This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.Doi Number

Definition and Analysis of Receiving System

Voltage Support Strength Factor
Chengjun Xia1, Senior Member, IEEE, Xuanping Li1, Shoutao Li1, and Baorong Zhou2
1
School of Electric Power, South China University of Technology, Guangzhou 510640, China
2
Electric Power Research Institute of CSG, Guangzhou 510663, China

Corresponding author: Xuanping Li (e-mail: leexp39@foxmail.com).

This work was supported in part by the National Natural Science Fund & China-State Grid Joint Fund for Smart Grid (U1766213), in part by Science and
Technology Program of China Southern Power Grid(ZBKJXM20170071)..

ABSTRACT Based on the analysis of the relationship between the voltage sensitivity of the receiving
system and the DC power stability, an evaluation index, Receiving System Voltage Support Strength Factor
(RVSF), is proposed in this paper for identifying the voltage support strength of the receiving system. The
threshold of RVSF is derived and used as the criteria for distinguishing the voltage support strength of the
receiving system. The critical Receiving System Voltage Support Strength Factor (CRVSF) can keep
constant in receiving systems which operate in different modes. Furthermore, the analytical expression of
RVSF is derived from the Jacobian matrix form of the incremental flow equation of the AC system. By
deriving the relationship between RVSF and power stability and considering the constraint of commutation
angle, it can be obtained that the receiving system is a very weak system when RVSF<0.60, a weak system
when 0.60<RVSF<2, and a strong system in case of RVSF>2. Finally, the performance of RVSF and other
widely used system strength indicators such as effective short circuit ratio (ESCR) and multi-infeed
effective short circuit ratio (MIESCR) are compared and analyzed. Compared with other indicators, it is
found that RVSF has a clear threshold when measuring the strength of different receiving systems.

INDEX TERMS Multi-infeed HVDC, power stability, receiving system voltage support strength factor
(RVSF), voltage sensitivity.

Ⅰ. INTRODUCTION
With the development of HVDC transmission technology,
lots of HVDC transmission projects have been put into
operation worldwide. After the implementation of the
"West-to-East Electricity Transmission" strategy in China,
a multi-infeed HVDC system has been formed in the load
centers of economically developed regions such as Eastern
and Southern China[1]. The increase in the capacity and
number of HVDC systems fed into the same AC grid has
raised the risk of safe and stable operation of the grid and
brings a series of challenges to the normal operation of the
grid[2]. Therefore, accurately measuring the strength of the
receiving power grid is crucial to ensure the normal
operation of the multi-infeed system[3]-[4].
Researchers have done a lot of research on the system
structure, control strategy and stable operation of DC
transmission and distribution[5]-[12].When evaluating the
voltage support strength of the receiving system, the
indicators based on the grid structure parameters are widely
used, like, in a single-infeed system, the short circuit ratio
(SCR), the effective short circuit ratio (ESCR) and the Q
effective short circuit ratio (QESCR)[15]. These indicators
have a clear physical concept, which is calculated by the
grid structure parameters and the active power of the DC
system. Based on the constraint of DC power stability, the
critical short circuit ratio (CSCR) can be derived and used
to distinguish the voltage support strength of the receiving
system. It is generally considered that the critical short
circuit ratio is 2, which means that the LCC-HVDC system
does not work properly under the rated condition when the
short circuit ratio is less than 2[15]-[16]. In order to
evaluate the voltage support strength of the receiving
system of the multi-infeed system, the CIGRE DC working
group proposed an index of multi-infeed short circuit ratio
(MISCR)[17]. The indicator considers the interaction
between different DC systems, but the selection of the
threshold lacks a rigorous theoretical derivation[4]. Based
on the multi-infeed system, the reference [18] considers the
effects of DC system reactive power compensation,

VOLUME XX, 2017 1

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

different operation modes and parallel AC lines, and applicability of RVSF indicators.
proposes multi-infeed effective short circuit ratio The rest of this paper is organized into seven sections.
(MIESCR), multi-infeed operation short circuit ratio Following the introduction, analysis of DC system power
(MIOSCR) and improved short circuit ratio (ISCR). stability is assessed in Section Ⅱ. The definition of RVSF and
However, these indicators do not fundamentally solve the the threshold for identifying the strength of the receiving
problem that the multi-infeed system voltage support system with RVSF are provided in Section Ⅲ, while
strength index lacks theoretical support and clear threshold. analytical calculation method of RVSF is presented in
Based on an equivalent single-infeed model of a multi- Section Ⅳ. In Section Ⅴ, verification of the effectiveness of
infeed HVDC system, the reference [19] proposed an index RVSF indicator is obtained. Comparison between RVSF and
called equivalent effective short-circuit ratio (EESCR). other system strength indicators is performed in Section Ⅵ
However, since the different modes of operation of DC while the conclusion of this paper is shown in Section Ⅶ.
system are not considered, the proposed equivalent model is
optimistic. In the reference [20] and [21], the relationship Ⅱ. DC SYSTEM POWER STABILITY
between the maximum transmission power and the MISCR
of multi-infeed HVDC system is studied. It is proposed that A. POWER STABILITY ANALYSIS MODEL
the actual strength of the multi-feed HVDC system is The structure of a DC system and its receiving system is
between the MISCR and the SCR. In the reference [3], the shown in Fig. 1. Pd and −Qd indicate active power and
generalized short-circuit ratio index is defined from the reactive power flowing into the converter bus. U and δ are
perspective of the eigenvalue, and the threshold of the the converter bus voltage and its phase angle. The converter
index is determined by the voltage stability criterion, which transformer ratio is denoted as τ. DC voltage and DC
can effectively evaluate the voltage support strength of the current are represented by Ud and Id, and the extinction
multi-infeed system. The shortcoming of this method is that angle is plotted as  while the commutation angle is
the steady-state operating characteristics of each DC system written as μ. Specially, the reactive power compensation
are assumed to be the same. Under different operation device on the converter bus is considered as part of the
modes of receiving system, the power transmission and receiving system.
control modes of DC systems are different, so that the For simplicity, the calculations in this paper are based on
steady-state operating characteristics cannot be generalized. the per unit value. the DC power is taken as the rated power
The reference [22] defines the operating impedance of capacity of the system. For the based voltages, the nominal
VSC-HVDC, and ESCR of the system is obtained. In the voltage at AC side and DC side are selected as base
reference [23], all DCs fed into the receiving system are voltages of AC side and DC side, respectively.
equivalent to an equivalent conventional DC, and the We can use the following equations to describe the AC
evaluation index named critical equivalent effective short- system shown in Fig. 1 [15].
circuit ratio(CEESCR) is derived. The indicators proposed
Id
in the above reference are based on the equivalent of DC
systems connected to the grid, and it is difficult to meet the Pd , −Qd  :1
Receiving
situation of flexible and variable receiving system. It is a Ud
system
question worth exploring that whether there is a new
indicator that not only has clear threshold but also reflects
U  ,
the voltage support strength of the receiving system well in DC system
the case that the receiving system operates flexibly. FIGURE 1. Receiving system model of AC-DC power system.
A voltage support strength index based on DC power
stability and the voltage sensitivity for receiving system is
Pd =CU 2 ( cos 2 − cos(2 + 2  ) ) (1)
proposed in this paper. The main contributions of this paper
are as follow: Qd =CU ( 2 + sin 2 − sin(2 + 2  ) )
2
(2)
1) By deriving the relationship between DC power
I d =KU ( cos  − cos( +  ) ) (3)
stability and voltage sensitivity, the receiving system
voltage support strength factor (RVSF) is defined to U d = Pd I d (4)
characterize the voltage support strength of the U =f ( Qd , Pd ) (5)
receiving system.
2) Deriving the threshold of RVSF as the identification P = Pd (6)
criteria for the strength of the receiving system. Q = −Qd (7)
3) The analytical expression of RVSF is derived by the
modified Jacobian matrix of the receiving system. Equation (1)-(4) are quasi-steady-state equations for DC
4) Comparing RVSF with other system strength systems. (5) represents the characteristic equation of the
indicators, verifying the superiority and wide receiving system. P and Q in (6) and (7) are the active

VOLUME XX, 2017 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access
power and reactive power of the converter bus flowing into
the receiving system. (1)-(7) contain a total of 11 variables,
P, Q, Pd, Qd, Ud, Id, U, γ, μ, C and K. Among these variables,
C and K are coefficients related to the commutation
impedance, the converter transformer ratio and the base
value conversion. When equipped with a perfect filtering
device, the converter bus voltage is sinusoidal, and the
commutation impedance is the short-circuit reactance of the
converter transformer[7]. In the case where the converter
transformer tap is not considered, C and K are constants
only related to the converter transformer parameters and the
base value conversion. The value of C and K can be
obtained as follow
C =3nST ( 4 P u 2)
dN k
Where, St is the capacity of converter transformer. n is
the number of thyristor valve groups, and u k stands for the
short-circuit voltage ratio of the converter transformer.
The remaining nine variables, P, Q, Pd, Qd, Ud, Id, U, γ
and μ, are the operating state variables of the DC system.
Each set of values of these variables corresponds to one
operating point of the DC system. At the rated operating
point, the nominal values of these variables are expressed
as Pdn =1, UN =1, IdN =1 and  =  N . Substituting the above
nominal values and the parameter C into the (1)-(4), we can
derive the QdN, UdN, μN and K. It is worth mentioning that
this is one way to easily obtain the value of K which is a
constant related to the parameters of the converter
transformer, and it keeps constant even if the operating
point changes.
There are seven constraint equations in (1)-(7), so two
operating state variables are independent in the nine
variables. Once any two of these variables are determined,
the operating state of the DC system can be obtained. These
two independent variables can be selected differently
according to different control modes and analysis purposes.
B. POWER STABILITY
Many studies have shown that power stability is a major
factor limiting the normal operation of DC systems[8].
According to the analysis in Section II-A, we choose  N
and Id as independent variables. In the constant extinction
1.0
Pd
0.8
0.6
0.5 1.0 1.5
Id
FIGURE 2. Maximum power curve of HVDC system.
VOLUME XX, 2017
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
angle control mode, taking  = N = min and gradually
increasing the DC current Id, the curve of the DC
transmission power Pd changing with DC current Id can be
obtained, and the curve is shown in Fig. 2.
As shown in Fig. 2, when the DC system is operating in
the area of dPd/dId >0, the DC transmission power increases
with the increase of the DC current, and the DC system can
maintain power stability. When the DC system operates in
the area of dPd/dId <0, the power decreases with the
increase of the DC current, and it is difficult for the DC
system to maintain the power stability[13]-[14]. When
dPd/dId =0, the DC system reaches the maximum
(8)
transmission power and it is the limit of power stability. At
a certain voltage support strength of the receiving system,
the rated operating point of the DC system coincides with
the point of maximum available power (MAP). At this
point we can get that
dPd
=0 (9)
dI d I =1
d
When (9) satisfies the condition, the voltage support
capability of the receiving system is just enough to support
the DC system to operate stably at the rated operating point.
At this time, the voltage support capability of the receiving
system is defined as the critical receiving system voltage
support strength. If the voltage support capability of the
receiving system continues to decrease, and even becomes
less than the critical receiving voltage support strength, the
DC system will be difficult to operate normally under the
rated conditions.
C. THE RELATIONSHIP BETWEEN THE VOLTAGE
SENSITIVITY OF THE RECEVING SYSTEM AND THE
POWER STABILITY OF THE DC SYSTEM
There is  = 0 when the inverter side of the DC system is
in the control mode of constant extinction angle . When the
DC current order ΔId is increased, the changes of other
operating state variables are ΔPd, ΔQd, ΔP, ΔQ, ΔU, ΔUd
and Δμ. Because P = Pd and Q = Qd , the small signal
total differential form of (5) can be written as
U U
U = − Qd + Pd (10)
Q P
Where, U / Qd represents the sensitivity of the converter
bus voltage to the reactive power when the active power is
constant. U / Pd stands for the sensitivity of the converter
bus voltage to the active power when the reactive power is
constant. (10) establishes the relationship between the
2.0 sensitivity of the receiving system and the change of the
converter bus voltage. When the inverter side is under the
9
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

control mode of constant extinction angle, the total operating parameters of the DC system and the parameter C.
differential form of (1)-(4) is derived as (11)-(14). According to the calculation method in Section II-A, the
rated operating point of the system depends on parameters
Pd P
Pd = U + d  (11) C and  N . Without considering the voltage regulation of the
U  converter transformer tap, parameter C is constant. If the
Qd Q parameter  N is constant, the U Q corresponding to the
Qd = U + d  (12)
U  critical terminal support strength remains unchanged.
I d I
I d = U + d  (13) Ⅲ. RECEVING SYSTEM VOLTAGE SUPPORT
U  STRENGTH FACTOR
Pd
=0 (14) A. DEFINITION OF RECEVING VOLTAGE SUPPORT
I d STRENGTH FACTOR
The DC system consumes a lot of reactive power during
Where, operation, and the reactive power consumption changes
 Pd when the DC power fluctuates. The measurement of the
 U =2CU ( cos 2 − cos(2 + 2 ) ) voltage support capability of the receiving system is largely
 a measure of the voltage stability of the converter bus when
 Pd =2CU 2 sin(2 + 2 ) the reactive power of the receiving system changes. Define
  the Receiving-system Voltage Support Strength Factor

 Qd (RVSF) as
 U =2CU ( 2 + sin 2 − sin(2 + 2 ) )
Q
 (15) RVSF = (19)
 Qd =2CU 2 (1 − cos(2 + 2 ) ) U P = PdN
Q =− QdN
 

 I d =K ( cos  − cos( +  ) )
Where, P and Q are the active power and reactive power
 U injected into the converter bus, respectively. Then at this
 I time, Q U is the ratio of small reactive power change
 d =KU sin (  +  ) injected into the converter bus to the voltage change of the
 
converter bus when system operates at the rated operating
Through simplifying (10)-(14), we can get point and it does not consider the characteristics of the DC
system on the converter bus. The DC system on the
U converter bus can be replaced with constant power at its
+ba
dPd Q rated state, and then RVSF can be obtained by applying the
= (16)
dI d c + d U + e U reactive disturbance in the simulation. Obviously, RVSF
Q P can also be expressed as
-1
When the system is running at the rated power point,  U 
RVSF =   (20)
there are PdN=1, UN=1, IdN=1, then  Q 
a = −4CQdN sin ( 2 N + 2  N ) + 4C (1 − cos ( 2 N + 2  N ) ) RVSF can reflect the voltage support strength of the

b = −2C sin ( 2 N + 2  N ) receiving system. When the reactive power is changed in
 the injection receiving system, the larger the RVSF is, the
c = − K sin (  N +  N ) (17)
more stable the converter bus voltage is, the greater the

d = −2 KQdN sin (  N +  N ) + 2C (1 − cos ( 2 N + 2  N ) ) ability of the receiving system to support the voltage is.

e = 2 K sin (  N +  N ) − 2C sin ( 2 N + 2  N ) B. RVSF THRESHOLD
The RVSF obtained when Pd / I d is defined as a Critical
Substituting (17) into (16), there is
Receiving-system Voltage Support Strength Factor
U sin ( 2 N + 2  N ) (CRVSF).
= (18)
Q 2QdN sin ( 2 N + 2  N ) − 2 (1 − cos ( 2 N + 2  N ) )
2 (1 − cos ( 2 N + 2 N ) ) − 2QdN sin ( 2 N + 2  N )
CRVSF= (21)
Because QdN is related to parameter C, the value of sin ( 2 N + 2 N )
U / Q corresponding to the critical receiving system
voltage support strength can be determined by the rated

VOLUME XX, 2017 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
As can be seen from the analysis in Section II, the value
of CRVSF is determined by parameters C and  N .Based on
the Cigre benchmark model, the typical values of the
converter transformer parameters are used. There are n=2,
γN=15°, St =0.59179 and uk=18%. And then we can get that
C=1.3515. The CRVSF=0.603 can be obtained by taking
the typical value γN=15°of constant extinction angle control.
It is worth noting that we do not limit the parameters and
structure of the receiving system during the calculation of
CRVSF, so CRVSF is able to satisfy the diversity of
structure and components of the receiving system. It can
also be regarded as a criterion used to distinguish weak
receiving system and very weak receiving system.
E  P, Q Pd , −Qd  :1
Zeqeq
U 
FIGURE 3. Simplified model of single-infeed HVDC system.
In 12-pulse converter, the limit of the commutation angle
μ is 30°. When the strength of the receiving system
supporting voltage is weak, the point of MAP appears
before the operating point of μ=30° in the process of
increasing the current Id. When the strength of the receiving
system supporting voltage is strong, the point of MAP
appears after the operating point of μ=30°, so DC system
will not suffer power instability. Since the operating point
of μ=30°is not near the rated operating point, it is difficult
to directly establish the relationship between the RVSF and
the variables at the operating point. To determine this
relationship, we must rely on a structure of the receiving
system. For simplicity, the receiving system was simplified
to a Thevenin equivalent circuit to form a single-infeed
system as shown in Fig. 3. Where, E,  , Zeq and θeq are
equivalent parameters of the AC system. P and Q represent
the active power and reactive power flowing into the
equivalent AC system from the converter bus.
As shown in Fig. 3, the relationship between Q / U ,
P / U and the system equivalent impedance parameters
can be obtained by the power flow equation of the AC
system.
Z eq 2
Z eq sin  eq +
U U2
= (22)
Q Z eq 2
U−
U3
(Pd
2
+ Qd
Z eq 2
Z eq cos  eq −
U U2
= (23)
P Z eq 2
U−
U3
(Pd
2
+ Qd
VOLUME XX, 2017
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
10.1109/ACCESS.2019.2940409, IEEE Access
The RVSF is changed along with the magnitude of the
impedance Zeq. A plurality of different impedance angles
θeq are calculated, finally the point of MAP is coincident
with the operating point of μ=30°, and RVSF of the
receiving system approaches 2. Therefore, the criterion for
distinguishing the weak receiving system from the strong
receiving system is RVSF=2.
C. RVSF CRITERIA FOR DISTINGUISHING SYSTEM
When the DC system feeds power into the AC system, there
is an interaction between AC and DC systems. The higher
the equivalent impedance of AC system, the smaller the
short-circuit capacity of the system, and the more
significant the interaction between AC and DC systems.
When the system strength changes from weak to strong,
Ud RVSF=0.60 and RVSF=2 are two special points. When
RVSF<0.60, the IMAP corresponding to MAP is less than the
 ， rated current Id=1. Since the system strength is unable to
support the DC system to operate at the rated operating
point, the power transfer limit of LCC is determined by the
MAP, which is smaller than rated power. When
RVSF=0.60, the rated operating point coincides with the
point of MAP. At this time, the system strength can support
the DC system to operate at the rated operating point of
Id=1, and the power limit is equal to the MAP. When
RVSF>0.6, the IMAP is greater than the rated current Id=1.
The system strength is adequate to support the DC system
to operate at the rated point. Since a 12-pulse converter
usually requires its commutation angle to operate in a range
of less than 30 degrees, in some cases the commutation
angle imposes constraints on the power transfer limit. When
0.6<RVSF<2, the IMAP is always less than the Iμ
corresponding to μ=30°, but greater than the rated current
Id=1. In the case, the commutation angle does not affect the
power transfer limit, so the power transfer limit is
determined by the MAP. When RVSF>2, the IMAP appears
later than the Iμ that the commutation angle imposes
constraints on the power transfer limit. And the power
transfer limit is determined by the transfer power at μ =30°.
The definitions of three systems are given below.
When the power transfer limit of the LCC is less than the
rated power, but equal to MAP, we call the receiving
system a very weak system, and its strength is described as
RVSF<0.6.
Qd When the power transfer limit is greater than the rated
power and equal to the MAP, we call the receiver system a
2
) weak system, and its strength is described as 0.6<RVSF<2.
When the power transfer limit is greater than the rated
power, but limited by the commutation angle of 30° and
cannot reach MAP, we call the receiver system a strong
Pd
system, and its strength is described as RVSF>2.
Therefore, the RVSF index based on the sensitivity of
2
) reactive power to voltage can give a clearer boundary to the
system strength without considering the specific topology
9
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

and characteristics of grid components according to the  PR      

    J P J PU
following criteria.    JM   =     U 
(25)
Very weak AC system: RVSF≤0.60  QR   U   J Q J QU
Weak AC system: 0.60<RVSF≤2  and J QU
Where JM is the modified Jacobian matrix. J PU  are
Strong AC system: RVSF>2 the correction matrix of JPU and JQU. ΔPR and ΔQR are the
active power disturbance and reactive power disturbance
Ⅳ. ANALYTICAL CALCULATION METHOD OF RVSF injected into the commutating bus. For nodes i, ΔPR and
Fig. 4 shows a hybrid multi-infeed HVDC system model ΔQR can be expressed as follows.
consisting of LCC-HVDC and VSC-HVDC. Using RVSF QiR = Qi − Qid − Qil
to evaluate the receiving system, the DC system of the  (26)
converter bus is replaced by a constant power source, PiR = Pi − Pid − Pil
excluding the characteristics of the DC system on the Where, ΔPi and ΔQi are the changes in active and reactive
commutating bus. The voltage amplitude and phase angle power flowing out of the node on the AC line. ΔPid and
of the equivalent AC system i are Ei and  i . The ΔQid are the amount of active power and reactive power
injected into the node of the DC system. ΔPil and ΔQil are
impedance amplitude and angle of the equivalent
the variation of active and reactive power injected into the
impedance are and θi. The commutated bus bar i and j are
node of the parallel load. Correcting the diagonal elements
connected by a tie line having an impedance amplitude of
of JPU and JQU, we can get
Zij and an impedance angle of θij.
Pid Pil
E11 Pa1,Qa1 Pd1 ,−Qd1  (i ,i ) = J PU (i ,i ) −
J PU − (27)
U i U i
Z11 U11 Bc1 Qid Qil
 (i ,i ) = J QU (i ,i ) −
J QU − (28)
LCC-HVDC1 U i U i
Z12 12 Where Pid / U i and Pil / U i are the sensitivity of the
Z13 13
E22 Pd2 ,−Qd2 active power to the node voltage of the DC system and the
Pa2,Qa2
parallel load respectively; Qid / U i and Qil / U i are the
sensitivity of the reactive power of the DC system and the
Z22 Bc2
U22 parallel load to the node voltage.
Z23 23 LCC-HVDC2 For the commutation bus i that needs to calculate the
RVSF, since the DC system is equivalent to a constant
E33 Pa3,Qa3 Pd3 ,Qd3
power load, thus
 Pid
Z33 U 3  3  U = 0
 i
 (29)
 Qid = 0
VSC-HVDC

FIGURE 4. Simplified model of hybrid multi-infeed HVDC system.  U i

The load connected in parallel on the commutating bus is
For the AC system shown in Fig. 4, the incremental
only the parallel reactive power compensation device on the
power flow equation is
commutation bus of LCC-HVDC. On the commutating bus,
 P      J P J PU     the sensitivity of the active and reactive power of the load to
 Q   J  U  =  J   (24)
     Q J QU   U  the voltage is
Where, ΔP and ΔQ are the amount of active and reactive  Pil
power flowing out of the node through the AC line. Δδ and  U = 0
 i
ΔU are the variation of the voltage phase angle and  (30)
amplitude. ΔP, ΔQ, Δδ and ΔU are column vectors having  Qil = 2 B U
n elements, and J is a 2n×2n Jacobian matrix.  U i ci i

In addition to the DC equivalent of the converter busbar
being constant power, it is also necessary to be considered
that the power input from other DC systems to the AC
system when calculating the RVSF. In this case, the matrix
J needs to be corrected. Since the delivery power of the DC
system is independent of the phase angle of the AC system,
only JPU and JQU need to be corrected. After correcting the
Jacobian matrix, the following power flow equation is
obtained:
Where Bci is the admittance of the parallel reactive power
compensation device on the commutating bus i.
The active power and reactive power of the VSC-HVDC
is related to its control mode. When it is in the constant
active power and the constant reactive power control mode,
there is

VOLUME XX, 2017 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

 Pid includes DC systems such as Tianguang, Chusui, Puqiao,

 U = 0 Gaozhao, Xing'an and Niucong. All DC systems are in
 i
 (31) rated state, and the rectifier side control mode is constant
 Qid = 0 current control, while the inverter side control mode is
 U i constant extinction angle control. According to the method
If the VSC-HVDC is in constant power and constant AC of obtaining reference value in this paper, the parameters C
voltage control mode, there is and  N of each DC system can be obtained. The rated
 Pid operating point of the DC system can be calculated by the
 =0 method of section II-A. The  N of each DC system in the
 U i (32)
 U = 0 model both are 15º, and the values of the parameter C are
 i shown in Tab. I.
Assume that ΔPR =0 exists In the simulation, DC current order is increased by
 0     0.01pu, and then the value of dPd/dId is calculated, which is
 Q  = J M  U  (33) called simulated value. The DC system is equivalent to a
 R  
constant power load. Reducing the active power and
We simplify (32) and get the relationship between ΔQR reactive power of 0.01pu in the simulation program to
and ΔU as follows
obtain U / P and U / Q . Substituting U / P and
QR = J M' U (34)
U / Q into (16), dPd/dId can be obtained and defined as
'
Where, J M = J QU
'
− J Q J P−1 J PU
'
. Assume that the studied theoretical value. Obviously, if the analysis is correct, the
LCC-HVDC is fed into i node. Therefore, the voltage simulated value of dPd/dId should be equal to the theoretical
support strength of the i node is value.
The simulated and theoretical values of dPd/dId for each
RVSFm = J M' ( m,m) (35)
DC systems at the rated operating point and their
corresponding RVSF are shown in Fig. 5. As we can see,
Ⅴ. VERIFICATION OF THE EFFECTIVENESS OF RVSF the simulated and theoretical values of dPd/dId for each DC
INDICATOR systems are approximately equal. It can be seen from the
A. VERIFICATION OF THE RELATIONSHIP BETWEEN
simulation results that the relationship between RVSF and
VOLTAGE SENTIVITY AND POWER STABILITY OF THE power stability can be fully verified in the real-world grid
RECEVING SYSTEM system. Since the simulation example uses a very complex
The analysis in Section III of this paper is based on the real-world grid system with six DC systems, this conclusion
derivation of the relationship between the voltage can also be extended to other power systems.
sensitivity of the receiving system and the DC power in 0.90
Section II-C. The following is a verification of these
theories in practical system examples. 0.85 Xing'an
Tianguang Gaozhao
TABLE I 0.80
Niucong
Puqiao
dPd dId

VALUES OF C IN DIFFERENT HVDC SYSTEMS

0.75 Chusui
HVDC System C
0.70
Chusui 1.49
theoretical value
0.65 simulated value
Puqiao 1.49

Niucong 1.77 0.60

5 10 15

Gaozhao 1.91 RVSF

FIGURE 5. Comparison of dPd/dId from two methods under rated
Xing'an 1.91 operating conditions.

Tianguang 1.77 B. VERIFICATION OF RVSF THRESHOLD

The example uses the PSD-BPA model of China
Southern Power Grid. The model considers the effects of
automatic excitation device and governor, and contains a
wealth of power system components such as STATCOM
and TCSC. The load uses a ZIP static load model that
considers the frequency characteristics. The receiving
system in the model is the Guangdong power grid, which
The LCC-HVDC system of Fig. 3 was built on
PSCAD/EMTDC to verify the value of the CRVSF derived
in the previous section. We keep the AC system impedance
angle θ=75°, and adjust the AC system impedance Zeq and
the equivalent power supply voltage Em to make the system
run at the rated state of Pd =1 and Un =1. As shown in the
Fig. 6 below, when Zeq = 0.4619, the solution is calculated

VOLUME XX, 2017 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access

that Em = 0.9870. In this system, there is dPd/dId =0 when dPd/dId is calculated by applying a small amount of
Id=1.00. change of 0.01pu to Id1. The calculation of the SCR is
The DC system is equivalent to a constant power load of shown in (36), and the analytical formula of the MIESCR is
the same capacity, and then the system is disturbed by shown in (37).
cutting a 0.01pu reactive load. We calculate the ratio of ESCR = Z11−1 (36)
reactive power variation to voltage change on the bus to get
the theoretical value. Compared with the theoretical value MIESCR = ( Z11 Pd 1 + Z12 Pd 2 )−1 (37)
CRVSF=0.603, the simulated value of RVSF is 0.611 that Where, Z11, Z12, and Z22 are elements in the node impedance
the error is 1.33%. matrix of the AC system shown in Fig. 4, respectively. The
1.001
Pd3 is changed from -0.55 to 0.55, and the step size is 0.01.
1.000 Zeq=0.4500
Zeq=0.4619 The correspondence between RVSF, ESCR, MIESCR and
0.999 Zeq=0.4700 dPd/dId are shown in Fig. 7 (a), (b) and (c), respectively. As
0.998 we know from Fig. 7, when Pd3 changes from -0.55 to 0.55,
P d/(p.u.)

0.997 case 4 is a weak system, but does not become a very weak
0.996 system. The critical values of each index corresponding to
0.995
case1~3 are shown in Tab. Ⅳ. As shown in Fig. 7 and Tab.
Ⅳ, neither ESCR nor MIESCR has obvious correlation
0.994
with dPd/dId in the hybrid multi-feed system which means
0.993
0.94 0.96 0.98 1 1.02 1.04 1.06 that these two indicators lack a clear critical threshold
Id/(p.u.) corresponding to dPd/dId=0. However, it can be found that
FIGURE 6. Power curve under different system support strength. RVSF has a good correlation with dPd/dId=0 in the hybrid
multi-feed system, and when dPd/dId=0, there is always
Ⅵ. COMPARATIVE ANALYSIS OF STRENGTH CRVSF=0.60, indicating that the CRVSF can accurately
EVALUATION INDICATORS measure the critical value of the voltage support strength of
This section compares RVSF, ESCR, MIESCR and the receiving system. These prominent advantages of RVSF
dPd/dId=0. Tab. Ⅲ shows the four sets of AC impedance make it possible to effectively evaluate the receiving
parameters and the corresponding control modes of VSC- system strength of the hybrid multi-infeed system.
HVDC. We mainly consider two kinds of control modes: 0.4 0.4
constant active power/constant reactive power (referred to 0.2 0.2

dPd1 /dId1
dPd1 /dId1

as P/Q) and constant active power/constant alternating

0.0 0.0
voltage (referred to as P/V). And then, RVSF, ESCR,
MIESCR and dPd/dId are calculated separately. −0.2 −0.2
The following describes the calculation method of RVSF −0.4 −0.4
0.0 0.4 0.8 1.2 0.6 1.0 1.4 1.8 2.2
when Pd3 is different. The AC impedance parameters are
RVSF ESCR
known as shown in Tab. Ⅲ. The active power of the VSC- (a) (b)
HVDC at rated is Pd3N=0.5. Since VSC-HVDC exchanges
0.4
less reactive power with the commutation bus during rated
operation, it is assumed that the reactive power of the VSC- 0.2
dPd1 /dId1

HVDC at rated is zero, that is Qd3N=0. When Pd3 is changed, 0.0

the two LCC-HVDC systems are operated under the rated −0.2
state where Id1=Id2=1 is always kept unchanged. −0.4
Considering that VSC-HVDC has overload and power flow 0.2 0.6 1.0 1.4 1.8
reversal capability, Pd3 can range from -0.55 to 0.55. The MIESCR
(c)
control modes of VSC-HVDC given in Tab. Ⅲ. is known,
and then RVSF of the receiving system of LCC-HVDC1 FIGURE 7. Correspondence between RVSF, ESCR, MIESCR and dPd/dId.
can be calculated. TABLE Ⅳ
CRITICAL VALUES COMPARISON OF INDICATORS
TABLE Ⅲ
PARAMETERS OF FOUR CASES Case 1 Case 2 Case 3

Case Z1 Z2 Z3 Z12 Z13 Z23 Control mode CESCR 1.76 0.84 2.04

Case 1 0.9 1.1 1.0 0.4 0.5 0.25 P/Q CMIESCR 1.01 0.52 1.18

Case 2 1.5 1.1 1.3 0.8 1.4 1.0 P/V RVSF 0.60 0.60 0.60

Case 3 0.8 0.8 0.8 0.5 0.4 0.5 P/Q

Ⅶ. CONCLUSION
Case 4 0.8 0.8 0.6 0.5 0.5 0.5 P/Q

VOLUME XX, 2017 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2940409, IEEE Access
In this paper, a sensitivity-based strength index of
supporting voltage, the receiving system voltage support
strength factor (RVSF), is proposed. Then the relationship
between RVSF and power stability criterion, dPd/dId, is
derived. The RVSF calculated based on dPd/dId =0 is
defined as the critical voltage support strength factor of the
receiving system, that is called CRVSF. When LCC-HVDC
system connects to different receiving systems, the CRVSF
remains constant, which can satisfy the various operating
modes of the receiving system. The proposed criteria for
identifying the strength of receiving system are as follow.
Very weak system: RVSF ≤ 0.60
Weak system: 0.60<RVSF≤2
Strong system: RVSF>2
In the hybrid multi-infeed HVDC system, considering
the influence of VSC-HVDC transmission power and its
different control modes, the Jacobian matrix of the
receiving system is modified to obtain the analytical
expression of RVSF.
The simulated value of RVSF is obtained from the model
built in PSCAD/EMTDC. Compared with the theoretical
calculation value of RVSF, the error of the simulation value
is extremely small, and the validity of CRVSF=0.603 is
verified. A comparative analysis of RVSF with ESCR and
MIESCR shows that there is a strong correlation between
RVSF=0.603 and dPd/dId=0. In different receiving systems,
CRVSF remains unchanged, while ESCR and MIESCR do
not have clear thresholds.
Considering the sensitivity of voltage to reactive power,
RVSF is a new way to evaluate the voltage support strength
of the receiving system. Exploring other features of RVSF
and combining with the characteristics of the power system
to apply these research results, will be carried out in further
research work.
REFERENCES
[1] D. Huang, Y. Shu, J. Ruan and Y. Hu, “Ultra High Voltage
Transmission in China: Developments, Current Status and Future
Prospects,” Proceedings of the IEEE, vol. 97, no. 3, pp. 555-583,
March 2009.
[2] Z. Wei, J. Liu, W. Fang, Junxian Hou and Z. Xiang, “Commutation
Failure analysis in single- and multi-infeed HVDC systems,” 2016
IEEE PES Asia-Pacific Power and Energy Engineering Conference
(APPEEC), Xi'an, 2016, pp. 2244-2249.
[3] F. Zhang et al., “Generalized short circuit ratio for multi-infeed LCC-
HVDC systems,” 2017 IEEE Power & Energy Society General
Meeting, Chicago, IL, 2017, pp. 1-5.
[4] Lin Weifang, Tang Yong, Bu Guangquan and Shao Yao, “Voltage
stability analysis of multi-infeed AC/DC power system based on
multi-infeed short circuit ratio,” 2010 International Conference on
Power System Technology, Hangzhou, 2010, pp. 1-6.
[5] Y. Liu and Z. Chen, "A Flexible Power Control Method of VSC-
HVDC Link for the Enhancement of Effective Short-Circuit Ratio in
a Hybrid Multi-Infeed HVDC System," in IEEE Transactions on
Power Systems, vol. 28, no. 2, pp. 1568-1581, May 2013.
[6] Shilpa G and P. Manohar, "Hybrid HVDC system for multi-infeed
applications," 2013 International Conference on Emerging Trends in
Communication, Control, Signal Processing and Computing
Applications (C2SPCA), Bangalore, 2013, pp. 1-5.
VOLUME XX, 2017
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
[7] D. L. H. Aik and G. Andersson, “Power stability analysis of multi-
infeed HVDC systems,” IEEE Transactions on Power Delivery, vol.
13, no. 3, pp. 923-931, Jul 1998.
[8] H. Xiao, Y. Li, J. Zhu and X. Duan, “Efficient approach to quantify
commutation failure immunity levels in multi-infeed HVDC
systems,” IET Generation, Transmission & Distribution, vol. 10, no.
4, pp. 1032-1038, 3 10 2016.
[9] B. Franken and G. Andersson, “Analysis of HVDC converters
connected to weak AC systems,” IEEE Transactions on Power
Systems, vol. 5, no. 1, pp. 235-242, Feb. 1990.
[10] D. L. H. Aik and G. Andersson, “Voltage stability analysis of multi-
infeed HVDC systems,” IEEE Transactions on Power Delivery, vol.
12, no. 3, pp. 1309-1318, Jul 1997.
[11] Chen Xingwang and Guan-Lin, “Research on limitation of the multi-
infeed short circuit ratio,” 2016 IEEE PES Asia-Pacific Power and
Energy Engineering Conference (APPEEC), Xi'an, 2016, pp. 712-
715.
[12] Chen Xiuyu, Han Mingxiao, Liu Chongru, Impact of control modes
on voltage interaction between multi-infeed AC-DC system
[J] ． Automation of Electric Power Systems,2012,36(2):58-63(in
Chinese).
[13] Y. Xia, W. Wei, M. Yu, X. Wang and Y. Peng, "Power Management
for a Hybrid AC/DC Microgrid With Multiple Subgrids," in IEEE
Transactions on Power Electronics, vol. 33, no. 4, pp. 3520-3533,
April 2018.
[14] Y. Xia, M. Yu, X. Wang and W. Wei, "Describing Function Method
Based Power Oscillation Analysis of LCL-Filtered Single-Stage PV
Generators Connected to Weak Grid," in IEEE Transactions on
Power Electronics, vol. 34, no. 9, pp. 8724-8738, Sept. 2019.
[15] IEEE Guide for Planning DC Links Terminating at AC Locations
Having Low Short-Circuit Capacities, in IEEE Std 1204-1997, vol.,
no., pp.1-216, Jan. 21 1997.
[16] Xu Zheng. Characteristics of HVDC connected to weak AC systems part1:
HVDC transmission capability[J]. Power System Technology, 1997,
21(1): 12-16(in Chinese).
[17] D. Lee Hau Aik and G. Andersson, “Influence of load characteristics
on the power/voltage stability of HVDC systems. II. Stability margin
sensitivity,” IEEE Transactions on Power Delivery, vol. 13, no. 4, pp.
1445-1452, Oct 1998.
[18] Y. Zhang, Y. Su and L. Guan, “Evaluation voltage supporting strength of
receiving-end power grid of HVDC transmission system in environment
of AC/DC parallel power system and asynchronously interconnected
power grid,” Power System Technology, 2012, vol. 36, no. 1, pp. 90-95,
Jan 2012.
[19] “Systems with multiple DC infeed,” Technical Brochure no.364 by
CIGRÉ Working Group B4.41, Dec. 2008.
[20] D. H. A. Lee and G. Andersson, “An Equivalent Single-Infeed
Model of Multi-Infeed HVDC Systems for Voltage and Power
Stability Analysis,” IEEE Transactions on Power Delivery, vol. 31,
no. 1, pp. 303-312, Feb. 2016.
[21] X. Chen, A. M. Gole and M. Han, “Analysis of Mixed
Inverter/Rectifier Multi-Infeed HVDC Systems,” in IEEE
Transactions on Power Delivery, vol. 27, no. 3, pp. 1565-1573, July
2012.
[22] X. Ni, A. M. Gole, C. Zhao and C. Guo, "An Improved Measure of
AC System Strength for Performance Analysis of Multi-Infeed HVdc
Systems Including VSC and LCC Converters," in IEEE Transactions
on Power Delivery, vol. 33, no. 1, pp. 169-178, Feb. 2018.
[23] D. L. Hau Aik and G. Andersson, "Impact of renewable energy
sources on steady-state stability of weak AC/DC systems," in CSEE
Journal of Power and Energy Systems, vol. 3, no. 4, pp. 419-430,
Dec. 2017.
9