Enhancement of the Flickermeter for Grid-Connected Wind Turbines

Abstract

Distributed generators connected to the power system usually produce voltage fluctuations. For wind turbines connected to a grid, large changes in wind speed can cause voltage flicker at the point of common coupling. The measurement of voltage flicker caused only by wind turbines is difficult. The wind turbine under test is usually connected to a medium voltage point, in which other fluctuating loads may produce significant voltage disturbances at the wind turbine terminal where the measurement is made. Although the IEC 61400-21-1 standard specifies a method to evaluate voltage flicker caused by wind turbines, because of the complex algorithm and process of the IEC standard, there is currently a lack of measurement equipment that meets the IEC standard. In addition, some countries that use other voltage flicker standards, such as ΔV₁₀, do not have suitable flicker measurements for wind turbines. Therefore, this study proposes an enhanced version of the IEC 61400-21-1 standard, which integrates the ΔV₁₀ method, so that the proposed measurement system complies with the IEC and ΔV₁₀ standards. In this study, the voltage flicker measurement system is successfully implemented, which can help engineers to predict the voltage flicker by wind turbines and assess whether a region or grid is suitable for installing wind turbines. Therefore, it can provide wind turbine companies with a quick assessment of voltage flicker to comply with the certification process.

1. Introduction

Wind energy is a safe, clean, and endless renewable energy source. Because wind energy has no fuel and air pollution problems, in the face of the current climate abnormalities and global warming, wind energy technology has attracted more and more attention from the world [1,2]. Although wind power is a sustainable energy, when the wind power generator is merged with or disconnected from the existing electric power system, it will cause voltage fluctuations and short-circuit current in the grid, thereby affecting power quality. Furthermore, the inherently unstable wind speed can also cause the wind turbines to produce voltage fluctuations, called voltage flicker [3,4,5,6]. Voltage flicker will not only cause unstable lighting of lamps, but also cause health problems and electronic equipment failure [7]. Therefore, measuring voltage flicker becomes an important issue of power quality. The flickermeter standard IEC 61000-4-15 was developed by the International Electrotechnical Commission and has been widely used worldwide for decades [7]. It can evaluate the flicker severity level of power supply voltage as the values of P_st (short-term flicker severity) and P_lt (long-term flicker severity). However, the IEC 61000-4-15 flickermeter is not suitable for directly measuring the voltage flicker caused by wind turbines, because the tested wind turbine is usually connected to a medium voltage point, and other fluctuating loads may generate obvious voltage perturbations in the grid during measurement. Thus, the IEC announced the standard of IEC 61400-21-1 in 2019 [8] for evaluating the voltage flicker caused by grid-connected wind turbines, which is partially based on the IEC 61000-4-15 standard.

Wind turbine companies need a certification process for selling wind turbines. Gutierrez et al. [9] presented a flicker measurement system for wind turbines’ certification based on the IEC standard. The system consisted of two modules: a signal register to store the voltage and current time-series data and a process algorithm in accordance with the IEC standard. The IEC standard method is based on measuring the flicker of a single wind turbine and then using this measurement result to calculate the flicker emission from a number of wind turbines. Therefore, Barahona et al. [10] investigated the validity and accuracy of the standard method in IEC 61400-21, i.e., the early version of IEC 61400-21-1, for evaluating the flicker emission from multiple wind turbines. Dexiong Li [11] revealed that the fictitious grid circuit provided in the IEC standard should not connect a current source with an inductor in series. Thus, two new fictitious grid circuits were proposed to improve the deficiency. Khan et al. [12] worked on the deviation on flicker evaluation of wind turbines associating with the disturbance of harmonics and interharmonics. However, because of the complexity of the algorithm and process of the IEC 61400-21-1 standard, there is currently a lack of measurement equipment that meets the standard.

Another flicker measurement approach called “Equivalent 10-Hz Voltage Flicker (ΔV₁₀)”, developed by the Central Research Institute of the Electric Power Industry (CRIEPI) [13,14] in 1978, is widely used in Japan, Brazil, France, Italy, Argentina, and Taiwan. According to the specification of the approach, the voltage flicker severity is measured as the term of ΔV₁₀ value, and the maximum value of the evaluation index ΔV₁₀ should not exceed 0.45% as the basis for regulation. However, the CRIEPI does not provided a method to evaluate the voltage flicker caused by wind turbines. Although the works in Gherasim [15], Redondo [16,17], and Manas [18] implemented the flickermeter based on the IEC standard for wind turbines, the ΔV₁₀ standard has not been considered. Thus, in this work, we added the ΔV₁₀ method into the IEC 61400-21-1 standard to propose an enhanced version of the two-mode IEC voltage flicker measurement for wind turbines, which is suitable for countries using the IEC standard or the ΔV₁₀ standard.

To implement this proposed method, MATLAB Simulink is used to build a model of a doubly fed induction generator (DFIG) with grid connection to generate current and voltage time-series data at the wind turbine terminals to simulate the voltage flicker on the virtual grid without other voltage fluctuation loads. A measurement system based on the IEC 61400-21-1 combined with ΔV₁₀ standard was implemented to assess the voltage flicker caused by wind turbines under continuous operation. It can help wind turbine companies with a quick certification of power quality for wind turbine installations. The proposed system can evaluate whether a region or grid is suitable for installing wind turbines. It can also predict the voltage flicker that wind turbines may cause in advance.

2. Methodology

The measurement of voltage flicker caused by wind turbines is a complex issue. As mentioned above, the point of common coupling (PCC) where the measurement of wind turbines is taken also connects to other fluctuating loads producing voltage flicker. Thus, neither the IEC 61000-4-15 nor the ΔV₁₀ flickermeter connected to the PCC can directly measure the voltage flicker caused by wind turbines. In addition, the severity of voltage flicker depends on the characteristic of wind turbine itself, the changing wind speed, the grid impedance phase angle, and the short-circuit apparent power of the grid [3]. Currently, only the standard of IEC 61400-21-1 provides an assessment of the voltage flicker of grid-connected wind turbines. Thus, we added the ΔV₁₀ method to the algorithm of IEC 61400-21-1 to evaluate the voltage flicker in wind turbines in terms of P_st and ΔV₁₀. This section represents the proposed method based on the IEC standard combined with the ΔV₁₀ method.

As shown in Figure 1, a DFIG wind turbine model is built by MATLAB Simulink to simulate a continuously operating wind turbine. Then, the measurement data of wind speed, generated current, and voltage can be obtained in time series form. According to the standard, the wind speed is sampled at least at 1 S/s (one sample per second). Because of the Nyquist rate, the current i_m(t) and voltage u_m(t) of the wind turbine terminal are sampled at 2 kS/s, owing to the low pass filter with an 800 Hz cut-off frequency in data acquisition. The time-series data of i_m(t) and u_m(t) are then analyzed according to the algorithm specified in the IEC 61400-21-1 and 61000-4-15 standards. At the end of the analysis, the voltage flicker severity values of P_st and ΔV₁₀ can be determined. Figure 2 shows a three-phase grid-connected DFIG generic wind turbine model with current and voltage data acquisition designed in Simulink. The parameters shown in Figure 2 and their values are listed in

Table 1. Parameters of the DFIG wind turbine and their values.

Parameters	Values
Wind speed sampling rate	1 S/s
Sampling rate of wind turbine terminal voltage um and current im	2 kS/s
Wind speed range	3–15 m/s
DFIG rated active power	2 MW
Distribution system	1.4 kV

. In this work, the DFIG wind turbine is 2 MW and connected to an 11.4 kV distribution system. The wind speed is usually set from 3 m/s to 15 m/s, which are the lowest wind speed at which the wind turbine can be started and the highest wind speed at which the wind turbine can operate safely, respectively.

There is a three-phase grounding transformer in the 11.4 kV feeder to protect the distribution system. The DFIG wind turbine includes a wound rotor induction generator and an AC/DC/AC IGBT-based PWM converter. The stator winding is directly connected to the grid, but the rotor is fed at variable frequency through the PWM converter. The optimal wind turbine speed producing maximum power energy is proportional to the wind speed. For lower wind speeds, the rotor is rotating at subsynchronous speed, while for high wind speeds, it is rotating at hypersynchronous speed.

The proposed flickermeter modified from the IEC 61400-21-1 standard is illustrated in Figure 3, in which each block is described in the following paragraphs. Besides IEC 61000-4-15 standard to calculate the preliminary flicker level, ΔV₁₀ measurement is also adopted in the proposed method to fit more countries, as shown in the thick-line box in Figure 3.

2.1. Step 1: Data Acquisition of i_m(t) and u_m(t)

The standard requires groups of at least five time-series data of voltage and current in 10 min span per wind speed bin for three phases. The requirement of a 10 min data span is based on the IEC 61000-4-15 to calculate P_st for each 10 min voltage span.

Table 2. Number of current and voltage test data to be captured.

	3 ϕ	Wind Speed Bin (m/s)	Number of 10 Min Data
current im(t) voltage um(t)	R	3–<4	at least 5 sets
		4–<5	at least 5 sets
		….	….
		14–<15	at least 5 sets
	S	3–<4	at least 5 sets
		4–<5	at least 5 sets
		….	….
		14–<15	at least 5 sets
	T	3–<4	at least 5 sets
		4–<5	at least 5 sets
		….	….
		14–<15	at least 5 sets

shows the required groups of test data.

2.2. Step 2: Fictitious Grid

The voltage measured in the terminal of wind turbines is often affected by other fluctuating loads in the grid. Therefore, the standard proposes an equivalent circuit of a fictitious grid, as shown in Figure 4, to transfer the measured data u_m(t) to u_fic(t) so that the flicker from other fluctuating loads is excluded, and the transferred data are then suitable for entering the IEC 61000-4-15 algorithm, block 2.

In Figure 4, an ideal phase-to-neutral voltage source with the instantaneous value u_o(t) is represented in the fictitious grid, and a grid impedance is given by a resistor R_fic in series with an inductor L_fic. The wind turbine is represented by the current generator i_m(t), which is the measured current. This model gives a simulated value u_fic(t) of instantaneous phase-to-neutral voltage according to Equations (1) and (2), which is implemented in block 1 of Figure 3.

$${{\displaystyle u}}_{fic}(t)={{\displaystyle u}}_0\left(t\right)+{{\displaystyle R}}_{fic}\times {{\displaystyle i}}_m\left(t\right)+{{\displaystyle L}}_{fic}\frac{{{\displaystyle di}}_m\left(t\right)}{dt}$$

(1)

$${\mathit{u}}_0\left(\mathit{t}\right)=\sqrt{\frac{2}{3}}{\mathit{U}}_{\mathit{n}}\sin \left({\alpha }_{\mathit{m}}\left(\mathit{t}\right)\right)$$

(2)

where U_n represents the rms value of the nominal voltage and α_m(t) is the fundamental phase angle of u_m(t). The values of R_fic and L_fic are derived from the grid impedance phase angle ψ_k = 30°, 50°, 70°, and 85°, which are determined by the IEC 61400-21-1. R_fic and L_fic shall be selected to obtain the appropriate grid impedance phase angle ψ_k according to the Equation (3).

$$\tan ({{\displaystyle {\psi }_k}}_{})=\frac{2\pi \times f_g\times L_{fic}}{{{\displaystyle R_{fic}}}_{}}=\frac{{{\displaystyle X}}_{fic}}{{{\displaystyle R}}_{fic}}$$

(3)

where f_g is the nominal frequency (50 or 60 Hz). Once the series data of u_fic(t) are deter- mined, they can be sent to block 2 in Figure 3 for flicker calculation in accordance with the IEC 61000-4-15 standard.

2.3. Step 3: IEC 61000-4-15 Flickermeter and ΔV₁₀ Calculation

The IEC standard of flickermeter provides a model to simulate the behavior of the lamp–eye–brain response when voltage slight variation causes lamp flickering [7]. The model attempts to assess the annoying level subject to the flickering of voltage and lighting. The flickermeter outputs the short-term flicker severity P_st to characterize the subjective phenomenon. The IEC specification recommends to take P_st = 1 as the upper-limit of regulation, called unit flicker. For P_st < 0.7, the flicker usually is insensible; for P_st > 1.3, the flicker will make people feel uncomfortable [11].

Entering the time-series data u_fic(t) into the input of block 2 will obtain the values of P_st,fic or ΔV_10,fic for each phase and for different grid impedance angle ψ_k. The implementation of block 2, i.e., a general flickermeter, has been completed in previous works [19,20], so the details of this general flickermeter are not described again here.

On the other hand, for the countries using the ΔV₁₀ standard, the block 2 in Figure 3 can be switched to ΔV₁₀ mode. In ΔV₁₀ mode, all of the magnitudes (ΔV_n) of the voltage fluctuation for flicker frequencies could be determined by the fast Fourier transform (FFT). The next stage is to look up the equivalent effect (or sensitivity coefficient) of all frequency components with respect to 10 Hz, as shown in Figure 5. Then, using these coefficients to weight the corresponding frequency components, one can obtain the value of ΔV₁₀ expressed as Equation (4).

$$\Delta V_{10}=\frac{\sqrt{{\displaystyle \sum _{n=1}{(a_n\cdot \Delta V_n)}^2}}}{V}$$

(4)

where ΔV₁₀ indicates the severity level of flicker; V is the rated main voltage; and ΔV_n and a_n represent the fluctuation amplitude of specified frequency and its sensitivity coefficient, respectively. Normally, the value of ΔV₁₀ is regulated under 0.45% [13].

2.4. Step 4: Normalization

As shown in block 3 of Figure 3, the step is to determine the flicker coefficient c(ψ_k) of a wind turbine as a function of the grid impedance phase angle ψ_k and wind speed distribution. Each P_st,fic is normalized to a flicker coefficient c(ψ_k) by Equation (5).

$$c\left({\psi }_{\mathit{k}}\right){=\mathit{P}}_{\mathit{st},\mathit{fic}}\frac{{\mathit{S}}_{\mathit{k},\mathit{fic}}}{{\mathit{S}}_{\mathit{n}}}$$

(5)

where S_n is the rated apparent power of a wind turbine and S_k,fic represents the short-circuit apparent power of the fictitious grid, which can be determined from Equation (6).

$${\mathit{S}}_{\mathit{k},\mathit{fic}}=\frac{{\mathit{U}}_{\mathit{n}}{}^2}{\sqrt{{\mathit{R}}_{\mathit{fic}}{}^2{+(2\pi \mathit{f}}_{\mathit{g}}{\mathit{L}}_{\mathit{fic}}{)}^2}}$$

(6)

2.5. Step 5: Weighting

To scale the measured frequency of occurrence of the flicker coefficients to correspond with the assumed wind speed distribution, a weighting factor shall be determined for each wind speed bin (block 4). The procedure to find the weighting factor is as follows. The assumed frequency of occurrence f_y,i of wind speeds within the i’th wind speed bin shall correspond to a Rayleigh distribution [3,21]:

$${\mathit{f}}_{\mathit{y},\mathit{i}}=\mathit{exp}\left[-\frac{\pi }{4}{\left(\frac{{\mathit{v}}_{\mathit{i}}-0.5}{{\mathit{v}}_{\mathit{a}}}\right)}^2\right]-\mathit{exp}\left[-\frac{\pi }{4}{\left(\frac{{\mathit{v}}_{\mathit{i}}+0.5}{{\mathit{v}}_{\mathit{a}}}\right)}^2\right]$$

(7)

where v_i is the midpoint of the i’th wind speed bin and v_a is the annual average wind speed. The actual occurrence frequency f_m,i of measured flicker coefficients within the i’th wind speed bin is obtained by Equation (8).

$${\mathit{f}}_{\mathit{m},\mathit{i}}=\frac{{\mathit{N}}_{\mathit{m},\mathit{i}}}{{\mathit{N}}_{\mathit{m}}}$$

(8)

where N_m,i is the number of measured flicker coefficient values within the i’th wind speed bin and N_m is the total number of flicker coefficient values. Then, the weighting factor w_i for each wind speed bin is determined in the Equation (9).

$${\mathit{w}}_{\mathit{i}}=\frac{{\mathit{f}}_{\mathit{y},\mathit{i}}}{{\mathit{f}}_{\mathit{m},\mathit{i}}}$$

(9)

The weighted accumulated distribution P_r(c < x) of the measured flicker coefficient values is found by Equation (10).

$${\mathit{P}}_{\mathit{r}}\left(\mathit{c}<\mathit{x}\right)=\frac{{{\displaystyle \sum }}_{\mathit{i}=1}^{{\mathit{N}}_{\mathit{bin}}}{\mathit{w}}_{\mathit{i}}{\mathit{N}}_{\mathit{m},\mathit{i},\mathit{c}<\mathit{x}}}{{{\displaystyle \sum }}_{\mathit{i}=1}^{{\mathit{N}}_{\mathit{bin}}}{\mathit{w}}_{\mathit{i}}{\mathit{N}}_{\mathit{m},\mathit{i}}}$$

(10)

where N_{m,i,c < x} is the number of flicker coefficient values less than or equal to the value x within the i’th wind speed bin and N_bin is the total number of wind speed bins. Finally, the flicker coefficient c(ψ_k,v_a) is determined as the 99th percentile of the weighted accumulated distribution of the flicker coefficient values, which can be obtained by calculating P_r(c < x) and reading the 99th percentile from that. The determined flicker coefficient c(ψ_k,v_a) shall be reported and recorded in block 5.

2.6. Step 6: Calculation of P_st and ΔV₁₀

The 99th percentile flicker emission from a single wind turbine under continuous operation is calculated by the following:

$${\mathit{P}}_{\mathit{st}}{=\mathit{c}\mathrm{(}\psi }_{\mathit{k}}{,\mathit{v}}_{\mathit{a}}\mathrm{)}\frac{{\mathit{S}}_{\mathit{n}}}{{\mathit{S}}_{\mathit{k}}}$$

(11)

where c(ψ_k,v_a) is the flicker coefficient of the wind turbine for the given grid impedance phase angle ψ_k and for the given annual average wind speed v_a; S_n is the rated apparent power of the wind turbine; and S_k is the short-circuit apparent power at the PCC. The flicker coefficient c(ψ_k,v_a) for other values of ψ_k and v_a can be produced by applying linear interpolation. For multiple wind turbines, the flicker emission of them can be assessed from Equation (12).

$${\mathit{P}}_{\mathit{st}\Sigma }=\frac{1}{{\mathit{S}}_{\mathit{k}}}\sqrt{{\displaystyle \sum }_{\mathit{i}=1}^{{\mathit{N}}_{\mathit{wt}}}{{(\mathit{c}}_{\mathit{i}}\left({\psi }_{\mathit{k}}{,\mathit{v}}_{\mathit{a}}\right){\text{×}\mathit{s}}_{\mathit{n},\mathit{i}})}^2}$$

(12)

where c_i(ψ_k,v_a) is the flicker coefficient of individual wind turbine; N_wt is the number of wind turbines connected to the PCC; and S_n,i is the rated apparent power of the individual wind turbine.

In the other mode, the voltage flicker emission for the ΔV₁₀ standard can also be determined through the same algorithm from Equations (4) to (12). The symbols used in the text and their meanings are listed in

Table 3. Symbols used in the text and their meanings.

Symbols	Meaning
Pst	Short-time flicker severity
Pst,fic	Pst in fictitious grid
ΔV10	Equivalent 10 Hz voltage flicker severity
im(t)	Captured current
um(t)	Captured voltage
ufic(t)	Transferred voltage in fictitious grid
Rfic	Equivalent resistor in fictitious grid
Lfic	Equivalent inductor in fictitious grid
Un	rms value of the nominal voltage
αm(t)	Fundamental phase angle of um(t)
ψk	Grid impedance phase angle
fg	Nominal frequency
c(ψk)	Flicker coefficient
Sn	Rated apparent power of a wind turbine
Sk,fic	Short-circuit apparent power of the fictitious grid
va	Annual average wind speed
wi	Weighting factor for the ith wind speed bin
Pr	Weighted accumulated distribution
Sn,i	Rated apparent power of the ith wind turbine

.

3. Experimental Results

Series data of collected current i_m(t) of a grid-connected wind turbine are shown in Figure 6, in which the generated current is not stable owing to wind speed changing. The current is sampled at 2 kHz for 10 min. The data of i_m(t) are then entered to block 1 of Figure 3 to obtain the series data of u_fic(t). In order to validate the proposed meas-urement system, the built DFIG wind turbine model is utilized to generate 180 sets of i_m(t) data in 10 min span per wind speed bin for three phases according to Table 2.

The wind turbine parameters and virtual grid parameters are shown in

Table 4. Parameters of The Wind Turbine and Fictitious Grid.

Rfic	4.5 Ω	Sn	2 MW
Lfic	0.007 H	Sk	90 MW
Sk,fic	100 MW	Un	22.8 kV
Ψk	30°, 50°, 70°, 85o	va	6, 7.5, 8.5, 10 m/s

.

After inputting these 180 sets of i_m(t) data, the flicker coefficient output c(ψ_k) of block 3 in Figure 3 with ψ_k = 30° is determined and shown in Figure 7. The flicker coefficient can be regarded as an indicator for preliminary assessment of the severity of flicker. It can be seen from the figure that the flicker coefficient is roughly proportional to the wind speed. In addition, owing to the inherent characteristics of wind speed, the higher the wind speed, the greater the deviation of the flicker coefficient.

The weighted flicker coefficient c(ψ_k,v_a) for different wind speeds and grid impedance phase angle, i.e., the output of step 5 mentioned above, is listed in

Table 5. Weighted flicker coefficients for different wind speeds and grid impedance phase angle in the P_st model.

ψk (°)	30	50	70	85
Va (m/s)	C (ψk,va) in the Pst model
6	10.47	9.26	7.77	6.04
7.5	11.59	10.25	8.60	6.69
8.5	11.66	10.31	8.65	6.73
10	11.93	10.55	8.85	6.89

and

Table 6. Weighted flicker coefficients for different wind speeds and grid impedance phase angle in the ΔV₁₀ model.

ψk (°)	30	50	70	85
Va (m/s)	C (ψk,va) in the ΔV10 model
6	4.14	3.11	1.63	0.42
7.5	4.33	3.21	1.71	0.44
8.5	4.50	3.34	1.78	0.45
10	4.63	3.44	1.83	0.47

for the P_st model and ΔV₁₀ model, respectively. The final outputs P_st and ΔV₁₀ are determined by Equation (11) and shown in Figure 8 and Figure 9, respectively.

It is seen that the wind turbine operating under the grid impedance phase angle ψ_k = 30°–85° has acceptable P_st and ΔV₁₀ values. The acceptable P_st value is P_st < 1 and the acceptable ΔV₁₀ value depends on the country, as listed in

Table 7. The acceptable ΔV₁₀ values.

	Voltage Flicker Limits ΔV10 (%)
France, Italy	Max. < 0.3
Japan	Max. < 0.45, Avg. < 0.32
Taiwan	Max. < 0.45
Brazil, Argentina, Iran	Max. < 0.5

[13]. In addition, the figures show that, the higher the wind speed, the greater the value of P_st and ΔV₁₀, because a high wind speed will cause the wind speed and power generation to be more unstable. Furthermore, the higher the grid impedance phase angle, the lower the voltage flicker caused by the wind turbine. This is because the fictitious network with wind turbines shown in Figure 4 is actually an RL low-pass filter with a cut-off frequency of R/L, so the high grid impedance phase angle will produce a lower cut-off frequency to suppress the voltage variation.

If the grid impedance phase angle and annual average wind speed are not given in Table 4 and Table 5, the IEC standard claims that linear interpolation can be used to obtain the weighted flicker coefficients and determine the P_st and ΔV₁₀ values. For example, assuming that the grid impedance phase angle ψ_k = 35° and the wind speed v_a = 6 m/s, linear interpolation can be used to determine that the c(ψ_k,v_a) of P_st mode and ΔV₁₀ mode are 10.17 and 3.88, respectively. Then, P_st = 0.23 and ΔV₁₀ = 0.09 are determined in the end. This section fully demonstrates the entire calculation process of P_st and ΔV₁₀ with average wind speed and grid impedance phase angle as parameters. The pre-evaluated P_st and ΔV₁₀ values can help engineers assess the feasibility of the combination of wind turbine specifications, grid impedance, and average wind speed at the location of the wind turbine. For countries using ΔV₁₀ standard, this is a reliable solution for evaluating the voltage flicker caused by wind turbines.

4. Conclusions

Renewable energy is receiving more and more attention, and the power quality problems caused by renewable energy are also an issue worth investigating. This study presents an enhanced flickermeter to measure the voltage flicker caused by grid-connected wind turbines in accordance with the standard IEC 61400-21-1 (P_st) and the standard CRIEPI (ΔV₁₀), so that countries using the ΔV₁₀ standard can also evaluate the flicker emission by wind turbines based on the ΔV₁₀ standard. In this work, given the wind turbine specification, grid apparent power and grid impedance, and wind speed data, the voltage flicker by wind turbines can be pre-determined in terms of P_st and ΔV₁₀. Therefore, the proposed system can help engineers in the countries using ΔV₁₀ to adopt the IEC standard during the transition period. During the measurement, the proposed system can distinguish the flicker coefficient c(ψ_k,v_a) with different grid impedance phase angles and wind speeds. It can be seen from the experimental results that a higher wind speed will lead to the worst flicker coefficient, and a lower grid impedance phase angle will result in a greater severity of flicker.

In addition, the greater the wind speed, the greater the deviation of the flicker coefficient. The deviation is caused by the harmonics and inter-harmonics existing in the fictitious grid, so the measurement accuracy might be slightly reduced. The solution to this problem is to increase the filter operation in the flickermeter input. The proposed measurement system is successfully implemented using MATLAB, and can output P_st and ΔV₁₀ values according to the engineer’s needs. It can provide wind turbine companies with a quick pre-evaluation of voltage flicker to comply with the certification process before installing the wind turbines. Moreover, the pre-assessment results can be used to confirm the specifications of installing a voltage stabilizer for wind turbines, or to assess the appropriate location of the grid connected to the wind turbine for less voltage flicker caused. Another renewable energy, photovoltaic panels, is also an unstable power source that might cause voltage flicker. The voltage flicker produced by the mixing of photovoltaic panels and wind turbines is worthy of discussion in future work.