Aalborg Universitet
Improved Reliability of Single-Phase PV Inverters by Limiting the Maximum Feed-in
Power
Yang, Yongheng; Wang, Huai; Blaabjerg, Frede
Published in:
Proceedings of the 2014 IEEE Energy Conversion Congress and Exposition (ECCE)
DOI (link to publication from Publisher):
10.1109/ECCE.2014.6953385
Publication date:
2014
Document Version
Early version, also known as pre-print
Link to publication from Aalborg University
Citation for published version (APA):
Yang, Y., Wang, H., & Blaabjerg, F. (2014). Improved Reliability of Single-Phase PV Inverters by Limiting the
Maximum Feed-in Power. In Proceedings of the 2014 IEEE Energy Conversion Congress and Exposition
(ECCE) (pp. 128-135). IEEE Press. https://doi.org/10.1109/ECCE.2014.6953385
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
- Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
- You may not further distribute the material or use it for any profit-making activity or commercial gain
- You may freely distribute the URL identifying the publication in the public portal -
Take down policy
If you believe that this document breaches copyright please contact us at vbn@aub.aau.dk providing details, and we will remove access to
the work immediately and investigate your claim.
© 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all
other uses, in any current or future media, including reprinting/republishing this material for
advertising or promotional purposes, creating new collective works, for resale or redistribution to
servers or lists, or reuse of any copyrighted component of this work in other works.
Digital Object Identifier (DOI):
Proceedings of the IEEE Energy Conversion Congress and Exposition (ECCE 2014), Pittsburgh, PA, USA,
14-18 September, 2014.
Improved Reliability of Single-Phase PV Inverters by Limiting the Maximum Feed-in Power
Yongheng Yang
Huai Wang
Frede Blaabjerg
Suggested Citation
Y. Yang, H. Wang, and F. Blaabjerg, "Improved reliability of single-phase PV inverters by limiting the
maximum feed-in power," in Proc. IEEE Energy Convers. Congr. and Expo., 2014, pp. 128-135.
Improved Reliability of Single-Phase PV Inverters
by Limiting the Maximum Feed-in Power
Yongheng Yang, IEEE Student Member, Huai Wang, IEEE Member, Frede Blaabjerg, IEEE Fellow
Department of Energy Technology
Aalborg University
Pontoppidanstraede 101, Aalborg East DK-9220, Denmark
yoy@et.aau.dk, hwa@et.aau.dk, fbl@et.aau.dk
Abstract— Grid operation experiences have revealed the neces-
sity to limit the maximum feed-in power from PV inverter systems
under a high penetration scenario in order to avoid voltage and
frequency instability issues. A Constant Power Generation (CPG)
control method has been proposed at the inverter level. The CPG
control strategy is activated only when the DC input power from
PV panels exceeds a specific power limit. It enables to limit the
maximum feed-in power to the electric grids and also to improve
the utilization of PV inverters. As a further study, this paper
investigates the reliability performance of the power devices
(e.g. IGBTs) used in PV inverters with the CPG control under
different feed-in power limits. A long-term mission profile (i.e.
solar irradiance and ambient temperature) based stress analysis
approach is extended and applied to obtain the yearly electrical
and thermal stresses of the power devices, allowing a quantitative
prediction of the power device lifetime. A study case on a 3 kW
single-phase PV inverter has demonstrated the advantages of the
CPG control in terms of improved reliability.
I. INTRODUCTION
With a spectacular growth rate of PhotoVoltaic (PV) instal-
lations, challenging issues like overloading of the grid due
to the peak power generation of PV systems have recently
gained much attention [1]–[3]. In the case of a very large-
scale adoption of PV systems, advanced control strategies like
power-ramp limitation and absolute power control, which are
currently e.g. required for wind power systems in Denmark,
should also be transitioned and strengthened into the next-
generation PV systems [1], [4]–[9]. As a power limiting
control, a Constant Power Generation (CPG) control by lim-
iting maximum feed-in power has been proposed in [9], and
witnessed as an effective way to eliminate overloading. When
it is compared to the solutions of expanding the power grid
infrastructure or integrating energy storage systems to tolerate
the peak power [4]–[11], the CPG control might be a more
economically viable strategy, since it only contributes to a
limited energy yield reduction in a real case, where typically
the peak power generation is very rare.
In addition, the CPG control allows a reduction of the
thermal stresses on the power devices (e.g. IGBTs), since
the power losses inducing temperature rises inside the power
devices will be changed, when the PV system enters into CPG
mode from Maximum Power Point Tracking (MPPT) mode
and vice versa. As a consequence, the thermal stresses will
affect the reliability of the PV system. However, there is still
a lack of quantitative analysis on the potential reliability im-
provement enabled by the CPG control, besides the mitigation
of overloading at a high penetration level. Moreover, even for
real-field applications, where limiting peak power control was
not initially included, the CPG control can still be applied for
potentially extending the lifetime of existing PV inverters by
only software algorithm modifications. Seen from this point,
it is interesting to justify the long-term performance of PV
inverters from both reliability and economic viability (i.e. a
trade-off between the lifetime extension and the overall energy
yield reduction), and thus find the optimal power limitation
level in terms of cost-of-energy [4], [9], [11].
Regarding the reliability of PV inverters, it has become of
intense importance and involves multiple disciplines [5], [7],
[12]–[19]. The lifetime prediction research on power devices
is transitioning from handbook-based approaches [18], [19]
to more physics-based methods, which require in-depth un-
derstanding of various failure mechanisms and thus dedicated
lifetime models, e.g. an analytical based Confin-Mason model
[12]–[14], [16]. Among these failure factors, thermal stresses,
depending on the mission profile as well as the inverter
operating conditions, have been the most observed ones in
PV systems (both inverters and capacitors) [17], [20]. Hence,
the varying operation conditions due to the intermittent nature
of solar energy has been one of the challenges to perform
reliability analysis in PV systems. Currently, most of the ex-
isting reliability prediction methods for the lifetime estimation
of power devices in PV inverters only consider either short-
term mission profiles [13], [14] or long-term mission profiles
with a low data-sampling frequency, where the effects of
small temperature cycles are not considered [19]. Moreover,
the widely used lifetime models unfortunately consider only
a few failure modes, e.g. the junction temperature cycle
amplitude and the mean junction temperature [16], [19]–[22].
However, improving the lifetime estimation accuracy requires
an elaborated analysis of a long-term mission profile, and also
a detailed reliability model.
In view of the above issues, a mission profile based reli-
ability analysis approach has been proposed in [17], which
is extended and applied to the PV systems with the MPPT-
CPG control in this paper. This reliability approach takes a
real-field yearly mission profile with a high sampling rate
978-1-4799-5776-7/14/$31.00 ©2014 IEEE
39 6WULQJV
3
39
%RRVW
L
39
Y
39
Y
J
L
J
4
'&
'&
'&
$&
&
GF
,QYHUWHU )LOWHU
Y
GF
3:0
LQY
3:0
E
,QYHUWHU
&RQWURO
6XSHUYLVRU\ FRPPDQG 3
OLPLW
*ULG
3
R
=
J
3&&
&
6 7
D
Y
GF
%RRVW &RQWURO
0337 &3*
&
39
&RPPXQLFDWLRQ
0LVVLRQ
3URILOHV
/
L
/
J
&
I
3
OLPLW
Fig. 1. A two-stage single-phase grid-connected PV system with MPPT
and CPG control considering mission profiles.
(200 ms) into consideration, and the mission profile has been
decomposed into the ones of different time scales, i.e. short-
term mission profiles and long-term mission profiles. The
resultant mission profile at a large time scale is analyzed
using a rain-flow counting algorithm. The MPPT-CPG control
method has been applied to a 3 kW single-phase PV system.
The temperature loading profiles, including thermal cycles at
fundamental frequency induced by short-term mission profiles
and the cycles with large periods mainly due to long-term
mission profiles, offer the possibility to quantitatively calculate
the consumed life and thus an estimation of the lifetime with
a reliability model. The application of the extended reliability
analysis approach presented in § III shows that, a PV system
with CPG control, which only leads to a limited energy yield
reduction, can contribute not only to unloading of the grid but
also to improved reliability of the power converters.
II. S
YSTEM DESCRIPTION AND OPERATION
The PV system considered in this paper is a single-phase
system as shown in Fig. 1. The boost converter offers the
flexibilities of MPPT and active power control (e.g. CPG
control) [9], and extends the operational time of the PV
inverter when the solar irradiance level is very low. The PV
inverter can be transformerless to maintain a high efficiency.
In this paper, a full-bridge topology with a bipolar modulation
scheme is adopted, since the bipolar modulation scheme can
effectively mitigate leakage currents, which is required by PV
integration standards. A hybrid control scheme of MPPT and
CPG control allows further to increase the penetration level.
The CPG control can be implemented by a) integrating energy
storage systems like a battery, b) managing the power at the
secondary control level, and c) modifying the conventional
MPPT algorithms [4], [8], [9].
The CPG control by modifying the MPPT algorithm is
adopted in this paper for the single-phase PV systems due to its
simplicity. The control structure of a two-stage PV system with
the CPG control is shown in Fig. 2. The operation principle of
a PV system with the MPPT-CPG hybrid control scheme can
be described as follows. When the available PV output power
P
PV
exceeds the power limitation P
limit
, the system goes into
the CPG mode with a constant power generation of the PV
strings, which is controlled by a proportional controller (k
cpg
).
When P
PV
≤ P
limit
, the PV system operates in MPPT mode
0337
Y
SY
L
SY
N
PSS
2SHUDWLRQ 0RGH
3
SY
3
OLPLW
N
FSJ
3:0
E
3
SY
3
OLPLW
3
SY
3
OLPLW
Y
GF
N
S
N
L
V
&XUUHQW
&RQWUROOHU
3:0
LQY
Y
GF
Y
J
3KDVH
/RFNHG /RRS
L
J
D
3, &RQWUROOHU
L
J
E
Fig. 2. Control diagram of a single-phase PV system with CPG ability: (a)
boost control diagram and (b) PV inverter control system.
3
ORVV
7
M
3
LQ
39
0RGHOV
0LVVLRQ
3URILOHV
&
7
M
/LIHWLPH
/LIH
0RGHO
3
RXW
7
M
/LIHWLPH
D
E
0337&3*
0LVVLRQ 3URILOHV
(OHFWULFDO 6\VWHP
&
*ULG
3RZHU
&RQYHUWHUV
7KHUPDO
0RGHO
/LIH
0RGHO
/RRNXS 7DEOH
EDVHG
6LPXODWLRQ 0RGHO
Fig. 3. Mission profile based lifetime analysis approach for the power
switching devices: (a) detailed structure for short-term mission profiles and
(b) look-up table based analysis structure for long-term mission profiles.
with a peak power injection to the grid from the PV strings. A
proportional controller k
mpp
is used to regulate the PV panel
current. It can be seen that the hybrid control scheme requires
minor and simple control algorithm modifications instead of
complicated hardware adjustments (e.g. with energy storage
systems), which means that it does not increase the total
implementation cost. In respect to the current controller, a
good power quality of the injected grid current should be
maintained in terms of low total harmonic distortions [23].
Considering this issue, a Proportional Resonant (PR) controller
[23], [24] has been adopted as the current controller in Fig. 2.
In both operation modes, the DC-link voltage v
dc
is controlled
through a Proportional Integrator (PI) controller to follow the
reference command, v
∗
dc
.
III. M
ISSION PROFILE BASED RELIABILITY ANALYSIS
Improving the reliability of the power electronics based PV
system has been an intense topic [25] in order to integrate cost-
effective solar PV energy into the grid. The mission profile
has been witnessed as one determining factor of the failure
in power converters [19], [20], [26], [27]. Thus, a mission
profile based lifetime analysis approach [17] is extended in
the following section considering both short-term and long-
term mission profile effects.
A. Mission Profile based Lifetime Analysis Approach
Fig. 3 shows the extended mission profile based reliability
analysis approach. This reliability analysis approach can be
0LVVLRQ 3URILOH
WLPH
W
V
0LVVLRQ 3URILOH
WLPH
0LVVLRQ 3URILOH
WLPH
GHFRPSRVH
'HWDLOHG 5HOLDELOLW\
$QDO\VLV 6WUXFWXUH
)LJ D
/RRNXS 7DEOH
EDVHG 5HOLDELOLW\
$QDO\VLV 6WUXFWXUH
)LJ E
7KHUPDO /RDGLQJ
WLPH
7KHUPDO /RDGLQJ
WLPH
W
V
ǻ7
M
7
MP
ǻ7
M
7
MBPD[
7
MP
7
MBPLQ
ǻ7
M
7
MP
/LIH 0RGHO
W
RQ
ǻ7
M
7
MP
W
RQ
òW
RQ
&\FOH
&RXQWLQJ
7
MP
/LIH 0RGHO
ǻ7
M
7
MP
W
RQ
(VWLPDWHG
/LIHWLPH
EDVHG RQ
0LQHU¶VUXOH
2ULJLQDO SURILOH
6KRUWWHUP
0)
W
V
0)
0)
0)
ǻ6
Fig. 4. Proposed mission profile decomposition procedure of the extended reliability analysis approach for temperature loading translation.
adopted for analysis of mission profiles at different time scales,
and thus predict the lifetime of IGBTs. For short-term mission
profiles, the temperature loading profile (junction temperature)
can directly be obtained from Fig. 3(a). However, for a long-
term mission profile with a high data-sampling rate (e.g. 200
ms), it will be time-consuming, or even impossible, to capture
the full temperature loading profile. Thus, look-up tables are
adopted to accelerate the evaluation process as it is shown in
Fig. 3(b), which requires decomposing the mission profile at
different time scales.
A decomposition procedure is proposed as shown in Fig. 4,
where the original mission profile is decomposed with a period
of t
s
under an assumption that in this short period the mission
profile of t
s
is constant and that the junction temperature can
go into steady state within the time of t
s
. Consequently, in each
time interval of t
s
, the mission profile (e.g. MF
1
and MF
2
) can
be treated as a short-term mission profile, where the analysis
approach shown in Fig. 3(a) is applicable. Notably, under the
decomposed short-term mission profile, the thermal cycles are
mainly at fundamental frequency with identical cycle period,
t
on
, e.g. t
on
= 0.02 s in a 50 Hz power grid, as exemplified
in Fig. 5. However, as it is shown in Fig. 4, there is a stress
difference (e.g. the stress difference ΔS between MF
1
and
MF
2
) among those short-term mission profiles, and this will
also introduce temperature stresses on the power devices, as
shown in Fig. 5. Therefore, a long-term mission profile is
reconstructed using the average stress from short-term mission
profiles (e.g. MF
1
and MF
2
). Finally, a look-up table based
approach shown in Fig. 3(b) can be applied to extract the
long-term thermal loading profile.
B. Temperature Loading Interpretation
After the decomposition of the long-term mission profile,
the temperature loading profiles appearing in the power de-
vices should be appropriately extracted or interpreted accord-
ing to the lifetime model. For example, the Coffin-Manson
model [12]–[14], [22] indicates that the number of cycles
to failure (N
f
) is only dependent on the temperature cycles,
including cycle amplitude (ΔT
j
) and mean junction tempera-
ture (T
jm
). Those values can be obtained under a short-term
'HYLFH MXQFWLRQ WHPS &
7LPH V
ǻ7
M
òW
RQ
7
MP
ǻ7
M
W
RQ
7
MP
ǻ7
M
W
RQ
7
MP
Fig. 5. Temperature loading example of the power devices in the case of
solar irradiance variations (T
a
= 50
◦
C).
mission profile, as it is shown in Fig. 3(a) and Fig. 4, while
for a long-term mission profile, counting algorithms are used
to extract the temperature loading profile information. There
are many cycle counting algorithms reported, e.g. level cross-
ing counting, rain-flow counting, and simple range counting
methods [13]–[15], [26], which can be used to appropriately
interpret the thermal loading profile according to a dedicated
lifetime model. Then, the lifetime can be calculated with the
extracted information. However, it has been found that N
f
is
also affected by the cycle period (t
on
), bond-wire aspect ratio
(ar), and the diode (f
d
) [21]. Hence, a detailed lifetime model
has been introduced in [21], and it can be given by,
N
f
= AΔT
α
j
(ar)
β
1
ΔT
j
+β
0
f(t
on
)exp
E
a
k
B
T
jm
f
d
(1)
with
f(t
on
)=
C +(t
on
)
γ
C +1
in which A, α, β
0
, β
1
, γ and C are the model parameters
that can be obtained by means of curve-fitting using numerical
simulation or experimental results (accelerating tests) [16]. k
B
is the Boltzmann constant, and E
a
is the activation energy. The
values of those parameters and also the test conditions for an
IGBT module are shown in Table I.