Professional Documents
Culture Documents
www.elsevier.com/locate/solener
a
Departamento de Ingeniera Electronica, de Sistemas Informaticos y Automatica, Universidad de Huelva, Spain
b
Departamento de Fsica Aplicada, II, Universidad de Malaga, Spain
Received 16 May 2005; received in revised form 1 March 2006; accepted 15 June 2006
Available online 24 August 2006
Abstract
The operating point of a photovoltaic generator that is connected to a load is determined by the intersection point of its characteristic
curves. In general, this point is not the same as the generators maximum power point. This dierence means losses in the system per-
formance. DC/DC converters together with maximum power point tracking systems (MPPT) are used to avoid these losses. Dierent
algorithms have been proposed for maximum power point tracking. Nevertheless, the choice of the conguration of the right converter
has not been studied so widely, although this choice, as demonstrated in this work, has an important inuence in the optimum perfor-
mance of the photovoltaic system. In this article, we conduct a study of the three basic topologies of DC/DC converters with resistive
load connected to photovoltaic modules. This article demonstrates that there is a limitation in the systems performance according to the
type of converter used. Two fundamental conclusions are derived from this study: (1) the buckboost DC/DC converter topology is the
only one which allows the follow-up of the PV module maximum power point regardless of temperature, irradiance and connected load
and (2) the connection of a buckboost DC/DC converter in a photovoltaic facility to the panel output could be a good practice to
improve performance.
2006 Elsevier Ltd. All rights reserved.
Keywords: Photovoltaic module; DC/DC converter; IV curve; Maximum power point tracker; Losses
0038-092X/$ - see front matter 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2006.06.006
32 J.M. Enrique et al. / Solar Energy 81 (2007) 3138
Nomenclature
" " . # #
type of converter used, which signicantly aects the pho- V IRS V
IRnPS
q nS nP AKT nS
tovoltaic systems performance. I nP I L I s e 1 1
RP
MPPT is used in PV power systems to force the PV mod-
ule operating at MPP. In this way the PV module produces P I V 2
the maximum power output. For this operating point, it " " . # #
q V P RS
nS V nP AKT
V
VP RS
overcomes the disadvantages of high initial installation costs P nP V I L I s e 1
nS nP
3
and low energy conversion eciency. Previously-used meth- RP
ods of achieving MPPT include: (1) incremental conduc-
dP
tance (IncCond); (2) perturbation and observation (P&O); 0 4
dV MPP
(3) neural network and (4) curve-tting (Hua et al., 2003).
At present there are numerous works aimed at designing The nP and nS parameters indicate the number of cells
MPPT systems (Bahgat et al., 2004; Enslin et al., 1997; Gar- connected in parallel and in series, respectively; RP and R
ca and Alonso, 2000; Hua et al., 2003; Kitano et al., 2003; S, are the intrinsic parallel and series resistances associated
Masoum et al., 2002; Neto et al., 2000; Schilla et al., 2000; to the panel; K is the Boltzman constant (1.38 1023 J/K)
Veerachary et al., 2002, 2003; Yu et al., 2004), where the e- and q is the charge on an electron. Factor A determines the
ciency of each of them is shown and comparatives of the dif- deviation of the characteristics of an ideal pn junction,
ferent methods of MPP tracking are established under and IS is the reverse saturation current, which presents a
dierent operating conditions. However, the choice of the dependence on the panel temperature. IL represents the
appropriate DC/DC converter for the implementation of current (photo-current) generated by solar radiation (G).
both the MPPT system and its integration in the facility Such a current shows a linear relation with regard to radi-
array has not been explicitly studied, despite its aecting sig- ation and temperature.
nicantly the optimum operation of the photovoltaic system. Eq. (1) (considering the dependence of its parameters on
The aim of this work is to make a comparative of the T and G) provides the so-called IV curves of a photovol-
photovoltaic system performance using the three basic taic panel, and the multiplication result of both magnitudes
topologies of DC/DC converters and MPPT tracker, so provides the supplied power: Eqs. (2) and (3). This curve
that it may be possible to make a decision on the best con- changes depending on the incident irradiance and the cell
guration to be used. This work is divided into the follow- temperature. Each curve presents a maximum power point
ing sections: Sections 2 and 3 present some characteristics (MPP, point of coordinate VP), which provides the optimal
and properties of photovoltaic modules and DC/DC con- operation point for an ecient use of the panel (Hohm and
verters, especially as regards the input impedance that they Ropp, 2002; Hua and Shen, 1998).
present under certain operating conditions. The analysis The MPP is calculated solving Eq. (3) with the condition
and results for each conguration are shown in Sections 4 (4). This calculation is tedious and slow, since these expres-
and 5. Finally, some conclusions are drawn in Section 6. sions do not have an analytical solution, and therefore,
they have to be solved by numerical methods (i.e., New-
2. Theoretical models of solar arrays tons method). Other two important points of this curve
are the open-circuit voltage (Voc) and the short-circuit cur-
A simplied exponential expression (Gow and Manning, rent (Isc). The voltage in an open circuit represents the
1999) describes the relationship between voltage (V) and maximum voltage given by the panel to a zero current
current given by a module, Eq. (1). (without load), while the short circuit current represents
J.M. Enrique et al. / Solar Energy 81 (2007) 3138 33
Fig. 2. DC/DC converters commonly used and their input resistance. (a) Buck Converter; (b) boost converter; (c) buckboost converter; (d) input
resistance versus d in CCM; (e) input resistance versus d in CCM and (f) input resistance versus d in CCM.
r!2
RL 4K
Ri-DCM 1 1 2 P RL 9
4 d
r!2
RL 4K
lim Ri-DCM lim 1 1 2 10
d!1 d!1 4 d
In DCM K 6 K crit 1 d, then:
0 s12
RL @ 41 dA
lim Ri-DCM 6 lim 1 1 RL 11
d!1 d!1 4 d2
Fig. 3. Location of the operation point of a photovoltaic module. From (9) and (11) we have:
RL
lim Ri-CCM lim 1 6 lim Ri-DCM RL 12
d!0 d!0d2 d!1
RL Being the expressions of Ri continuous in d, for a scan-
lim Ri-CCM lim 2 RL 7
d!1 d!1 d ning of the converters duty cycle d 2 [0,1], Ri takes values
r!2 that belong to the interval [RL,1), being RL the load resis-
RL 4K
lim Ri-DCM lim 1 1 2 1 8 tance. If RMPP does not belong to the set of values allowed
d!0 d!0 4 d for Ri, the capture of MPP will not be possible, thus
J.M. Enrique et al. / Solar Energy 81 (2007) 3138 35
Fig. 4. Chart of MPP tracking with buck DC/DC converter. Fig. 5. Chart of MPP tracking with boost DC/DC converter.
Fig. 9. Maximum power point impedance trajectory RMPP(t) for the Fig. 11. Maximum power point power trajectory PMPP(t) and power
SX60 (BP) module for a clear day in Malaga (Spain). supplied P(t) to the 5 X and 20 X loads, with buck converter between the
photovoltaic module and the load.
Fig. 10. Maximum power point power trajectory PMPP(t) and power Fig. 12. Maximum power point power trajectory PMPP(t) and power
supplied P(t) to the 5 X and 20 X loads, without DC/DC converter supplied P(t) to the 5 X and 20 X loads, with boost converter between the
between the photovoltaic module and the load. photovoltaic module and the load.
vation P&O (Hohm and Ropp, 2002; Hua and Shen, Finally, when a buckboost converter is used the
1998; Hussein et al., 1995). PMPP(t) and P(t) trajectories are graphically equal, with
Fig. 10 shows the trajectories of the power supplied by the values of 0.999 for the MPP-tracking eciency. Ri can take
load and the MPP power for the two dierent values of RL. any value with this converter. This allows the photovoltaic
It is observed that when the panel is directly connected to the solar system to reach the MPP regardless of the existing
resistive load, without inserting any DC/DC converter, the irradiation level and RL, achieving a higher MPP-tracking
system will only operate at the maximum power point when eciency. Note that the MPP can be tracked for any RL
RMPP and RL match (see Fig. 9). If a buck converter is value, regardless of its relationship with RMPP.
inserted between the panel and the load (Fig. 11), we can In Table 3, a comparative of the MPP-tracking e-
observe that the system is only able to follow the maximum ciency provided by each of the congurations for the con-
power point for not very high irradiation values (depending cerned day of study is given. Observe that in all cases, the
on RL), i.e., when the maximum power point impedance
RMPP is relatively high. At maximum solar irradiation Table 3
hours, RMPP reaches its minimum values, and so the system MPP-tracking eciency obtained for each DCDC converter congura-
is unable to achieve the MPP. This is even more evident that tion and load
the higher RL is in relation to RMPP. When it is used a boost Load Without Buck Boost Buckboost
converter, (Fig. 12), the system is able to reach the maximum converter converter converter converter
(%) (%) (%) (%)
power point only at maximum irradiation hours (low
RMPP), with a remarkable loss of MPP-tracking eciency RL = 5 X 88.5 97.2 91.2 99.9
RL = 20 X 40.2 40.3 99.7 99.9
at the initial and nal hours of the day.
38 J.M. Enrique et al. / Solar Energy 81 (2007) 3138
conguration with buckboost converter is the one that photovoltaic module by neural networks. Renew. Energy 29 (3), 443
presents the highest eciency. 457.
Enrique, J.M., Duran, E., Sidrach, M., Andujar, J.M., 2005. A new
approach to obtain IV and PV curves of PV panels by using DCDC
6. Conclusions converters. In: 31st IEEE Photovoltaic Specialists Conference, January
37, 2005.
In this work we aimed at revealing the importance of the Enslin, J.H.R., Wolf, M.S., Snyman, D.B., Swiegers, W., 1997. Integrated
correct choice of the DC/DC converter in a photovoltaic photovoltaic maximum power point tracking converter. IEEE Trans.
Ind. Electron. 44 (6).
facility in order to obtain its highest MPP-tracking e- Garca, F., Alonso, M.C., 2000. On the modelling of a maximum power
ciency. In this article we demonstrate that only the buck point tracking system. In: 16th European Photovoltaic Solar Confer-
boost DC/DC converter is able to manage the facility to ence, May 2000, Glasgow, UK.
follow the photovoltaic panel maximum power point at Gow, A., Manning, C.D., 1999. Development of a photovoltaic array
all times, regardless of cell temperature, solar global irradi- model for use in power-electronics simulation studies. IEE Proc.
Electric Power Appl. 146 (2), 193200.
ation and connected load. Hohm, D.P., Ropp, M.E., 2002. Comparative study of maximum power
It is important to remark that the result obtained in the point tracking algorithms. Progress in Photovoltaics: Research and
analysis is independent from the MPP tracking system, i.e., Applications, November 2002.
however ecient this system may be, the DC/DC converter Hua, C., Shen, C., 1998. Study of maximum power tracking techniques
conguration imposes restrictions on it that it cannot sidestep. and control of DC/DC converters for photovoltaic power system. In:
29th Annual IEEE PESC. IEEE Computer Society Press, New York,
Despite the fact that the study carried out in this work pp. 8693.
is theoretical, it is important to note that from a practical Hua, C., Lin, J., Tzou, H., 2003. MPP control of a photovoltaic energy
approach, the buck and boost converters are the most e- system. ETEP 13 (4), July/August 2003.
cient topologies for a given price. While voltage exibility Hussein, K.H., Muta, I., Hoshino, T., Osaka, M., 1995. Maximum
varies, buckboost and Cuk (Cuk is a type of structure photovoltaic power tracking: an algorithm for rapidly changing
atmospheric conditions. IEE Proc. Gener. Transnm. Distrib. 142 (1),
derived from the buckboost topologies) converters are 5964.
always at eciency or, alternatively, price disadvantage. Jingquan, C., Erickson, R., Maksimovic, D., 2001. Averaged switch
Nevertheless, there are already congurations of buck modeling of boundary conduction mode DC-to-DC converters. In:
boost and Cuk converters where both the MOSFET and The 27th Annual Conference of the IEEE, Industrial Electronics
the inductor are of a very low resistance, achieving ecien- Society, IECON 01, vol. 2, pp. 844849, 29.
Kitano, T., Matsui, M., Xu, D., 2003. An MPPT control scheme for a PV
cies as regards input power higher than 95% and hardly 2 system utilizing limit cycle operation and its system design. Electron.
or 3% lower than the buck and boost topologies. Eng. Jpn. 143 (2).
According to the performed analysis, we dare to suggest Masoum, M.A.S., Dehbonei, H., Fuchs, E.F., 2002. Theoretical and
that a good practice could be including a buckboost DC/ experimental analyses of photovoltaic systems with voltage- and
DC converter in photovoltaic solar facilities at the PV current-based maximum power-point tracking. IEEE Trans. Energy
Conver. 17 (4).
array output and then connecting, after the converter, the Neto, L., Cardoso, B.J., Rocha, S., 2000. Control integrated maximum
rest of the facility elements (load). This practice guarantees power point tracking methods. In: 16th European Photovoltaic Solar
the photovoltaic panel maximum power point tracking for Conference, May 2000, Glasgow, UK.
any solar irradiation, cell temperature and load conditions, Schilla, T., Hill, R., Pearsall, N., Forbes, I., Bucher, G., 2000. Develop-
which could undoubtedly redound to the facilitys higher ment, simulation and evaluation of digital maximum power point
tracker models using PVNETSIM. In: 16th European Photovoltaic
system eciency. Solar Conference, May 2000, Glasgow, UK.
Singer, S., 1991. The application of loss-free resistors in power processing
References circuits. Power Electron. IEEE Trans. 6 (4), 595600.
Tse, K.K., Ho, M.T., Chung, H.S.-H., Hui, S.Y.R., 2002. A comparative
Akbaba, M., Alataawi, M.A.A., 1995. A new model for IV character- study of maximum-power-point tracker for photovoltaic panels using
istics of solar cell generators and its applications. Sol. Energy Mater. switching frequency modulation. IEEE Trans. Power Electron. 17 (6),
Sol. Cells 37, 123132. November.
Akbaba, M., Qamber, I., Kamal, A., 1998. Matching of separately excited Tse, K.K., Ho, M.T., Chung, H.S.-H., Hui, S.Y.R., 2004. A novel
DC motors to photovoltaic generators for maximum power output. maximum power point tracker for PV panels using switching-
Sol. Energy 63, 375385. frequency modulation scheme. IEEE Trans. Ind. Electron. 51 (2).
Andujar, J.M., Enrique, J.E., Duran, E., Bohorquez, M.A.M., 2004. Veerachary, M., Senjyu, T., Uezato, K., 2002. Feedforward maximum
Sistema para la generacion automatica de curvas IV, PV y power point tracking of PV systems using fuzzy controller. IEEE
monitorizacion de modulos fotovoltaicos. In: Proc. De las XXV Trans. Aero. Electron. Syst. 38 (3), July.
Jornadas de Automatica. Ciudad Real, Espana. ISBN: 84-688-7460-4. Veerachary, M., Senjyu, T., Uezato, K., 2003. Maximum power point
Appelbaum, J., 1986. Starting and steady-state characteristics of DC tracking of coupled inductor interleaved boost converter supplied PV
motors powered by solar cell generators. IEEE Trans. Energy Conver. system. IEE Proc. Electron. Power Appl. 150 (1).
EC-1, 1725. Yu, G.J., Jung, Y.S., Choi, J.Y., Kim, G.S., 2004. A novel two-mode
Bahgat, B.G., Helwa, N.H., Ahamd, G.E., ElShenawy, E.T., 2004. MPPT control Algorithm based on comparative study of existing
Estimation of the maximum power and normal operating power of a algorithms. Sol. Energy 76, 455463.