To interconnect the synchronous machine model of Eq. 4-66 with the rest of the system, the dq0 voltages and currents must be converted to the corresponding phase abc quantities using Park’s transformation. Assume that the machine voltage [vabc(t)] vector consists of a single harmonic component of order h:
The corresponding dq0-voltage vector is obtained by Park’s transformation (Eq. 4-58) as
Therefore, one harmonic abc voltage introduces three harmonics in terms of dq0 coordinates. The machine’s dq0 model can now be used to obtain the dq0 currents. From Eq. 4-66:
Transferring Eq. 4-69 back into the abc coordinates (using the inverse Park transformation), and with the condition [D]T [D0] = 0, we get
Based on this equation, the harmonic voltage of order h, [Vabc(h)], can generate harmonic currents of order (h – 2) and (h + 2) in abc coordinates (Fig. 4.28). This is known as frequency conversion. Note that (Fig. 4.28)
• Negative-sequence voltage [D][Vabc(h)] generates harmonic currents of order (h + 2) and h.
• Positive-sequence voltage [D]C [Vabc(h)] generates harmonic currents of order (h – 2) and h.
• Zero-sequence voltage [Do][Vabc(h)] generates only the harmonic current of the same order.
• There are three particular harmonic voltages that can generate harmonic currents of the same order.
All these observations are consistent with the revolving magnetic field theory [68].
The current source [Inl(h)] of the synchronous machine harmonic model (Fig. 4.27) consists of two components: [If(h)] due to the frequency-conversion phenomena and [Is(h)] due to the saturation effects. These currents will be computed separately.
As mentioned before, Eq. 4-70 also indicates that there are three particular harmonic voltages that can generate harmonic currents of the same order. Assuming that the voltage vector [vabc(t)] includes all harmonic components, the harmonic terminal voltages that result in currents of the same harmonic order h can be grouped and described in phasor form as [12]
If we substitute
then Eq. 4-71 can be rewritten as
This equation defines the machine model shown in Fig. 4.27, where
• [Vabc(h)] = [Vk(h)] – [Vm(h)].
• [Y(h)] correlates harmonic voltages to harmonic currents of the same order.
• [If(h)] describes the harmonic coupling due to frequency-conversion effects. This current vector is generated from the conversion of voltages of different harmonic orders.
Following a similar approach, the equivalent DC voltage source of the dq0-machine model at harmonic order h = 0 becomes a set of positive-sequence fundamental frequency voltage sources in abc coordinates. As explained earlier, the values of these voltages are determined in conjunction with load-flow constraints.
Therefore, matrix operations are sufficient to compute the admittance matrix [Y(h)] and the equivalent current source [If(h)] needed for the harmonic machine model:
• Compute [Y(h)] according to Eq. 4-72 whenever the machine admittance matrix needs to be added to the network admittance matrix in the frequency scan process; and
• Compute [If(h)] according to Eq. 4-73 where [Vabc(h)] contains the machine voltages obtained from the load-flow solution of the previous iteration.
Saturation of the stator and rotor cores of a synchronous machine has a significant impact on the machine’s operation and the corresponding operating point [68]. Other saturation factors (e.g., cross-coupling between d- and q-axes) are assumed to be negligible and are not considered in the model of Fig. 4.27.
It is shown in [12] that the influence of saturation can be represented as a current source in parallel with the dq0-machine model (Eq. 4-66):
where [Idq0–s(h)] is a known current vector representing the effect of saturation. For a given set of machine terminal voltages, vector [Idq0–s(h)] is obtained through a process of subiterations [12].
With saturation effects included in dq0 coordinates, Park’s transformation can now be applied to obtain the machine’s model in abc coordinates. The processing of the first part of Eq. 4-75 has been explained in the previous section. The second part (e.g., [Idq0–s(h)]) will be discussed next.
Using Park transformation matrix (Eq. 4-61), the dq0-harmonic currents of order h appear in abc coordinates as
It can be seen that the dq0 currents [Idq0–s(h)] also introduce three separate harmonic components of order (h – 1), h, and (h + 1) in the abc coordinates. Assuming that [Idq0–s(h)] includes harmonics of all orders, the particular dq0 harmonics that produce harmonic currents of the same order in abc coordinates can be grouped together in phasor form as
This is the current that is needed to represent the effects of saturation in abc coordinates.
The above equation defines the final machine model used for harmonic load-flow studies. The approach of computing [Is(h)] can be summarized as follows:
• Compute the saturation current [Idq0–s(h)] using the subiteration process described in reference [12] or any other desired approach; and
• Compute [Is(h)] according to Eq. 4-77 for all harmonics of interest.
The complete machine model with both the frequency-conversion and saturation effects included can now be described by
The negative-sequence impedance of a synchronous machine is important for simulation under unbalanced conditions. Its value is different depending on whether negative-sequence currents or negative-sequence voltages are used to determine the impedance [68]. This inconsistency is caused by the frequency conversion process. In this example the harmonic model of Fig. 4.27 (implemented in a multiphase harmonic load flow [65]) is used to run two test conditions that are commonly accepted to define the machine’s negative-sequence impedance:
• I-definition: synchronous machine is connected to a negative-sequence current source and the negative-sequence terminal voltage is recorded; and
• V-definition: synchronous machine is connected to a negative-sequence voltage source and the negative-sequence terminal current is recorded.
The impedance resulting from these tests is the ratio of the fundamental frequency voltage to the fundamental frequency current obtained by harmonic analysis. The program was run for three different levels of harmonic truncation (Table E4.12.1).
Table E4.12.1
Values of Negative-Sequence Impedance for Synchronous Machine [12] where all impedance values are in ohms
Test type | Exacta | Harmonic truncationb | ||
h = 1 | h = 1, 3 | h = 1, 3, 5 | ||
I-definition | 0.0655 + j0.3067 | 0.0857 + j0.2165 | 0.0655 + j0.3066 | 0.0655 + j0.3066 |
V-definition | 0.0858 + j0.2165 | 0.0856 + j0.2166 | 0.0859 + j0.2164 | 0.0859 + j0.2164 |
a Based on analytical formulas.
b h = 1: only the fundamental frequency component included; h = 1, 3: h = 1 modeling plus the third harmonic; h = 1, 3, 5: h = 1, 3 modeling plus the fifth harmonic.
The following observations can be derived from this table:
• The difference between the machine impedances calculated from the I- and V-definitions is caused by the 3rd harmonic. The contribution of the 5th harmonic is negligible;
• If harmonics are not included, the proposed machine model corresponds to the impedance of V-definition;
• In the general case of unbalanced machine operation, the machine’s negative-sequence impedance is neither the one from the V-definition nor the one from the I-definition. The correct machine response can only be fully evaluated with harmonics included. The combination of the load-flow constraints with harmonic solutions is therefore the best approach in such situations; and
• Agreement between simulation and exact results justifies the validity of the synchronous machine harmonic model (Fig. 4.27).
A profile of the harmonic content of the synchronous machine currents and voltages is calculated (using the harmonic model of Fig. 4.27) for two test machines. The voltage and current harmonics in a machine will be influenced by the conditions imposed by the external network. To simplify the interpretation of the results, a straightforward operating condition with grounded Y synchronous machines and unbalanced R, L, and C loads is considered (Fig. E4.13.1 without harmonic-current sources). The loads are adjusted so that the machine’s current imbalance at fundamental frequency is around the normal tolerance of 10%. The results (using two synchronous machines with different positive- and negative-sequence currents) are shown in Table E4.13.1 for the voltage harmonics and for the frequency-conversion-induced equivalent current source If(h). Significant harmonics are observed even for cases when the machines are operated within an acceptable range of imbalance. The effects of these harmonics could easily be magnified if voltage resonance takes place. Note that the harmonic magnitudes decrease quickly for higher orders. From the results of Tables E4.12.1 and E4.13.1, it may be concluded that only 3rd and 5th harmonics need to be included in synchronous machine’s representation for harmonic studies.
Table E4.13.1
Harmonic Profiles of Two Synchronous Machines (Unbalanced Operating Conditions) [12] where all Magnitudes are in (pu) and all Angles are in (°)
h = 1 | h = 3 | h = 5 | h = 7 | |||||
Machine | Item | Magnitude | Angle | Magnitude | Angle | Magnitude | Angle | Magnitude |
No. 1 | Va | 1.103 | 1.1 | 0.0148 | 88.3 | 0.0002 | 148.4 | 0 |
NPa = 9.35% | Vb | 1.025 | –121.5 | 0.0131 | –36.5 | 0.0002 | 23.8 | 0 |
Vc | 1.022 | 120.2 | 0.0131 | –151.5 | 0.0002 | –90.2 | 0 | |
If-a | 0.027 | 167.3 | 0.0225 | –152.5 | 0.0002 | –99.5 | 0 | |
If-b | 0.027 | –72.7 | 0.0225 | 87.4 | 0.0002 | 140.4 | 0 | |
If-c | 0.027 | 47.3 | 0.0225 | –32.4 | 0.0002 | 20.4 | 0 | |
No. 2 | Va | 1.080 | 1.3 | 0.0345 | 92.1 | 0.0011 | 157.2 | 0 |
NPa = 12.41% | Vb | 1.041 | –121.6 | 0.0313 | –33.0 | 0.0009 | 32.4 | 0 |
Vc | 1.030 | 120.3 | 0.0315 | –148.2 | 0.0010 | –82.7 | 0 | |
If-a | 0.025 | 155.2 | 0.0704 | –151.3 | 0.0016 | –94.8 | 0 | |
If-b | 0.025 | –84.7 | 0.0696 | 88.3 | 0.0016 | 145.2 | 0 | |
If-c | 0.025 | 35.2 | 0.0696 | –30.9 | 0.0016 | 25.2 | 0 |
a NP: ratio of the negative- to positive-sequence current at fundamental frequency.
It is believed that the flow of the third harmonic in lines can be prevented by connecting the three-phase equipment in delta. This assumption is tested here by assuming that the winding of the first synchronous generator (Application Example 4.13) is to be connected in delta. The results of this test (third harmonic currents for the Δ and Y-connected machines) are presented in Table E4.14.1
Table E4.14.1
Third Harmonic Currents in Grounded Y- and Δ-Connected Synchronous Generators [12] where all Magnitudes are in (pu) and all Angles are in (°)
Phase A | Phase B | Phase C | |||||
Magnitude | Angle | Magnitude | Angle | Magnitude | Angle | ||
V3 | Y–g | 0.0148 | 88.3 | 0.0130 | –36.5 | 0.0131 | –151.5 |
Δ | 0.0091 | 81.7 | 0.0071 | –40.7 | 0.0078 | –168.4 | |
I3 | Y–g | 0.0090 | 48.3 | 0.0125 | –71.7 | 0.0108 | 170.1 |
Δ | 0.0055 | 41.6 | 0.0068 | –76.0 | 0.0065 | 153.3 |
. Even though the third harmonic components are about 40% smaller in the Δ connection, they do not cancel completely. The reason is
The third harmonic generated by the frequency-conversion process in a synchronous machine is of negative sequence and not of zero sequence, and therefore cannot be eliminated by a Δ connection.
In the model of Fig. 4.27, the saturation curve is approximated in a piecewise-linear manner between the data (MMF, λ) = (0.9 pu, 0.9 pu), (1.5 pu, 1.2pu), (2.7 pu, 1.8 pu). Table E4.15.1
Table E4.15.1
Harmonic Profiles of Two Synchronous Machines with and without Saturation (under Unbalanced Operating Conditions) [12] where all Magnitudes are in (pu) and all Angles are in (°)
h = 1 | h = 3 | h = 5 | ||||||
Magnitude | Phase | Magnitude | Phase | Magnitude | Phase | Difference (%)a | ||
NSb | Va | 1.103 | 1.1 | 0.0148 | 88.3 | 0.00020 | 148.4 | |
SAT | Va | 1.103 | 1.1 | 0.0120 | 76.3 | 0.00014 | 115.8 | 19.0 |
NS | Vb | 1.025 | –121.5 | 0.0130 | –36.5 | 0.00017 | 23.8 | |
SAT | Vb | 1.025 | –121.5 | 0.0106 | –48.4 | 0.00012 | –8.8 | 18.6 |
NS | Vc | 1.022 | 120.2 | 0.0131 | –151.5 | 0.00017 | –90.2 | |
SAT | Vc | 1.022 | 120.2 | 0.0106 | –163.7 | 0.00013 | –122.9 | 18.6 |
NS | Ia | 0.0027 | 167.3 | 0.0225 | –152.5 | 0.00022 | –99.5 | |
SAT | Ia | 0.0022 | 168.5 | 0.0182 | –164.5 | 0.00014 | –131.3 | 19.0 |
NS | Ib | 0.0027 | –72.7 | 0.0224 | 87.4 | 0.00022 | 140.4 | |
SAT | Ib | 0.0022 | –71.4 | 0.0182 | 75.5 | 0.00014 | 108.6 | 18.6 |
NS | Ic | 0.0027 | 47.3 | 0.0225 | –32.4 | 0.00022 | 20.4 | |
SAT | Ic | 0.0022 | 48.5 | 0.0182 | –44.6 | 0.00014 | –11.3 | 18.7 |
a Relative difference of the 3rd harmonic magnitude.
b NS: without saturation; SAT: with saturation.
lists the harmonic profiles of the machine with and without saturation. There are large differences between the cases with and without saturation (the last column of Table E4.15.1). The differences in the phase angles are even more noticeable. These results suggest that
Saturation can have an important influence on harmonic power-flow analysis.
The effects of nonlinearities of synchronous machines are further investigated by including nonlinear loads, represented as 3rd and 5th current injections to the generator-load system (Fig. E4.13.1). Three case studies are performed:
• Case 1: with no machine nonlinearities,
• Case 2: with only frequency conversion, and
• Case 3: with both frequency conversion and saturation.
The results (Table E4.16.1) further confirm the need for detailed modeling of machine’s nonlinearities.
Table E4.16.1
Voltages of Machine 1 (Defined in Table E4.13.1) in the Presence of Nonlinear Loads (Fig. E4.13.1) [12] where all Magnitudes are in (pu) and all Angles are in (°)
Phase A | Phase B | Phase C | |||||
Magnitude | Phase | Magnitude | Phase | Magnitude | Phase | Case study | |
h = 1 | 1.099 | 0.2 | 1.023 | –120.8 | 1.028 | 120.5 | 1 |
1.100 | 0.3 | 1.024 | –120.9 | 1.026 | 120.5 | 2 | |
1.000 | 0.3 | 1.024 | –120.8 | 1.026 | 120.5 | 3 | |
h = 3 | 0.0607 | 69.4 | 0.0304 | –171.5 | 0.0536 | –68.4 | 1 |
0.0777 | 67.3 | 0.0283 | –155.3 | 0.0572 | –82.9 | 2 | |
0.724 | 66.4 | 0.0295 | –156.7 | 0.0526 | –79.7 | 3 | |
h = 5 | 0.0336 | 69.7 | 0.0450 | –59.2 | 0.0379 | –173.5 | 1 |
0.0445 | 80.1 | 0.0539 | –50.0 | 0.0458 | –173.6 | 2 | |
0.0417 | 76.8 | 0.0517 | –52.9 | 0.0439 | –177.4 | 3 |
Most synchronous machines have relatively large radial air-gap lengths, which range from a few millimeters to tens of centimeters. This is so because the synchronous reactances (Xd, Xq) must be small enough to guarantee a good dynamic (transient, subtransient) behavior. Although permanent-magnet machines have a mechanical air gap of less than one millimeter up to a few millimeters, their electrical air gap is always larger than one millimeter because the radial air-gap height (length) of the permanent magnet behaves magnetically like air with the permeability of free space μo because the recoil permeability μR of excellent permanent-magnetic material (e.g., NdFeB) is about μR ≈ 1.05μo. Therefore, an imperfect mechanical mounting – resulting in a static-rotor eccentricity – or a bent shaft – resulting in dynamic-rotor eccentricity – will not significantly alter the magnetic coupling of the stator windings and the rotating magnetic flux will be about circular. Here one should mention that permanent-magnet machines cannot be easily assembled for ratings larger than a few tens of kilowatts because a mounting gear will be required, which maintains a uniform concentric air gap during the mounting of the rotor.
The modified winding-function approach (MWFA) accounting for all space harmonics has been used for the calculations of machine inductances under rotor eccentricities [69]. Machines can fail due to air-gap eccentricity caused by mechanical problems including the following:
• inaccurate positioning of the rotor with respect to the stator,
• worn bearings,
• stator-core movement, and
• bent rotor shaft.
Electrical asymmetries contribute to harmonic generation such as
• Damper windings (amortisseur) are incomplete, that is, the amortisseur has a different construction between poles (q-axis) as compared with that on the pole d-axis; or
• Eddy currents in the pole faces of salient-pole machines contribute to subtransient response;
• Heating of the rotor due to nonuniform amortisseur and nonuniform eddy-current generation around the rotor circumference may cause shaft bending/deflection; and
• Incomplete shielding of the field winding of a machine fitted with nonuniform amortisseurs and subjected to harmonic disturbance.
Nonsinusoidal air-gap flux wave effects – due to any of the above-listed reasons – can be accounted for by representing the mutual inductances between the field and stator windings, and those between the stator windings, by trigonometric series as a function of the rotor angle θ. It is found that the inclusion of nonsinusoidal inductance variation causes a broad spectrum such that injection of a single-harmonic current results in voltages containing all odd harmonic orders including the fundamental. The extent to which the energy is spread across the spectrum depends on the pole shape, the numbers of stator and rotor slots, and the amortisseur design. The spectral spread is a second-order effect compared to the generation of the associated voltage at frequency (h ± 2). Harmonic behavior is also affected by nonsinusoidal inductance variation with rotor position and by the extent to which induced damper winding currents shield the field winding from the harmonic-gap flux waves. Rotor eccentricity in machines can be of two forms:
• static air gap eccentricity, and
• dynamic air gap eccentricity.
With the static air-gap eccentricity, the position of the minimum radial air-gap length is fixed in space and the center of rotation and that of the rotor are the same. Static eccentricity can be caused by oval stator cores or by the incorrect relative mounting of stator and rotor. In case of dynamic air-gap eccentricity, the center of rotation and that of the rotor are not the same and the minimum air gap rotates with the rotor. Therefore, dynamic eccentricity is both time and space dependent. Dynamic eccentricity can be caused by misalignment of bearings, mechanical resonance at critical speeds, a bent rotor shaft, and wear of bearings [69].
MATLAB/Simulink [70] can be used for the simulation of machine variables. All machine inductances employed for the simulation are expressed in their Fourier series based on the MWFA. Two different cases, namely, a noneccentric and an eccentric rotor case, for this analysis are investigated. In the first case, it is assumed that dynamic air-gap eccentricity is not present, and in the second one 50% dynamic air-gap eccentricity is introduced to investigate the effect of the eccentric rotor on stator-current signatures. To analyze the stator-current signatures of these two cases, FFT (fast Fourier transform) of the current signals are performed for all cases. The stator currents show the existence of the 5th (300 Hz), 7th (420 Hz), 11th (660 Hz), 13th (780 Hz), 17th (1020 Hz), and 19th (1140 Hz) harmonics even when the rotor is not eccentric. Implementing 50% eccentricity will cause the stator and rotor current induced harmonics to increase when compared to those generated without rotor eccentricity as is listed in Table 4.3.
Table 4.3
Relative Percentage of Stator Harmonics Due to 50% Dynamic Eccentricity
5th | 7th | 11th | 13th | 17th | 19th |
22.8% | 12.4% | 20.9% | 28.4% | 47.1% | 36.9% |
The stator current harmonics exist because the interaction of the magnetic fields caused by both the stator and rotor windings will produce harmonic fluxes that move relative to the stator and they induce corresponding current harmonics in the stationary stator windings. These harmonic fluxes in the air gap will increase as the rotor dynamic eccentricity increases, and consequently the current harmonic content increases too. The 3rd harmonic and its multiples are assumed not to exist in the stator windings because it is a three-phase system with no neutral connection. However, third-harmonic components are induced in the rotor-field winding. Because the stator and rotor-current signatures of synchronous machines have changed, either or both signatures can be utilized for detecting dynamic eccentricity. However, it is more practical to implement the stator-current signature analysis for the condition monitoring than the rotor-current signature analysis because some synchronous machines have brushless excitation [49], which will make the rotor current inaccessible. In summary, one can conclude that synchronous machines are not very prone to the effects caused by rotor eccentricity due to their large air gaps.
The generation of shaft flux [71] due to mechanical and magnetic asymmetries as well as solid-state switching is an important issue for large synchronous generators. Bearing currents are predominantly experienced in PWM AC drives [72–76] due to the high-frequency switching ripples resulting in bearing electroerosion. Techniques for measurement of parameters related to inverter-induced bearing currents are presented in [77].
The most important conclusions of this section are:
• Synchronous machines generate harmonics due to
2. saturation, and
3. unbalanced operation.
• Current harmonics due to nonlinear loads should be limited based on IEEE-519 [78].
• Unbalanced load-flow analysis without taking into account harmonics is not correct due to the ambiguous value of the negative-sequence impedance of synchronous machines. Correct results can only be obtained by including harmonics in a three-phase, load-flow approach.
• Only 3rd and 5th harmonics need to be included for the machine’s representation in harmonic analysis. The 3rd harmonic is of the negative-sequence and not of the zero-sequence type. Hence, it cannot be eliminated by Δ transformer connections. The same is true for transformer applications with DC current bias [79].
• Machine saturation can have noticeable effects on the harmonic distribution. These effects are more significant with respect to the harmonic phase angles.
• Rotor eccentricity has a minor influence on harmonic generation due to the large air gap of synchronous machines.
• Shaft fluxes and bearing currents can have detrimental impacts.
In the past and at present synchronous generators represent an integral part of a power system. Although the tendency exists to move from a central power station approach to a distributed generation (DG) system with renewable energy sources in the megawatt range, there will be always the requirement of frequency control, which is best performed by a central power station that can absorb load-flow fluctuations, as the so-called frequency leader. Renewable plants will be mostly operated at their maximum power output and therefore cannot provide additional (real or reactive) power if the load increases beyond their assigned output powers. In addition, such plants have an intermittent output, which cannot be controlled by dispatch and control centers. This is to say that also in future power systems the power quality will depend on the reliable operation of synchronous generators driven by (natural-) gas turbines.
This chapter starts out with the introduction of the synchronous machine model based on dq0 coordinates. To visualize steady-state saturated magnetic fields, numerical solutions are presented for no-load, full-load, and short-circuit conditions. Fields for permanent-magnet machines and switched-reluctance machines are presented as well. Thereafter, the two possible reference systems (e.g., consumer and generator) are addressed. The application examples relate to the calculation of synchronous reactances, the investigation of various fault conditions including reclosing, and the calculation of the amortisseur current for subtransient faults such as line-to-line, line-to-neutral, and balanced three-phase short circuits. The design of synchronous machines and permanent-magnet machines for wind-power and hybrid drive applications, respectively, is discussed. Design guidelines for synchronous machines are presented, and the calculation of magnetic forces – based on the Maxwell stress – is included. The performance of synchronous machines under the influence of harmonics is explained based on models used in application examples. The employment of such models for harmonic power flow analyses is mandatory. Finally, approaches for analyzing static and dynamic eccentricities, shaft fluxes, and bearing currents are outlined.
A nonsalient-pole synchronous machine has the following data: Xs = (Xd = Xq) = 1.8 pu, Ra = 0.01 pu, Iphase = 1.0 pu and VL-N = 1.0 pu.
a) Sketch the equivalent circuit using the consumer reference system.
b) Draw the phasor diagram (to scale) based on the consumer reference system for a lagging (underexcited) displacement power factor of cos φ = 0.90.
c) Sketch the equivalent circuit based on the generator reference system.
d) Draw the phasor diagram (to scale) employing the generator reference system for a lagging (overexcited) displacement power factor of cos φ = 0.90.
A salient-pole synchronous machine has the following data: Xd = 1.8 pu, Xq = 1.5 pu, Ra = 0.01 pu, Iphase = 1.0 pu, VL-N = 1.0 pu.
a) Sketch the equivalent circuit based on the consumer reference system.
b) Draw the phasor diagram (to scale) using the consumer reference system for a lagging (underexcited) displacement power factor of cosφ = 0.70.
c) Sketch the equivalent circuit based on the generator reference system.
d) Draw the phasor diagram (to scale) employing the generator reference system for a lagging (overexcited) displacement power factor of cosφ = 0.70.
Synchronous machines with two by 30° displaced stator windings (winding #1 and winding #2) are used to improve the voltage/current wave shapes so that the output voltages of a generator become more sinusoidal. This is similar to the employment of 12-pulse rectifiers versus 6-pulse rectifiers. To accomplish this, stator winding #1 feeds a Y–Y connected transformer and stator winding #2 supplies a Δ–Y connected transformer as illustrated in Fig. P4.3. For a S = 1200 MVA, VL-L = 24 kV, f = 60 Hz, p = 2 turbogenerator the machine parameters are in per unit (pu): Xd = 1.05, X12d = 0.80, Xq = 0.95, X12q = 0.73, Xffd = 2.10, ko = 0.5, k1 = 0.638, r = Ra = 0.00146, rf = Rf = 0.0007.
Draw the phasor diagram (to scale) for rated operation and a displacement power factor of cosφ = 0.8 lagging (overexcited, generator reference system).
A three-phase synchronous generator is initially unloaded and has a rated excitation so that Ef = 1.0 pu. At time t = 0, a three-phase short-circuit is applied. The machine parameters are direct (d) axis: Xd = 1.2 pu, Xf = 1.1 pu, Xm = 1.0 pu, Ra = 0.005 pu, Rf = 0.0011 pu; quadrature (q) axis: Xq = 0.8 pu, Xm = 0.6 pu, Ra = 0.005 pu.
a) Plot the first-half cycle of the torque T3phase_short-circuit due to the three-phase short-circuit neglecting damping and assuming that there is no amortisseur.
b) At which angle φmax (measured in degrees) occurs the maximum torque Tmax (measured in per unit)?
a) For the machine parameters of Problem 4.4 calculate and plot the line-to-line torque TL-L for the first half-cycle when phases b and c are short-circuited, and determine at which angle φmax (measured in degrees) the maximum torque (measured in per unit) occurs.
b) Plot the induced open-circuit voltage of phase a, that is, ea (t) (measured in per unit) as a function of the phase angle φ (measured in degrees) for the first half-cycle.
a) For the machine parameters of Problem 4.4 calculate the total (fundamental and harmonics) line-to-neutral short-circuit current ia for the first half-cycle when phase a and the neutral are short-circuited, and determine at which angle φmax (measured in degrees) the maximum current iL-N_max (measured in per unit) occurs.
b) Plot the fundamental short-circuit current of phase a ia1 (measured in per unit) for the first half-cycle as a function of the phase angle φ (measured in degrees).
a) For the machine parameters of Problem 4.4 calculate and plot for the first half-cycle the out-of-phase synchronizing torque.
b) Determine at which angle φmax (measured in degrees) the maximum torque (measured in per unit) occurs.
For the machine parameters of Problem 4.4 calculate and plot for the first half-cycle the steady-state torque T0, the synchronizing torque TS, and the damping torque TD. The per-unit transient time constant of the d-axis is Td′ = 242 pu and the per unit voltages Vpu = V0 = Vl-n = Vl-n0 = 1.0 pu may be used.
The starting of synchronous generators without amortisseur (damper) winding by induction-motor action is not advisable, as will be demonstrated in this problem. Calculate for a synchronous machine with the parameters Xd = 1.2 pu, Xq = 0.8 pu, Xf = 1.1 pu, Xm = 1.0 pu, Ra = 0.005 pu, and Rf = 0.0011 pu the starting and accelerating torque Tstarting as a function of the slip s. To do this, start out from the dq0-coordinate system and transform it to the fbo [81] coordinate system. Table P4.9.1 presents some of the results. It is the task of this problem to fill in the missing values of this table.
Table P4.9.1
Starting and Accelerating Torque of a Synchronous Machine without Amortisseur Developed by Induction-Motor Action as a Function of Slip s (pu)
s (pu) | 1.0 | 0.9 | 0.8 | 0.7 | 0.6 | 0.5 | 0.4 | 0.3 | 0.2 | 0.1 | 0 |
Tstarting (pu) | 0.0172 | 0.0685 |
As can be seen from the few entries of this table, a synchronous machine cannot develop any significant induction motor torque without amortisseur winding. Most large synchronous generators have a very rudimentary amortisseur and must not be started by induction-motor action. Small synchronous machines with ratings of less than 1 MW can be equipped with a full amortisseur winding and can be started by induction-motor action.
For a hybrid automobile drive a permanent-magnet motor – as a part of a brushless DC machine – is to be analyzed with the following data:
An encoder is used to sense the rotor speed/position. For your design it will be helpful to rely on the flux patterns and machine cross section of Fig. P4.10a,b. It is recommended to use as permanent-magnet material NdFeB.
b) What is the rated synchronous speed ns_rat?
c) Compute the (maximum) height (radial one-sided length) of the permanent magnets hmag_max.
d) Calculate the maximum flux density of this machine.
e) Determine the leakage inductance Lℓ (from stator phase a to stator phase b) and the winding resistance Ra of one stator phase.
f) Determine the current density Jstator, provided the stator winding copper fill factor is kcu = 0.7, and the average stator-slot width is about the same as the average stator-tooth width.
Find the Fourier coefficients of the current wave shapes of Fig. P4.11 that represent the input MOSFET current (upper wave shape) and the current of a brushless DC motor (lower wave shape). Select a certain sampling interval, e.g., Δt = 0.025 ms, perform a Fourier analysis, and show that for a minimum approximation error the Nyquist criterion is satisfied.
A block diagram of two interconnected areas of a power system (e.g., area #1 and area #2) is shown in Fig. P4.12. The two areas are connected by a single transmission line. The power flow over the transmission line will appear as a positive load to one area and an equal but negative load to the other, or vice versa, depending on the direction of power flow.
a) For steady-state operation show that with Δω1 = Δω2 = Δω the change in the angular velocity (which is proportional to the frequency f) is
and
b) Determine values for Δω (Eq. P12-1), ΔPtie (Eq. P12-2), and the new frequency fnew, where the nominal (rated) frequency is frated = 60 Hz, for these parameters: R1 = 0.05 pu, R2 = 0.1 pu, D1 = 0.8 pu, D2 = 1.0 pu, ΔPL1 = 0.2 pu, ΔPL2 = –0.3 pu.
c) For a base apparent power Sbase = 1000 MVA compute the power flow across the transmission line.
d) How much is the load increase in area #1 (ΔPmech1) and area #2 (ΔPmech2) due to the two load steps?
e) How would you change R1 and R2 in case R1 is either a wind or solar power plant operating at its maximum power point (and cannot accept any significant load increase due to the two load steps) and R2 is a coal fired plant that can be overloaded, and serve as a frequency leader?
Figure P4.13 illustrates the block diagram of governor, prime mover (steam turbine), and rotating mass and load of a turbogenerator set. For the frequency change (Δω) per generator output power change (ΔP),
the load change (ΔPL) per frequency change
angular momentum of steam turbine and generator set M = 4.5,
base apparent power Sbase = 500 MVA,
governor time constant TG = 0.01 s,
valve charging time constant TCH = 1.0 s,
load reference set point = 1.0.
a) Derive for Fig. P4.13 Δωsteady state by applying the final value theorem. You may assume the load reference set point(s) = 0 and where s is the Laplace operator.
b) List the ordinary differential equations and the algebraic equations of the block diagram of Fig. P4.13.
c) Use either Mathematica or MATLAB to establish steady-state conditions by imposing a step function for load reference set point(s) = and run the program with a zero step-load change ΔPL = 0 for 5 s. After 5 s impose a step load change of to find the transient response of Δω(t).
d) Use Mathematica or MATLAB to establish steady-state conditions by imposing a step function for load reference set point and run the program with a zero step-load change ΔPL = 0 for 5 s. After 5 s impose a step load change of to find the transient response of Δω(t).
Figure P4.12 shows the block diagram of two generators interconnected by a tie line (transmission line). Data of generation set (steam turbine and generator) #1:
for frequency change (Δω1) per generator output power change
the load change (ΔPL1) per frequency change
step-load change:
angular momentum of steam turbine and generator set: M1 = 4.5,
base apparent power: Sbase = 500 MVA,
governor time constant: TG1 = 0.01 s,
valve charging time constant: TCH1 = 1.0 s,
(load reference set point)1 = load ref1(s) = 0.8 pu.
Data of generation set (steam turbine and generator) #2:
for frequency change (Δω2) per change in generator output power
the load change (ΔPL2) per frequency change (Δω2):
step load change:
angular momentum of steam turbine and generator set: M2 = 6,
base apparent power: Sbase = 500 MVA,
governor time constant: TG2 = 0.02 s,
valve charging time constant: TCH2 = 1.5 s,
(load reference set point)2 = load ref2(s) = 0.8 pu.
Data for tie line: T = 377/Xtie with Xtie = 0.2 pu.
a) List the ordinary differential equations and the algebraic equations of the block diagram of Fig. P4.12.
b) Use either Mathematica or MATLAB to establish steady-state conditions by imposing a step function for load reference set point(s)1 = load ref1 (s) = pu, load reference set point (s)2 = load ref2 (s) = pu and run the program with a zero step-load changes ΔPL1 = 0, ΔPL2 = 0 for 10 s. After additional 5 s impose step-load change and after additional 2 s impose pu to find the transient responses Δω1(t) and Δω2(t).
Two synchronous generators are connected through three-phase transformers to a transmission line, as shown in Fig. P4.15. The ratings and reactances of the generators and transformers are
1. Generators 1 and 2: 600 MVA, 34.5 kV;
2. Transformer T1 rating is 600 MVA, (34.5 kV connected in Δ)/(345 kV connected in Y);
Xtransformer leakage = 10% = 0.10 pu. Note that the Y of transformer T1 is grounded.
3. Transformer T2 rating is 600 MVA, (34.5 kV connected in Δ)/(345 kV connected in Y);
Xtransformer leakage = 20% = 0.20 pu. Note that the Y of transformer T2 is not grounded.
4. On a chosen base of Sbase = 600 MVA, VL-L base = 345 kV in the transmission-line circuit, the line reactances are X1line = X2line = 10% = 0.10 pu and X0line = 30% = 0.30 pu.
a) Draw each of the three (positive-, negative-, zero-) sequence networks.
b) Determine the zero-, positive-, and negative-sequence bus impedance (reactance) matrices by means of the Zbus building algorithm, as described in Chapter 8 of [83].
c) Apply the matrix reduction technique by Kron, as outlined in Chapter 7 of [83].
d) Determine numerical value (subtransient) for the line-to-ground fault current when the fault occurs at bus 2 of Fig. P4.15.
e) Determine numerical value (subtransient) for the line-to-line fault current when the fault occurs at bus 2 of Fig. P4.15.
One of the major problems in using wind and solar energy is energy storage. Wind or solar energy may not be available when it is needed. A report from a wind farm in New Mexico states that the wind farm could lose as much as 60 MW within a minute. For example, there are several scenarios of how the power change of 60 MW per minute can be mitigated through complementary albeit more expensive power sources: one of them is the combination of a compressed-air power plant with a supercapacitor plant for bridging the time from when the wind power plant output decreases (60 MW per minute) to when the compressed-air energy storage (CAES) plant can take over. A CAES plant requires a start-up time of about 6 minutes. To bridge this 6 minute start-up time required for a 100 MW compressed-air power plant, supercapacitors or ultracapacitors are proposed to provide up to 100 MW during a 6 minute interval amounting to a required energy storage of 10 MWh. Inverters fed from supercapacitors can deliver power within milliseconds to the power system replacing the lost power of 60 MW per minute almost instantaneously. This combination of compressed-air power plant and supercapacitors or ultracapacitors as bridging energy sources can be employed for peak-power operation as well as for improving power quality by preventing short brownouts or blackouts.
Figure P4.16.1 depicts the block diagram of such a supercapacitor plant consisting of wind turbine, mechanical gear, synchronous generator, three-phase rectifier, supercapacitor bank, three-phase inverter, transformer, and power system.
a) Design a controlled three-phase rectifier in the tens of kilowatts range. Figure P4.16.2 shows the three-phase rectifier of Fig. P4.16.1 with six diodes and one self-commutated switch, an insulated-gate bipolar transistor (IGBT). The nominal phase voltages of the generator can be assumed to be van = 1200 V sin ωt, vbn = 1200 V sin(ωt – 120°), and vcn = 1200 V sin(ωt – 240°). The IGBT is gated with a switching frequency of 3 kHz and you may assume a duty cycle of δ = 50%. Cf1 = 200 μF, Lf1 = 90 mH, Cf2 = 50 μF, ideal diodes D1 to D6 and Df, vgs = 40 V magnitude, Rsd = Rss = 10 Ω, Csd = Css = 0.1 μF, Cf = 1000 μF, Lf = 1 mH, and Rload = 10 Ω.
1) Perform a PSpice analysis and plot output voltage vload(t), generator phase current iaphase, and the generator line-to-line voltage vab.
2) Subject the generator phase current to a Fourier analysis.
3) Determine the displacement power factor of the generator.
4) Determine the total power factor of the generator.
b) Design PWM (pulse-width-modulated) three-phase current controlled voltage-source inverter feeding power into the utility system via Ygrounded/Δ transformer in the tens of kilowatts range. The inverter circuit of Fig. P4.16.3 is to be analyzed with PSpice where the DC voltage is VDC = 600 V, AC output voltage of the inverter is VACL-L = 360 V, and the switching frequency of the IGBTs is fswitch = 7.2 kHz. Cfilter = 1000 μF, Lw = 1 mH, Rw = 10 mΩ, Lf = 45 μH, Cf = 10.3 μF, Rf = 10 mΩ, Lsyst = 265 μH, Rsyst = 50 mΩ. The power system phase voltages are vagrid(t) = 7200 V sin ωt, vbgrid(t) = 7200 V sin(ωt – 120°), and vcgrid(t) = 7200 V sin(ωt – 240°). Note the voltages vasyst(t), vbsyst(t), and vcsyst(t) are the power system voltages vagrid(t), vbgrid(t), and vcgrid(t) referred to the Y- or inverter side of the Y/Δ transformer.
1) Determine the transformation ratio of the Y/Δ transformer between inverter and power system; you may assume ideal transformer conditions.
2) Using PSpice calculate and plot input and output voltages and currents of the inverter, provided the output voltage is a sinusoid as given by the power system.
3) Subject the output current of the inverter to a Fourier analysis.
4) Determine the overall costs of this power plant if a specific cost of $4000/kW installed capacity is assumed. Note that a coal-fired plant has a specific cost of $2000/kW installed capacity.
Voltage and current ripples due to PWM current-source inverters cause voltage stress in windings of machines. The configuration of Fig. E3.14.1 represents the first two turns of the winding of an induction motor, which is assumed to be identical to the winding of a synchronous motor. Each turn can be represented by four segments having an inductance Li, a resistance Ri, a capacitance to ground (frame) Ci, and some interturn capacitances Cij and inductances Lij. Figure E.3.14.2 represents a detailed equivalent circuit for the configuration of Fig. E3.14.1.
To simplify the analysis neglect the capacitances Cij and the inductances Lij. This leads to the circuit Fig. E3.14.3, where the winding is fed by a PWM current source. One obtains the 15 differential equations as listed below.
The parameters are
R1 = R2 = R3 = R4 = R5 = R6 = R7 = R8 = 25 μΩ; L1 = L3 = L5 = L7 = 1 mH, and L2 = L4 = L6 = L8 = 10 mH; C1 = 0.2 μF for the cable of 10 m length; C3 = C5 = C7 = 0.7 pF, C2 = C4 = C6 = C8 = 7 pF, C15 = C37 = 0.35 pF, C26 = C48 = 3.5 pF; G1 = G3 = G5 = G7 = 1/(5000 Ω), G2 = G4 = G6 = G8 = 1/(500 Ω); and L15 = L37 = 0.005 mH, L26 = L48 = 0.01 mH. Z is either 1 μΩ (short-circuit) or 1 MΩ (open circuit).
Assume a PWM triangular current-step function is(t), as indicated in Fig. E3.14.4. Using Mathematica or MATLAB compute all state variables for the time period from t = tcalculate = 0 to 2.1 ms. Plot the voltages (v1 – v5), (v2 – v6), (v3 – v7), and (v4 – v8) from t = 0 to tplot = 2 ms. Note that the computing time t = tcalculate = 2.1 ms must be larger than the plot time t = tplot = 2 ms.