SPIE Newsroom

10.1117/2.1200704.0612
Sizing of grid-connected
photovoltaic systems
Jayanta Mondol
A validated simulation model can be used to investigate the effect of
photovoltaic/inverter sizing ratio on performance.
Grid-connected photovoltaic (PV) systems feed electricity di-
rectly to the electrical network, operating parallel to the conven-
tional electric source. The simplest grid-connected system, such
one with low-voltage for residential use, contains a PV array
and an inverter unit. In high-voltage applications, the system re-
quires transformers and appropriate power switching and pro-
tection devices. Grid-connected PV systems generate clean elec-
tricity near the point of use, without transmission and distribu-
tion losses or the need for batteries. Their performance depends
on local climate, the orientation and inclination of the PV array,
and inverter performance.
The output of a grid-connected PV system depends on the
PV/inverter sizing ratio (R
s
)
1
, dened as the ratio of PV array
capacity at standard test conditions to the inverters rated in-
Figure 1. Inverter efciency as a function of fractional load, dened
as the ratio of input power (P
PV
) to the inverters rated capacity
(P
inv,rated
).
Figure 2. C
E,N
as a function of sizing ratio for different PV surface
slopes and PV tracking strategies.
put capacity. Properly matching PV and inverter rated capaci-
ties improves grid-connected system performance. Optimal siz-
ing depends on local climate, surface orientation and inclina-
tion, inverter performance, and the PV/inverter cost ratio (T).
Under low insolation (incident solar power), a PV array gener-
ates power below its rated capacity, leading to inverter opera-
tion at partial load. Inverter efciency drops with part-load op-
eration: it also becomes sub-optimal when a signicantly under-
sized inverter is made to operate mainly in conditions of over-
load, which result in energy loss. Figure 1 illustrates inverter ef-
ciency under both overload and partial-load operation.
Using a validated
2
TRNSYS
3
simulation model, we studied
the effects of PV orientation, inclination, inverter characteristics,
insolation, and T on R
s
. Parameters of a grid-connected PV sys-
tem located in Northern Ireland
4
supplied the inputs.
5
The opti-
mum PV/inverter sizing ratio can be examined in terms of ener-
getic performance (C
E
)the annual total PV system output per
PV rated capacityand economic performance(C
C
), the annual
Continued on next page
10.1117/2.1200704.0612 Page 2/3
SPIE Newsroom
total PVsystemoutput per total capital cost.
5
For different sizing
ratios, C
E
and C
C
were normalised with respect to the corre-
sponding maximum values, with the resulting normalised val-
ues denoted as C
E,N
and C
C,N
, respectively. The optimumsizing
ratio is determined when the values of C
E,N
and C
C,N
are equal
to one. PV systems with three sun tracking strategies were also
considered.
The optimum sizing ratios for 0

, 45

, and 90

surface slopes
are 1.4, 1.2, and 1.5, respectively (Figure 2). The incident insola-
tion on a vertical or a horizontal surface is lower than on a 45

tilted surface, so vertical and horizontal PV arrays operate be-

low their rated capacities. A lower value of R
s
for a 45

tilted
surface therefore improves inverter performance at partial load
operation. Figure 3 shows the variation of C
E,N
as a function
of PV surface slope and sizing ratio for three surface azimuth
angles.
When T decreases, maximum system performance results in
lower R
s
(Figure 4). That means when the inverter cost increases
relative to the PV cost, an undersized inverter would optimise
the system economic performance. When T increases, the sys-
temenergy loss associated with an undersized inverter is not off-
set by the lower system cost. R
C,max
for tracking systems varies
from 1.11.6 for cost ratios between one and 14.
Figure 3. C
E,N
as a function of PV surface slope and sizing ratio for
three surface azimuth angles: 0

, 45

, and 90

.
Figure 4. R
C,max
as a function of T for three different PV surface slopes
and tracking strategies.
Figure 5. R
C,max
for different European locations as a function T.
Figure 5 shows R
C,max
as a function of T for ve selected Eu-
ropean locations. In Almeria, one of the high insolation sites,
R
C,max
varies from 1.11.5 when T ranges from one to 14, while
in London, a low insolation site, the variation is from 1.81.2.
Under the current market scenario, for a high efciency in-
verter system, PV/inverter sizing ratio for high and low insola-
tion locations should be within 1.11.2 and 1.31.4, respectively.
For a low efciency inverter system, the optimum sizing should
be within 1.21.3 (high insolation) and 1.41.5 (low insolation).
Optimum PV system economic performance for an undersized
inverter is achieved when the relative cost of the inverter is
higher than the cost of the PV. The optimum sizing ratio is less
inuenced by insolation conditions when inverter cost is rela-
tively much lower than PV cost.
Continued on next page
10.1117/2.1200704.0612 Page 3/3
SPIE Newsroom
Author Information
Jayanta Mondol
School of the Built Environment
University of Ulster
Jordanstown, Northern Ireland
Jayanta Mondol, an Academic Research Fellow, works at the
Centre for Sustainable Technologies, School of the Built Environ-
ment at the University of Ulster, UK. His research areas include
solar photovoltaics, solar thermal, clean power plant technolo-
gies, and household energy-efciency improvement.
References
1. K. Peippo and P. D. Lund, Optimal sizing of solar array and inverter in grid-
connected photovoltaic systems, Solar Energy Materials and Solar Cells 32, pp. 95
114, 1994.
2. J. D. Mondol, Y. G. Yohanis, M. Smyth, and B. Norton, Long-term validated
simulation of a building integrated photovoltaic system, Solar Energy 78, pp. 163
176, 2005.
3. Klein et al., University of Wisconsin, Madison, Wisconsin, USA, 2000.
4. J. D. Mondol, Y. G. Yohanis, M. Smyth, and B. Norton, Long-term performance
analysis of a grid-connected building integrated photovoltaic system in northern
ireland, Energy Conversion and Management 47, pp. 2925294, 2006.
5. J. D. Mondol, Y. G. Yohanis, and B. Norton, Optimal sizing of pv array and
inverter for grid-connected photovoltaic systems., Solar Energy 80, pp. 15171539,
2006.
c 2007 SPIEThe International Society for Optical Engineering