A PLL-Free Grid-Forming Control With Decoupled

Functionalities for High-Power Transmission System

Applications
Taoufik Qoria, Ebrahim Rokrok, Antoine Bruyere, Bruno Francois, Xavier
Guillaud

To cite this version:

Taoufik Qoria, Ebrahim Rokrok, Antoine Bruyere, Bruno Francois, Xavier Guillaud. A PLL-Free Grid-
Forming Control With Decoupled Functionalities for High-Power Transmission System Applications.
IEEE Access, 2020, 8, pp.197363 - 197378. �10.1109/access.2020.3034149�. �hal-03703440�

HAL Id: hal-03703440

https://hal.science/hal-03703440
Submitted on 23 Jun 2022

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Received October 15, 2020, accepted October 24, 2020, date of publication October 27, 2020, date of current version November 11, 2020.
Digital Object Identifier 10.1109/ACCESS.2020.3034149

A PLL-Free Grid-Forming Control With Decoupled

Functionalities for High-Power Transmission
System Applications
TAOUFIK QORIA, EBRAHIM ROKROK , (Student Member, IEEE),
ANTOINE BRUYERE, (Member, IEEE),
BRUNO FRANÇOIS, (Senior Member, IEEE),
AND XAVIER GUILLAUD , (Member, IEEE)
University of Lille, Arts et Metiers Institute of Technology, Centrale Lille, Yncrea Hauts-de-France, ULR 2697 - L2EP - Laboratoire d’Electrotechnique et
d’Electronique de Puissance, 59000 Lille, France
Corresponding authors: Taoufik Qoria ([email protected]) and Ebrahim Rokrok ([email protected])
This work was supported in part by the project HVDC Inertia Provision (HVDC Pro), financed by the ENERGIX program of the Research
Council of Norway (RCN) under Project 268053/E2, and in part by the industry partners; Statnett, Statoil, Réseau de Transport
de l’Electricité (RTE), and Elia.

ABSTRACT This paper proposes a grid-forming control strategy with a generic implementation approach
to meet the transmission system requirements asking for the massive integration of power electronic devices
into the power systems. In this context, several grid-forming controls have been proposed in the literature
either with or without a Phase-Locked Loop (PLL). The PLL-based techniques allow decoupling the different
control functionalities (i.e. inertia emulation, frequency support, active power setpoint tracking in steady-
state) while the PLL-free schemes, which aim to avoid the PLL drawbacks, create a compulsory coupling
between the control functionalities. The proposed grid-forming control in this paper is able to decouple
the control functionalities without any dedicated PLL, which makes it more advantageous compared to
what have been already proposed in the literature. Since the power converters are exposed to the small
and large grid events, the presented control has been tested in both situations. For the small grid events,
a simplified small-signal model is developed to assess the active power and frequency dynamics. In case
of large grid events, a current limitation algorithm is included to the control in order to protect the power
converter. To deal with the transient stability issues linked to the current limitation and enhance the converter
performance during the post-fault, a method based on adaptive inertia constant is proposed. To validate the
overall approach, time-domain simulations (in Matlab-Simulink) and experimentations are performed.

INDEX TERMS Active power regulation, current limitation, fast frequency response, grid-forming control,
inertial effect, transient stability.

I. INTRODUCTION strong AC voltage with a given magnitude and frequency in

The beginning of the renewable energy story had a moderate such a way that the converter can simply inject a controlled
influence on the power electronic converters market. Very power. This control strategy of the converter is well-known
quickly, the connection of wind turbines and photovoltaic as the grid-following control. Grid-following converters have
through power electronic converters became compulsory. several limitations:
In addition, some new transmission capacity had to be added – Inability in stand-alone operation mode.
in order to transmit the renewable energy from the place – Stability issues in connection to the weak grids and in
where it is generated to the place where it is consumed and faulty conditions [1], [2], which are mainly linked to
also to improve the electricity market operation. the Phase-Locked Loop (PLL) dynamics and controllers
Today, a major share of the power electronic converters are interactions.
controlled under the presumption that they are connected to a – The increase of grid-following converters results in the
decrease of the total system inertia [3], [4].
The associate editor coordinating the review of this manuscript and To come up against these challenges, the way to con-
approving it for publication was Zhouyang Ren . trol power converters has to be changed from following
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 197363
 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

the grid voltage to forming the grid voltage. Contrary to a

grid-following converter that behaves as a current source,
a grid-forming converter operates as a voltage source.
Unlike the grid-following control whose concept is quite
well-accepted, the grid-forming control is still on debate
and several variants (e.g. Virtual Synchronous Genera-
tor (VSG) [5]–[7], Synchronverter [8]–[12], Virtual Syn-
chronous Machine (VSM) [13]–[19], VISMA [20] and
FIGURE 1. Generic topology of the VSC.
droop control with lead-lag unit [5], [21]) are proposed
in the literature. All these concepts adopt the idea of
controlling the converter to mimic an actual Synchronous 1) A new grid-forming control scheme with the ability
Machine (SM) [22], [23]. Other control laws allowing the to decouple different control functionalities without a
grid-forming capability have been proposed recently: Vir- dedicated PLL.
tual Oscillator (VOC) [24]–[29], matching control [30] and 2) Simplified theoretical models to analyze the dynamic
Inducverter [31]. It has been proved that the VOC is equiva- behavior of the system. Time-domain simulations sup-
lent to a conventional droop control from a small-signal point port these models.
of view [25], [26], [32]. However, this control may be more 3) A generic approach for the implementation of the pro-
advantageous in case of voltage unbalance and distortion posed grid-forming control scheme on a 2-level VSC
due to its non-linear characteristic [25], [29]. The matching by including all the control functionalities; specifically
control is another method to emulate the SM’s behavior by current limitation algorithms.
creating a link between the DC bus voltage and the frequency 4) A method based on an adaptive inertia constant to
at the AC side through the control. The main drawback of improve the transient stability of the implemented
such controller is the all-in-one feature. It means that the DC grid-forming control.
voltage control, grid-frequency control, power sharing and To prove the effectiveness of the proposed control, dif-
the active power control are managed with only one control ferent test cases with a 5kW small-scale experimental plat-
gain that prevents optimizing all these functionalities at the form under normal and abnormal conditions have been
same time. tested.
Some of these controllers such as VSG and VSM use a The remaining of this paper is organized as follows.
PLL to decouple the power control from the frequency con- Section II presents the origin of the grid-forming control
trol capability. Whereas, the droop control, virtual oscillator, and introduces the controller, which drives the angle of the
matching control, Inducverter are not using a PLL. Actually voltage modulated by the converter. In section III, the full
the PLL-free controls are preferred to avoid the numerous sta- implementation of the grid-forming control on a 2-level VSC
bility issues caused by the PLL [33], [34] and possible inter- is explained. Section IV is devoted to the analysis of the
action between the PLL modes and power controller modes active power and frequency dynamics around an operating
[35]. However, with these control techniques, a coupling point with the implemented control. In section V, the transient
between the different control functionalities has been noticed. stability issue is addressed and a method to enhance the stabil-
In case of a grid frequency variation these controllers cannot ity is proposed. The experimental validation of the proposed
guarantee the steady state active power tracking because of control is given in section VI. Finally, section VII concludes
the inherent frequency support function, which is induced by the paper.
these types of grid-forming control [33].
To the best of our knowledge, until now, no control tech- II. BASICS ON THE GRID-FORMING CONTROL
nique that allows decoupling the control functionalities with- The main aim of a Voltage Source Converter (VSC) is to
out any dedicated PLL has been proposed. This decoupling is convert an AC power to a DC power and vice-versa. In what
very important since the owner of the power electronic device follows, the VSC is assimilated to a controllable three-phase
can decide if the converter participates to primary frequency voltage source: vma , vmb , vmc , which are modulated from
support or not. a DC bus voltage. The control system defines a set of
In this paper, a grid-forming control scheme embedding the three-phase signals v∗ma , v∗mb , v∗mc , in such a way that the aver-
decoupled capabilities is developed. The proposed control is age values of vma , vmb , vmc during the switching time of the
PLL-free. It is demonstrated that it meets the main require- transistors are equal to the corresponding reference signals
ments of a large transmission system such as inertial effect, (see Fig. 1).
good robustness in case of a grid impedance variation, and Some assumptions are made to simplify the analysis:
management of the voltage at the Point of Common Coupling • It is supposed that the VSC is connected to the grid with a
(PCC). It is also shown that with a simple modification on transformer, which is modeled with its series impedance
the inertia constant during a fault, this control reveals a good Rc + jXc . In the first step, the Lf Cf filter that is usually
transient stability. The contributions of the paper are listed as used for 2- Level VSCs is neglected in order to simplify
follow: the analysis.

197364 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 4. Power control with a double integrator.

FIGURE 2. Single-phase quasi-static representation of the VSC connected

to the grid. In the real life, the control is implemented in the time
domain. Let’s define ωm , the modulated voltage angular fre-
quency (in per-unit) and ωg , the frequency at the PCC. ωb is
the base frequency in rad/s. θm and θg are the time-domain
angles for the modulated voltage and the PCC voltage vg ,
respectively. It results in:

θm (t) = ωm ωb t + δm , (2)
θg (t) = ωg ωb t + δg . (3)

The Electromagnetic Transient (EMT) implementation of

the active power control is illustrated in Fig. 3-b where ψ is
supposed to be small enough so that sin(ψ) ≈ ψ.
Several types of controller can be chosen to control the
active power. Since the disturbance θg is linearly varying with
the time in steady state, a simple integrator is not enough
FIGURE 3. Active power control: (a) phasor representation (b) EMT to cancel the error between the active power and its refer-
representation. ence [32]. A double integrator is able to cancel this difference
in steady state (Fig. 4). Moreover, it brings an inertial effect
since it induces a proportional relation between the power
• The converter is connected to a power transmission sys- variation and the derivative of the frequency ωm :
tem, and therefore, the resistance is much smaller than
the reactance. 2H d 2 θm dωm
1P = P∗ (t) − P (t) = = 2H . (4)
• The grid is modeled by a Thevenin equivalent com- ωb dt 2 dt
posed of a voltage source veabc , an inductance Lg and a This equation is similar to the swing equation in a syn-
resistance Rg . chronous machine where, in per unit, the torque variation is
nearly equal to the power variation. From the block diagram
A. ACTIVE POWER CONTROL WITH THE MODULATED shown in Fig. 4, it is possible to calculate P with respect to
VOLTAGE ANGLE P∗ and θg :
The general organization of the active power control in a
2H 2
grid-forming converter can be explained with some simple 1 ωb s
considerations based on a quasi-static model of the grid P= 2H 2
P∗ − 2H 2
θg , (5)
ωb K c s +1 ωb Kc s + 1
described in Fig. 2.
The injected active power by the power converter in where Kc = Vm Vg /Xc . Equation (5) shows that this system
per-unit can be expressed as follows: is an oscillator. A damping effect has to be added to bring an
Vm Vg acceptable dynamic behavior.
P= sin (ψ) , ψ = δm − δg , (1)
Xc
B. PROPORTIONNAL AND INTEGRAL ACTION TO
where Vm and Vg are the RMS values of the modulated GENERATE THE INTERNAL ANGLE
voltage and the PCC voltage, respectively. The angles δm and A proportional action kp on the active power can be added
δm are the corresponding phasor angles. The base power Sb1 to damp the oscillation as depicted in Fig. 5. It results in a
is equal to the nominal power of the converter. specific formulation of the PI controller, called IP controller
Since it is not possible to control a difference of angles, in this paper. The active power transfer function is expressed
the effective control signal is δm . Fig. 3-a shows a general as follows:
scheme for the active power control. The output of the control 2H 2
is δm∗ , a reference angle for δ . Fig. 3-a presents also a 1 ωb s
m P= 2H 2
P∗ − 2H 2
θg . (6)
quasi-static system model. ωb Kc s + 2Hkp s + 1 ωb Kc s + 2Hkp s + 1

VOLUME 8, 2020 197365

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 5. Proposed scheme with an IP controller.

FIGURE 7. Scheme for generation of the reference voltage magnitude.

FIGURE 6. Equivalence between VSM and PI-controller.

FIGURE 8. A 2-Level VSC with an LC filter connected to the grid.

The gain kp is calculated with respect to a specific damping

coefficient ζ . From the original formulation, it is possible C. GENERATION OF REFERENCE VOLTAGE MAGNITUDE
to modify the scheme proposed in Fig. 5 in order to reveal The second constituent of the grid forming control is the
a ‘‘Virtual Synchronous Machine’’ like formulation. Indeed, management of the voltage magnitude. In steady state its
from (1), it can be deduced that: reference is determined by the given setpoints v∗md and v∗mq as
0 0
Vm Vg ωm − ωg ωb


ωm − ωg ωb

illustrated in Fig. 7. When choosing to align the voltage with
P= = Kc . (7) the d-axis, then v∗mq is set to 0. However, during a transient,
Xc s s 0
maintaining a constant voltage would induce some oscilla-
The internal frequency from the proposed control depicted tions due to the line’s dynamics, which are not represented in
in Fig. 5 is expressed as: the quasi-static formulation (presented in Fig. 5) [32], [36].
This effect is noticed especially in a transmission system
1 when there is not enough resistive effect to damp these natural
ωm = P∗ − P − kp P.


(8)
2Hs oscillations. Hence a transient damping resistor (Rv ) has to
be added in the control. ωf is the cut-off frequency of the
Replacing the term kp P in (8) by (7) yields: high-pass filter [32], [36].
1
kp Kc (ωm − ωg )ωb Since the connection impedance Xc (0.15 p.u.) is much
ωm = P∗ − P − . (9) smaller than for a synchronous machine in steady state
2Hs s
(≈2.5 p.u.) [37], the voltage at the PCC is stiffly controlled.
Equation (9) can be re-written as follows: Hence, an Automatic Voltage Regulator (AVR) is not com-
pulsory. If needed, a local reactive droop control can be
2Hs.ωm = P∗ − P + 2Hkp Kc (ωg − ωm )ωb .


(10) added as it is shown in Fig. 7, where nq and TQ are the
reactive droop gain and the time constant of the reactive loop,
This new mathematical formulation of the internal fre- respectively.
quency is graphically illustrated in Fig. 6-a. It has to be
emphasized that this control cannot be implemented like that. III. IMPLEMENTATION OF THE GRID-FORMING CONTROL
It is just another way to present the control in order to show ON THE VSC
its similarity with the VSM presented in Fig. 6-b. Although The previous section has presented the way to generate the
the general organization is similar, there is a fundamental dif- reference voltage with an ideal VSC. This is the core of the
ference since the estimation of the grid frequency ω˜g needed control, but other functionalities have to be implemented in
in the VSM [18] is replaced with the grid frequency ωg itself. order to get to the proper performances. In case of a 2-level

197366 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 9. Implementation of the grid-forming 2-Level converter based on IP controller.

FIGURE 10. Full implementation of grid-forming control with current limitation algorithms.

VSC with an LC filter (see Fig. 8), the control has to be It is important to mention that the system in Fig. 8 including
adjusted in order to take the internal current and voltage loops the control structure of Fig. 9 is a non-linear system. The non-
into account. Moreover, the control has to embed a current linearity is mainly due to the calculation of the active power
limitation algorithm in order to protect the converter against in the control side (P = egd igd + egq igq ), and the evolution of
overcurrent. E V
the active power with respect to the angle (P = Xcg+Xeg sinδeg ).
A. INTEGRATION OF THE INTERNAL LOOPS
Generally, for two-level converters, an LC filter is required to B. CURRENT LIMITATION CONTROL
eliminate the high-frequency harmonics caused by the PWM The core of the grid forming control is based on the gen-
signals. When the LC filter is considered, the active power eration and tracking of an AC voltage source. In case of
control is then achieved by acting on the phase of the voltage some events in the grid some over currents may be induced
eg with respect to vg . Then, this voltage is substituted to vm in in the power electronic converter. Fig. 10 shows two gen-
all the previous analysis. The internal angle θm and frequency eral approaches to implement a current limitation for the
ωm are replaced with θeg and ωeg respectively. The voltage v∗m grid-forming control: the Virtual Impedance (VI) [39]–[41]
is calculated from the voltage and current control loops [38] and the Current Saturation Algorithm (CSA) [42]–[44]. Both
(see Fig. 9). algorithms are recalled:

VOLUME 8, 2020 197367

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

1) CURRENT SATURATION IV. ANALYSIS OF THE PROPOSED GRID-FORMING

The saturation scheme is defined by the following function: CONTROL DYNAMICS AROUND AN OPERATING
 q POINT


 i ∗
sdqS
= i ∗
sdq
if i2sd + i2sq < ImaxSAT Two main events are analyzed in this section. First, a step
i∗sq
 " #


 i∗ = I
 q on the active power reference is applied in order to evaluate
max SAT cos arctan 2( ) if i2sd + i2sq > ImaxSAT the overall dynamics of the system and its robustness. Then,
sdS
i∗s
 d # the frequency response of the system either with or without
i∗sq

 "
implementation of the frequency support in the converter
 q
∗
i2sd + i2sq > ImaxSAT


 i = I max SAT sin arctan 2( ) if
 sqS
sd
i∗ control is studied.
(11)
A. ACTIVE POWER DYNAMICS
ImaxSAT stands for the maximum allowable current magni- From Fig. 5, the dynamic of the active power around an
tude of the converter, whereas i∗sdq denotes the saturated d-q operating point with respect to the power setpoint and the grid
SAT
currents. angle is expressed as follows:
2H 2
2) VIRTUAL IMPEDANCE 1 ωb s
1P = 1P∗ − 1θg .
The virtual impedance algorithm is activated only when the 2H 2 2H 2
ωb Kc s +2H kp s+1 ωb Kc s +2H kp s+1
current exceeds its nominal value In = 1p.u. Otherwise,
(17)
it will be disabled. The expressions the virtual impedance
(XVI and RVI ) are given in (12) and (13). Equation (17) can be simplified by expressing θg with
( respect to the active power. By neglecting the losses,
kpRVI σX /R δI if δI > 0
XVI = (12) the active power flow between the voltage terminals vm and vg
0 if δI 6 0, is the same as one between the terminals vg and ve . Therefore,
RVI = XVI /σX /R , (13) based on (7) it can be deduced:
Kc (1ωm − 1ωg )ωb Kg (1ωg − 1ωe )ωb
q
where δI = Is − In and Is = i2sd + i2sq is the converter 1P = = , (18)
current magnitude. The parameters kpRVI and σX /R are defined s s
respectively as the virtual impedance proportional gain and where Kg = Vg Ve /Xg . Since the grid frequency ωe is assumed
the virtual impedance ratio. to be constant, 1ωe ≈ 0. Therefore, (18) can be simplified
The parameter kpRVI is tuned to limit the current magnitude as:
to a suitable level ImaxVI during the overcurrent in steady Kg ωb
1P = 1ωg = Kg 1θg . (19)
state, while σX /R ensures a good system dynamics during s
the overcurrent and keeps a higher X /R ratio. The tuning By replacing 1θg in (17) by its expression in (19), the sim-
method of these parameters is explained in [33], [39]. When plified active power formula derived from Fig. 5 is obtained:
the virtual impedance is activated, the AC voltage references 1
are given by: 1P = 2H 1 1 2
1P∗ . (20)
(
ωb Kc + Kg )s + 2H k p s + 1
gd − δegd ,
e∗gd = e∗0 gq − δegq ,
∗
e∗gq = e∗0 ∗
(14)
The damping ratio ζ and natural frequency ωn can be calcu-
where, lated by following formulas:
δe∗gd = RVI isd − XVI isq ,
s s
(15) H ωb K c K g ωb Kc Kg
ζ = kp , ωn = . (21)
δe∗gq = RVI isq + XVI isd . (16) 2(Kc + Kg ) 2H (Kc + Kg )
Each algorithm can be implemented independently, or both From (21), the damping gain kp can be designed to achieve
of them can be hybridized to take the benefit of each one the desired dynamics of the active power based on ζ and ωn
at the same time as it is explain in [45]. The CSA allows for this simplified model.
a better current limiting during the first moments after the To illustrate the accuracy of the developed simplified
fault, whereas, VI cannot catch the first current peaks but it is small-signal model in (20), it is compared to the linearized
able to guarantee a better transient stability after fault clear- state-space model in (22) and non-linear model. The lin-
ance compared to the CSA. Therefore, their hybridization earized state-space model will also be used for robustness
allows an accurate current limitation and acceptable transient analysis in the next subsection.
stability.
Until here, the main functionalities of the grid-forming 1ẋ = A1x + B1u. (22)
control have been explained. The dynamic behavior of the
with:
converter has to be tested for different types of events T
egd egq isd isq igd igq δeg δωeg

(i.e. the behavior around an operating point and the perfor- x= ,
mance in case of large transient). . . . δeg ζdRv ζqRv Cvd Cvq Ccd Ccq

197368 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

TABLE 1. System and control parameters.

FIGURE 12. System pole map under various SCR.

FIGURE 11. Comparison between the nonlinear full dynamic model,

linearized model and proposed simplified model for SCR = 3.

FIGURE 13. Time domain simulation under various SCR.

uT = e0∗ P∗ .
 
gd0 Ve
• egd , egq , isd , isq , igd , igq : states of the LC filter.
• δeg , δωeg : states in the active power control loop.
• δeg : state of the filter applied to the reactive power
measurement.
R R
• ζd v , ζq v : states of the transient damping resistor.
• Cvd , Cvq , Ccd , Ccq : corresponding states of the internal
voltage and current control loops. FIGURE 14. Studied system with the line-tripping scenario.
The A and B matrices are given in the appendix. System
and control parameters are presented in Table1.
In Fig. 11, an active power set-point change of 1P∗ =
0.8 p.u is applied. The Short Circuit Ratio (SCR) is – The system is stable for a wide range of SCR.
pu
equal to 3 for this test case (SCR = 1/Xg [46]). The It should be noted that SCR = 1.2 is the critical SCR that
obtained results show the high fidelity of the proposed lin- allows transferring the rated power in steady state.
earized models in (20) and (22) compared to the non-linear Because of the high non-linearity of the system under
model. very weak grid conditions, the stability proof stated
in Fig. 12 is confirmed through time-domain simulations in
B. ROBUSTNESS AGAINST GRID IMPEDANCE VARIATION Fig. 13.
To assess the robustness of the proposed control against the The static variation of the SCR that is previously performed
SCR variation, some illustrative simulations are performed in should be supported by the real test cases that induce a
this subsection by considering the grid case of Fig. 8. SCR change. Line tripping in power transmission systems is
Figure 12 shows the eigenvalues trajectory of the system very likely and it leads to an increase of the interconnection
for both linearized state-space model and the simplified one impedance, and thus, a very fast decrease of the SCR.
with respect to the SCR variation. Two remarks can be derived This test case is performed on a grid impedance-switching
from this result: scenario illustrated in Fig. 14. When both impedances (Xg1 ,
– The dominant modes linked to the active power dynam- Rg1 and Xg2 , Rg2 ) are connected, SCR = 3, when Xg2 , Rg2 is
ics from the state-space model and simplified one are disconnected, the SCR decreases to 1.2. The corresponding
matched for various SCR. values of these impedances are given in Table 1.

VOLUME 8, 2020 197369

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 15. Active power response in case of line tripping.

FIGURE 17. AC voltage-controlled VSC connected to an equivalent AC grid

with variable frequency.

the inertia constant, the lead time constant and the lag time
constant, respectively.

1) SIMPLIFIED SYSTEM MODELING

FIGURE 16. PCC voltage in case of line tripping.
First, a simplified small-signal analytic model of the system
in Fig. 17 is built. The idea is to achieve a qualitative dynamic
analysis of this system in order to explain the active power
In Fig. 15, P∗ is initially set to 0.8 p.u. and both and frequency behavior. The converter behavior is driven
impedances are connected. At t = 1 s the impedance Xg2 , by (8). When studying the system around an operating point,
Rg2 is tripped. The result shows that the system remains it yields:
stable during this very strong disturbance and recovers its
equilibrium point within one second. 2Hs1ωm = −1P − 2H kp s1P
The decrease of the SCR results in a decrease of the = −1P − 2H kp Kc (1ωm − 1ωg )ωb . (23)
voltage magnitude at PCC ( Fig. 16). However, this voltage
The dynamics of the inertial AC source is represented by the
variation is acceptable since it meets the today’s grid code
differential equation:
requirements.
2Heq 1ωe s = 1Pm − 1Pg . (24)
C. FREQUENCY SUPPORT By neglecting the system losses:
In the previous sections, the VSC was connected to an infinite
bus, where the frequency is assumed to be constant. However, 1PSI
g + 1P = 1Pload ,
SI
(25)
an appropriate way to evaluate the inertial effect provided where 1PSIg = Sb2 1Pg and 1P = Sb1 1P. Converting this
SI
by the grid-forming converter is to test it on a variable fre- equation in Sb2 base results in:
quency grid. The goal is to study the frequency behavior
Sb 1Pload
with the proposed control and to characterize its inertial 1Pg + 1P 1 = . (26)
contribution through theoretical analysis and the time-domain Sb2 Sb2
simulations. To determine the dominant frequency response, it is assumed
Fig. 17 presents a VSC connected to an AC system formed that:
by a linear resistive load and an equivalent AC grid with a 1ωm = 1ωg = 1ωe . (27)
variable frequency. The equivalent AC grid consists of an
inertial AC voltage source in series with its impedance, which The validity of this assumption will be shown in the next
is driven by a model representing the equivalent dynamic subsection through time-domain simulations. By taking (27)
behavior of a large system. It consists of a swing equation into acount, (23) can be simplified as follows:
and a lead-lag filter, which is a simplified model of the turbine 2Hs1ωe = −1P. (28)
dynamics [21]. A droop control is added to support the grid
frequency. By putting 1P from (28) and 1Pg from (24) in (26),
AC Grid Notations: PSI the following equation that describes the grid frequency
g and Pg are the active power cal-
culated from the grid side in per-unit and SI, respectively. ωe dynamics is obtained:
is the grid frequency. Sb2 is the base power of the equivalent Sb 1Pload
2(H 1 + Heq )s1ωe = 1Pm − . (29)
AC grid. RSG is the droop control gain. Heq , TN and TD are Sb2 Sb2

197370 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 18. The simplified representation of the grid frequency dynamics.

TABLE 2. Parameters of the variable frequency grid.

FIGURE 19. Active power dynamics with respect to a load change.

Equation (29) can be illustrated graphically by the block

diagram in Fig. 18. This demonstrates the high similarity
between the proposed control and the swing equation of a
synchronous machine [37], [47]. Then, the dynamics of the
closed loop system can be derived:
1 RSG (1 + TD s)
1ωe = − 1Pload , (30)
Sb2 1 + as + bs2
Sb1
where a = TN + 2(Heq + Sb2 H )RSG and b = 2(Heq +
Sb1 FIGURE 20. Frequency dynamics with respect to a load change.
Sb2 H )RSG TD .Based on (30) the Rate of Change of Fre-
quency (RoCoF) expression for a load step (1Pload /s) can
be determined: droop control is not considered in this test case. There-
fore, the load is fully handled by the AC grid in steady state.
1Pload
lim s2 1ωe fb = Sb
fb , (31) The increase of the inertia constant decreases the RoCoF as
s→∞
2(Heq + Sb1 H )Sb2 demonstrated by (31) and improves the frequency nadir as
2
expected. (e.g., H = 5 s and H = 2.5 s correspond to
where fb is the base frequency in (Hz). By putting 1ωe RoCoF = 1.9 Hz/s and RoCoF = 2.5 Hz/s, respectively).
from (28) in (30), the converter active power with respect to The outer frequency droop control for the VSC depicted by
the load change is obtained as follows: the green block in Fig. 17 is an optional functionality that can
1 2H RSG (1 + TD s)s be added in order to allow the power converter participating
1P = 1Pload . (32) to the frequency support in steady-state. The droop control is
Sb2 1 + as + bs2
given by the following formula:
1 ∗
2) SMALL-SIGNAL MODEL VALIDATION AND ANALYSIS P∗ =(ω − ωm ). (33)
To validate the proposed simplified small-signal model R
(in (30) and (32)), it is compared to the non-linear model where ω∗ and R are the frequency setpoint and droop gain
in Fig. 17 for two cases: H = 5 s and H = 2.5 s. of the converter, respectively. By considering (33), the sim-
The parameters of the variable frequency AC grid are given plified representative model in Fig. 18 is modified by the
in Table 2. Time-domain simulations are given in Fig. 18 and one in Fig. 21. This simplified model is compared to the
Fig. 19, respectively, where a 400 MW load change is applied non-linear model in Fig. 22 to prove its accuracy.
to system.
These simulation results confirm the accuracy of the sim- V. RESPONSE TO LARGE EVENTS AND TRANSIENT
plified model to reproduce the dominant poles. Moreover, STABILITY
in Fig. 19 when the load is connected, the inertial support of To address the transient stability of the proposed control with
the grid-forming VSC based-PI controller strongly increases respect to a large disturbance, according to Fig. 23, a three
the VSC active power only during the transient since the phase bolted fault with different fault durations Tf is applied

VOLUME 8, 2020 197371

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 23. Studied system with the fault scenario.

These methods require additional fault detection algorithms

and triggering conditions. Additionally, as mentioned in [36],
the fault recovery based on these methods could only be
achieved after a considerable re-tuning of the voltage con-
FIGURE 21. The simplified representation of the grid frequency dynamics troller and the power controller to avoid the AC voltage col-
including the droop control. lapse. In the following lines, a method is proposed to enhance
the transient stability with no need for a PLL or control mode
switching.
Inspiring from the operation of a synchronous machine,
a high inertia constant results in a better transient stability.
Contrary to the SM where the inertia constant is imposed by
the mechanical characteristics, the inertia constant is tunable
with a grid-forming VSC, thus, an adaptive inertia algorithm
is proposed in this section to limit the frequency deviation
only when the current limitation algorithms are activated.
In other words, the proposed algorithm does not intervene in
normal operation.
The proposed solution in (34) consists in modifying the
inertia constant with respect to the AC voltage reference
(e0∗ 0∗
gd0 , egq0 in Fig. 10) generated from the virtual impedance.
If the current does not reach its maximum allowable value
Imax VI , then, the generated d-q voltages δe∗gdq from the virtual
impedance are equal to zero. While, when a fault occurs
and Is > Imax VI , the AC voltage reference is decreased to:
q
(1 − δe∗gd )2 + (δe∗gq )2 , where the value of e0∗
gd0 is set to 1 and
0∗
egq0 is set to 0. Therefore, the inertia constant is modified with
FIGURE 22. Active power and frequency dynamics with respect to a load the following formula:
change including the droop control (H = 5 s).
H0
H=q , (34)
(1 − δe∗gd )2 + (δe∗gq )2
at the departure of one of the lines and it is cleared by tripping
the faulty line, which results in a phase shift. where H0 is the initial constant in normal operation.
The results of this test case are gathered in Fig. 24. The The contribution of the proposed algorithm to the transient
current is well limited during the fault as expected, however, stability is demonstrated through the same test case presented
depending on the fault duration; the system may lose the in Fig. 24. The simulation results including the proposed
synchronism. The fault clearance in 150ms results in a stable algorithm are given in Fig. 25. One can notice that the system
resynchronization, while, when the fault duration is higher, now remains stable for both fault durations.
the angle diverges, which yields a transient instability. More To demonstrate the effectiveness of the proposed control in
details on the transient stability analysis of this phenomenon case of another large event, a phase shift that is linked to a line
can be found in [41], [45], [48]–[50]. reclosing is performed. The results are given in Fig. 26. In this
Referring to [45], the transient instability occurs when case, initially, the line that is modeled with X2 , R2 is assumed
no dedicated control algorithm is implemented to limit to be disconnected, then, it is connected at t = 1 s resulting
the angle deviation during large transients (i.e., when in a change of the SCR from 1.2 to 3, which engender an
the power mismatch is not controlled during the fault). important current transient that has been well limited thanks
Some papers propose to switch from grid-forming to the to the current limitation algorithm. After the disturbance,
PLL-based grid-following control in order to maintain the the system stably recovers its equilibrium point within 1s.
synchronism and to ensure a transient stability [36], [51]. Since the voltage magnitude at the PCC is not moving,

197372 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 27. Mockup presentation.

FIGURE 24. Response of the converter to the line-tripping event with

time duration.

FIGURE 28. Functional scheme.

TABLE 3. Mockup parameters.

FIGURE 25. Enhancement of the transient stability through the adaptive

inertia algorithm.

FIGURE 26. Phase shift of 30◦ linked to a line reclosing.

the inertia parameter is kept nearly constant as presented

in (34).
FIGURE 29. Active power change.
VI. EXPERIMENTAL VALIDATION
The aim of this section is to validate experimentally the the- The 2-Level VSC (1) is supplied by an ideal 600V DC
oretical developments. The experimental bench is illustrated voltage source (4) and connected to a high bandwidth AC
in Fig. 27. amplifier (5) through an LCL filter (3) as depicted in Fig. 28.

VOLUME 8, 2020 197373

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

FIGURE 30. Result of experimentation - (a): 100% voltage sag, (b): 40◦ phase shift.

197374 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

 ωb ωb 
 0 ω 0 ω b 0 − 0 
Cf Cf

 

ω

 

b
ω
 
−ω 0 0 0 0
 
0 b

 


 C f


ω ω ω ω ω
   

 b b b Rf b b


− k k − k C ω − k + 0 k 0
 
0

 pc pv pc f pc pc


 L Lf Lf Lf L 
 f f 
ωb ωb ωb Rf ωb ωb
 
A11 =
 kpc Cf ω0 − kpc kpv 0 − kpc + 0 kpc 


 Lf Lf Lf Lf Lf 


ω ω



 b Rc +R g b


0 0 0 − ω0 ωb

 

 


 L C +L g L C +L g



ω ω



b Rc +R g b

−ω0 ωb
 
0 0 0 −

 

 
LC +Lg LC +Lg 6×6
 

 0 0 0 0 0 0 0 0 0 


 0 0 0 0 0 0 0 0 0 

ωb ωb ωb ωb

 

 

 0 k pc k pv kpc kiv 0 0 k ic 0 k pc k pv 0 

L L L L
 
f f f f
 
ωb ωb ωb ωb

 

 
A12 = 0 0 0 kpc kiv kpc kiv 0 kic 0 kpc kpv
Lf Lf Lf Lf
Ve sin δeg0 ωb

 

 
0 0 0 0 0 0 0 0

 

 
 LC + Lg
 

δ ω
 
 V
 e
 cos eg0 b


0 0 0 0 0 0 0 0
 

 
LC + Lg 6×9
 
 −kp Igd0 −kp Igq0 0 0 −kp Egd0 0 

 


 I gd0 I gq0 E gd0


− − 0 0 − 0

 

 


 2H 2H 2H 



 nq Igq 0 n q I gd 0 n q E gd0


− 0 0 0

 

 


 T Q T Q T Q



−1 0 0 0 0 0

 


 

A21 = 0 −1 0 0 0 0
−Cf ω0
 
−kpv −1 0 1 0

 

 
 Cf ω0
 
−k 0 −1 0 1
 
pv
 

ω
 
ω


R R + R
 
 R V b v c g b

ω ω
 

 − 0 0 0 −R v 0 b


L + L L + L
 
C g C g

 

ω ω


R R R + R
 
 V b v c g b 
Rv ω0 ωb
 

 0 − 0 0 

LC + Lg LC + Lg 9×6
ω
 

 0 b 0 0 0 0 0 0 0 

0 0 0 0 0 0 0 0 0 

 

 


 −1 


 0 0 0 0 0 0 0 0 

T
 
Q

 

 
0 0 1 0 0 0 0 1 0

 


 

0 0 0 0 0 0 0 0 1
 
A22 =

 0 0 kpv kiv 0 0 0 kpv 0  
0 0 0 0 0 0 0 0 k
 

 pv  
Rv Ve sin δeg0 ωb

 

 

 − 0 0 0 0 0 0 −ω f 0 

L + L
 
C g

 

Rv Ve cos δeg0 ωb

 

 
−

 0 0 0 0 0 0 0 −ωf  
LC + Lg

9×9
0 0 ωf
 
0 0 0 0 0 0 0 0 0 0 0 0
cos δeg0 ωb sin δeg0 ωb

 

 

0 0 0 0 − 0 0 0 0 0 0 0 0 0
BT = L C + L g L C + L g
1

 


0 0 0 0 
0 0 0 0 0 0 0 0 0 0
2H 3×15

VOLUME 8, 2020 197375

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

Table 3 presents the mockup parameters. The amplifier is APPENDIX

used to emulate the AC system (i.e.; 300V ph-ph) as well as to The A and B matrices of the linearized state-space model are
generate the events discussed in this paper (i.e.; 100% voltage given in the following lines:
sag emulating a 3-phase bolted fault and a phase shift). The  
2-Level VSC is controlled with a dSPACE dS1005 (2) with a A11 A12
1ẋ = 1x + B1u, (35)
40 µs time step. The switching frequency of the converter is A21 A22
fsw = 10 kHz.
The electrical quantities displayed in experimentation are where, A11 , A12 , A21 , A22 , and BT as shown at the bottom
respectively the grid voltages vg in (abc) frame, the VSC of the previous page.
output current Is in (abc) frame, the grid current Ig in (abc)
frame and the active power. Three test cases are performed: REFERENCES
[1] European Union, ‘‘Commission Regulation (Eu) 2016/631,’’ Off. J. Eur.
– 0.5 p.u. active power change. Union, no. 14, p. 68, Apr. 2016.
– A 100% grid voltage sag. [2] A. Adib, B. Mirafzal, X. Wang, and F. Blaabjerg, ‘‘On stability of voltage
– A phase shift of 40◦ source inverters in weak grids,’’ IEEE Access, vol. 6, pp. 4427–4439, 2018,
doi: 10.1109/ACCESS.2017.2788818.
[3] A. Ulbig, T. S. Borsche, and G. Andersson, ‘‘Impact of low rotational
A. ACTIVE POWER CHANGE inertia on power system stability and operation,’’ 2013, arXiv:1312.6435.
At t = t0 , P∗ = 0.1 p.u, then, a 0.5 p.u power step is [Online]. Available: http://arxiv.org/abs/1312.6435
[4] ENTSO-E. (Mar. 2017). High Penetration of Power Electronic Interfaced
applied. Fig. 29 shows the active power dynamics with the Power Sources (HPoPEIPS). [Online]. Available: https://consultations.
proposed control. The active power tracks its reference with- entsoe.eu/system-development/entso-e-connection-codes-
out overshoot within 600ms, which is very close to what has implementation-guidance-d-3
[5] J. Liu, Y. Miura, and T. Ise, ‘‘Comparison of dynamic characteristics
been specified in Fig. 11. A difference between the simulation between virtual synchronous generator and droop control in inverter-based
and the experimentation can simply be explained by the fact distributed generators,’’ IEEE Trans. Power Electron., vol. 31, no. 5,
that the X /R ratio in a small-scale converter is smaller, which pp. 3600–3611, May 2016, doi: 10.1109/TPEL.2015.2465852.
[6] B. Li and L. Zhou, ‘‘Power decoupling method based on the diagonal com-
affects a little bit the dynamic performance of the system. pensating matrix for VSG-controlled parallel inverters in the microgrid,’’
Energies, vol. 10, no. 12, p. 2159, Dec. 2017, doi: 10.3390/en10122159.
B. 100% VOLTAGE SAG [7] J. Driesen and K. Visscher, ‘‘Virtual synchronous generators,’’ in Proc.
IEEE Power Energy Soc. Gen. Meeting-Convers. Del. Electr. Energy 21st
Initially, P∗ = 0.7 p.u. Subsequently, a 400 ms voltage sag is Century, Jul. 2008, pp. 1–3, doi: 10.1109/PES.2008.4596800.
applied as shown in Fig. 30-a. When the fault occurs, the AC [8] Q.-C. Zhong and G. Weiss, ‘‘Synchronverters: Inverters that mimic
voltage drop induces an increase in the VSC output current, synchronous generators,’’ IEEE Trans. Ind. Electron., vol. 58, no. 4,
pp. 1259–1267, Apr. 2011, doi: 10.1109/TIE.2010.2048839.
which is limited beyond its maximum allowable current mag- [9] Q.-C. Zhong, P.-L. Nguyen, Z. Ma, and W. Sheng, ‘‘Self-synchronized
nitude, 1.2 p.u (12 A). Once the fault is cleared, the system synchronverters: Inverters without a dedicated synchronization unit,’’ IEEE
stably recovers its equilibrium point. Trans. Power Electron., vol. 29, no. 2, pp. 617–630, Feb. 2014, doi: 10.
1109/TPEL.2013.2258684.
[10] Z. Ma, Q.-C. Zhong, and J. D. Yan, ‘‘Synchronverter-based control
C. 40◦ PHASE SHIFT strategies for three-phase PWM rectifiers,’’ in Proc. 7th IEEE Conf.
Ind. Electron. Appl. (ICIEA), Jul. 2012, pp. 225–230, doi: 10.1109/
In this subsection, the power reference is first set to ICIEA.2012.6360727.
P∗ = 0.5 p.u and a phase-shift of 40◦ is then applied, [11] R. Rosso, J. Cassoli, G. Buticchi, S. Engelken, and M. Liserre, ‘‘Robust
as illustrated in Fig. 30-b. The current rises to its maximum stability analysis of LCL filter based synchronverter under different grid
conditions,’’ IEEE Trans. Power Electron., vol. 34, no. 6, pp. 5842–5853,
allowable value, then, the system recovers its equilibrium Jun. 2019, doi: 10.1109/TPEL.2018.2867040.
point depending on the dynamics of the controller. [12] R. Rosso, S. Engelken, and M. Liserre, ‘‘Robust stability analysis of syn-
chronverters operating in parallel,’’ IEEE Trans. Power Electron., vol. 34,
no. 11, pp. 11309–11319, Nov. 2019, doi: 10.1109/TPEL.2019.2896707.
VII. CONCLUSION [13] W. Wu, Y. Chen, A. Luo, L. Zhou, X. Zhou, L. Yang, Y. Dong, and
In this paper, a grid-forming control strategy has been pro- J. M. Guerrero, ‘‘A virtual inertia control strategy for DC microgrids analo-
posed and its effectiveness under different grid test cases gized with virtual synchronous machines,’’ IEEE Trans. Ind. Electron.,
vol. 64, no. 7, pp. 6005–6016, Jul. 2017, doi: 10.1109/TIE.2016.2645898.
has been demonstrated. To analyze the dynamic performance [14] S. D’Arco and J. A. Suul, ‘‘Equivalence of virtual synchronous machines
of the proposed control, a small signal model has been and frequency-droops for converter-based MicroGrids,’’ IEEE Trans.
developed and verified. Moreover, the transient stability was Smart Grid, vol. 5, no. 1, pp. 394–395, Jan. 2014, doi: 10.1109/TSG.2013.
2288000.
addressed and an adaptive inertia constant algorithm has
[15] Y. Chen, R. Hesse, D. Turschner, and H.-P. Beck, ‘‘Comparison of methods
been implemented in order to guarantee the system stability for implementing virtual synchronous machine on inverters,’’ in Proc. Int.
against large events that may occur in power system. It can Conf. Renew. Energy Power Qual. J., Apr. 2012, pp. 734–739, doi: 10.
24084/repqj10.453.
be said that this simple PI controller with an adaption of
[16] H.-P. Beck and R. Hesse, ‘‘Virtual synchronous machine,’’ in Proc. 9th Int.
the parameters during a short circuit is extremely robust and Conf. Electr. Power Qual. Utilisation, Oct. 2007, pp. 1–6, doi: 10.1109/
it meets the main requirements asked by the transmission EPQU.2007.4424220.
system operators. Other perspectives consist in extending the [17] Y. Chen, R. Hesse, D. Turschner, and H.-P. Beck, ‘‘Improving the grid
power quality using virtual synchronous machines,’’ in Proc. Int. Conf.
proposed control for an MMC topology and testing it under Power Eng., Energy Electr. Drives, May 2011, pp. 1–6, doi: 10.1109/Pow-
unbalanced faulty conditions as well. erEng.2011.6036498.

197376 VOLUME 8, 2020

T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

[18] S. D’Arco, J. A. Suul, and O. B. Fosso, ‘‘A virtual synchronous machine [36] L. Zhang, L. Harnefors, and H.-P. Nee, ‘‘Power-synchronization con-
implementation for distributed control of power converters in smartgrids,’’ trol of grid-connected voltage-source converters,’’ IEEE Trans. Power
Electr. Power Syst. Res., vol. 122, pp. 180–197, May 2015, doi: 10.1016/j. Syst., vol. 25, no. 2, pp. 809–820, May 2010, doi: 10.1109/TPWRS.
epsr.2015.01.001. 2009.2032231.
[19] M. Yu, A. J. Roscoe, C. D. Booth, A. Dysko, R. Ierna, J. Zhu, [37] P. Kundur, Power System Stability and Control. New York, NY, USA:
and H. Urdal, ‘‘Use of an inertia-less virtual synchronous machine McGraw-Hill, 1994.
within future power networks with high penetrations of converters,’’ in [38] T. Qoria, F. Gruson, F. Colas, X. Guillaud, M.-S. Debry, and T. Pre-
Proc. Power Syst. Comput. Conf. (PSCC), Jun. 2016, pp. 1–7, doi: 10. vost, ‘‘Tuning of cascaded controllers for robust grid-forming voltage
1109/PSCC.2016.7540926. source converter,’’ in Proc. Power Syst. Comput. Conf. (PSCC), Jun. 2018,
[20] R. Hesse, D. Turschner, and H.-P. Beck, ‘‘Micro grid stabilization using pp. 1–7, doi: 10.23919/PSCC.2018.8443018.
[39] A. D. Paquette and D. M. Divan, ‘‘Virtual impedance current limiting
the virtual synchronous machine (VISMA),’’ Renew. Energy Power Qual.
for inverters in microgrids with synchronous generators,’’ IEEE Trans.
J., vol. 1, no. 7, pp. 676–681, Apr. 2009, doi: 10.24084/repqj07.472.
Ind. Appl., vol. 51, no. 2, pp. 1630–1638, Mar. 2015, doi: 10.1109/TIA.
[21] T. Qoria, F. Gruson, F. Colas, G. Denis, T. Prevost, and X. Guillaud,
2014.2345877.
‘‘Inertia effect and load sharing capability of grid forming converters [40] X. Lu, J. Wang, J. M. Guerrero, and D. Zhao, ‘‘Virtual-impedance-
connected to a transmission grid,’’ in Proc. 15th IET Int. Conf. AC DC based fault current limiters for inverter dominated AC microgrids,’’ IEEE
Power Transmiss. (ACDC), Coventry, U.K., 2019, p. 6. Trans. Smart Grid, vol. 9, no. 3, pp. 1599–1612, May 2018, doi: 10.1109/
[22] N. Soni, S. Doolla, and M. C. Chandorkar, ‘‘Improvement of transient TSG.2016.2594811.
response in microgrids using virtual inertia,’’ IEEE Trans. Power Del., [41] T. Qoria, F. Gruson, F. Colas, G. Denis, T. Prevost, and X. Guillaud,
vol. 28, no. 3, pp. 1830–1838, Jul. 2013, doi: 10.1109/TPWRD.2013. ‘‘Critical clearing time determination and enhancement of grid-forming
2264738. converters embedding virtual impedance as current limitation algorithm,’’
[23] J. Zhu, C. D. Booth, G. P. Adam, A. J. Roscoe, and C. G. Bright, ‘‘Inertia IEEE J. Emerg. Sel. Topics Power Electron., vol. 8, no. 2, pp. 1050–1061,
emulation control strategy for VSC-HVDC transmission systems,’’ IEEE Jun. 2020, doi: 10.1109/JESTPE.2019.2959085.
Trans. Power Syst., vol. 28, no. 2, pp. 1277–1287, May 2013, doi: 10. [42] N. Bottrell and T. C. Green, ‘‘Comparison of current-limiting strategies
1109/TPWRS.2012.2213101. during fault ride-through of inverters to prevent latch-up and wind-up,’’
[24] D. Groß, M. Colombino, J.-S. Brouillon, and F. Dörfler, IEEE Trans. Power Electron., vol. 29, no. 7, pp. 3786–3797, Jul. 2014,
‘‘The effect of transmission-line dynamics on grid-forming dispatchable doi: 10.1109/TPEL.2013.2279162.
virtual oscillator control,’’ 2018, arXiv:1802.08881. [Online]. Available: [43] M. G. Taul, X. Wang, P. Davari, and F. Blaabjerg, ‘‘Current limiting
http://arxiv.org/abs/1802.08881 control with enhanced dynamics of grid-forming converters during fault
[25] M. Ali, A. Sahoo, H. I. Nurdin, J. Ravishankar, and J. E. Fletcher, ‘‘On the conditions,’’ IEEE J. Emerg. Sel. Topics Power Electron., vol. 8, no. 2,
power sharing dynamics of parallel-connected virtual oscillator-controlled pp. 1062–1073, Jun. 2020, doi: 10.1109/JESTPE.2019.2931477.
and droop-controlled inverters in an AC microgrid,’’ in Proc. IECON- [44] Q.-C. Zhong and G. C. Konstantopoulos, ‘‘Current-limiting droop control
45th Annu. Conf. IEEE Ind. Electron. Soc., Lisbon, Portugal, Oct. 2019, of grid-connected inverters,’’ IEEE Trans. Ind. Electron., vol. 64, no. 7,
pp. 3931–3936, doi: 10.1109/IECON.2019.8927510. pp. 5963–5973, Jul. 2017, doi: 10.1109/TIE.2016.2622402.
[26] Z. Shi, H. I. Nurdin, J. E. Fletcher, and J. Li, ‘‘Similarities between [45] T. Qoria, F. Gruson, F. Colas, X. Kestelyn, and X. Guillaud, ‘‘Current
virtual oscillator controlled and droop controlled three-phase invert- limiting algorithms and transient stability analysis of grid-forming VSCs,’’
ers,’’ in Proc. IEEE 18th Int. Power Electron. Motion Control Conf. Electr. Power Syst. Res., vol. 189, Dec. 2020, Art. no. 106726, doi: 10.
(PEMC), Budapest, Hungary, Aug. 2018, pp. 434–439, doi: 10.1109/ 1016/j.epsr.2020.106726.
[46] E. Rokrok, T. Qoria, A. Bruyere, B. Francois, and X. Guillaud, ‘‘Effect
EPEPEMC.2018.8521928.
of using PLL-based grid-forming control on active power dynamics under
[27] M. Colombino, D. Groß, J.-S. Brouillon, and F. Dörfler, ‘‘Global phase
various SCR,’’ in Proc. IECON-45th Annu. Conf. IEEE Ind. Electron. Soc.,
and magnitude synchronization of coupled oscillators with application to
vol. 1, Oct. 2019, pp. 4799–4804, doi: 10.1109/IECON.2019.8927648.
the control of grid-forming power inverters,’’ 2017, arXiv:1710.00694. [47] T. Qoria, ‘‘Grid-forming control to achieve a 100% power electronics inter-
[Online]. Available: http://arxiv.org/abs/1710.00694 faced power transmission systems,’’ Ph.D. dissertation, ENSAM, Paris,
[28] D. Raisz, T. T. Thai, and A. Monti, ‘‘Power control of virtual oscil- France, 2020.
lator controlled inverters in grid-connected mode,’’ IEEE Trans. Power [48] L. Huang, H. Xin, Z. Wang, L. Zhang, K. Wu, and J. Hu, ‘‘Transient sta-
Electron., vol. 34, no. 6, pp. 5916–5926, Jun. 2019, doi: 10.1109/TPEL. bility analysis and control design of droop-controlled voltage source con-
2018.2868996. verters considering current limitation,’’ IEEE Trans. Smart Grid, vol. 10,
[29] T. T. Thai, D. Raisz, A. Monti, F. Ponci, and A. Ahmadifar, ‘‘Voltage har- no. 1, pp. 578–591, Jan. 2019, doi: 10.1109/TSG.2017.2749259.
monic reduction using virtual oscillator based inverters in islanded micro- [49] H. Xin, L. Huang, L. Zhang, Z. Wang, and J. Hu, ‘‘Synchronous
grids,’’ in Proc. 18th Int. Conf. Harmon. Qual. Power (ICHQP), Ljubljana, instability mechanism of P-f droop-controlled voltage source converter
Slovenia, May 2018, pp. 1–6, doi: 10.1109/ICHQP.2018.8378896. caused by current saturation,’’ IEEE Trans. Power Syst., vol. 31, no. 6,
[30] M. Z. Jacobson, M. A. Delucchi, M. A. Cameron, and B. V. Mathiesen, pp. 5206–5207, Nov. 2016, doi: 10.1109/TPWRS.2016.2521325.
‘‘Matching demand with supply at low cost in 139 countries among 20 [50] X. Wang, M. G. Taul, H. Wu, Y. Liao, F. Blaabjerg, and
world regions with 100% intermittent wind, water, and sunlight (WWS) L. Harnefors, ‘‘Grid-synchronization stability of converter-based
for all purposes,’’ Renew. Energy, vol. 123, pp. 236–248, Aug. 2018, resources—An overview,’’ IEEE Open J. Ind. Appl., vol. 1, pp. 115–134,
doi: 10.1016/j.renene.2018.02.009. 2020, doi: 10.1109/OJIA.2020.3020392.
[31] M. Ashabani, F. D. Freijedo, S. Golestan, and J. M. Guerrero, ‘‘Inducvert- [51] M. Ndreko, S. Rüberg, and W. Winter, ‘‘Grid forming control scheme for
ers: PLL-less converters with auto-synchronization and emulated iner- power systems with up to 100% power electronic interfaced generation:
tia capability,’’ IEEE Trans. Smart Grid, vol. 7, no. 3, pp. 1660–1674, A case study on great Britain test system,’’ IET Renew. Power Gener.,
May 2016, doi: 10.1109/TSG.2015.2468600. vol. 14, no. 8, pp. 1268–1281, Jun. 2020, doi: 10.1049/iet-rpg.2019.0700.
[32] T. Qoria, Q. Cossart, C. Li, X. Guillaud, F. Gruson, and X. Kestelyn,
‘‘Deliverable 3.2: Local control and simulation tools for large transmission TAOUFIK QORIA received the M.S. degree in
systems,’’ Migrate Project WP3, Paris, France, 2018, p. 89. electrical engineering for sustainable development
[33] G. Denis, ‘‘From grid-following to grid-forming: The new strategy from the Lille 1 University of Science and Tech-
to build 100% power-electronics interfaced transmission system with
nology, Villeneuve d’Ascq, France, in 2016. He is
enhanced transient behavior,’’ Ph.D. dissertation, Ecole centrale de Lille,
currently pursuing the Ph.D. degree with the
Villeneuve-d’Ascq, France, 2017.
ENSAM ParisTech in the Laboratory of Electrical
[34] F. Milano, F. Dorfler, G. Hug, D. J. Hill, and G. Verbic, ‘‘Foundations
and challenges of low-inertia systems (Invited Paper),’’ in Proc. Power Engineering and the Power Electronics of Lille
Syst. Comput. Conf. (PSCC), Dublin, Ireland, Jun. 2018, pp. 1–25, doi: 10. (L2EP), Lille, France. Since 2016, he has been
23919/PSCC.2018.8450880. working on the modeling and control of the power
[35] E. Rokrok, T. Qoria, A. Bruyere, B. Francois, and X. Guillaud, ‘‘Clas- inverters in HVDC and HVAC applications. He is
sification and dynamic assessment of droop-based grid-forming con- also working as a Contractual Researcher with the Ecole Centrale de Lille.
trol schemes: Application in HVDC systems,’’ Electr. Power Syst. Res., His current research interest includes the massive integration of power
vol. 189, Dec. 2020, Art. no. 106765, doi: 10.1016/j.epsr.2020.106765. electronic devices in power transmission systems.

VOLUME 8, 2020 197377

 T. Qoria et al.: PLL-Free Grid-Forming Control With Decoupled Functionalities for High-Power Transmission System Applications

EBRAHIM ROKROK (Student Member, IEEE) BRUNO FRANÇOIS (Senior Member, IEEE) is
received the B.Sc. and M.Sc. degrees in elec- currently a Full Professor with the Centrale Lille
trical engineering from the Isfahan University Institute and the Laboratory of Electrical Engi-
of Technology, in 2014 and 2016, respectively. neering and Power Electronics Laboratory, Lille.
He is currently pursuing the Ph.D. degree with the His research interests include the advanced energy
Ecole Centrale de Lille, Laboratory of Electrical management of power systems, architectures and
Engineering and the Power Electronics of Lille control systems of future electrical networks,
(L2EP), Lille, France. This Ph.D. is supported smart grids, and uncertainty impact in operation of
by the project "HVDC Inertia Provision" (HVDC electrical systems.
Pro), financed by the ENERGIX Program of the
Research Council of Norway (RCN). His research interests include modeling
and control of the power converters for HVDC applications. XAVIER GUILLAUD (Member, IEEE) has been
working as a Professor with the Lille Labora-
tory of Electrical Engineering and Power Elec-
tronics (L2EP), Lille, France, since 2002. First,
he worked on the modeling and control of power
ANTOINE BRUYERE (Member, IEEE) received electronic systems. Then, he studied the inte-
the Ph.D. degree in electrical engineering in 2009. gration of distributed generation and renewable
He spent ten years in automotive industry, at the energy in power systems. Nowadays, he is study-
Valeo Powertrain Electronics Product Group, ing the high-voltage power electronic converters
working as an Expert in automotive electrification. in transmission systems. He is leading the devel-
In 2016, he became an Assistant Professor at the opment of an experimental facility composed of actual power electronic
Centrale Lille Institute, Laboratory of Electrical converters interacting with virtual grids modeled in a real-time simulator.
Engineering of Lille. His research interests include He is involved in several projects in the field of power electronics including
renewable energy integration on power-grid, using European grid projects and different projects with French companies.
power-electronics converters.

197378 VOLUME 8, 2020