Figure 7.9 shows a typical i bus of an n-bus system. This i bus is connected to three other buses m, k, and l. The load connected to bus i demands a complex power Sload,i.
The total complex (apparent) power leaving bus i via the lines to buses k, l, m is
Slines,i=V˜i[I˜k+I˜l+I˜m]∗,
(7-65)
where* denotes the complex conjugate and ỹij are the line admittances:
Therefore, for the PQ buses we have the mismatch powers
ΔWi=Sload,i+Slines,i
(7-71)
=Sload,i+[(rowiofY¯¯¯bus)V¯¯¯bus]∗V˜i,
(7-72)
where Sload,i is the load (demand) complex (apparent) power at bus i (Sload,i = Pi + jQi). Breaking the mismatch power ΔWi into a real part ΔPi and an imaginary part ΔQi, we have the real and reactive mismatch powers, respectively:
ΔPi=Pi+Real{[(rowiofY¯¯¯bus)V¯¯¯bus]∗V˜i}
(7-73)
and
ΔQi=Qi+Imag{[(rowiofY¯¯¯bus)V¯¯¯bus]∗V˜i}
(7-74)
with Pi and Qi being the real and reactive load (demand) powers at bus i.
where yij and θij are the magnitude and the phase angle of the (i,j)th entry of the admittance matrix, respectively, and Vj = |V˜j| and δj are the voltage rms magnitude and phase angle at bus j.
Based on the equations given, the conventional fundamental power flow algorithm is the computation of the bus voltage vector for a given system configuration and linear loads. The Newton–Raphson approach (Eq. 7-64) will be used to force the mismatch power to zero (Eq. 7-57). The iterative solution procedure for the fundamental Newton–Raphson power flow algorithm is as follows (Fig. 7.10):
•Step 1: Construct the admittance matrix Y¯¯¯bus (Eq. 7-41).
•Step 2: Make an initial guess for x¯ (e.g., 1.0 pu volts and zero radians for all bus voltages, Eq. 7-42).
•Step 3: Evaluate mismatch power (ΔW¯¯¯¯(x¯), Eq. 7-56). If it is small enough, then stop.
•Step 4: Establish Jacobian J¯¯ (Eqs. 7-30 and 7-82) and calculate Δx¯ξ (Eq. 7-62) using either matrix inversion (for problems with a few buses) or upper (lower) triangular factorization combined with backward/forward substitution.
•Step 5: Update the bus vector voltage (Eq. 7-63) and go to Step 3.
It is recommended to use upper (lower) triangular factorization combined with backward/forward substitutions rather than matrix inversion in computing the solution because an iterative method is used where the coefficient matrix is updated from iteration to iteration.
7.3.5 Application Example 7.8: Computation of Fundamental Admittance Matrix
A simple (few buses) power system (Fig. E7.8.1) is used to illustrate in detail the Newton–Raphson load flow analysis. All impedances are in pu and the base apparent power is Sbase = 1 kVA. Compute the fundamental admittance matrix for this system.
Solution to Application Example 7.8
For this example y14 = –ỹ14, where ỹ14 is the admittance between bus 1 and bus 4 and y14 is an entry of the bus admittance matrix Ybus. Note the minus sign.
Note that Y¯¯¯bus is singular; that is, (Y¯¯¯bus)−1 cannot be found, because there is no admittance connected to the ground bus. This is unimportant for the Newton-Raphson approach because for forming of the Jacobian the entries of Y¯¯¯bus are used and Y¯¯¯bus is not inverted.
Comment about Singularity
•Considering
[I¯bus]=[Y¯¯¯bus]−1[V¯¯¯bus],
without any connection (admittance) from any bus to ground (Fig. 7.11), Y¯¯¯−1bus is singular. However, if Y¯¯¯bus contains load information (admittance to ground) then it will not be singular.
•Considering
[Δx¯]=[J¯¯]−1[ΔW¯¯¯¯],
if [J¯¯] contains load information (P3, Q3, Fig. 7.12), then [J¯¯] is not singular.
7.3.6 Application Example 7.9: Evaluation of Fundamental Mismatch Vector
For the system of Fig. E7.8.1, assume bus 1 is the swing bus (δ1 = 0 and |V˜1| = 1.0 pu). Then we can make an initial guess for bus voltage vector x¯0 = (δ2, |V˜2|, δ3, |V˜3|, δ4, |V˜4|)t = (0, 1, 0, 1, 0, 1)t. Note, the superscript 0 means initial guess. Compute the mismatch power vector for this initial condition.