1

Modeling Grid Connection for Solar and Wind

Energy
P. J. van Duijsen, Simulation Research, The Netherlands

Frank Chen, Pitotech, Taiwan

The PV inverters are categorized depending on the PV power

Abstract—Modeling of grid connected converters for solar and plant configuration.
wind energy requires not only power electronics technology, but
also detailed modeling of the grid synchronization and 50 – 500Watt:
modulation techniques. Control of active and reactive power in Here mostly one solar panel is used where the inverter is
both single and three phase grid connections can be achieved by
integrated into the solar module. Typically used for feeding
quadrature controllers, analogous to field oriented control in
electrical drives. Modulation strategies, loss determination and
on remote locations of Measurement Stations or traffic
thermal cycling, as well as life time estimation are important lights.
factors that can be studied into detail during simulation. 500Watt – 2kW:
For small rooftop plants, one solar string is used.
Index Terms—Modeling, Simulation, Wind, Solar, Grid 1.5kW – 6kW:
Connection, Grid Synchronisation Larger rooftop plants are configured either as solar string or
solar Multi String.
I. INTRODUCTION 6kW – 100kW:
Grid connected converters are required to transfer harvested Three-phase grid connection is applied here with solar multi
green energy from wind and solar systems into the main grid. string or arrays.
The importance of the single-phase grid connection for PV and > 100kW:
wind power systems should not be underestimated. It is one of Three phase grid connection and central configured grid
the key components when it comes to stable, and efficient inverters.
power transfer from the solar or wind power system into the
grid. Not only grid-synchronization, also EMI problems, The first grid connected inverters were based on Silicon
harmonic regulations and efficiency are important design Controlled Rectifiers (SCR) technology which were also
issues that have to be solved. limited in control and came with a high harmonic content,
making the use of bulky inefficient filters necessary.
Although it looks fairly simple at first sight, the transfer of DC With the introduction of Mosfets for the lower power area and
power from the solar or wind power system to the grid can be IGBT’s for the high power applications, the control of grid
categorized into various solutions. Depending on the size of side inverters became more advanced.
the PV-system and the arrangement of the solar panels, the The primary concern in grid converter design is the efficiency,
following combination of solar panels is possible: mainly driven by the high cost of solar energy. This resulted in
large variation of PV grid converter. Compared to motor drive
Solar Panel: inverters, the grid-connected converter is more complex as it
Parallel and series connected solar cells. includes extra features such as MPP control, and grid
Solar Module: synchronization. The power electronics design is more
Parallel and series connected solar cells including bypass complicated compared to motor drives, because of the grid
diodes. connection and safety. EMI problems are of special interest
Solar Array: especially those associated with leakage currents from the
Parallel and series connection of solar modules. solar panel to the earth.
Solar string: For electric motor control two-phase and three-phase bridge
Series connection of solar panels. inverters are the standard technology. However for PV systems
Solar Multi String: different topologies exist, especially for single-phase grid
Parallel connection of solar strings. converters.
The first main concern regarding the design of grid connected
converters is efficiency, due to the costs of solar produced
energy. Secondly, since the lifetime of PV panels typically
goes beyond 20 years, also the lifetime of the grid-connected
 2

inverter becomes a design issue. This means extra effort

regarding the thermal design. The first attempt to improve
efficiency is to remove the transformer, either from the low
voltage as from the high voltage side. This improves the total
efficiency with 2 to 3 % since core-losses were eliminated.

II. MULTILEVEL MODELING AND SIMULATION

Multilevel modeling and simulation in the field of green and
renewable energy is a broad topic. In the first place the Figure 1: Low Voltage High-Frequency transformer based.
systems are very diverse, like wind and solar, but also the
physical background of each type of green energy system The second approach is to have the transformer at the high
varies greatly. For the modeling and simulation, this leads to voltage sideas shown in figure 2.
two observations that have to be taken into consideration:

• Physical background of the underlying system

• System or detailed of the model of systems under
investigation

First, for example, wind energy requires knowledge on

electromagnetic energy conversion, while solar requires
knowledge on semiconductor devices. Control in solar systems
is mostly a sort of a smart search algorithm that finds the Figure 2: High Voltage Low-Frequency transformer based.
optimum electric load for a solar module, while in wind power
systems, the control is clearly dependent on the wind speed. Here the transformer is larger in volume compared to the low-
Generally speaking, various technologies are used when voltage transformer design, because the operational frequency
working with green energy systems. is much lower (50 Hz or 60 Hz).
Secondly the multilevel character of green energy applications A typical topology for a low frequency high voltage side
is very clear. The components can de modeled as simple transformer based converter is shown in figure 3[Ertl]. Here
system blocks with clearly defined functions, but on the other the boost converter creates a sinusoidal modulated DC voltage
hand multilevel models including all details can model them. that is connected via the coupled inductors to the main grid.
For example, generator modeling in FEM and detailed
semiconductor models in solar modules, compared to the
modeling of algorithms in Maximum Power Point controllers.
For grid connected converters the multilevel modeling concept
is used extensively for the main purpose of making a
distinction between the model for the grid converter and the
model of the grid synchronization.

III. SOLAR POWER GRID CONVERTERS

A. Transformer based PV design

Single-phase grid connection is not as simple as it would look
at first sight. Regulation on injected harmonics and the
efficiency of the inverter are important factors. Inverters that
have a galvanic isolation by means of a transformer have a low
efficiency. They are employed in many commercial grid-
converters and can be classified as low-voltage side
transformer and high-voltage side transformer. The basic
topology for a low voltage side transformer based converter is
shown in figure 1.

The DC-DC convert is mostly of the type Forward of Flyback, Figure 3: Transformed based design
since they provide galvanic isolation through their coupled
inductors. Here the transformer is a high frequency device Although isolation comes naturally with the transformer-based
(20kHz) and therefore the volume of transformer is smaller designs, they have a number of disadvantages, from which
than a comparable low frequency (50Hz) transformer. efficiency is the main drawback.
 3

The main disadvantage of the galvanic isolation is the core- The leakage current is too large for practical applications. In
loss in the coupled inductors and therefore the transformer-less the transformer based designs this leakage current didn’t exist
grid converters were developed. because of the galvanic isolation of the transformer. One
method to get rid of the leakage current is the H5 converter
B. Transformer-less based PV designs
[Victor M. et al] as shown in figure 7.
To overcome the inefficiency of the transformer, a high
voltage has tot be generated directly by the solar cells or by a
boosting DC-DC converter.
Especially with large solar strings, the generated voltage might
be high enough to be used directly without a boosting DC-DC
converter. However, keep in mind that a DC-DC converter is
still of practical use for the MPP controller, as it gives a stable
DC bus voltage for the grid connected inverter.
A second note is on the boosting converter, which has a
limited voltage transfer level.

Figure 7: H5 Full Bridge converter

In the H5 converter, an extra switch is included that

disconnects the DC input during the zero voltage state of the
output of the converter. As result, the leakage current reduces
from close to 1Amperes to less 20mA on average, as shown in
figure 8. Compared to the leakage current for a full bridge
converter as shown in figure 6, this is a remarkable reduction.

Figure 4: Full Bridge converter

Figure 4 shows a full bridge converter connected to a solar

multi string. The output voltage of the solar multi string is such
high, that the inverter can directly use it to inject currents into
the main grid.

Figure 8: H5 Full Bridge converter leakage current

C. Grid Connect Solar converter with MPP

This example shows the control of a grid-connected solar
system. Figure 9 shows the circuit model including the MPP
Figure 5: Full Bridge converter Grid injected current control and grid-connection control.
A detailed second level circuit model that includes load
The waveform of the grid-injected current is displayed in dependent loss and temperature dependency models the solar
figure 5 and shows a low harmonic content. There is however module. A boost converter that regulates the MPP for the solar
one serious drawback of the full bridge converter, being a module electrically loads the solar module. The boost
leakage current through the solar panels to the earth. converter is also modeled as a second level circuit model. The
MPP controller is a first level system model that calculates the
derivative of the power as a function of the voltage of the solar
module. Together with the first level system model for the PI
controller, the amount of power harvested by the solar module
is maximized. The last part is the grid connection. Here a first
level system model for the inverter and control is used. This
simulation shows the mixture between the various levels in
modeling.

Figure 6: Full Bridge converter leakage current

 4

Since all models are connected in a circuit level model, the u-i Scope 1 in the Caspoc simulation in figure 9, shows the u-i
interactions can be studied as well as the MPP control part. characteristic of the solar module. Small circles indicate the
The single-phase grid converter is a very simple system level characteristic, while the arrow points to the current point of
model that only controls the current delivered to the single- operation. Scope 2 shows the grid-side current, which is in
phase grid. In a more detailed study, the complete single-phase phase with the grid voltage. Scope 6 shows the amount of
inverter with controlling algorithm and modulation strategy as power delivered by the solar module and the amount of power
well as grid synchronization could be applied. However for delivered to the grid.
studying the behavior of the MPP controller, this simplified
system level model suffices.

Figure 9:Solar module with boost converter, grid connection and MPP control

varying wind speed, the control directly adjusts the amount

IV. WIND POWER MODELING of power drawn from the wind turbine generator and also
controls the amount of power delivered into the main grid.
A. Wind Turbine Maximum Power Control A second level circuit model models the wind turbine,
The wind energy example shows an overall system where the ratio between the wind tip speed and rotor
simulation of a wind power system. A more detailed second angular speed determines the amount of power the wind
level circuit model can replace every single component in turbine can deliver. This model is a minimum requirement
this first level system simulation. In this example, only the when investigating the behavior for varying wind speed.
wind speed to rotor angular speed ratio (tip speed ratio The generator is modeled as a first level system model, as
[TSR], λ) is modeled in a circuit level model. Also third the primary goal is to study the stability of the overall
level component models or even a mix between first, second control system. The wind turbine includes the TSR
and third level models would be possible in this type of dependent power coefficient Cp(λ,β), where λ is the TSR
simulations. and β is the pitch angle of the rotor blades. The
characteristic of Cp(λ,β) is shown in scope 2 in figure 10
Figure 10 shows a grid connected wind turbine system with and this curve can be specified either by analytical
optimum power control, modeled in Caspoc. The maximum equations, or by a look-up table from measurement results.
amount of power a wind turbine can deliver is equal to kω3. The amount of mechanical power Pm is displayed in scope 1
Therefore the control electrically loads the wind turbine as function of the angular rotor speed. Inside the “Maximum
generator such, that it subtracts exactly kω3 Watts from the Power Control”, the electrical power is calculated and
wind turbine and thus keep the system in equilibrium. For controls the combined system level Generator/Inverter
 5

model. The left side of this model is the shaft of the voltage and current, mechanical and electrical power. The
generator, while the right side is the electrical connection to scopes clearly show the dependency of the amount of power
the grid. The remaining scopes in figure 10 show from top delivered to the grid on the wind speed.
to bottom, wind speed and angular shaft speed, grid-side

Figure 10: Wind turbine with grid connection and maximum power control

control of the generator via power electronics, but also the

B. Three-phase wind power grid connection
calculation of losses of the power electronics has to be
Only the squirrel cage induction machine in a wind modeled in detail. Furthermore the modulation principle has
generator is capable of delivering power directly to the grid a severe impact on the generation of harmonics on the grid
without any power electronics. For control purposes the side.
generator systems of wind turbines and the DC output from A grid connection could be made using a hysteresis
the solar modules are connected to the grid via power controller. The harmonics produced by the inverter are
electronics. The grid connection serves merely two main rather large and unpredictable since the base switching
purposes; scaling and power factor control. frequency is not fixed.

Important when modeling wind power and solar systems are

the detailed power electronics converters. Not only the
 6

Figure 11: Grid connection with Space Vector Modulation (SVM)

[4] Bauer, P & Duijsen, PJ van (2009). Simulating losses
Instead of a hysteresis controller, a PQ controller combined and semiconductor junction temperature in power
with Space Vector Modulation (SVM) is be used, to electronics. Bodo's power systems, 9, 36-38.
generate the three phase current injected into the grid. [5] B. Klimpke; A Hybrid Magnetic Field Solver Using a
Here the grid voltages as well as the injected currents are Combined Finite Element/Boundary Element Field
measured and using a PI controller, the error between them Solver; Sensors Magazine; May 2004
is minimized. [6] van Duijsen, P.; Leuchter, J.; Bauer, P.; Lifetime
estimation with thermal models of semiconductors;
V. CONCLUSIONS Energy Conversion Congress and Exposition (ECCE),
Multilevel modeling allows the combination of various 2010 IEEE; pp 978 - 985
levels of models to be interconnected. Depending on the [7] Duijsen, PJ van, Bauer, P & Leuchter, J (2010).
required simulation models system models can be employed Thermal models for semiconductors. In S Mircevski &
for overall system behavior and can be interconnected with D Boroyevich (Eds.), 2010 14th International Power
detailed component models. Concentrating on the minimum Electronics and Motion Control Conference (pp. 23-
required levels in a model reduces overall simulation time 28). Ohrid, Macedonia: IEEE.
and does not overcomplicate the total model. [8] Bauer, P & Duijsen, PJ van (2009). Sensorless control
for electrical and hybrid electric vehicles. Bodo's power
systems, 11, 50-52.
[9] Chao, D.C.-H.; van Duijsen, P.J.; Hwang, J.J.; Chin-
REFERENCES Wen Liao; Modeling of a Taiwan fuel cell powered
[1] Bauer, P & Duijsen, PJ van (2005). Challenges and scooter; International Conference on Power
advances in simulation. In s.n. (Ed.), Proceedings of Electronics and Drive Systems, 2009. PEDS 2009. , pp:
the 36th power electronics specialists conference (pp. 913 – 919
1030-1036). Piscataway, USA: IEEE. [10] Ertl H., Kolar J.W., Laborübung Leistungselektronik,
[2] Bauer, P & Duijsen, PJ van (2008). Power electronics TU Wien
simulations. International review on computers and [11] Victor M. et al., US Patent Application, Pub No.:
software (IRECOS), 3(3), 307-314. US2005/0286281 A1, Pub. Date: 29 Dec 2005
[3] Duijsen, P van, Bauer, P & Chen, F (2006). Modeling [12] Caspoc, Simulation Research, www.caspoc.com
and simulation for wind energy. In s.n. (Ed.),
Proceedings of the Taiwan power electronics
conference & exhibition (pp. 1087-1092). Taipei:
Taiwan Power Electronic Association/IEEE Taipei
Chapter.