3.1.3. Turning the Load Off

Before the load is turned off the whole network operates in the steady state. Then, at the moment *t* = 1400 s, the load is turned off (Figure 6).

**Figure 6.** Load turning off process. Load voltage: waveform 1, and load current: waveform 2.

Just after the load turning off moment the load current stops immediately. However, because of the energy flow buffering the flow of source current has been extended for the next period *T*, i.e., for the time 1400 s–1420 s (Figures 7 and 8). As the result the whole circuit achieves zero steady state, i.e., with no load current/power, and UPQC is ready to the next compensation if load is turned on again.

**Figure 7.** Load turning off process. Load voltage: waveform 1, and source current: waveform 2.

**Figure 8.** Load turning off process. DC-link capacitor voltage: waveform 1, and load conductance signal: waveform 2.

### *3.2. Compensation for Source Voltage and Current Distortions*

Beside compensation for voltage deformation from higher harmonics the UPQC should maintain sinusoidal load voltage even if source voltage is influenced by irregular or unexpected distortions. Source voltage swells and sags, flicker, and pulse-type distortions can be enumerated in this context.

In Section 3.2. the load consists of two branches in parallel. The first one contains a thyristor power controller, consists of a 15 Ω resistor in series with two thyristors in antiparallel connection operating with phase angle equal to π/4. The second one consists of 15 Ω resistor in series with 32 mH inductor. This load branch introduces reactive power into the network.

#### 3.2.1. Compensation for Source Voltage Swell

The source voltage is composed initially of fundamental frequency component of rms 230 V and also of two harmonics: 32 V/250 Hz and 32 V/350 Hz. Then, beginning from *t* = 120 ms there is a swell of the source voltage fundamental component by 50%. The most important waveforms and electrical quantities are shown in Figures 9–11, and then in Tables 5 and 6.

The load operates in the steady state when at instant *t* = 80 ms UPQC's shunt and series converters are activated simultaneously. For the first period *T* of UPQC operation, i.e., for time period 80 ms–100 ms, the load is powered practically solely with the use of energy stored in the DC-link capacitor, i.e., without drawing energy from the supply source. This cause decrease of DC-link capacitor voltage. Then, the first non-zero magnitude of conductance signal (Equation (5)) can be obtained at the very end of this time period, i.e., at instant *t* = 100 ms, waveform 3 in Figure 10.

**Figure 9.** Source voltage swell compensation. Source voltage: waveform 1, and load voltage: waveform 2.

**Figure 10.** Source voltage swell. Load current, source current and conductance signal: waveforms 1, 2 and 3.

**Figure 11.** Source voltage swell. DC-link capacitor voltage and series converter current: waveforms 1 and 2.

**Table 5.** Basic parameters characterizing load work before and during source voltage swell. The parameter definitions are the same as for Table 3.


**Table 6.** Basic parameters characterizing UPQC-and-load subcircuit operation for the voltage swell. The parameter describing is the same as for Table 2.


At *t* = 120 ms the amplitude of source voltage fundamental component rises from 325 V up to 487 V, Figure 9. Since for the control method considered there is no time needed for analysis of source voltage spectrum this source voltage increase can be compensated immediately. The load voltage amplitude rises, but only to 349 V, Figure 9 and Table 5.

In order to compensate for this voltage increase a voltage correction is generated using energy stored in the DC-link capacitor. Depending on the voltage swell magnitude the current of the converter side of the series converter can reach significant values. This can result in significant increase of energy loss in this converter (Figure 11) and compare *PLoad* and *PUPQC*<sup>+</sup>*Load* in Tables 5 and 6.

### 3.2.2. Compensation for Source Voltage Sag

The load operates in the steady state when the UPQC is turning on at *t* = 80 ms. At *t* = 120 ms the amplitude of source voltage fundamental frequency component decreases to 163 V. Due to the UPQC action the load voltage can be maintained on the amplitude about 295 V, Table 7 and Figure 12.

**Table 7.** Basic parameters characterizing load work before and during source voltage sag. The parameter describing is the same as for Table 3.


**Figure 12.** Source voltage sag compensation. Source voltage: waveform 1 and load voltage: waveform 2.

The source voltage drop is compensated with the use of energy drown from the supply source. This is performed as follows. The deficiency of source energy, caused by voltage sag, is balanced using of energy stored in the DC-link capacitor. As the result its voltage decreases. It causes an increase in the conductance signal (Equation (5)) and an increase in amplitude of the source current reference (Equation (6)). Source current rises increasing source active power (Figure 13). This "additional" source power is utilized to increase load voltage to be near its nominal magnitude.

Similarly to the case of voltage swell the compensation for voltage sag may require large current of the series converter. This imply a high magnitude of variable component of the DC-link capacitor voltage. If the instantaneous capacitor voltage falls close to the source instantaneous voltage, then UPQC may lose the possibility of correct operation. Therefore it can be said that in the analyzed case UPQC operates at the limit of its capabilities. This is illustrated in Figure 14, waveform 1.

**Figure 13.** Source voltage sag compensation. Source current: waveform 1, load current: waveform 2, and conductance signal: waveform 3.

**Figure 14.** Source voltage sag compensation. DC-link capacitor voltage and load series converter current: waveform 1 and 2, respectively.

It should be also noticed a problem of energetic cost of compensation for the source voltage sag. During compensation high current of the series converter causes a large dissipation of energy in its power circuitry. Such power loss can be estimated on the basis of a comparison of the active power at the load terminals against the active power at the UPQC's input terminals: compare the parameter *Pload* in Table 7 with the parameter *PUPQC*<sup>+</sup>*Load* in Table 8.


**Table 8.** Basic parameters describing UPQC-and-load subcircuit action before and during the source voltage sag. The parameters shown were defined in Table 2.

#### 3.2.3. Compensation for Source Voltage Fluctuations

Compensation for source voltage fluctuations within the range of magnitude of 0.9 to 1.1 of its nominal value and at frequency of 8 Hz is considered in this section. Such voltage fluctuation may be classified as flicker. Flickers cause a number of adverse effects in electrical circuits operation, for example for electric motors action electronic devices operation or lighting installations.

For the discussed UPQC control method, the way and effect of reducing flicker-type of source voltage fluctuations is identical with the method of compensating for source voltage swells and sags. In addition, because of smaller disturbances in source voltage amplitude they are easier for compensation. However, the flicker-type voltage oscillations can be considered as repetitive run in a narrow range of low frequencies. Therefore, it seems to be convenient to present here a distinctive way of the UPQC action which can be useful for such periodical-like voltage or current disturbances. In particular, the option of choosing a convenient value for the *Tst* parameter, see Equations (1) and (5), is utilized here. This possibility has been used to increase the inertia in UPQC operation against changes in load active power. Voltage fluctuations cause changes in this power. In general, the greater the magnitude of the *Tst* parameter the closer the conductance signal (Equation (5)) run with respect to its multi-period mean. As a result the source current (and load voltage) can be stabilized, so that flicker is easier to be reduced in the whole grid.

It should be noted, that if increase the *Tst* time the DC-link capacitor operates with lowered voltage. For this reason the condition (Equation (7)) may not be met. This can reduce UPQC dynamics or even cause UPQC operation to failure.

The UPQC operation when the time *Tst* is set to be equal to the source voltage period *T,* and then when it has been increased to 3*T* is analyzed. The conductance signal (Equation (5)) and DC-link capacitor voltage are shown in the same scale in Figure 15, respectively. It can be observed that for the *Tst* parameter increased to 3*T* fluctuation of conductance signal is reduced: Waveform 3 in contrast with waveform 1. Simultaneously the UPQC operates with lowered DC-link capacitor voltage: waveform 4 with respect to waveform 2.

**Figure 15.** Source voltage flicker compensation. DC-link capacitor voltage for *Tst* = *T* and for *Tst* = 3*T*: waveforms 2 and 4, respectively, and conductance signal for *Tst* = *T* and for *Tst* = 3*T*: waveforms 1 and 3, respectively.

The effect of flicker (and still existing harmonics) compensation for *Tst* = 3*T* is shown in Figure 16. The compensation can be considered sufficient. Taking into account that the load voltage amplitude is maintained to be constant, it can be concluded that the flicker compensation is sufficient.

**Figure 16.** Source voltage flicker compensation. Source voltage: waveform 1, and harmonics-compensated and amplitude-levelled load voltage: waveform 2.

#### **4. Studies for UPQC Extended Operation**

The extended functionality of UPQC is understood here as the possibility of using it also to control the energy exchange between all—Being influenced by given UPQC device action—Elements of the network. These elements may be of passive or active type or may be changeable from this point of view. There is no restriction on location in the network of these elements. They may be covered by UPQC extended action as well being located on the AC-side as on the DC-side of UPQC device. Therefore, beside compensating for undesirable components of grid voltage and current runs the UPQC can also serve as a local energy distribution center that can operate with high power factor. There is no change in UPQC circuitry parameters and in source voltage waveform components with respect to those introduced in Section 3. However, in order to highlight UPQC's extended capabilities the load is composed to be nonlinear, time variable and of changeable passive or active kind.

#### *4.1. UPQC Operation with Switched Passive*/*Active Work of AC-Side Load Elements*

In general, there are two main possibilities of energy flow management when some nominally passive elements of load become generators and the amount of energy being generated in the load exceeds energy being consumed there:


The case (a) may be realized in a full form, when all amount of the "excessive" energy is transmitted up stream to the source, or in a partial form, when some portion of the "excessive" energy is accumulated in the DC-link capacitor. Waveforms related to the (a) strategy are shown in a general outlook in Figures 17–19, and then, in more precise look and with detailed comments, are presented in Figures 20–22. Then, waveforms related to the (b) strategy are shown in general outlook in Figures 23 and 24, and then are detailed in Figures 25–29.

**Figure 17.** Source voltage: waveform 1, load voltage: waveform 2 and load current: waveform 3. Whole network action, general view.

**Figure 18.** DC-link capacitor voltage: waveform 1, conductance signal: waveform 2 and source current: waveform 3 with Y scale of 54 mS/div. Whole network action, case 4.1 (a).

**Figure 19.** DC-link capacitor voltage: waveform 1, conductance signal: waveform 2 and source current: waveform 3 with Y scale of 10 A/div. Whole network action, case 4.1 (b).

**Figure 20.** Load voltage: waveform 1, load current: waveform 2 and source current: waveform 3.

**Figure 22.** Load current: waveform 1 and source current: waveform 2.

**Figure 24.** DC-link capacitor voltage: waveform 1, source current: 2 and conductance signal: 3 with Y scale of 27 mS/div.

**Figure 25.** Source and load voltages: waveform 1 and 2. Theirs RMS and THD parameters are 348 V and 14.7%, and then 268 V and 7.7%, respectively.

**Figure 26.** Source and load voltages: waveform 1 and 2. Theirs RMS and THD parameters are 162 V and 32.7%, and then 219 V and 4.1%, respectively.

**Figure 27.** Source and load voltages: waveform 1 and 2. Theirs RMS and THD parameters are 235 V and 21.4%, and then 231 V and 4.3%, respectively.

**Figure 28.** Source and load voltages: waveform 1 and 2. Theirs RMS and THD parameters are 235 V and 21.7%, and then 231 V and 4.7%, respectively.

**Figure 29.** Source and load voltages: waveform 1 and 2. Theirs RMS and THD parameters are 162 V and 32.6%, and then 219 V and 4.2%, respectively.

Figure 20 shows load voltage and current, and then source current during the first 100 ms of whole network action. This time period corresponds with the same time interval in Figures 17–19, and then in Figures 23 and 24.

The waveform 2 in Figure 20 is formed to represent total current of several loads, where some of them are nonlinear and time variable. The current is highly distorted, having also an inductive and DC components. Initially, in the time period *T* of 0 ms–20 ms the load active and apparent powers are 2.8 kW and 3.0 kVA, respectively, and 2.7 kW and 2.9 kVA during time period *T* of 20 ms–40 ms when UPQC starts its action.

Then, during the time interval 100 ms–300 ms, a new load-sided element was activated and therefore the load current run changes, see waveform 1 in Figure 21 (and waveform 3 in Figure 17). The load current is now composed to be almost unrealistically strongly distorted in order to show high and extended performance of UPQC. In particular a relatively large negative constant component of 17–28 A appears in the load current spectrum, so the fundamental frequency component is inverted relative to the fundamental component of the source voltage. Thus, the load, taken as a whole, works now as a source of energy. Because the in-load generated power exceeds both the power consumed in the load and dissipated in the UPQC the instantaneous DC-link capacitor voltage rises above its initial magnitude. As a result, the conductance signal changes its sign to negative, see waveform 3 in Figure 18. Consequently, the source current, still being purely active, begins to be controlled by the UPQC's shunt converter in order to carry some amount of the in-load generated power to the source, waveform 2 in Figure 21.

Estimation of energy balance in the network when it is practically in the steady state (here in the time period 200 ms–220 ms, when there is no change in the load power and in the static DC-link capacitor voltage, see Figures 17 and 18) is as follows: in-load generated power: 7.85 kW, in-load consumed power: 2.72 kV, power transmitted from the load to the source: 5.07 kW, power dissipated in UPQC's shunt converter: 43 W and power dissipated in UPQC's series converter: 226 W. From this energy balance results that the in-load generated power feeds passive elements of the compensated load and covers UPQC's energy loss. The remaining "excessive" amount of in-load generated power is transmitted to the source.

After switching off the in-load generating element the DC-link capacitor voltage diminishes below its initial magnitude. For this reason the conductance signal becomes positive, consequently source current polarity has been inverted and load draws energy from the source again. This is shown in Figures 18 and 22.

During the time period about 100 ms–350 ms the UPQC operates as a compensator as well as an energy distributor. Note, during this time load voltage and source current were maintained to be sinusoidal, irrespective of energy flow direction (Figures 17 and 18).

#### 4.1.1. Split Storing/Distributing of the In-Load Generated Energy

As it was already stated the energy generated in the active part of load may be decomposed into the portion consumed immediately and into the "excessive" portion. In turn this "excessive" portion may be split into a piece to be transmitted immediately to the grid and a piece to be stored in the DC-link capacitor. In other words, a power limit may be imposed on energy transmission to the grid (or, if needed, a voltage limit may be imposed on maximal DC-link capacitor voltage).

Such possibility of limited back-transmission is illustrated in Figure 19. In this example the maximal back-transmission power is bounded to 2.7 kW. Because the in-load generated power is greater than the allowed power limit of the back-transmission the DC-link capacitor voltage (energy) rises. It lasts till the moment of switching-off the generating element that is located in the load. Then the "excessive" energy portion, which were stored in the capacitor, is discharging partially to the source and partially to UPQC—Being dissipated in its power circuitry. After achieving the balance between load and source active powers, what can be seen in Figure 19 about *t* = 380 ms, the source takes over powering the load.

#### 4.1.2. Full Storing of the In-Load Generated Energy

Characteristic waveforms of currents and voltages for the case of the full energy-storing mode (without upstream energy transmission) are shown in general outlook in Figures 23 and 24, and then some critical time areas are zoomed in Figures 25–29.

The waveform 1 in Figure 24 demonstrates the process of accumulating/discharging the in-load generated "excessive" energy in the DC-link capacitor. The energy accumulation process takes place in time periods 180 ms–340 ms and then 540 ms–620 ms, whereas the energy discharging can be seen during time periods 340 ms–520 ms and 620 ms–940 ms. These time intervals fall within the wider 180 ms–940 ms range in which UPQC's shunt converter blocks any current flow to-or-from the supply source. In other words, during the time 180 ms–940 ms the load works in an energetically autonomous way, that means without any energy drawing from or giving to the supply source. Power fluctuations of all elements of the network are buffered by UPQC. Simultaneously all UPQC's conventional compensation tasks are fully fulfilled.

The most critical areas occurs about 200 ms–300 ms and then about 550 ms–600 ms. There is a cumulation of strong source voltage harmonics distortions with voltage swell and later with voltage sag. There are also large load current deformations, including high energy negative pulses that generates "asymmetrical" active power, occurring only during positive half-waves of source voltage.

The 240 ms–260 ms *T* period was chosen to show the effect of UPQC action (Figure 25).

For this period the RMS and THD parameters of source voltage run have been reduced at load terminals from the "swelled" magnitude of 348 V down to 268 V and from 14.7% to 7.7%, respectively. Unfortunately, it can be said that despite a significant improvement of load voltage parameters, the voltage quality requirements have not been met.

Within the second critical area the 580 ms–600 ms *T* period was chosen to show the effect of UPQC action (Figure 26). For this period the RMS and THD parameters of source voltage waveform have been improved at load terminals from the "sagged" magnitude of 162 V up to 219 V and from 32.7% down to 4.1%, respectively. It can be said that these parameters may be considered satisfactory.

The source and load voltages for non-critical areas are presented in next three figures. The 340 ms–360 ms, 440 ms–460 ms and 680 ms–700 ms *T* periods were chosen to show these waveforms in Figures 27–29, respectively. The load voltage RMS and THD parameters can be accepted as sufficient from the voltage quality point of view. In particular, there is no significant difference in the content of harmonics when working with or without compensation for source voltage sag.

4.1.3. A Side Effect of Conductance Signal Control Method: Alleviating and Catching Energy of Source Voltage Spike Distortion

From time to time an impulsive voltage transient (voltage spike) can appear in the grid, e.g., as a result of lighting stroke. Lighting arresters can be used to stop the transient. Fortunately, it follows from the principle of the considered control method that UPQC can alleviate, or even store and then utilize some amount of energy of such voltage distortion.

The network operates in the steady state when a voltage spike appears at *t* = 105 ms, Figure 30. Parameters characterizing this spike are 3.3 kV in magnitude, 1.2 μs rise time, 10 μs peak voltage duration and 200 μs fall time.

**Figure 30.** Source voltage with spike distortion: waveform 1 and DC-link capacitor voltage: waveform 2.

Energy of this source spike increases amount of energy, which flows through UPQC input terminals, from 89.9 J during the stead state *T* period 80 ms–100 ms, up to 141.6 J for the next (i.e., hit by the spike) 100 ms–120 ms *T* period. Some amount of energy of the spike impacts the load immediately, what can be seen as load voltage distortion shown in waveform 2 in Figure 31. As the result, the energy consumed by the load rises from 88.2 J for the 80 ms–100 ms time period up to 125.4 J for the next one. However, some portion of energy of the spike has been caught by UPQC. This energy portion increases energy stored in the DC-link capacitor: note the difference in capacitor voltage between instants *t* = 120 ms and *t* = 100 ms in waveform 2 in Figure 30.

**Figure 31.** Source voltage: waveform 1 and load voltage: waveform 2.

Because UPQC buffers energy variations between the supply source and the load the energy of the pulse distortion, which reaches load terminals, is lower than the distortion of energy that appears on UPQC input terminals. This is 125.4 J with respect to 141.6 J, respectively. This energy difference increases electric charge stored in the DC-link capacitor and, at the same time, increases its static voltage from 581 V at *t* = 100 ms up to 584 V at *t* = 120 ms, waveform 2 in Figure 30. This "additional" voltage (or energy) decreases the conductance signal (Equation (1)) from 82.3 mS for *T* period 100 ms–120 ms down to 65.7 mS for the next *T* period 120 ms–140 ms, waveform 1 in Figure 32. This fall of conductance signal (Equation (1)) causes decreasing in source current amplitude from 26.6 A down to 21.3 A, waveform 2 in Figure 32 during the *T* period 120 ms–140 ms. In other words, during this *T* period UPQC uses the from-the-distortion energy to power both itself and the load.

**Figure 32.** Conductance signal: waveform 1 and source current: waveform 2.

Just after the moment at which energy of the pulse distortion stored in the DC-link capacitor is discharged, the network returns to the steady state, waveform 2 in Figure 30 and both waveforms in Figure 32.

#### 4.1.4. UPQC Operation During Switched Passive/Active Work of DC-Side Load

In order to extend the UPQC usefulness a load as well as a source of energy may be connected to the UPQC's DC-link capacitor. In such a case UPQC can perform some extra tasks. In particular, depending on direction of energy flow, UPQC can act as a high power factor rectifier, which can power DC-side loads with the use of AC-side generated energy, or as an inverter, which can supply AC-side loads with the use of energy generated by DC-side sources. Therefore, UPQC may also serve as an energy bridge and buffer that can control energy flow between all UPQC's AC-side and DC-side loads and sources. Figures 33 and 34 illustrate such extended mode of UPQC operation.

**Figure 33.** DC-side load current: waveform 1, DC-link capacitor voltage: waveform 2 and conductance signal: waveform 3 (Y scale for this signal is 25 mS/div).

**Figure 34.** AC-side source voltage: waveform 1, AC-side load voltage: waveform 2, AC-side load current: waveform 3, AC-side source current: waveform 4 and conductance signal: waveform 5, where Y scale for the conductance signal is 57 mS/div.

In Figure 33 UPQC operates in the steady state when a DC-side current of 10 A magnitude begins to charge the DC-link capacitor at the moment *t* = 60 ms, waveform 1 in Figure 33. Due to the increase in DC-link capacitor voltage, the conductance signal changes both its magnitude and sign to be negative, waveform 3 in Figure 33 and waveform 5 in Figure 34.

As a result energy of this charging DC-side current is transferring to the AC-side network. This energy can power AC-side loads and, concurrently, its surplus may be transmitted upstream to the grid. This to-the-grid energy transferring process starts at *t* = 80 ms, waveform 4 in Figure 34.

Then, at time instant *t* = 120 ms, the DC-side current begins to discharge the DC-link capacitor. The conductance signal is recalculated into a new magnitude that depends on sum of AC-side and DC-side active powers. As a result the UPQC, still compensating for voltage and current disturbances, operates concurrently as a high power factor rectifier, i.e., a rectifier with sinusoidal input current, waveform 4 in Figure 34.

Figures 35–39 characterize the UPFC work for when voltage/current runs to be compensated are highly and variously distorted. An additional element, which can operate either as a load or a source of energy and which power can vary in time both in magnitude and in sign, is connected on the DC-link capacitor. Such a network can be considered as generating a kind of worst case of voltage and current runs to be improved. Parameters of these voltage/current runs are specified in the comment that follows Figure 36.

The distorted AC-side source voltage and then the compensated AC-side load voltage are shown in Figure 35. The AC source voltage is distorted by the same higher harmonics and swell/sag disturbances compared to that shown in Figure 23, but this time the whole system is also influenced by energy consumption/generation by the DC-side circuitry. The current run of the DC-side load/source element is shown in Figure 36. Positive polarity of this current means that this element generates energy and vice versa.

**Figure 35.** Global view on AC-side source voltage: waveform 1 and on AC-side load voltage: waveform 2.

**Figure 36.** Global view on AC-side load current: waveform 1 and on DC-side load current: waveform 2.

**Figure 37.** Global view on DC-link capacitor voltage: waveform 1, AC-side source current: waveform 2 and conductance signal: waveform 3 with Y scale of 56 mS/div.

**Figure 38.** Critical time period 100 ms–300 ms. AC-load voltage: waveform 1 and AC-source current: waveform 2.

**Figure 39.** Critical time period 560 ms–760 ms. AC-load voltage: waveform 1 and AC-source current: waveform 2.

DC-link capacitor voltage, AC-side source current and conductance signal (shown as the envelope of source current) are presented in Figure 37. These signals are essential for discussing the UPQC extended operation. In particular, energy balancing between all sources and loads of the network as well as UPQC's "internal energy effort" in order to compensate for nonactive voltage and current components can be identified when analyzing the waveform of DC-link capacitor voltage. This voltage run contains the in-*T*-period oscillations that relate to compensation for nonactive components of AC-side current and voltage. In the same time the DC-link capacitor voltage increases or decreases statically, i.e., from *Tn* period to *Tn*+<sup>1</sup> one, when UPQC balances active powers of all loads and energy sources of the network.

The most critical conditions for UPQC action appear around 200 ms–300 ms and 500 ms–700 ms. They may be treated as a base for a general assessment of UPQC performance.

	- source voltage RMS/THD of 348 V/14.5% have been transformed into load voltage 247 V/3.8%, respectively;
	- load current THD of 82.3% and its DC component of magnitude −28.2 A have been transformed into source current THD and DC component of 1.9% and 0.0 A, respectively.
	- source voltage RMS of 162 V and THD of 33.2% have been transformed into load voltage RMS and THD of 218 V and 4.6%, respectively;
	- load current THD and DC component of 79.5% and −28.5 A has been transformed into source current THD and DC component of 2.6% and 0.1 A, respectively.

It can be stated that in both critical areas of the UPQC operation the disturbed input voltage and current runs have been compensated satisfactory.

#### **5. Conclusions**

The paper presents the possibility of compensation for nonactive voltage/current components with the use of compensators that are controlled using a conductance signal. In general, the conductance signal results from the active power of the compensated load. This signal can be calculated based on two variables to be measured:


By using the conductance signal as the reference for the compensator action it is possible to omit the technically complicated methods of analysis of voltage/current waveforms by their decomposition into plurality of components. The same idea of avoiding the harmonic analysis is visible here if one compares the considered method to the p-q instantaneous power theory and its use to compensators control. Depending on planned purpose of compensation, such a solution may well be an advantage or a disadvantage.

The conductance signal is obtained based on observation of energy balance in the circuit consisting of the source, the UPQC compensator and the load. Any energy imbalance between the power required by the load and supplied by the source is buffered by the UPQC. In other words the network aims the steady state under control of the UPQC.

Using the conductance signal control method extends the functionality of UPQC. There is the possibility of controlling the flow of energy between all active and passive components of the network. This can help to increase the efficiency of the network.

The use of conductance signal in order to control the UPQC action enables bi-directional energy transmission with unity power factor both from the source to the load and in the opposite direction when the load can also generate energy. Both UPQC's AC- and DC-side generated or consumed energy may be handled and exchanged in such bi-directional way. This opens the possibility of using UPQC also as a local energy buffering-and-distribution center that can be useful for smart microgrids, increasing their energy efficiency.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
