Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Design Hybrid Solar-Wind-Diesel Power Sys

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

ONE OPTIMAL SIZING METHOD FOR DESIGNING HYBRID SOLAR-WIND-DIESEL

POWER GENERATION SYSTEMS

Zhou Wei, Yang Hongxing


The Hong Kong Polytechnic University
Hung Hom, Kowloon,
Hong Kong, China
w.zhou@polyu.edu.hk, behxyang@polyu.edu.hk

increase overall energy output and reduce energy storage


ABSTRACT requirements significantly. However, with the increased
complexity, the optimum design of hybrid system becomes
This paper recommended an optimal design model for
complicated through uncertain renewable energy supplies
designing stand-alone hybrid solar-wind-diesel systems.
and load demand, non-linear characteristics of the
The optimum configuration ensures that the annualized cost
of the systems is minimized while satisfying the load components, and the fact that optimum configuration and
demand for 5-year period from 1996 to 2000, result in zero optimum control strategy of the system are interdependent.
load rejection with minimum cost. The eight decision
variables included in the optimization process are the PV Various optimization techniques, such as probabilistic
module number, PV module slope angle, wind turbine approach, graphical construction method and iterative
number, wind turbine installation height, battery number, technique, have been recommended by researchers [1-4] for
diesel generator (DG) power, DG starting and stopping renewable energy system designs. Besides these
points. The objective function (annualized cost of system)
optimization techniques for designing solar and/or wind
minimization is implemented by using genetic algorithms,
which has the ability to achieve the global optimum systems, also some diesel generator control strategies were
solution with relative computational simplicity. found [5] for the designing of power generation systems
including diesel generators.

1. INTRODUCTION In this paper, one optimum design method for hybrid


solar-wind-diesel system is developed. The minimization of
A drawback, common to solar and wind power generations,
system cost is achieved not only by selecting an appropriate
is their unpredictable nature and dependence on weather
system configuration, but also by finding a suitable control
and climatic changes. Both of these would have to be
strategy (starting and stopping point) of the diesel generator
oversized to make their stand alone systems completely
and suitable settings for it. The configuration and control
reliable. Therefore, depending on the requirement and the
availability of energy sources, more than two sources strategy of the hybrid system that meet the load
maybe combined, such as solar-wind-Diesel system. This requirements with minimum cost can be obtained by an
kind of hybrid system can attenuate individual fluctuations, optimization technique - the Genetic Algorithm (GA) [6].
1490 Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement

to tailor the PV array output. If a matrix of Ns×Np PV


2. MODEL OF THE HYBRID SYSTEM COMPONENTS modules is considered, the maximum power output of the
PV system can be calculated by:
Typical stand-alone hybrid solar-wind-diesel power
generation system (see Fig. 1) consists of PV array, wind PPV = N p ⋅ N s ⋅ Pmod ule (2)
turbine, diesel generator, battery bank, inverter, rectifier,
controller, and other accessory devices and cables. In order 2.2 Wind Turbine Performance Model
to predict the hybrid system performance, individual
components need to be modeled first and then their mix can
The wind power law [8] has been recognized as a useful
be evaluated to meet the load demand.
tool to transfer the anemometer data recorded at certain
levels to the desired hub center:

ξ
⎛H ⎞
v = vr ⎜ WT ⎟ (3)
⎝ Hr ⎠

The wind turbine power output can be simulated by:

⎧ v − vC
⎪ PR ⋅ (vC İ v İ vR )
⎪ v R − vC
⎪⎪
PWT = ⎨ PR (vR İ v İ vF ) (4)

Fig. 1: Diagram of the stand-alone hybrid system. ⎪
⎪0 (v İ vC and v ı vF )
⎪⎩
2.1 PV Array Performance Model

2.3 Battery Performance Model


PV module performance is highly influenced by weather
conditions, especially solar radiation and PV module
Most battery models focus on three different characteristics.
temperature. A simplified simulation model [7] is used to
The first and most commonly used model mainly focuses
estimate the actual performance of PV modules under
on the battery state of charge (SOC). The second type of
varying operating conditions.
model is the voltage model, which is employed to model
the terminal voltage. The third type of model is the lifetime
model used for assessing the expected lifetime of the
Voc ⎛ Voc ⎞
− ln ⎜ + 0.72 ⎟ battery.
Pmod ule =
nKT / q ⎝ nKT / q ⎠ ⋅ ⎛1 − Rs ⎞
Voc ⎜ ⎟
1+ ⎝ Voc / I sc ⎠ (1) Battery State-of-Charge (SOC)
nKT / q (1)
α γ
⎛G⎞ Voco ⎛T ⎞ For a perfect knowledge of the real SOC of a battery, it is
⋅ I sco ⎜ ⎟ ⋅ ⋅⎜ o ⎟
⎝ 0 ⎠ 1 + β ln Go
G ⎝T ⎠ necessary to know the initial SOC, the charge or discharge
G time and the current. Taking the losses into consideration,
the SOC of the battery can be simply calculated by:
PV modules represent the fundamental power conversion
unit of a PV system, but a single PV module has limited ⎛ σ ⋅ Δt ⎞ I bat ( t ) ⋅ Δt ⋅ηbat
SOC (t + 1) = SOC (t ) ⋅ ⎜1 − ⎟+ (5)
potential to provide power at high voltage or high current ⎝ 24 ⎠ ′
Cbat
levels. It’s then mandatory to connect PV modules in series
and in parallel in order to scale-up the voltage and current
4 PV TECHNOLOGIES, SYSTEMS AND APPLICATIONS 1491

where σ is the self-discharge rate which depends on the regardless of how much or how little it is used.
accumulated charge and the battery state of health [9]; C’bat
is the battery nominal capacity, Ah. ηbat is the battery Ybat = min (Ybat c , Ybat
, , f ) (9)
charging and discharging efficiency.

2.4 Diesel generator control strategy


(2) Battery Floating Charge Voltage

The battery floating charge voltage response can be For the studied system, the diesel generator is started at
modeled by the equation-fit method: times when the battery SOC falls below a certain level
(diesel generator starting point), and then the diesel
Vbat = A × ( SOC )3 + B × ( SOC ) 2 + C × SOC + D (6) generator runs at full power (or at a rate not exceeding the
maximum current that batteries are capable of absorbing) to
charge the batteries with any surplus power until the battery
where A, B, C and D are functions of the battery current I:
SOC reach the diesel generator stopping point. Determining
the best value for these set points (diesel generator
⎛ A ⎞ ⎛ a1 a2 a3 ⎞ 2
⎜ ⎟ ⎜ ⎟⎛ I ⎞ starting/stopping point) is the key to achieve an optimum
⎜ B ⎟ = ⎜ b1 b2 b3 ⎟ ⎜ ⎟
I (7) operation.
⎜ C ⎟ ⎜ c1 c2 c3 ⎟ ⎜ ⎟
⎜ ⎟ ⎜ ⎟ 1⎟

⎝ D ⎠ ⎝ d1 d2 d3 ⎠ ⎝ ⎠
3. ECONOMIC MODEL BASED ON ACS CONCEPT
By fitting the equations to battery performance data, the
parameters a1, a2, a3 …… d1, d2, d3 can be calculated using The economical approach, according to the concept of
Least Squares Fitting method. Annualized Cost of System (ACS), is developed to be the
best benchmark of system cost analysis in this study. The
(3) Battery Lifetime ACS is composed of the annualized capital cost Cacap, the
annualized replacement cost Carep, the annualized
Two independent limitations on the lifetime of battery maintenance cost Camain, and the annualized fuel cost CFuel:
banks (the battery cycle life Ybat,c and the battery float life
Ybat,f) are employed. ACS = Cacap ( PV + Wind + Bat + Tower + DG ) + Carep ( Bat )
+ Camain ( PV + Wind + Bat + Tower + DG ) + CFuel ( DG )
Battery cycle life Ybat,c is the length of time that the battery (10)
will last under normal cycles before it requires
replacement:
3.1 Annualized Capital Cost

1
Ybat c = (8) The annualized capital cost of each component is:
⎛ SOC1i − SOC2i ⎞
,
n
1
∑ ⎜⎜
DODi
⋅ ⎟
C ( DODi ) ⎟⎠
i= ⎝
1
Cacap = Ccap ⋅ CRF ( i, Yproj ) (11)

where n is the number of individual cycles that happens


where Ccap is the initial capital cost of each component,
during the typical year; SOC1i and SOC2i are the battery
US$; Yproj is the component lifetime, Year; CRF is the
starting and stopping SOC for the ith discharging process;
capital recovery factor:
DODi is the depth of discharge, it equals 1-SOC2i; C is the
number of cycles until end-of-life as specified in Y
i ⋅ (1 + i ) proj

datasheets. CRF ( i, Yproj ) = Y


(12)
(1 + i ) proj
−1
The battery float life Ybat,f is the maximum length of time
that the battery will last before it needs replacement The annual real interest rate i is related to the nominal
1492 Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement

interest rate i' (the rate at which you could get a loan) and total energy production originating from renewable
the annual inflation rate f by the equation given below. sources:

i′ − f f RE =
PRE
× 100% (17)
i= (13) PRE + PDG
1+ f

where PRE and PDG are the renewable energy and diesel
3.2 The Annualized Replacement Cost
power, kWh.

In the studied hybrid system, only the battery needs to be


If a renewable energy fraction target fRE’ is set, any system
replaced periodically during the project lifetime.
configurations, whose renewable energy fraction falls
below the prearranged target, will be rejected as infeasible.
Carep = Crep ⋅ SFF ( i, Yrep ) (14)

where Crep is the replacement cost of the component 5. OPTIMUM DESIGN WITH GENETIC ALGORITHM
(battery), US$; Yrep is the component (battery) lifetime,
Year; SFF is the sinking fund facto: Minimization of the objective function (ACS) is
implemented employing Genetic Algorithm (GA). The flow
i
SFF ( i, Yrep ) =
chart of the optimization process to obtain the optimum
Yrep
(15)
(1 + i ) −1 configuration and control strategies is illustrated in Fig. 2.

The system maintenance cost is deemed to be constant


every year for each component.

3.3 Fuel Cost

According to Skarstein and Uhlen [10], an assumed linear


characteristic between the fuel consumption cost and the
generator power level is a good approximation:

⎡ Popr ⎤
CFuel = ⎢ F0 + ( FR − F0 ) ⋅ ⎥ ⋅cf
(16)
⎣ PR ⎦

where CFuel is the fuel consumption cost of the diesel


generator, US$/h; cf is the fuel price, US$/l; F0 is the fuel
consumption of the diesel generator under no load
conditions, l/h; FR is the fuel consumption under rated
power conditions, l/h.

The configuration with the lowest Annualized Cost of


System (ACS) is taken as the optimal one from the
configurations that can guarantee the load requirements. Fig. 2: Flow chart of the optimum design model.

The initial assumption of system configuration and control


4. RENEWABLE ENERGY FRACTIONS strategy settings will subject to the following inequalities
constraints:
The renewable energy fraction is the portion of the system's
4 PV TECHNOLOGIES, SYSTEMS AND APPLICATIONS 1493

Min ( N PV , N wind , N bat ) ≥ 0 (18) requirement, the optimum configuration and control
strategy can be identified from the set of configurations.
H low ≤ HWT ≤ H high (19)

6. CONCLUSIONS
0 ≤ β ′ ≤ 90 (20)

Based on Genetic Algorithm (GA), which has the ability to


0 < DGSOC1 < DGSOC 2 < 1 (21)
attain the global optimum with relative computational
simplicity, one novel sizing method for hybrid
where DGSOC1 and DGSOC2 are the diesel generator starting
solar-wind-diesel system is developed in this paper to
and stopping point respectively.
calculate the system optimum configuration that satisfies
An initial population of 10 chromosomes, comprising the the load demand with minimum annualized cost of system.
1st generation, is generated randomly and the constraints Minimization of the system cost is achieved not only by
described by inequalities (18)–(21) are evaluated for each selecting an appropriate system configuration, but also by
chromosome. finding a suitable control strategy. The decision variables
included in the optimization process are the PV module
The PV array power output is calculated according to the number, PV module slope angle, wind turbine number,
PV array performance model by using the specifications of wind turbine installation height, battery number, diesel
PV module as well as ambient air temperature and solar generator (DG) type, DG starting and stopping points.
radiation conditions. The wind turbine performance
calculations need to take into account the effects of wind
turbine installation height. The battery bank, with a total 7. REFERENCES
nominal capacity C’bat (Ah), is permitted to discharge up to
a limit defined by the maximum depth of discharge (DOD), (1) Tina, G., Gagliano, S., Raiti, S. Hybrid solar/wind
which is specified by the system designer at the beginning
power system probabilistic modeling for long-term
of the optimization process. Battery SOC will then be
performance assessment. Solar Energy 2006; 80:
compared with the diesel generator starting and stopping
578-88.
points to control the on/off operation of the diesel
generator. (2) Borowy, B.S., Salameh, Z.M. Methodology for
optimally sizing the combination of a battery bank and
And then, the system design will be optimized by PV array in a wind/PV hybrid system. IEEE
employing Genetic Algorithm, which dynamically searches Transactions on Energy Conversion 1996; 11(2):
for the optimal configuration and control strategy to 367-73.
minimize the annualized cost of system (ACS). For each (3) Yang, H.X., Burnett, L., Lu, J. Weather data and
system configuration, the system’s renewable energy probability analysis of hybrid photovoltaic–wind power
fraction fRE will be examined for whether the required generation systems in Hongkong. Renew Energy 2003;
renewable energy fraction target fRE’ can be reached, i.e., fRE 28: 1813–24.
should be higher than fRE’. Then the lower cost renewable (4) Yang, H.X., Lu, L., Zhou, W. A Novel Optimization
energy fraction target satisfied configurations, will be
Sizing Model for Hybrid Solar-Wind Power Generation
subject to the following crossover and mutation operations
System 2007; 81(1):76 - 84.
of the GA in order to produce the next generation
(5) Ashari, M. and Nayar, C. V. An optimum dispatch
population until a pre-specified number of generations has
strategy using set points for a photovoltaic
been reached or when a criterion that determines the
convergence is satisfied. (PV)-diesel-battery hybrid power system. Solar Energy
1999; 66 (1): 1-9.
By achieving the lowest ACS while satisfying the fRE’ (6) Mitsuo Gen and Runwei Cheng. Genetic Algorithms
1494 Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement

and Engineering Design. Wiley-IEEE 1997. (9) Guasch, D., Silvestre S. Dynamic battery model for
(7) Zhou W, Yang HX and Fang ZH. A Novel Model for photovoltaic applications. Progress in Photovoltaics:
Photovoltaic Array Performance Prediction. Applied Research and Applications 2003; 11: 193-206.
Energy. 2007 (In press) (10) Skarstein O., Uhlen K. Design considerations with
(8) Paul Gipe. Wind energy comes of age. John Wiley and respect to long-term diesel saving in wind/ diesel
sons. 1995: P536. Wind Engineering 1989; 13: 72-87.

You might also like