Professional Documents
Culture Documents
Objectives
- Protection Design - Circuit Breaker Rating (fault level) - Transmission Planning - Operational Planning - Substation safety - Interference study - Circuit Rating Circuit Breaker (CB) rating and Fault level
Under normal (healthy) condition, the CB will experience a voltage Vrated Under fault, the CB has to clear a fault IF (much greater than normal load current). The CB rating is often expressed MVA Fault level = 3 Vrated IF Per unit fault level = ( 3 Vrated IF) /( 3 Vbase Ibase) =IF/Ibase=IF pu (pu fault current) because Vbase is Vrated
Assumptions 1. No shunt elements a) No shunt capacitor nor reactor b) Tx Tap ratio t = 1 c) B=0 (cable, line) 2. Iload = 0 3. Eg = V(=1 pu) for all nodes and machine.
Worked examples
A) 2-bus system (circuit as above) A 33kV bus-j of fault level of 250MVA is connected to bus-k by a 1.089ohm cable. Determine the fault current at bus-k. (Use 100MVA base.) At bus-j, IF=250/100pu=2.5pu, Xg=1/2.5=0.4pu Xt=1.089100/332=0.1pu At bus-k, IF=V/Xth = 1/(0.4+0.1)=1/0.5pu 2pu Since Ibase=Sbase/(3 Vbase)=100/(333) =1.75kA, IF=21.75=3.5kA B) 4-bus system (p173 Jones) On 100MVA base Thevenin XG=0.4pu for generator, XM=0.5pu for motor, and XL=0.92pu for each line By network reduction, the final Xth at bus D is 0.345pu (Fig. 6.2d) and IF=1/0.345=2.9pu, i.e. pu fault level at D = IF = 2.9pu 290MVA
Very often, the pu source reactance is obtained by XS=1/IF (assuming ES=1). Strictly speaking, it is only valid if IF2=0. The most appropriate method is to use fault infeed IF1 such that XS=1/I F1, instead of fault level IF.
If =
Eg Xg + Xt
V + Il Xg + Xt
I=YV
where, e.g. for a 2-node system,
Eg
Xg
Xt
= Io + Il = 3 f
I o = fault current with load ignored f Conclusion: The fault current equals to summation of fault current with load ignored and load current. Discrepancies Since Il is normally a fraction of If, the fault current appears negligible changes with load (e.g. between peak and light). e.g. If =5, Il (peak)=1, Il (light)=0.5, then If(peak)=6, If(light)=5.5 Reasons of discrepancy: - Number of generators neglected - Transformer tap neglected - Number of parallel circuits (e.g. Tx) neglected - Outages of some cables to reduce MVAr generation at mid-night
Y=
Ya + Yb Yb
Yb Yb + Yc
Yb Ya Yc
V = Y-1 I = Z I
V1 M Vj
Z11 L M Z j1
Z1k M Z jk M Z kk M
0 M =
Z1k M Z jk M Z kk M Ik
M M = Vk Z k1 L M M
Ik M 0
kth column
By Ohms law
Z eq =
Vk Z kk I k = = Z kk Ik Ik
-----------(1)
Open-circuit voltage Voc = prefault voltage = normal (loadflow) voltage Voc If Zeq can be determined, then fault current If = Z eq. + Z f
The voltage rise at node j due to Ik injection is Vj = Zjk Ik Note: only one column of Z-matrix is required.
If =
Zkk Ek If Zf
Ek Z kk + Z f
Ek Z kk for dead short
If =
(c)
Calculate the voltage during fault for all nodes For node j , the voltage rise due to a current Ik (= -If ) is
Ej
= Ik Zjk = If Zjk
from (1)
A Z= j
Actually If Zjk is a voltage drop because If is a negative current injection. By superposition theorem, the voltage at node j is: Vj = Ej + Ej = Ej If Zjk
Check: if k = j,
A B C D
---------- (2)
Vk = Ek If Zkk = Ek
Ek Zkk = 0 Z kk
(d)
Calculate branch current flow (including fault and load) Imn = (Vm Vn) Ymn ----------- (3)
The above also holds in forming the negative and zero sequence Z-matrix, except that the previous loadflow contains only positive sequence quantities. Therefore step (a) is modified to (a') Build the sequence Y-matrix (i.e. Y or Yo) for the circuit elements.
Both methods use Thevenin, but the simplified (conventional) approach using network reduction has to eliminate all the buses (except the faulty bus), in order to establish the Thevenin equivalent and then to obtain the fault current. Therefore, the voltage dip at healthy bus, the fault flow, the CB interrupting current nor the machine fault contribution cannot be calculated in this conventional method.