformer
model can be expressed as [9].
1 )
2 1
(
2
2 1
2
+ + +

s r r s r r
CV
methanal
q
H
in
q
(8)
Which describes the relationship between the input
gas methanol and the produced hydrogen.
Assuming constant temperature and oxygen
concentration, the fuel cell output voltage can be obtained
from the sum of Nemsts voltage, the activation over voltage
and the ohmic over voltage [9],
Vcell = E

act

ohmic (9)

Where

act = Bin(CIFC) (10)

ohmic = s
int
IFC (11)
Based on Eqs.(3) to (11), the developed model for fuel cell is
shown Fig. 4.
Fuel cell model. The fuel cell that is modeled here is a
48V, 27kW, PEM fuel cell. Each cell has the voltage of 0.8V
and each stack in fuel cell has a current capacity of 94A. The
fuel cell model consists of six parallel stacks and each stack
has 88 cells in series
E. Micro-Turbine Model
The operation principles of a micro-turbine are
similar to a steam turbine. Compressor, combustion chamber,
and a turbine are the three main components of a micro-
turbine. The air after being compressed in compressor forms a
combustible mixture by getting mixed with the injected fuel.
The mixture is ignited in combustor to produce heated air.
Then the heated air is expanded in turbine to drive the turbine
[8]. There are two major types of micro-turbine; single shaft
and split shaft [13]. In single shaft micro-turbine electric
generator and turbine are mounted in same shaft. In split shaft
micro-turbine, a gear box is used to couple the electric
generator to the micro-turbine. Model is developed for a
37kW, 300V micro-turbine [8].
Fig. 5 (a) shows a single shaft micro-turbine. In
single shaft micro-turbine compressor, turbine, and electric
generator are all on the same shaft. The turbine speed is in the
range of 50,000 to 120,000 rpm. The frequency of produced
voltage is in the range of 1500 to 4000 Hz [8]. A cyclo -
converter is required to reduce the frequency to 60 Hz. In split
shaft model, the generator shaft is connected to turbine shaft
through a gear box. The gear box is used to reduce the
generator speed to 3600 rpm. If the synchronous generator has
two poles, the frequency will be 60 Hz and there is no need
for a cyclo-converter.
Fig.5 (a) Single Shaft micro turbine schematic and
(b) split shaft micro turbine schematic.
A split shaft micro-turbine is modeled based on the
principle explained in [8]. In order to model the micro-turbine,
synchronous generator with speed and power controller is
used. A simplified synchronous machine model of Simpower
System is utilized as a synchronous generator driven by the
mechanical power produced by the turbine model.
Three controllers (speed, power, and load controls)
are used to control the synchronous machine for its operation
as a micro-turbine. The speed controller is defined by,

1
]
1

+
s
k
K
Ks
r ref n i
) (
(12)
The power controller is defined by,

1
]
1

+
s
i
K
f
K p
ref
p
n i
p ) (
(13)
Finally, the load controller is defined by,
max
3
1
1
1 max
L
t
K
s T
P L
in
L +
+

1
]
1

(14)
The minimum of

in, Pin, and Lin is defined as Cv which is the

output of a Low Value Gate (LVG) in the proposed model as
Cv = min{

in, Pin,Lin} (15)

The mechanical input power is,
Pm = p1 -

rDturbine (16)

where

1
1
]
1

+ +

1 )
2 1
(
2
2 1
1
1
s T T s T T
v
C P
(17)
P1 is limited by the minimum and maximum valve
position. Additionally, excitation voltage control is used in the
micro-turbine model in order to keep the synchronous
generator output voltage within a reasonable band. In this
control, the measured and reference phase to phase voltage
values of the generator voltage is compared. The difference is
then processed through a PI controller which determines the
excitation voltage to be applied to the generator.
A Proportional Integral controller (PI controller) is
employed to regulate the voltage of DC bus at 150 V. The
proportional gain is set to KP =0.01 and the integral gain
is set to Ki =100.
V. RESULTS
The overall system configuration is demonstrated in
Fig. 8.
Fig.6 Over all system configuration
Fuel cell (Fe) and micro-turbine (MT) are used as
two input sources for the multiple input positive buck-boost
(MIPBB) converter. The inductance of MIPBB converter is 2
H and the capacitance is 5mF . The switching frequency is
50 kHz. In order to analyze the dynamic behavior of the
system, step changes in load variations are investigated. This
system is designed for the load of 22.5 kw. At the time of step
load change, the output load is increased to 33.6 kW at time
0.3s (increases about 50% of demanding power) and reduced
to 11.25 kW at time o.7s (decrease about 50% of demanding
power).
Figs. 9 and 10 show the voltage and current of fuel
cell respectively. When the load increases, the voltage of fuel
cell is dropped about 2 volts and the current of fuel cell is
increased to provide more power to the load. The reverse is
hold for decrease in demanding power.
Fig. 11&12 shows the voltage and current of
microturbine. When the load increases, the current of micro-
turbine is increased instantly to provide the extra demanding
power at the load. Again, when the load decreases, the current
of micro-turbine is decreased to compensate the less power
that is needed at the load bus.
Fig. 13 and 14 shows the voltage and current at the
DC bus respectively. The voltage at DC bus is regulated at
150 V: In transient conditions such as step change In
load,there IS a very small oscillation in voltage. This figure
shows the robustness and effectiveness of the voltage mode
controller that is used. In transient condition, there is no
voltage drop or increase at the DC bus and the only change is
in the current of DC bus.
Fig. 15 shows the power of FC, MT, and output
power respectively. Initially, when the demanding power is
22..5 kW, 9.5 kW of this power (about 42% of the power) IS
supplied by FC and the rest (about 13.75 kW) ~f demanding
power is supported by MT. When the demanding power IS
increased to 33.6 kW, 12 kW of this power (about 36% of the
power) is supplied by FC and 22.25 kW is supported by MT.
When the demanding power is decreased to 11.25 kW, 6.65
kW of this power (about 59% of the power) is supplied by FC
and 4.7 kW is supported by MT. This figure shows how the
output power is shared between the FC and MT. A new
topology for telecom power system is introduced in this paper.
The new system was primarily powered by micro-sources, Le.
MT and FC. and 4.7 kW is supported by MT. This figure
shows how the output power is shared between the FC and
MT.
Conclusions
A new topology for telecom power system is
introduced in this thesis. The new topology has the advantages
of eliminating batteries, diesel generator, and transfer switch.
A multiple input positive output buck boost de/dc converter
has been introduced in this manuscript for utilizing
microsources (i.e., MT and FC). Modeling of PEM PC and
MT was described. Besides, the topology, operating
principles, and control system of a multiple input positive
buck boost converter were introduced, this converter is used
to connect multiple inputs with different voltage current
characteristics to a single output. The telecom station is
powered by the proposed topology, furthermore, using a dc/ac
inverter and appropriate filters, the proposed topology is
connected to a distribution network. The developed filters
exhibit excellent performance in the harmonic reduction of the
inverter output. Two control modes used for the multiple
input buck boost converter is also tested and results showed
the effectiveness and the success of the system. Dynamic
behavior of the proposed topology for load transients
demonstrated. Presented results show that the proposed
topology is an enabling technology for future telecom power
systems.
REFERENCES
1. M. Fraise and L. Nussbaum, "Environment friendly high quality, high
availability telecom power plant architecture," International
Telecommunication Energy Conference, 2002, pp. 463-469.
2. W. Allen and S. Natal, "Achieving ultra-high system availability in a
battery-less -48Vdc power plant," International Telecommunication
Energy Conference, 2002, pp. 287-294.
3. B. Maher, "High reliability backup for telecommunications using ultra
capacitors," International Telecommunication Energy Conference, 2004,
pp. 623-625.
4. M. Perry and S. Kotso, "A back-up power solution with no batteries,"
International Telecommunication Energy Conf., 2004, pp. 210-217
5. A.Kwasinski and P. Krein, "A microgrid-based telecom power system
using modular multiple-input dc-dc converters," International
Telecommunication Energy Conference, 2005, pp. 515-520.
6. F. Massardo, C. F. McDonald and T. Korakianitis, "Microturbine/fuel-
cell coupling for high-efficiency electrical power generation,"
Transaction ofthe ASME, vol. 124, Jan. 2002, pp.ll0-116.
7. Khaligh, "A multiple-input dc-dc positive buck-boost converter
topology," in Proc. IEEE Applied Power Electronics Conference,Austin,
TX, Feb. 2008.
8. M. Uzunoglu, O. Onar, M. Y. EL-Sharkh, N. S. Sisworahardjo, A.
Rahman, and M. S. Alam, "Parallel operation characteristics of PEM fuel
cell and microturbine power plants," Journal of power sources vol. 168,
no. 2, 2007, pp. 469-476.
9. M. Uzunoglu and M. S. Alam, "Dynamic modeling, design, and
simulation of a combined PEM fuel cell and ultracapacitor system for
stand-alone residential applications," IEEE Transactions on Energy
Conversion, vol. 21, no. 3, Sept. 2006, pp. 767-775.
10.E. Larmine and A. Dicks, Fuel cell systems explained, second edition,
John Wiley and sons, 2000.
11.M. Y. EL-Sharkh, A. Rahman, M. S. Alam, P. C. Byrne, A. A. Sakla, and
T. Thomas, "A dynamic model for a stand-alone PEM fuel cell power
plant for residential application," Journal ofpower sources, vol.
138,2004,pp.199-204.
12.Padulles, G. W. Ault, and J. R. McDonald, "An integrated SOFC plant
dynamic model for power systems simulation," Journal ofPower Sources,
vol. 86, 2000, pp. 495-500.
13.Y. Zhu and K. Tomsovic, "Development of models for analyzing the load-
flowing performance of microturbines and fuel cell," Electric Power
System Research, vol. 62, 2002, pp. 1-11.
14.M. Nagpal, A. Moshref, G. K. Morison, and P. Kundur, "Experience with
testing and modeling of gas turbines," Power Engineering Society Winter
Meeting, vol. 2, 2001, pp. 652-656.
15.P. T. Krein, Elements of Power Electronics, Oxford university press,
1998.
16.W. Ernst and J. Nerschook. Telecoms Networks: the new rules of
power, The fuel cell review, Vol. I issue 1, pp 25-28 (2004). Complete
article can be.
http//fcr.org..articles/features/1/6/5/1