Scribd Logo
Carica

Benvenuto in Scribd!

  • 23 C

    Caricato da

    api-3698505
    13 visualizzazioni10 pagine

    Titolo originale

    23-c

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Formati disponibili

    PDF o leggi online da Scribd

    Hai trovato utile questo documento?

    Questo contenuto è inappropriato?

    Segnala questo documento

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Scarica in formato PDF o leggi online su Scribd
    Segnala contenuti inappropriati
    13 visualizzazioni10 pagine

    23 C

    Caricato da

    api-3698505

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Scarica in formato PDF o leggi online su Scribd
    Segnala contenuti inappropriati
    di 10
    Eiscme POWER ENERGY SYSTEMS Electrical Power and Energy Systems 28 (20065) 109-118 ew elsevier.comMocateijepes FAD-ATC: A new method for computing dynamic ATC M. Eidiani **, M.H.M. Shanechi ° “Department of Electrical Engineering, Azad University of Bojnourd, Bejnowrd, Iran " Department of Elecirical Engineering, Ferdowsi University of Mashhad, Mashhad, tran Received 16 April 2004; recived in revised form 26 September 2005; accepted 19 November 2005 Abstract This paper presents a Fast and Accurate Dynamic method for ATC (FAD-ATC) calculations with considering transient stability analysis (TSA) and voltage stability analysis (VSA) termination criteria, This method can be used for contingency screening and ranking, Estimation of the determinant of Jacobian maitix is used for assessment of voltage stability. This method is compared with the following ‘nethods: energy difference between SEP (Stable Equilibrium Point) and UBP (Unstable Equilibrium Point), index of Chiang, and continuation power flow. ‘TSA method use Athay’s PEBS (Potential Energy Boundary Surface) and s new method named POMP. The POMP method follows the Point Of Maximum Potential (POMP) energy on post-fault system trajectory. This point is approximated by Taylor's expansion of second order. Since tis method does not use any convergent algorithm, itis more reliable than all existing methods that use UEP, The goals of this method are assessment of ATC with considering VSA and TSA and assessment of dangerous contingencies, The ideas are demonstrated on 2, 3, 7 (Cigre), 10, 30 (IEEE) and 145 (Iowa State) bus networks, © 2005 Elsevier Lid. All rights reserved. Keywords: ATC: VSA; TSA: Energy fanetion 1, Introduction Order 889 mandated each control area to compute ATC and post them on a communication system called Open Access Same-Time Information System (OASIS) [1]. ATC computing ‘methods can be divided into static and dynamic methods. Static methods can be divided into three categories: OPF (linear and nonlinear optimization), sensitivity analysis [2], and Continu- ation Power Flow (CPF) (3.4). In OPF, for each transaction, generations and loads are increased until allowable trans- mission power between two areas reaches maximum, In CPR, for each transaction, generations and loads are increased until one line reaches its MVA limit or other static terminated conditions. ‘Termination criteria in static methods are: transmission line flow constraints (thermal and static stability), diverging DC load flow, diverging AC load flow (including voltage collapse) and voltage limit on each bus (0.95 < Vi <.05). * Conesponding author. Address: Vakil-Abad 15, Usef Ghasemi 10, No. 1542, Floor One, PC 9188858466, Mashad, ran E-mail address: info @eidisni com (M. Eidian). URL: higvwoweidiani.com. (0142-06155 - see front matter ® 2005 Elsevier Ltd. All ights reserved, do: 10.1016f jjopes.2005,1 1,00 Simplicity, transparency, flexibility and rapidity (velocity or celerity) are the advantages of static methods, Dis- advantages and defects of these methods are optimistic, inaccuracy, and not considering all constraints. Dynamic method is a transient stability constrained by maximum allowable transfer (MAT). This method consists of sereen- ing a large number of contingencies and scrutinizing the dangerous ones. TSA constrained by maximum allowable transfer [5] is one of the best dynamic methods. The existing of stable equilibrium in post-disturbance system is only a necessary condition of system stability [6]. Thus, it is also important to censure that the system can safely make the transition from the pre- to post-disturbance operation point. Several methods for dynamic ATC evaluation are those based on second-kick (7), SMIB, dot product [8,9] and bifurcation theory [10]. Termination criteria in dynamic methods are transient instability, voltage instability (dynamic), and static termination conditions. Robustness and accuracy are the advantages of dynamic methods. Disadvantages and defects are: the ‘potentially harmful (dangerous)’ contingencies in base case Gist stable operating point) and ATC case (when system reaches terminated criterion) are not necessarily the same, voltage stability and transient stability are not considered simultaneously, and response is slow. 0 M,Bidiani, M.H.M. Shanechi / Electrical Power and Energy Systems 28 (2006) 109-118 Stable operating points in base case and ATC case are different. Therefore, we use contingency ranking in stressed stable operating point, In this paper, we improve ATC with TSA and VSA termination criteria. Estimation of the determinant of Jacobian matrix (EDOJ) is used for computing VSA index. This method is compared with energy difference between SEP and UEP U1} f index of Chiang [12], and continuation power flow BA) Det(J)=0 means that at least one real eigenvalue of J must be zero, Under certain modeling assumptions, a zero eigenvalue of J corresponds to lees of small disturbance Stability of the system [13,14], The determinant of Jacobian matrix may exhibit large discontinuities; therefore, the Jacobian determinant method is inadequate to predict the critical point. Estimation of the determinant of Jacobian matrix (EDOD) method (Section 2.1) is not only very accurate but also very fast. TSA is a very important termination criterion in computing ATC. PEBS [15], BCU [16] and second kick [7] are the most important methods in TSA with energy method. We use POMP for TSA termination eriterion and demonstrate that this method has a very good performance, In this paper, EDOJ method uses voltage stability constraint a upper limit of transient stability and computed transient stability limit by energy method is used as initial condition of simulation method. VSA methods are presented in Section 2. TSA methods are presented in Section 3, ATC computing methods are applied to one small system in Section 4. In Section 5, system equations and structure preserving energy functions [17] are derived. The main algorithm is presented in Section 6. In Section 7, ATC is determined for 7, 10, 30 (IEEE), and 145 bus (lowa State), 2, YSA methods Contiauation power flow [3.4], angular distance between the SEP and closest UEP in a Euclidean sense [18], energy difference between the SEP and closest UEP using an energy function [11], test function (7,) [12], and estimations of the determinant of Jacobian matrix are the most important methods for VSA. Computing the performance index proposed in [12] is casier than other proposed methods, which require computing eigenvalues, eigenvectores, singular values, energy functions, or condition numbers [12]. It should be pointed out that the proposed index in [12], does not require computing unstable load flow solutions (UEP). This method [12] has the following deficiencies; (1) determination of f, is time consuming and (2) the solution is not accurate. In this section, we propose a new approach for VSA. 2.1, Estimation of the Jacobian determinant Jacobian matrix can be divided into four sub matrixes A, B, C and D. It is easy to show that: AFT ews) = detain — ca) a = + Deis) = ie — cD ) Define: diag(A) = [ar az Gy Joa are diagonal elements of A, prod(diag(A)) =... ay¢tn2-.dyy . 1 I 1 Udi) = |r ae ig @ a 0. 0 mn 0 0 diag(diag(a)) = oe 0 se Ay Now we show: IXil>IXyI¥ij and X=A4,B,C and D 8 We have: = 0 VV ysin66,—3)—y,) aT 3 V,¥ysin(d,— 4,79) 4 20; ° Do Mi¥ysinld, 6-74) +2VjYysin(—y,) ts So = VY, sin(d, -W . 6,0 and ¥4=90, then: j= lAul > lAgl S) = [Dil > |Dyl By; and C;; can be computed similarly, Nove, we can propose new approximation for Jacobian matrix. We have: Det(J) = prod(diag(A))-prod(diag(D) —diag(C)-diag( I/diag(A))-diag(B)) 6) ‘This estimation of Jacobian determinant is calculated very fast and it can be used for computing ATC. We define 4 as: Py =Pa(L+A) FEN, (number of generators) o Py = Py +2) JSNp (number of demands) ‘That: 1=0= P= P (base power flow) I= hp P =P (Voltage stability limit), Where P?, and {are generation and demand at operating point. 2 is estimated Ms Bidiant, MEM. Shanechi / Electrical Power and Energy Systems 28 (2006) 109-118 dy quadratic equation: =—a(Det() +b a and b are unknown 8) ‘Two stable load flow solutions needed to compute a and 6. Two different load and generation conditions (in ATC direction) is used for computing a and b. Now there are two equations and two unknown parameters: co) ‘The experimental results and most simulation results show that this approximation is not as accurate as 4, of Chiang. Now, we use a new load flow solution to inerease the accuracy of new method and to calculate three unknown parameters (a, 6, ¢}: (oy VS

    Potrebbero piacerti anche