Professional Documents
Culture Documents
Dunstan 1954
Dunstan 1954
In the of the complete theoretical determination of increased to decrease their d-c short-circuit
case of old d-c systems it will enable us the rise of current to its maximum, and its current while d-c voltage regulators are in
to study them from the viewpoint of subsequent decay. operation to maintain voltage for normal
modernization of protective devices and to The responsibility for recommending a loads and overloads?
correct the weak spots as they are dis- short-circuit calculating procedure and an 5. Can industry establish for itself the
covered. application guide for interrupting devices design characteristics of small and large rec-
We extend to the authors our congratu- on d-c systems has been placed on a sub- tifiers so that the most economical combi-
lations on their excellent and practical committee of the Low Voltage Air Breaker nation of d-c circuit breakers and trans-
paper. Committee of the AIEE Committee on former reactance are selected and still satisfy
Switchgear. normal load characteristics?
There is still much to be learned about the 6. What are the short-circuit capabilities
W. R. Crites and A. G. Darling: Mr. actions of machines and other sources of and characteristics of metallic rectifiers? of
Eichenberg's comment concerning the short-circuit current. Among the more im- storage batteries? What are the outputs of
growth of d-c systems is very important. portant avenues of inv-2stigation which electrolytic cells when a fault occurs into
The increase in the use of d-c systems means could yield beneficial results, these are which they can feed?
that there is an increasing need for a stand- suggested: 7. In what terms can d-c interrupting
ardized method of calculating fault currents devices such as breakers, contactors and
in these systems. 1. A better understanding of d-c short- fuses be rated so that the application engi-
This paper is but one of several steps circuit currents from rectifiers feeding into neer will know the time that elapses before
toward an accepted practice for calculating external elements of resistance and induct- enough resistance is introduced into the cir-
fault currents in d-c systems. It could not ance. cuit to prevent the current from going to
have been written without the fundamental 2. When two or more sources of current higher values, and will also know the maxi-
theory given in the Linville paper and the feed into a common circuit, we need to mum current to which the breaker or con-
Linville and Ward papers. (See references know what relationship those currents take tactor or fuse may be subjected without re-
1 and 2 of the paper.) McClinton, Bran- with respect to each other as each tries to spect to the rate at which the current will
cato and Panoff checked the theory by test rise to its ultimate value. Does the pre- rise through the device?
and agreed, with some modification to the dominant current stay as predominant or
saturation factor, that the theory could become more so? These are some of the things that need to
represent actual conditions. (See reference 3. What happens to the rate of rise of be solved to design d-c systems intelligently
3.) current as the point of fault is moved in and economically.
The next step consisted in calculation of steps of increasing amounts of resistance or If the answers to these and similar ques-
machine characteristics, tabulating and inductance, or both, away from the termi- tions were known, a long step would have
plotting them as given in two reports to the nals of the source? This needs to be an- been taken toward the simplification of d-c
D-C Machinery Subcommittee.4,6 Snively swered for rectifiers as well as machines and fault calculation.
and Robinson showed how machine induct- in the case of rectifiers, the effect of chang- Besides the intent of the authors to indi-
ance could be obtained by calculation and ing the amounts and proportions of primary cate a simplified approach to obtaining the
by test. (See reference 5.) The character- and secondary resistance and inductance critical values of short-circuit current from
istics of some of the components of the ex- needs to be clarified. d-c motors and generators, it is their hope
ternal circuit are given in other works listed 4. Can adequate operating character- that, by so doing, broader and more inten-
in the references. The contribution which istics be obtained from rectifiers in their sive interest in the fault characteristics of
this paper makes is that of a simplification transformer or primary supply reactance is all d-c sources and circuits will be aroused.
these numbers. Medford, the third bus PASS 125 KW KVA6LDRA PROSPECT 125-_
OO5230KW 9 5700KVAR |_ KW KVAR KW 'YA
from the bottom in the left column, is the s 4850KW 5 10400KVAR 9- 4
340K 9900KVAR KW KVAR
starting end of the track as shown by its KW K NE 00W
numbers 1, 15, and 32. For convenience ~~~iN'KW KVAF/
R
in tracing the track the first few busses are O
60KW 5 71 0KAR 664 5 27001KW 9 IOOOKVAR 9 29450KW 9 1600KVAR
GRANTS-
Medford, Park Street, Grants Pass 66, PASS66 -;5~
~~~~~~~~K KVAWR260YR
_POPCT6 5 4750KW 9 11500KVAR
Gold Ray, Prospect 66, Prospect 125, KW KVA ,' 5 92450KW 9 6800KVAR ',
Days Creek, Dixonville, Toketee, Dixon-
ville, Days Creek, etc. KW KVAR
5 95150KW 9 7800KVAR
9 17500KW S 4500 KVAR
KW KVAR 5 17950KW 9 195001VAR
LONE PINE A2
ST.-' '75_ 17500KW 9 4500K!V,
Current Flows PARK McCLOUD
KW KA
KW KVAR 5 17950KW 9 19600KVAR
Fig. 2 shows the loads and generations KW KVAR|
expressed as currents, and the summation
of these currents flowing down the track. KVAR
5 46400KW 9 18200KVAR KW K(VA R
_ KW
|5 400KW 9 O1400KVAR KW KVAR
The line currents shown in Fig. 2 are MEDFORD COPC02 125- WEED 1255 '0
5 43700KW 9 17200KVAR
progressive totals of the load and gen- .5 48751)KW 5 10400KVAR
KW KYAR t 5 66700KW 9 9100KYAR
eration currents and are easily formed by
an addition run with load cards and line
cards properly interspersed. KW KVAR 595150KW 9 7800KVAR 5 920t)KW 9 3900KVAR
5 42650KW 5 26200KVAR
The currents in Figs. 2 and 3 are given FALL 'COPCO 2 66
-- ~~WUE
D666 - KW KVA R
to -a decimals in the per-unit system. CREEK '5 50000K<W 9 IOOOOKVAR
F 5 52500KW9 34000KVAR 5 5 800XKW 9 1200YAR
The 5's and 9's preceding the in-phase KW KVAR t KW KVAy I
~~~KW KVAR
and quadrature components of current
indicate signs, a 5 indicating a plus quan- 5 43900KW 5 58 00KVA R KW KVAR |
I KW KVAR |
tity and a 9 indicating a negative quan- KW KVA R i5 56900KW 9 200KVAR I
tity or a current flowing opposite the YREKA _-
I
LITT LE S HASTA' 1~CATEL 5 500K W 5 40001CYAR
5 1300 KW 9 6000KVAR 5 IOOKW 9 IOOOKVAR
arrow. Generally the quadrature com- X5 5001KW 5 4000KVAR
ponent of current flows in the opposite
direction from reactive power which is Fig. 2. Summation load currents
the reason for the numerous instances of
the components of current flowing in op-
posite directions. and the series voltages inserted to repre- reactive power. Reactive power flows
A voltage in each of the seven loops sent off-nominal transformer taps. and reactive loads should be corrected for
resulting from the current flows in Fig. 2 line charging, -but this step was omitted
is obtained by multiplying these current Voltage Calculation because the small charging current for
flows by the line impedances and sum- this particular network did not warrant
ming the products by loops. Off-nominal Voltages in the network are computed the extra half-hour of computing involved.
transformer ratios are represented by a serially starting with an assumed voltage The transformer off-nominal voltage
series voltage inserted at the trans- at some reference bus and following the ratio was taken care of by a series volt-
former, and these series voltages must also track summiing the impedance drops and age, and the current ratio is accounted for
be summed by loops. The loop voltages series voltages for transformer taps by correcting the load at the transformer
from the currents in Fig. 2 and from the throughout the network. This computa- low-voltage bus.
transformer ratios are multiplied by the tion is discussed subsequently as an illus- The generation, load and loss listed at
inverse of the impedance matrix resulting tration of card handling. the top of Fig. 4 was formed by addition
in a set of loop currents which will pro- runs of the load cards. Generation was
duce compensating voltages in each loop Completed Study plus load and load was minus load so that
to satisfy Kirchhoff's law that the sum- the total of all load cards was the loss,
mation of voltage about each loop must Fig. 4 shows the completed study in- the total of the plus load cards was the
equal zero. cluding check quantities. Voltages were generation, and the total of the minus
Fig. 3 shows the line currents resulting determined as discussed and are printed load cards was the load.
from combining the loop flows which are out in rectangular co-ordinates. The
listed at the bottom of Fig. 3 with the voltage printed opposite each transformer Checks
line currents in Fig. 2. Fig. 3 is a cur- is the series voltage inserted to represent
rent diagram showing the actual currents the off-nominal ratio. The bus voltages In running a network study it is im-
which would flow under the assumed were multiplied by the conjugates of the portant to have a way of checking that
schedule of load and generation currents, currents in Fig. 3 to form the power and the series of calculations have been cor-
5
18474 W 5
IF
GRNT 56190KW 519GOLDA 5 21744KW 529732KVAR 9354W58IKA
5 500KW 5 4000KVAR
U
LOOP CURRENTS
LOOF P 5 3 1157 8 I8926
LOOF P2 9 37830 9220107 YREKA
LGOOF P3 9 1430 5 7746
LOOF P4 5 37573 9 e5204
LOOP P5 0 46459 9 1639
LOOP PI6 5 6467 5 21275
LOOF P 7 9 54551 9 91140
BUS CHECK
VOLTAGE CHECK
MEDFORD 5 5 LONE P66 5
1 WEED
0 66 5 99193 9 1905
PARK ST 5 9 2 FALL CREEK 2
Fig. 3 (above). 9 WEED 125 5 5 2953
Current diagram GR PASS 66 5 5 COPCO66 2 96562
5 COPCO 125 5 97979 5 1944
GOLD RAY 9 2 5 COPCO 5
125
COPCO 66 5 960 19 5 1634
?-3 PROSP 66 5
9 WEED 125 1
2 FALL CREEK 5 957 15 5 1246
5
PROSP 125 5 MC CLOUD MEDFORD 5 99994 9 4
DAYS CREEK 5 2 5 WEED 66 1 LONE P 66 5 10 1452 5 767
DIXONVILLE 5 5 CASTELLA LONE P 125 5 9 1638 5
TOKETEE 5 5
9 3884
Fig. 4 (right). Load flow LIT SHASTA 2 DIXONVILLE 5 5
diagram PAS 125 5 2 5 YREKA I 99865 26160
OR CREEK 5 96619 5 19976
LONE P 125 9 15 2
VATS
MEDFORD 5 99999 9 4
rent deck 19 and the voltages representing
transformer ratios 13. The answer decks
are all voltage decks and include the
print voltage deck 22, the load voltage
deck 23, the sending voltage deck 24, the
receiving voltage deck 25, and the volt-
age check deck 32. The cards are sorted
in reverse track sequence so that volt-
age drops may be added instead of sub-
tracted. Since the track closes upon it-
MANUALLY PLACE self, the sorted deck may be cut at any
THE DECKS IN THE FOLLOWING SEQUENCE bus in the network as reference for the
I. 00 DECK (MULTIPLIER) voltage run. That is, the engineer can
2. 21 DECK (MULTIPLIER)
3. 13 DECK (M'CAND) set the voltage at any bus in the network,
4. 19 DECK (M'CAND) and all the other bus voltages will be
5. 22DECK(ANSWER) computed with reference to this lead-off
6. 23 DECK
(ANSWER) voltage.
7. 24 DECK
(ANSWER)
8. 25 DECK (ANSWER)
9. 32 DECK (ANSWER) Iterative Procedure
I
SORT The input data for the mechanized solu-
COLS. 76 TO 73 tion discussed herein are the loads and
PICK UP IN REVERSE ORDER generation expressed as currents, the
UNIT I REFERENCE transformer taps expressed as series volt-
MULTIPLIER THIRD VOLTAGE ages and a voltage at any specified point.
FIRST SECOND A machine operator can process these data
and return to the system engineer a dia-
UNIT MULTIPLIER AND REFERENCE VOLTAGE CARDS MAY BE PLACED AS gram similar to Fig. 4, which is an exact
SPECIFIED BY THE ENGINEER. PLACE THE CARDS AHEAD OF THE SPECI- checked solution for the input data.
FIED REFERENCE BUS THEN PLACE THE CARDS ORIGINALLY AHEAD OF
THE REFERENCE BUS ON THE BACK OF THE DECK However, for the engineer to express the
I loads and generation as currents and the
CALCU LATOR transformer taps as series voltages he
MULTIPLY BOARD must know or estimate the various bus
PROGRESSIVE TOTALS
CLEAR MACHINE voltages. Thus a study may have to be
I reworked to reach the desired loadings,
SORT transformer taps, and voltage levels.
TO SEPARATE DECKS This adjustment procedure may be easily
COLS. 26 a 25
I mastered by an engineer familiar with
his system.
Iterative procedure may also be used to
Fig. 5. Card-handling chart make a circuit change in a network. For
example if a line is to be added between
rectly performed. A voltage check and a it is probable that the multiplications of certain busses, the line current may be
bus check are listed at the bottom of the currents and voltages were correctly per- estimated and added to the receiving end
study in Fig. 4. The voltage check is a formed. bus, and deducted from the sending end
list of the voltages computed at busses bus, and a case run with the changed bus
a second or third time when summing Illustration of Card Handling loadings. The actual current in the
the voltage drops down the track. Thus added line equals the difference in ter-
all busses which have more than one track- One of the steps in the card-handling minal bus voltages multiplied by the line
sequence number in Fig. 1 appear in the procedure is illustrated in Fig. 5. This is conductance so that any case run with-
voltage check. The check voltages the step at which voltage throughout the out the line may be changed to a case
should compare with the voltage printed network is calculated. Fundamentally with the line by changing the loads at the
at the same bus in the circuit diagram this is a multiplication run wherein line line terminal busses. This concept ena-
except for round-off error. Inasmuch as impedances are multiplied by line cur- bles a system engineer to extend the use-
the voltages were calculated progressively, rents to form voltage drops which are fulness of a network which has been
it is probable that if a few bracketing progressively summed. Therefore, quan- coded for computation until there are a
voltages are correct, the intervening volt- tities which are to be added, such as the sufficient number of changes in the net-
ages are also correct. reference voltage and the series voltages work to justify a new coding and master
Each power and reactive power listed for transformer taps, are not added di- deck.
in Fig. 4 is the result of an independent rectly but are multiplied by a unit card to
multiplication of a bus voltage by the go in as products the same as the im- Conclusion
conjugate of an appropriate current. pedance drops. The multiplier decks
The bus check is the sum of the power are the line impedance deck 00 and the The digital load flow study appears to
and reactive power entering each bus and unit multipliers for transformer ratios 21. be attractive to those who have no net-
should equal zero. If this check holds, The multiplicand decks are the track cur- work analyzer available and to engineers