Active and reactive power control of a VSC-HVdc

Electric Power Systems Research 78 (2008) 1756–1763

Contents lists available at ScienceDirect

Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr

Active and reactive power control of a VSC-HVdc ¨ H.F. Latorre ∗ , M. Ghandhari, L. Soder Royal Institute of Technology, Electric Power Systems, Teknikringen 33, 10044 Stockholm, Sweden

a r t i c l e

i n f o

Article history: Received 16 February 2007 Received in revised form 5 February 2008 Accepted 9 March 2008 Available online 24 April 2008 Keywords: Control Lyapunov functions Modal analysis Power oscillations damping Transient stability Voltage support VSC-HVdc

a b s t r a c t Voltage source converter-based HVdc (VSC-HVdc) systems have the ability to rapidly control the transmitted active power, and also to independently exchange reactive power with transmissions systems. Due to these characteristics, VSC-HVdcs with a suitable control scheme can offer an alternative means to enhance transient stability, to improve power oscillations damping, and to provide voltage support. In this paper, a VSC-HVdc is represented by a simple model, referred to as the injection model. Based on this model, an energy function is developed for a multi-machine power system including VSC-HVdcs. Furthermore, based on Lyapunov theory (control Lyapunov function) and small signal analysis (modal analysis), various control strategies for transient stability and damping of low-frequency power oscillations are derived. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Utilization of power electronics-based controllable systems (or devices) in transmission systems has opened new opportunities for the power industry to optimize utilization of the existing transmission systems, and at the same time to keep high-system reliability and security. As a member of these controllable systems, voltage source converter-based HVdcs (VSC-HVdcs) have the ability to rapidly control the transmitted active power, and also to independently exchange reactive power with transmissions systems. Therefore, VSC-HVdcs with a suitable control scheme can offer an alternative means to enhance transient stability, to improve power oscillations damping, and to provide voltage support. In this paper, a VSC-HVdc is represented by a simple model, referred to as the injection model. This model is intended for analysis of load flow and electromechanical dynamics, and it is helpful for understanding the impact of the VSC-HVdc on power system stability. Furthermore, power system components including VSC-HVdcs are represented by their positive sequence systems (only the fundamental frequency components of the voltages and currents are considered). It is well known that by controlling the transmitted active power through the VSC-HVdc, it is possible to enhance the system tran-

∗ Corresponding author. Tel.: +46 87907767. E-mail addresses: [email protected] (H.F. Latorre), ¨ [email protected] (M. Ghandhari), [email protected] (L. Soder). 0378-7796/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.epsr.2008.03.003

sient stability, and to improve the low-frequency power oscillations damping. Based on control Lyapunov function (CLF) and modal analysis, various control strategies for controlling the transmitted active power are presented. Since a VSC-HVdc has the ability to rapidly and independently exchange reactive power with transmissions systems, a control algorithm for controlling the exchanged reactive power is also presented. It will be shown that the reactive power control has a significant contribution on system stability and damping. This control is more feasible for a heavily loaded power system and/or when the dc link in the VSC-HVdc system is out of service, but the converter stations are in operating conditions. 2. Injection model Let a VSC-HVdc system be installed between bus i and bus j in parallel with an ac transmission line as shown in Fig. 1. Due to the VSC-HVdc characteristics (i.e. independent control of active and reactive power exchange), each converter (seen from the ac bus to which the converter is connected) can be considered as an ideal sinusoidal voltage source whose magnitude (Uc ) and phase angle () can be controlled. Therefore, as shown in Fig. 2, the VSC-HVdc can be modeled as two controllable voltage sources in series with two reactances which represent the reactance of each converter transformer, respectively. In Fig. 2: U¯ ci = Uci eji = Uci cos i + jUci sin i = Uci1 + jUci2 , U¯ cj = Ucj ejj = Ucj cos j + jUcj sin j = Ucj1 + jUcj2

(1)

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

1757

given by Psio = bTi Ui (sin i Uci1o − cos i Uci2o ), Qsio = bTi [Ui2 − Ui (cos i Uci1o + sin i Uci2o )], Psjo = bTj Uj (sin j Ucj1o − cos j Ucj2o ) Fig. 1. VSC-HVdc in parallel with an ac transmission line.

Qsjo = bTj [Uj2 − Uj (cos j Ucj1o + sin j Ucj2o )]

(6)

Psi = bTi Ui (sin i Uci1 − cos i Uci2 ), Qsi = −bTi Ui (cos i Uci1 + sin i Uci2 ), Psj = bTj Uj (sin j Ucj1 − cos j Ucj2 ), Qsj = −bTj Uj (cos j Ucj1 + sin j Ucj2 )

(7)

Note however that

Fig. 2. Simple model of the VSC-HVdc.

Psjo = −Psio ,

Psj = −Psi

(8)

It should be noted that the proposed injection model and a model developed by ABB (which is intended for analysis of electromechanical dynamics) have been respectively applied to a test power system. The simulation results for different disturbances have shown that the test system dynamic behaviors with the injection model and the ABB developed model, respectively, have a good coincidence. Fig. 3. Injection model of the VSC-HVdc.

3. Modulation of active power where Uci1 , Uci2 , Ucj1 and Ucj2 are the control variables of the voltage sources. It is assumed that the dc voltage control keeps the dc voltage close to its rated voltage. Therefore, the losses of the converters are assumed constant, regardless of the current through the converters. The losses are consequently represented as a constant active load. However, the losses of the dc cables are neglected. Fig. 3 shows the injection model of the VSC-HVdc, where:

In this section, only the modulation of the active power is considered, i.e. Qsi = Qsj = 0 in Eq. (5). It is well known that by controlling the transmitted active power through the VSC-HVdc, it is possible to enhance the system transient stability, and to improve the low-frequency power oscillations damping. Based on CLF and modal analysis, various control strategies for controlling the transmitted active power are derived in this section.

Psi = bTi Ui (sin i Uci1 − cos i Uci2 ),

3.1. Control Lyapunov function

Qsi = bTi [Ui2 − Ui (cos i Uci1 + sin i Uci2 )],

Consider the uncontrolled system:

Psj = bTj Uj (sin j Ucj1 − cos j Ucj2 ), Qsj = bTj [Uj2 − Uj (cos j Ucj1 + sin j Ucj2 )]

(2)

with bTi = 1/XTi . Furthermore, the relation between buses i and j is given by Psj = −Psi

(3)

However, the reactive power is independently exchanged at each bus. The control variables in Eq. (1) can be divided into two terms, expressing the fixed variables (or set values) and modulated variables as follows: Uci1 = Uci1o + Uci1 ,

Uci2 = Uci2o + Uci2 ,

Ucj1 = Ucj1o + Ucj1 ,

Ucj2 = Ucj2o + Ucj2

(4)

Qsi = Qsio + Qsi ,

Psj = Psjo + Psj ,

Qsj = Qsjo + Qsj

˙ V(x) = V˙ uncontrolled = ∇V(x) · fo (x) ≤ 0 (10) Now consider the controlled system (affine system) defined by x˙ = f (x, u) = fo (x) +

nu 

(5)

where the subindex o denotes the steady-state parts (or set values), and P and Q correspond to the modulated parts which are

ui fi (x)

(11)

i=1

where ui is a control variable and nu is the number of controllable variables. Using Vuncontrolled as a Lyapunov (or an energy) function for the controlled system (11), the time derivative of this function along the trajectories of Eq. (11) is then given by



Thus, the expressions of the active and reactive power can also be rewritten in a similar way as Psi = Psio + Psi ,

x˙ = fo (x) (9) Let V(x) = Vuncontrolled be a Lyapunov (or an energy) function for Eq. (9), i.e. V(x) is positive definite and its time derivative is given by

˙ V(x) = ∇V(x) ·

fo (x) +

nu 



ui fi (x)

i=1

= ∇V(x) · fo (x) +

nu 

ui ∇V(x) · fi (x)

i=1

= V˙ uncontrolled + V˙ controlled ≤ V˙ controlled

(12)

1758

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

The objective of the CLF is to select ui in such way that ˙ Vcontrolled is non-positive. A comprehensive presentation of the CLF can be found in [1–5]. 3.1.1. Application to power systems Power systems are most naturally described by a set of differential-algebraic equations of the form: x˙ = f (x, y), 0 = g(x, y)

(13)

By virtue of the implicit function theorem, it can be shown that Eq. (13) is locally equivalent to a differential equation of the form (provided that ∂g/∂y is nonsingular): x˙ = f (x, h(x)) = fo (x) In [5,6] and references therein, for a power system in the COI reference frame with n-generators (one-axis model) and n + N-load buses with constant active loads the following energy function is given: ˜ = Vuncontrolled = V1 + ˜ E  , U, ) V(ω, ˜ ı, q

8 

V2k + Co

(14)

k=1

Thus, the time derivative of the energy function is

  T dV(x) dok ˙ 2 = V˙ uncontrolled = − Dk (ω ˜ k )2 −  (E qk ) xdk − xdk dt

1 Mk ω ˜ k2 , V1 = 2

n 

n

V21 = −

k=1

VHVdco =



2n+N

V22 =

PLk ˜ k ,

V23 =

 2n

V24 =

k=n+1

V25 = −

 2xdk−n

2 [E  qk−n

 

 − 2Eqk−n Uk

+ Uk2

cos(ı˜ k−n − ˜ k )],



Bkl Uk Ul cos(˜ k − ˜ l ),

k=n+1l=n+1 2n   xdk−n − xqk−n

V26 =

k=n+1

 4xdk−n xqk−n

n   Efdk Eqk

V27 = −

k=1

[Uj2 − 2Uj (cos ˜ j Ucj1o + sin ˜ j Ucj2o )]

 xdk − xdk

 [Uk2

V28 =

,

cos(2(ı˜ k−n − ˜ k ))],

− Uk2 n 

E  2qk

8 

k=1

V2k + VHVdco + Co

dVHVdco dV24 dV25 dV26 dV22 + + + + dt dt dt dt dt = (Pi + PLi + Psio )˜˙ i = 0, dVHVdco dV23 dV24 dV25 dV26 + + + + dt dt dt dt dt

(21)



˜ i

Ui

U˙ i =0 Ui

dVHVdco dV24 dV25 dV26 dV22 + + + + dt dt dt dt dt = (Pj + PLj + Psjo + Plosses )˜˙ j = 0, dVHVdco dV24 dV25 dV26 dV23 + + + + dt dt dt dt dt = (Qj + QLj + Qsjo )

 ) 2(xdk − xdk

(20)

Note that the modulated (or control) variables in Eq. (4) are not considered in the above energy function, i.e. Uci1 = Uci2 = Ucj1 = Ucj2 = 0. The presence of the VSC-HVdc does not alter the negativeness HVdc of (19), since at bus i and bus j, V˙ uncontrolled is given by

2n+N 2n+N

1 2

2

= (Qi + QLi + Qsio ) 1

(19)

k=1

dUk ,

Uk

k=n+1

bTj

= V1 + VHVdc uncontrolled

2n+N

k=n+1

k=1

Thus, the energy function for a power system including an uncontrolled VSC-HVdc is given by



  Q Lk

k=1

bTi 2 [Ui − 2Ui (cos ˜ i Uci1o + sin ˜ i Uci2o )] 2 +

Pmk ı˜ k ,

k=1

n

which is non-positive. Having an uncontrolled VSC-HVdc (based on the injection model) in the system, the construction of the energy function follows the procedure given in [7], which results in



where

n

U˙ j Uj

(22)

˜ j

Uj

=0

(23)

Thus ˜ dω/dt, ˜ and similarly for Using the notation [dV/dt]ω˜ for ∂V/∂ω ˙ the other states, V(x) becomes

HVdc V˙ uncontrolled

 dV 

The introduction of the modulated parts of the VSC-HVdc implies that Uci1 , Uci2 , Ucj1 and Ucj2 are not fixed anymore, since Uci1 = Uci2 = Ucj1 = Ucj2 = 0. Thus, with modulated

1

dt

 dV

22

dt

 dV

23

dt

 dV

27

dt

ω ˜

+

 dV

dV24 dV26 + + dt dt dt 21



dV24 dV25 dV26 + + + dt dt dt +

dV24 dV25 dV26 + + dt dt dt

+

dV28 dV24 + dt dt



˜

 ı˜

=

 U

=

n 

Eq

n 

=−

k=1

=−

Dk (ω ˜ k)

2

(15)

k=1



(Pk + PLk )˜˙ k = 0



(Qk + QLk )

 Tdok

 xdk − xdk

(E˙  qk )

2

U˙ k =0 Uk

= V˙ uncontrolled

(16)

(17)

(18) Fig. 4. Feedback control system.

(24)

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

1759

Fig. 5. Active power control.

HVdc

Obviously, V˙ controlled becomes non-positive if Psi = k(˜˙ i − ˜˙ j ) = kfij = kf

(29)

where k is a positive gain and fij is the difference of the frequency of the buses where the VSCs are connected. Eq. (29) gives the CLF-based modulation of active power to enhance the system stability and damping. 3.2. Modal analysis The purpose of the modal analysis is to design a supplementary regulator with certain selected input signals to provide positive damping in the system. Linearizing the system described by Eq. (13) around its equilibrium point, the dynamic of the power system is then described by Fig. 6. Reactive power control.

x˙ = Ax(t) + Bu(t),

(or control) variables we are dealing with a controlled system similar to (11). Now, using the energy function (21) for the controlled system, the time derivative of this function along the trajectories of this controlled system is given by ˙ V(x) =

HVdc V˙ uncontrolled

Y(t) = Cx(t) The transfer function of (30) is given by [8,9]

HVdc

Y(s)  Ri = u(s) s − i n

G(s) =

HVdc + V˙ controlled

(30)

i=1

HVdc

where V˙ uncontrolled is defined by (24) and V˙ controlled is defined by HVdc V˙ controlled = −Psi ˜˙ i − Psj ˜˙ j

which is resulted from

 dV

22

dt

+

dV24 dV25 dV26 dVHVdc + + + dt dt dt dt

(25)

 ˜

= −Psi ˜˙ i − Psj ˜˙ j (26)

 dV

23

dt

+

dV24 dV25 dV26 dVHVdc + + + dt dt dt dt

 U

U˙ j U˙ j U˙ U˙ = −Qsi i − Qsj =0 i −0 =0 Uj Uj Ui Ui

(27)

since only active power is modulated. HVdc The objective of control Lyapunov function is to keep V˙ controlled non-positive. By virtue of Eq. (8), Eq. (25) can be rewritten as HVdc V˙ controlled = −Psi (˜˙ i − ˜˙ j )

(28)

Fig. 7. Test power system.

Fig. 8. Proposed scheme control of the VSC-HVdc.

1760

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

where Ri is the residue of G(s) and is expressed as Ri = CVir Vil B where Vir and Vir are the right and left eigenvectors, respectively, of the eigenvalue i . As shown in Fig. 4, adding a feedback control (whose transfer function is H(s)) into the open-looped system G(s), it can be shown that (with s = i ) [10] i = Ri H(i )

oscillations damping), and the selected input signal is either the magnitude of a selected line current (iline ) or the difference between the frequencies of two selected buses (f ). The tuning procedure of the parameters of H(s) follows the residue technique presented in [10]. Fig. 5 shows the modal-analysis-based modulation of active power to improve the system stability and damping. 4. Modulation of reactive power

(31)

Eq. (31) implies that the position of a selected eigenvalue (or mode) can be changed by tuning the parameters of the transfer function H(s) based on the information from the residue Ri . Based on Fig. 4 and Eq. 31, the active power through the VSCHVdc will be modulated to improve the damping of the selected poorly damped (or undamped) mode(s). The block diagram of the H(s) (used in this paper) is shown in Fig. 5 which consists of a wash-out filter, a lead-lag filter and a gain. The output signal of this feedback control is termed POD (which stands for power

Fig. 9. Modulation of active power.

One of the main characteristics of a VSC-HVdc is its ability to rapidly and independently exchange reactive power with transmissions systems. In order to exploit this feature, a control algorithm is applied to modulate the reactive power exchange. This control algorithm is divided into an automatic voltage regulator (AVR), and a supplementary control to improve the power oscillations damping (POD). The supplementary control is more feasible when the dc-link cannot transmit active power, but the converters are in functions to exchange reactive power. The derivation of the POD signal follows the same procedure of the modulation of active power described in Section 3.2. The block diagram control is shown in Fig. 6.

Fig. 10. Modulation of active power (heavily loaded system).

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

5. Numerical example 5.1. Power system The well-known test system shown in Fig. 7 is used to study the impact of a VSC-HVdc on the system stability and damping. The system data can be found in [9]. The generators are modeled with one field winding, one damper winding in d-axis and two damper windings in q-axis. The active and reactive components of the loads have constant current and constant impedance characteristics, respectively. The generators are equipped with AVRs, PSSs and turbine governors. All simulations are performed by using SIMPOW [11] and the results are plotted in MATLAB. 5.2. Overall control of the VSC-HVdc The overall control scheme of the VSC-HVdc used in this paper is shown in Fig. 8. Besides the modulation of active and reactive power to improve the stability of the power system, a load flow control (PLF ) is also included in the control scheme.

Fig. 11. Modulation of active power and AVR (heavily loaded system).

1761

Knowing Psi , Qsi and Qsj , the control variables Uci1 , Uci2 , Ucj1 and Ucj2 can be easily calculated from Eq. (1), (4) and (5). The input signals of the modal-analysis-based active power control, Fig. 5, is either f = f7 − f9 = f79 , or the magnitude of the current through the line between bus 6 and bus 7, i67 . The supplementary control for the modulation of reactive power, Fig. 6, also considers i67 as the input signal. This supplementary control is used when there is no transfer of active power through the dc-link. 5.3. Only modulation of active power Initially the system is set to transfer a total of 450 MW from bus 7 to bus 9, 159 MW of which are transferred through the HVdc. The fault case consists of a three phase short circuit at bus 8. The fault is cleared after 100 ms by opening one of the transmission lines between bus 7 and bus 8. There is no modulation of reactive power. Fig. 9 illustrates the simulation results. Both control strategies provide very good damping. However, the oscillations die out faster with the control based on modal analysis.

Fig. 12. System response to modulation of PLF .

1762

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

trol. However, the system is stable when CLF-based control strategy and the modal-analysis-based control with f79 as input signal are used. But, the post-fault steady-state voltage magnitude at bus 8 decreases to about 0.88 p.u. Next step is to raise the voltage at bus 8 to a (possible) desired value U8des . It can be done by manually (and stepwise) changing the set value of the transmitted power (Psio ) or by modulating PLF with input signal (U8des − U8 ). Fig. 12 shows the results of the proposed modulation which is activated at t = 30 s, and U8des = 0.95 p.u. As shown in the figure, by increasing the transmitted power (Psi ) through the VSC-HVdc, the reactive power losses in the ac lines decrease and thereby less reactive power compensation by the VSC-HVdc is needed. Therefore, the injected reactive power by VSC-HVdc (−Qsi ) decreases when the transmitted power is increasing. This results in raising the voltage

at bus 8, and also in Psi2 + Qsi2 ) below its keeping the converter apparent power (Si = limit, i.e. Simax . 5.5. Only modulation of reactive power In this case, the total power flow from bus 7 to bus 9 is 600 MW, 159 MW of which are transferred through the HVdc. The applied disturbance is a permanent interruption of power flow through the dc-link. However, it is assumed that both converters are available to exchange reactive power with the transmission system. Fig. 13 shows the simulation results with no supplementary control, i.e. kqi = kqj = 0 (see Fig. 6), and with a supplementary control in the first converter, i.e. kqi = 0, respectively. As shown in the figure, due to this disturbance and with no supplementary control, an unstable inter-area mode is excited in the system. However, by applying a supplementary control to the reactive power modulation of the first converter, this mode becomes stable and well damped. 6. Conclusions

Fig. 13. Modulation of reactive power.

Next, the total transferred power from bus 7 to bus 9 is increased to 750 MW. The transfer of power through the HVdc remains equal to 159 MW. The same fault case is applied. The simulations results in Fig. 10 show that the CLF-based control strategy and the modal-analysis-based control with f79 as input signal allow the system to remain in synchronism during the first oscillations and additionally provide damping once the first swing stability has been achieved. However, the system becomes unstable later mostly due to the lack of reactive power support. Note that the system experiences first swing instability when i67 is used as input signal into the modal-analysis-based control. 5.4. Modulation of active and reactive power (AVR) The modulation of reactive power, Fig. 6, is now included for the heavily loaded system. The gains kqi and kqj are set to zero, i.e. only voltage support (AVR) is provided. Fig. 11 shows the simulations results for the same fault. As shown in the figure, the system still experiences first swing instability when i67 is used as input signal into the modal-analysis-based con-

It has been shown that VSC-HVdcs with a suitable control scheme can offer an alternative means to enhance transient stability, to improve power oscillations damping, and to provide voltage support, since they have the ability to rapidly control the transmitted active power, and also to independently exchange reactive power with transmissions systems. Based on a simplified power system model and injection model, an energy function for a power system including VSCHVdcs has been proposed (Eq. (21)). Using the proposed energy function, a CLF-based control strategy for modulation of the active power of a VSC-HVdc has been derived to enhance transient stability and damping. However, this control strategy has been applied to a detailed power system model with successful results. It has been shown that the proposed CLF-based and modalanalysis-based control strategies have been effective to enhance transient stability and damping (Fig. 9). However, for the case with heavily loaded system, it has been shown that only modulation of active power is not sufficient to stabilize the system (Fig. 10). Therefore, the modulation of the reactive power of the VSC-HVdc must be considered. Figs. 11 and 13 clearly show that the reactive power control has a significant contribution on system stability and damping. Figs. 10 and 11 show that the modal-analysis-based active power control is more effective and robust with the input signal f = f79 than with the input signal i67 . The reason might be that f = f79 (probably) contains global information, but i67 only contains local information. Figs. 12 and 8 and the simulation results show that a VSC-HVdc with a multi-choice control scheme has the ability to optimize uti-

H.F. Latorre et al. / Electric Power Systems Research 78 (2008) 1756–1763

lization of the existing transmission systems, and at the same time to keep high-system reliability and security. 7. Future work The future work will include application of VSC-HVdcs (with more detailed model) to large and more complex power systems, and also application of PMU-based signals as input signals to the proposed control strategies. References [1] Z. Artstein, Stabilization with relaxed controls, Nonlinear Anal. Theor. Meth. Appl. 7 (11) (1983) 1163–1173. [2] E. Sontag, A universal construction of Artstein’s theorem on nonlinear stabilization, Syst. Control. Lett. 13 (1983) 1163–1173. [3] V. Jurdjevic, J.P. Quinn, Controllability and stability, J. Diff. Eq. 28 (1978) 381–389. [4] M. Ghandhari, G. Andersson, M. Pavella, D. Ernst, A control strategy for controllable series capacitor in electric power systems, Automatica (2001) 37. [5] M. Ghandhari, G. Andersson, I.A. Hiskens, Control Lyapunov functions for controllable series devices, IEEE Trans. Power Syst. 16 (4) (2001). [6] M.A. Pai, Energy Function Analysis for Power System Stability, Kluwer Academic Publishers, 1989.

1763

[7] M. Pavella, P.G. Murthy, Transient Stability of Power Systems, Theory and Practice, John Wiley & Sons, 1994. [8] IEEE Power System Engineering Committee, Eigenanalysis and Frequency Domain Methods for System Dynamic Performance, IEEE Publication 90TH0292-3-PWR, New York, 1989. [9] P. Kundur, Power System Stability and Control, McGraw-Hill, New York, 1994, p. 813. [10] F.L. Pagola, I.J. Perez-Arriaga, G.C. Verghese, On sensitivities, residues and participation. Application to oscillatory stability analysis and control, IEEE Trans. Power Syst. 4 (1) (1989). [11] STRI, SIMPOW—A Digital Power System Simulator, ABB Review No. 7, 1990. H´ector F. Latorre received the M.Sc. degree in Electrical Engineering from Royal ´ Institute of Technology, Stockholm, Sweden, in 2002. He worked at Interconexion ´ Electrica S.A. (ISA), Colombia, 9 years, in the area of design of substations. He is currently Ph.D. student at the Royal Institute of Technology (KTH). Mehrdad Ghandhari received the M.Sc., Tech. Lic. and Ph.D. degrees in Electrical Engineering from Royal Institute of Technology, Stockholm, Sweden, in 1995, 1997, and 2000, respectively. He is currently Assistant Professor at the Royal Institute of Technology (KTH). Lennart S¨ oder was born in Solna, Sweden in 1956. He received his M.Sc. and Ph.D. degrees in Electrical Engineering from the Royal Institute of Technology, Stockholm, Sweden in 1982 and 1988, respectively. He is currently a professor in Electric Power Systems at the Royal Institute of Technology. He also works with projects concerning deregulated electricity markets, distribution systems, risk analysis and integration of wind power.