diff options
author | Angelo Rossi <angelo.rossi.homelab@gmail.com> | 2023-06-21 12:04:16 +0000 |
---|---|---|
committer | Angelo Rossi <angelo.rossi.homelab@gmail.com> | 2023-06-21 12:04:16 +0000 |
commit | b18347ffc9db9641e215995edea1c04c363b2bdf (patch) | |
tree | f3908dc911399f1a21e17d950355ee56dc0919ee /benchmarks/dcn22.dat |
Initial commit.
Diffstat (limited to 'benchmarks/dcn22.dat')
-rw-r--r-- | benchmarks/dcn22.dat | 720 |
1 files changed, 720 insertions, 0 deletions
diff --git a/benchmarks/dcn22.dat b/benchmarks/dcn22.dat new file mode 100644 index 0000000..8b31b6b --- /dev/null +++ b/benchmarks/dcn22.dat @@ -0,0 +1,720 @@ +BEGIN NEW DATA CASE
+C BENCHMARK DCNEW-22
+C Illustrate modeling of unsymmetric series-RL that uses compensation as
+C described in the July, 1997, newsletter story. Begin with uncoupled
+C branches of 1 ohm resistance. The answer is obvious by inspection. The
+C [R] matrix is diagonal matrix equal to the unit matrix, and [X] is zero.
+C In terms of symmetrical components, Zo = Z1 = Z2 (a second half of the
+C data this identical but alternative representation to produce the same
+C known answer). With each phase having 2 ohms resistance split in half,
+C and with balanced, sinusoidal sources having 2 volts, the currents are
+C equal to half the source voltage: sinusoidal of amplitude 1. Frequency
+C of the 3-phase sources at SENDA, SENDB, and SENDC have been reduced to
+C nearly zero so the 1st step of phase "a" is very close to the peak. Step
+C 1 is very close to answer at time zero: Va = 1 volt, Vb = Vc = 1/2 volt
+C A total of 7 subcases are involved.
+C UTPF update of 20 June 2007 allows ATP data to override .PL4 file type
+C choices that normally are made within STARTUP. Illustrate use here :
+C FMTPL4 L4BYTE NEWPL4 Next card FORMAT ( 16X, A6, 2X, 2I8 )
+C CONCATENATE. 1 0 <Force .PL4 type> key text anywhere
+C For 2 or more stacked subcases, this declaration must be in the first
+C and must be in the original data file. Here, the C-like choice is
+C mandatory in order that it be in place for subcase 4, which will use
+C $DEPOSIT to change NEWPL4 again, to 2 (for Pisa). That only works
+C if a C-like or Pisa file then is in effect. One can not use $DEPOSIT
+C to change a formatted .PL4 file to Pisa format in this way. So, the
+C preceding ensures that the 4th will work; it ensures that some user
+C will not rely upon values in STARTUP that might be incompatible.
+ .0001 .0005
+ 1 -1 1 0 1 -1
+ 5 5 20 20 100 100
+ SENDA ENDA 1.0 1
+ SENDB ENDB SENDA ENDA 1
+ SENDC ENDC SENDA ENDA 1
+ ENDA ENDB 1.E7
+ ENDB ENDC ENDA ENDB
+ ENDC ENDA ENDA ENDB
+91ENDA MODEL Z0Z1Z2 1.0
+91ENDB MODEL Z0Z1Z2 1.0
+91ENDC MODEL Z0Z1Z2 1.0
+ SENDA RECA 1.0 1
+ SENDB RECB SENDA RECA 1
+ SENDC RECC SENDA RECA 1
+ RECA RECB 1.E7
+ RECB RECC RECA RECB
+ RECC RECA RECA RECB
+91RECA MODEL [R][L]
+91RECB MODEL [R][L]
+91RECC MODEL [R][L]
+ 1.0 0.0 0.0
+ 0.0 0.0 0.0
+ 0.0 1.0 0.0
+ 0.0 0.0 0.0
+ 0.0 0.0 1.0
+ 0.0 0.0 0.0
+BLANK card follows the last branch card
+BLANK line terminates the last (here, nonexistent) switch
+14SENDA 2.0 0.1 0.0 { 1st of 3 sources. Note balanced,
+14SENDB 2.0 0.1 -120. { three-phase, sinusoidal excitation
+14SENDC 2.0 0.1 120. { with no phasor solution.
+BLANK card follows the last source card
+ SENDA RECA ENDA SENDB RECB ENDB SENDC RECC ENDC
+C Step Time SENDA RECA ENDA SENDB RECB ENDB SENDC RECC ENDC SENDA
+C ENDA
+C SENDB SENDC SENDA SENDB SENDC
+C ENDB ENDC RECA RECB RECC
+C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
+C 0.0 0.0 0.0 0.0 0.0
+C 1 .1E-3 2.0 .999999848 .999999848 -.99989117 -.49994551 -.49994551 -1.0001088 -.50005434 -.50005434 1.00000015
+C -.49994566 -.50005449 1.00000015 -.49994566 -.50005449
+C 2 .2E-3 1.99999998 .999999842 .999999842 -.99978234 -.49989109 -.49989109 -1.0002176 -.50010875 -.50010875 1.00000014
+C -.49989124 -.5001089 1.00000014 -.49989124 -.5001089
+BLANK card ending node voltage outputs
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+C 2nd of 7 subcases. Progress to symmetrical components having
+C Z1 = Z2 so the answer can be shown to be identical to that
+C using Type-51,52,53 modeling with sequence impedances. In
+C fact, there are 2 identical, uncoupled networks driven from
+C the same balanced 3-phase sources at SENDA, SENDB, and SENDC.
+ .0001 .050
+ 1 1 1 0 1 -1
+ 5 5 20 20 100 100
+ SENDA ENDA 0.3 1.0 1
+ SENDB ENDB SENDA ENDA 1
+ SENDC ENDC SENDA ENDA 1
+ ENDA ENDB 1.E7 { Balanced, interphase leakage gives 3-phase
+ ENDB ENDC ENDA ENDB { (rather than 3, single-phase) Z-thevenin.
+ ENDC ENDA ENDA ENDB { 3 coupled phases are required by "Z0Z1Z2"
+91ENDA MODEL Z0Z1Z2 0.3 1.0 { Sequence Ro, Lo in [ohms, mHenry]
+91ENDB MODEL Z0Z1Z2 0.1 0.5 { Sequence R1, L1 in [ohms, mHenry]
+91ENDC MODEL Z0Z1Z2 0.1 0.5 { Note Z2 = Z1 so [Z] is symmetric
+C Next, build a copy of this, but using the old (Type-51,52,53) modeling:
+ SENDA RECA 0.3 1.0 1
+ SENDB RECB SENDA RECA 1
+ SENDC RECC SENDA RECA 1
+ RECA RECB 1.E7
+ RECB RECC RECA RECB
+ RECC RECA RECA RECB
+51RECA 0.3 1.0 { Ro, Lo in [ohms, mHenry]
+52RECB 0.1 0.5 { R1, L1 in [ohms, mHenry]
+53RECC
+BLANK card follows the last branch card
+BLANK line terminates the last (here, nonexistent) switch
+14SENDA 2.0 50. 0.0 { 1st of 3 sources. Note balanced,
+14SENDB 2.0 50. -120. { three-phase, sinusoidal excitation
+14SENDC 2.0 50. 120. { with no phasor solution.
+BLANK card follows the last source card
+ SENDA RECA ENDA
+C Step Time SENDA RECA ENDA SENDA SENDB SENDC SENDA SENDB SENDC
+C ENDA ENDB ENDC RECA RECB RECC
+C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
+C 1 .1E-3 1.99901312 .664144463 .664144463 .065757077 -.0310889 -.03466818 .065757077 -.0310889 -.03466818
+C 2 .2E-3 1.99605346 .658837702 .658837702 .195443292 -.09061195 -.10483134 .195443292 -.09061195 -.10483134
+C 3 .3E-3 1.99112393 .652991396 .652991396 .321457199 -.14486663 -.17659057 .321457199 -.14486663 -.17659057
+BLANK card ending node voltage outputs
+C 500 .05 -2. -.59687593 -.59687593 -2.0936819 3.18312082 -1.0894389 -2.0936819 3.18312082 -1.0894389
+C Variable maxima : 2.0 .664144463 .664144463 3.23520786 3.50393608 3.23541237 3.23520786 3.50393608 3.23541237
+C Times of maxima : .02 .1E-3 .1E-3 .0428 .0092 .0361 .0428 .0092 .0361
+C Variable minima : -2. -.60247492 -.60247492 -3.3079993 -3.2352894 -3.4484065 -3.3079993 -3.2352894 -3.4484065
+C Times of minima : .01 .0496 .0496 .0127 .0394 .0059 .0127 .0394 .0059
+ CALCOMP PLOT
+ 144 5. 0.0 50. -1.0 1. RECA ENDA { These 2 phase-a voltages should be equal
+ 194 5. 0.0 50. -4.0 4.0BRANCH { Following phase-a currents should be equal
+ SENDA RECA SENDA ENDA
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+C 3rd of 7 subcases generalizes the preceding. For the Z0Z1Z2 model,
+C Z2 is changed so it no longer is equal to Z1. Since Type-51,52,53
+C modeling no longer can provide a comparison, this half is replaced by
+C the more general [R][L] alternative that requires the user to supply
+C full matrices. Although input precision is limited to 8 digits to the
+C right of the fixed decimal point, agreement is close (6 or more digits)
+ .0001 .050
+ 1 1 1 0 1 -1
+ 5 5 20 20 100 100
+ SENDA ENDA 0.3 1.0 1
+ SENDB ENDB SENDA ENDA 1
+ SENDC ENDC SENDA ENDA 1
+ ENDA ENDB 1.E7
+ ENDB ENDC ENDA ENDB
+ ENDC ENDA ENDA ENDB
+91ENDA MODEL Z0Z1Z2 0.3 1.0
+91ENDB MODEL Z0Z1Z2 0.1 0.5
+91ENDC MODEL Z0Z1Z2 .101 0.8
+ SENDA RECA 0.3 1.0 1
+ SENDB RECB SENDA RECA 1
+ SENDC RECC SENDA RECA 1
+ RECA RECB 1.E7
+ RECB RECC RECA RECB
+ RECC RECA RECA RECB
+91RECA MODEL [R][L]
+91RECB MODEL [R][L]
+91RECC MODEL [R][L]
+C Correct the data 1 April 2002. Following a change to USERNL, the
+C phase-domain X or L has changed. Previously, we were using units
+C of mH, which relied on STATFR of STARTUP for the conversion.
+C This is improved by the definition of XOPT = 60 Hz (see below),
+C which frees the computation from dependence of STARTUP. Also, we
+C switch to input of X in ohms (rather than L in mH):
+$DISABLE { For the hysterical record, retain the old matrix on comments:
+ .1670000000 .0665866025 .0664133975 { R(1,1), R(1,2), R(1,3)
+ .7666666667-.1720084679 .4053418013 { L(1,1), L(1,2), L(1,3)
+ .0664133975 .1670000000 .0665866025 { Row 2 of [R]
+ .4053418013 .7666666667-.1720084679 { Row 2 of [L]
+ .0665866025 .0664133975 .1670000000 { Row 3 of [R]
+-.1720084679 .4053418013 .7666666667 { Row 3 of [L]
+$ENABLE { Done showing old (and wrong) matrix [L] in mHenry. Begin new:
+C Preceding disabled data was from years past. Correct this 1 April 2002.
+$UNITS, 60.0, 60.0, { Define frequency XOPT = 60 for the impedance computation
+C The .DBG will includes the following information. We use full precision of
+C the matrix values, although blanks and leading zeros have been removed:
+C USERNL begins with Lo, L1, L2 [H] = 1.00000E-03 5.00000E-04 8.00000E-04
+C Converted to Xo, X1, X2 = 3.76991E-01 1.88496E-01 3.01593E-01
+C 3x3 phase-domain impedance matrix in ohms follow. For each row I, X(I,J) is below R(I,J). w = 3.769911E+02 rad/sec.
+ .1670000000 .0991483886 .0338516114 { R(1,1), R(1,2), R(1,3) in ohms
+ .2890265241 .0436936220 .0442709723 { X(1,1), X(1,2), X(1,3) in ohms
+ .0338516114 .1670000000 .0991483886 { Row 2 of [R]
+ .0442709723 .2890265241 .0436936220 { Row 2 of [X]
+ .0991483886 .0338516114 .1670000000 { Row 3 of [R]
+ .0436936220 .0442709723 .2890265241 { Row 3 of [X]
+C $UNITS, -1.0, -1.0, { Done with ohms and micromhos at 60 Hz, so restore original
+C The preceding line is not tolerated because the associated warning
+C message is issued during the first time step. If we cancel the XOPT
+C definition, there will be no trace during the first time step. So,
+C without changing the answer, we leave XOPT = 60 to suppress the
+C warning message that will be seen in output of preceding 2 subcases.
+BLANK card follows the last branch card
+BLANK line terminates the last (here, nonexistent) switch
+14SENDA 2.0 50. 0.0 { 1st of 3 sources. Note balanced,
+14SENDB 2.0 50. -120. { three-phase, sinusoidal excitation
+14SENDC 2.0 50. 120. { with no phasor solution.
+BLANK card follows the last source card
+ SENDA RECA ENDA
+C First 3 output variables are electric-network voltage differences (upper voltage minus lower voltage);
+C Next 6 output variables are branch currents (flowing from the upper node to the lower node);
+C Step Time SENDA RECA ENDA SENDA SENDB SENDC SENDA SENDB SENDC
+C ENDA ENDB ENDC RECA RECB RECC
+C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
+C 1 .1E-3 1.99901312 .784095685 .784095685 .05984815 -.02824862 -.03159953 .05984815 -.02824862 -.03159953
+C 2 .2E-3 1.99605346 .776335531 .776335531 .178012015 -.08221824 -.09579377 .178012015 -.08221824 -.09579377
+C 3 .3E-3 1.99112393 .768107458 .768107458 .29308232 -.13116828 -.16191404 .29308232 -.13116828 -.16191404
+BLANK card ending node voltage outputs
+C 500 .05 -2. -.57981099 -.57981099 -2.1430566 3.2140066 -1.07095 -2.1430566 3.2140066 -1.07095
+C Variable maxima : 2.0 .784095685 .784095685 3.2728624 3.57534262 3.28836555 3.2728624 3.57534261 3.28836555
+C Times of maxima : .02 .1E-3 .1E-3 .0427 .0091 .0161 .0427 .0091 .0161
+C Variable minima : -2. -.58387034 -.58387034 -3.4132778 -3.2730642 -3.3604773 -3.4132778 -3.2730642 -3.3604773
+C Times of minima : .01 .0496 .0496 .0126 .0394 .0059 .0126 .0394 .0059
+ CALCOMP PLOT
+ 144 5. 0.0 50. -1.0 1. RECA ENDA
+ 194 5. 0.0 50. -4.0 4.0BRANCH
+ SENDA RECA SENDA ENDA
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+C 4th of 7 subcases is unrelated to the preceding three. Instead, it
+C illustrates op amp (operational amplifier) modeling as first requested
+C by Masahiro Kan of Toshiba Corp. in Japan. See October, 1997, newsletter
+C As of 30 Aug 97, the op amp (Type-20 source) is ignored during the
+C phasor solution.
+C 19 March 2001, expand to illustrate Pisa-format .PL4 file for time
+C simulation. Use began with HARMONIC FREQUENCY SCAN in the new 15th
+C subcase of DCNEW-21. But correct operation for normal time simulation
+C also should be confirmed, so modify this existing 4th subcase for such
+C a test. Expand dT and T-max so screen plot is half of a recognizable
+C sinusoid:
+$DEPOSIT, NEWPL4=2 { Use SPY DEPOSIT to override .PL4 file type given in STARTUP
+$DEPOSIT, LUNIT4=4 { Use SPY DEPOSIT to override minus sign specified in STARTUP
+$DEPOSIT, NOPOST=1 { Use SPY DEPOSIT to override PostScript choice in STARTUP
+$DEPOSIT, NOHPGL=1 { Use SPY DEPOSIT to override the HP-GL choice in STARTUP
+$DEPOSIT, NOGNU=1 { Use SPY DEPOSIT to override the GNUPLOT choice in STARTUP
+C To prove that Pisa-format code is being used, it is easy to turn on debug
+C printout. Use here is like that pioneered in DCNEW-21 for HFS. But there
+C are differences. Whereas HFS automatically used LUNIT4 = +4 regardless
+C of what STARTUP indicated (usually value -4), this is not true for time
+C simulation, so the preceding definition of LUNIT4 is necessary in order
+C to force ATP to read from the disk-stored .PL4 file during plotting. About
+C overlay number, the .PL4 header is created in overlay 15 rather than 11. In
+C the .DBG file, look for the names HEADPI and LU4BEG to see Pisa-related
+C data values. To minimize size of the .DBG file, PostScript, HP-GL, and
+C GNUPLOT output are turned off (the final 3 $DEPOSIT lines above).
+C Turn off diagnostic 22 April 2007 as it disfigures .LIS of Mingw32 ATP:
+C DIAGNOSTIC { HEADPI in 15 } 9 { LU4BEG in 28 } 9
+PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision for 8 output columns
+C .000200 .0005 { Original simulation only advanced 3 time steps
+ .001 .010 { 10 time steps. Large dT is legal since network is resistive
+ 1 1 1 1
+ GEN SEND 1.0
+ SEND 1.0 0.0
+ OPAMP 1.0 1
+BLANK card ending branch cards.
+BLANK card ending switch cards.
+14GEN 2.0 50. 0.0 -1.
+C <____> <____><____> Gain occupies columns 11-20
+C BUS1 Gain BUSK BUSM By definition, V-1 = Gain * ( V-k - V-m)
+20OPAMP 10.GEN SEND { If V-k or V-m is zero, leave name blank)
+BLANK card terminating all EMTP source cards.
+ GEN SEND OPAMP
+BLANK card terminating all output requests.
+ CALCOMP PLOT { To demonstrate use of Pisa-format .PL4 file, try screen plot
+ 144 1. 0.0 10. SEND OPAMP { 2 sinusoidal signals differ by factor 10
+BLANK card terminating all plot cards.
+BEGIN NEW DATA CASE
+C 5th of 7 subcases really should be an extension of the following DCNEW-23
+C since these further support the 12th subcase of DCN23.DAT But the
+C following disk file already has 15 subcases as the addition is made on
+C 5 April 2000. Symmetrical component data of Type-51, 52, 53 branches is
+C the subject. Compensation is not being used. Consider a single, lumped,
+C 3-phase, series R-L branch having this imbalanced data:
+C Zo = 3.5 * Z1 = 10.5 + j 14.0 Ohm
+C Z1 = 3.0 + j 4.0 Ohm
+C Z2 = 0.5 * Z1 = 1.5 + j 2.0 Ohm
+C The far end will be terminated this way: phase "a" will be open whereas
+C phases "b" and "c" will be connected together. I.e., this is a normal
+C line-to-line fault. Data comes from Orlando Hevia in Santa Fe, Argentina
+C as copied by WSM on 3 April 2002. Mr. Hevia drove the line section using
+C a balanced, 3-phase voltage source. Although he included a phasor soution
+C for initial conditions, WSM drops this because it might be confusing (ATP
+C does not yet correctly represent unsymmetric [R] or [L] during the phasor
+C solution). Starting from zero is no problem, however, as the solution
+C settles into the steady state smoothly within a cycle or two. Just to be
+C sure, 5 cycles are simulated, and extrema are limited to the final 1.25 of
+C these. This is necessary to ignore the transient of energization. There
+C will be 3 identical tests. Each has different data, but gives identically
+C the same answers to the 8 or 9 digits of dT-loop output. Orlando Hevia
+C wrote a FORTRAN program to produce the solution exactly, in closed form.
+C This produced the following results: POLAR Magnitude Degrees
+C IB 230.9401 -143.1301
+C IC 230.9401 36.8699
+C ATP now will confirm this magnitude using simulation.
+DIAGNOSTIC { Cancel diagnostic printout of preceding subcase
+PRINTED NUMBER WIDTH, 11, 2, { Request maximum precision (for 8 output columns)
+BEGIN PEAK VALUE SEARCH, 0.075, { Ignore 1st 3.75 cycles, until in steady state
+ .000100 .100 { dT is relatively large to speed simulation. Only 1K steps
+ 1 -1 1 0 1 -1
+ 5 5 20 20 100 100
+$UNITS, 50.0, 0.0 { Begin inductance data in ohms at XOPT = 50 Hz
+C 1st of 3 line sections follows. SEQ at sending end indicates sequence data:
+51GENA OPEN1 MODEL Z0Z1Z2 10.5 14.00 { Ro and Xo, both in ohms
+52SEQB FLT1 3.0 4.00 { R1 and X1, both in ohms
+53SEQC FLT1 1.5 2.00 { R2 and X2, both in ohms
+$UNITS, 0.0, 0.0 { Return to inductance data in mHenry; XOPT = 0
+C 3x3 phase-domain [R] in ohms and [L] in mHenries follow. For each row I, L(I,J) is below R(I,J). w = 3.14159265E+02 rad/sec.
+C 5.0000000000 2.1726497308 3.3273502692
+C 21.2206590789 13.0496847320 10.2930402549
+C 3.3273502692 5.0000000000 2.1726497308
+C 10.2930402549 21.2206590789 13.0496847320
+C 2.1726497308 3.3273502692 5.0000000000
+C 13.0496847320 10.2930402549 21.2206590789
+C The present comments are diagnostic output from use of MODEL Z0Z1Z2. Such
+C output always will be found in the .DBG file. From it, produce the following
+C equivalent representation of [R] in ohms and [L] in mHenries:
+C 2nd of 3 line sections follows. PHS at sending end indicates phase domain:
+51GENA OPEN2 MODEL [R][L]
+52PHSB FLT2
+53PHSC FLT2
+ 5.000000000 2.172649731 3.327350269 { R(1,1), R(1,2), and R(1,3) in ohms
+ 21.22065908 13.04968473 10.29304025 { L(1,1), L(1,2), and L(1,3) in mHenries
+ 3.327350269 5.000000000 2.172649731 { Row 2 of [R]
+ 10.29304025 21.22065908 13.04968473 { Row 2 of [L]
+ 2.172649731 3.327350269 5.000000000 { Row 3 of [R]
+ 13.04968473 10.29304025 21.22065908 { Row 3 of [L]
+C Finally, data for 3rd of 3 line sections comes from Orlando. This is _not_
+C the same circuit, note, since [R] & [L] are not the same. The diagonal
+C elements are the same, but off-diagonals are not. This 3rd circuit differs
+C from the preceding 2, but it gives identically the same fault current. It
+C was constructed by Mr. Hevia by transferring impedance between Z1 and Z2
+C while maintaining the sum fixed. For either a L-L fault of this subcase or
+C a 1-L-G fault of the following one, fault current depends on the sum of
+C Z1 and Z2 but not on either Z1 or Z2 independently. Consider these
+C parameters, which seem exact to the limits of single precision:
+C Z of SEQ branch Z of OPH branch
+C Z0 parameter 10.5 + j 14.00 10.500 + j 14.0
+C Z1 parameter 3.0 + j 4.00 2.325 + j 13.0
+C Z2 parameter 1.5 + j 2.00 2.175 - j 7.0
+C But voltage of the faulted node will be different. Note the huge difference
+C between X1 and X2. Yet, they sum equally: 4 + 2 = 13 - 7.
+51GENA OPEN3 MODEL [R][L]
+52OPHB FLT3
+53OPHC FLT3
+ 5.0000000 -3.0235027 8.5235027
+ 21.2206591 11.8091947 11.5335303
+ 8.5235027 5.0000000 -3.0235027
+ 11.5335303 21.2206591 11.8091947
+ -3.0235027 8.5235027 5.0000000
+ 11.8091947 11.5335303 21.2206591
+BLANK card ending branches
+C Following switches measure line currents in the 3 circuits. Since phase "a"
+C is open, this is not of enough interest to warrant output. For the record,
+C let's document how close to zero such currents are, however. Salford EMTP
+C shows (prior to removal of the 3 switches from phase "a":
+C Step Time GENA GENA GENA
+C SEQA PHSA OPHA
+C 0 0.0 0.0 0.0 0.0
+C 1 .1E-3 -.367E-15 .3632E-15 .1798E-15
+C 2 .2E-3 -.284E-16 .1234E-15 -.271E-15
+C Variable max: .1121E-14 .1246E-14 .1845E-14
+C Times of max: .098 .0894 .0764
+C Variable min: -.122E-14 -.107E-14 -.229E-14
+C Times of min: .09 .0789 .0879
+C To conclude, no switches for phase "a". There remain 3 switches for phase
+C "b" (these 3 are ordered first, followed by the 3 for phase "c". This
+C way, corresponding outputs of the dT loop are contiguous, which encourages
+C easy comparison.
+ GENB SEQB MEASURING 1
+ GENB PHSB MEASURING 1
+ GENB OPHB MEASURING 1
+ GENC SEQC MEASURING 1
+ GENC PHSC MEASURING 1
+ GENC OPHC MEASURING 1
+BLANK card ending switches
+14GENA 1000. 50. -0.00 { Balanced, 3-phase excitation that
+14GENB 1000. 50. -120. { is _not_ present during the steady
+14GENC 1000. 50. -240. { state. There is no phasor solution.
+BLANK card follows the last source card
+C Column headings for the 9 EMTP output variables follow. These are divided among the 5 possible classes as follows ....
+C First 3 output variables are electric-network voltage differences (upper voltage minus lower voltage);
+C Next 6 output variables are branch currents (flowing from the upper node to the lower node);
+C Step Time FLT1 FLT2 FLT3 GENB GENB GENB GENC GENC GENC
+C SEQB PHSB OPHB SEQC PHSC OPHC
+C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
+C 1 .1E-3 -495.9539 -495.9539 -500.178 .14077359 .14077359 .14077359 -.1407736 -.1407736 -.1407736
+C 2 .2E-3 -491.6694 -491.6694 -501.478 .55967714 .55967714 .55967714 -.5596771 -.5596771 -.5596771
+C 3 .3E-3 -487.1449 -487.1449 -503.861 1.2498171 1.2498171 1.249817 -1.249817 -1.249817 -1.249817
+ FLT1 FLT2 FLT3 { Show that the fault voltage of OPH (FLT3; 3rd of 3) differs
+BLANK card ending node voltage outputs
+C 500 .05 333.3359 333.3359 -572.6502 184.74924 184.74924 184.74924 -184.7492 -184.7492 -184.7492
+C 600 .06 -333.3373 -333.3373 572.64118 -184.7477 -184.7477 -184.7477 184.7477 184.7477 184.7477
+C 700 .07 333.33716 333.33716 -572.642 184.74784 184.74784 184.74784 -184.7478 -184.7478 -184.7478
+C 800 .08 -333.3372 -333.3372 572.64195 -184.7478 -184.7478 -184.7478 184.74783 184.74783 184.74783
+C 900 .09 333.33717 333.33717 -572.642 184.74783 184.74783 184.74783 -184.7478 -184.7478 -184.7478
+C 1000 0.1 -333.3372 -333.3372 572.64196 -184.7478 -184.7478 -184.7478 184.74783 184.74783 184.74783
+C Variable max: 333.33717 333.33717 977.26344 230.90147 230.90147 230.90147 230.90147 230.90147 230.90147
+C Times of max: .09 .09 .097 .088 .088 .088 .098 .098 .098
+C Variable min: -333.3372 -333.3372 -977.2634 -230.9015 -230.9015 -230.9015 -230.9015 -230.9015 -230.9015
+C Times of min: .08 .08 .087 .098 .098 .098 .088 .088 .088
+C Note about preceding currents. The maximum of 230.90147 and minimum of
+C -230.9015 are nearly negatives of each other. This is a sign that the
+C steady state is really steady (equality in the 7th decimal digit). But
+C Orlando Hevia's exact solution of 230.9401 implies some error? Why? The
+C substantial dT (discretization error). As dT is decreased, the ATP
+C answer approaches Orlando Hevia's. Using dT = 2 microseconds, one can
+C produce exact agreement to all 7 printed digits.
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+C 6th of 7 subcases is a modification of the preceding 5th. Repeat
+C the preceding subcase, except for a different fault at the far end.
+C Termination will be as follows: phase "a" will be grounded whereas
+C phases "b" and "c" are left open. I.e., this is a normal single-line-
+C to-ground fault. Data comes from Orlando Hevia in Santa Fe, Argentina
+C Another change is this: since only a single phase carries current,
+C there is no shortage of space in the output vector. So, add a 4th
+C alternative: MODEL Z0Z1Z2 with inductance in mHenries. About the
+C correct answer, Orlando Hevia wrote a FORTRAN program to produce
+C the solution exactly, in closed form. This produced the following
+C results: POLAR Magnitude Degrees
+C IA 120.0000 -53.1301
+C Using ATP simulation, we now will confirm the 120-amp peak current.
+PRINTED NUMBER WIDTH, 12, 2, { Request maximum precision (for 8 output columns)
+POWER FREQUENCY 50.0 { Remove dependence on STARTUP value
+BEGIN PEAK VALUE SEARCH, 0.075, { Ignore 1st 3.75 cycles, until in steady state
+ .000100 .100 { dT is relatively large to speed simulation. Only 1K steps
+ 1 -1 1 0 1 -1
+ 5 5 20 20 100 100
+$UNITS, 50.0, 0.0 { Begin inductance data in ohms at XOPT = 50 Hz
+C 1st of 4 line sections follows. SEQ at sending end indicates sequence data:
+51SEQA MODEL Z0Z1Z2 10.5 14.00 { Ro and Xo, both in ohms
+52GENB OPEN1 3.0 4.00 { R1 and X1, both in ohms
+53GENC OPEN2 1.5 2.00 { R2 and X2, both in ohms
+$UNITS, 0.0, 0.0 { Return to inductance data in mHenry; XOPT = 0
+C 3x3 phase-domain [R] in ohms and [L] in mHenries follow. For each row I, L(I,J) is below R(I,J). w = 3.14159265E+02 rad/sec.
+C 5.0000000000 2.1726497308 3.3273502692
+C 21.2206590789 13.0496847320 10.2930402549
+C 3.3273502692 5.0000000000 2.1726497308
+C 10.2930402549 21.2206590789 13.0496847320
+C 2.1726497308 3.3273502692 5.0000000000
+C 13.0496847320 10.2930402549 21.2206590789
+C The present comments are diagnostic output from use of MODEL Z0Z1Z2. Such
+C output always will be found in the .DBG file. From it, produce the following
+C equivalent representation of [R] in ohms and [L] in mHenries:
+C 2nd of 3 line sections follows. PHS at sending end indicates phase domain:
+51PHSA MODEL [R][L]
+52GENB OPEN3
+53GENC OPEN4
+ 5.000000000 2.172649731 3.327350269 { R(1,1), R(1,2), and R(1,3) in ohms
+ 21.22065908 13.04968473 10.29304025 { L(1,1), L(1,2), and L(1,3) in mHenries
+ 3.327350269 5.000000000 2.172649731 { Row 2 of [R]
+ 10.29304025 21.22065908 13.04968473 { Row 2 of [L]
+ 2.172649731 3.327350269 5.000000000 { Row 3 of [R]
+ 13.04968473 10.29304025 21.22065908 { Row 3 of [L]
+C Finally, data for 3rd of 3 line sections comes from Orlando. See preceding
+C subcase for comments about it. Z1 and Z2 differ while the sum is fixed.
+C So, the fault current will be the same, but not the fault voltage, note.
+51OPHA MODEL [R][L]
+52GENB OPEN5
+53GENC OPEN6
+ 5.0000000 -3.0235027 8.5235027
+ 21.2206591 11.8091947 11.5335303
+ 8.5235027 5.0000000 -3.0235027
+ 11.5335303 21.2206591 11.8091947
+ -3.0235027 8.5235027 5.0000000
+ 11.8091947 11.5335303 21.2206591
+C 4th of 4 line sections follows. MH at sending end indicates L in mHenries.
+C This 4th alternative is the same as the first except X in ohms has been
+C converted to L in mHenries by multiplying by 10 / Pi = 3.18309886184 This
+C gives Lo, L1, L2 = 44.5633840657 12.7323954474 6.36619772368
+$VINTAGE, 1, { Switch to wide format, with R and L read as 2E16.0
+C 3456789012345678901234567890123456789012345678901234567890
+C RRRRRRRRRRRRRRRRLLLLLLLLLLLLLLLL
+51MHA MODEL Z0Z1Z2 10.5 44.5633840657 { Ro in ohms and Lo in mH
+52GENB OPEN7 3.0 12.7323954474 { R1 in ohms and L1 in mH
+53GENC OPEN8 1.5 6.36619772368 { R2 in ohms and L2 in mH
+$VINTAGE, 0, { Done with wide format; return to old (narrow) format
+BLANK card ending branches
+C Following switches measure line currents in the 3 lines. Since phase "a"
+C is the only one connected, ignore switches for phases "b" and "c":
+ GENA SEQA MEASURING 1
+ GENA PHSA MEASURING 1
+ GENA OPHA MEASURING 1
+ GENA MHA MEASURING 1
+BLANK card ending switches
+14GENA 1000. 50. -0.00 { Balanced, 3-phase excitation that
+14GENB 1000. 50. -120. { is _not_ present during the steady
+14GENC 1000. 50. -240. { state. There is no phasor solution.
+BLANK card follows the last source card
+C OPEN1 OPEN3 OPEN5 OPEN7 { Output fault voltages (unfaulted phase b, anyway)
+C Note. Activate the preceding comment card to demonstrate that the solution
+C of the OPH circuit really is different from the other 3. I.e., the voltage
+C at OPEN5 will differ from the voltages at OPEN1, OPEN3, and OPEN7.
+C The fault currents are identical, but fault voltages will differ. This is
+C because the sum of Z1 and Z2 agrees, but neither Z1 nor Z2 does.
+C Next 4 output variables are branch currents (flowing from the upper node to the lower node);
+C Step Time GENA GENA GENA GENA
+C SEQA PHSA OPHA MHA
+C 0 0.0 0.0 0.0 0.0 0.0
+C 1 .1E-3 2.32761034 2.32761034 2.32761033 2.32761034
+C 2 .2E-3 6.92518039 6.92518039 6.92518039 6.92518039
+C 3 .3E-3 11.4064981 11.4064981 11.4064981 11.4064981
+BLANK card ending node voltage outputs
+C 600 .06 71.992366 71.992366 71.9923659 71.992366
+C 700 .07 -71.992425 -71.992425 -71.992425 -71.992425
+C 800 .08 71.9924194 71.9924194 71.9924193 71.9924194
+C 900 .09 -71.99242 -71.99242 -71.99242 -71.99242
+C 1000 0.1 71.9924199 71.9924199 71.9924198 71.9924199
+C Variable maxima: 119.979925 119.979925 119.979925 119.979925
+C Times of maxima: .083 .083 .083 .083
+C Variable minima: -119.97993 -119.97993 -119.97993 -119.97993
+C Times of minima: .093 .093 .093 .093
+C Note about preceding extema. The maximum of 119.979925 and minimum of
+C -119.97993 are nearly negatives of each other. This is a sign that the
+C steady state is really steady (equality in the 8th decimal digit). But
+C Orlando Hevia's exact solution of 120.0000 indicates some error. Why?
+C The substantial dT (discretization error). As dT is decreased, the ATP
+C answer approaches Orlando Hevia's. For all dT, all 4 maxima agree. Thus
+C only a single value need be shown, as a function time-step size dT:
+C 100 usec ===> 119.979925
+C 50 usec ===> 119.998403
+C 20 usec ===> 119.999337
+C 10 usec ===> 119.999920
+C 5 usec ===> 119.999967
+C 2 usec ===> 119.999997
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+C 7th of 7 subcases is the same as the 5th except that the sources have
+C been activated during the steady state. This subcase has a phasor
+C solution --- now possible as this subcase is added on 5 August 2008.
+C Note that the final time steps of the two subcases are identical.
+C BEGIN PEAK VALUE SEARCH has been removed since there no longer is any
+C need for allowing the solution to "settle" into the steady-state. In
+C place of 5 cycles, simulate just 2 --- plenty to demonstrate sinusoidal
+C signals (see addition of 2 plot cards after CALCOMP PLOT).
+PRINTED NUMBER WIDTH, 11, 2, { Request maximum precision (for 8 output columns)
+ .000100 .040 { dT is relatively large to speed simulation. Only 400 steps
+ 1 1 1 1 1 -1
+ 5 5 20 20 100 100
+$UNITS, 50.0, 0.0 { Begin inductance data in ohms at XOPT = 50 Hz
+C 1st of 3 line sections follows. SEQ at sending end indicates sequence data:
+51GENA OPEN1 MODEL Z0Z1Z2 10.5 14.00 { Ro and Xo, both in ohms
+52SEQB FLT1 3.0 4.00 { R1 and X1, both in ohms
+53SEQC FLT1 1.5 2.00 { R2 and X2, both in ohms
+$UNITS, 0.0, 0.0 { Return to inductance data in mHenry; XOPT = 0
+51GENA OPEN2 MODEL [R][L]
+52PHSB FLT2
+53PHSC FLT2
+ 5.000000000 2.172649731 3.327350269 { R(1,1), R(1,2), and R(1,3) in ohms
+ 21.22065908 13.04968473 10.29304025 { L(1,1), L(1,2), and L(1,3) in mHenries
+ 3.327350269 5.000000000 2.172649731 { Row 2 of [R]
+ 10.29304025 21.22065908 13.04968473 { Row 2 of [L]
+ 2.172649731 3.327350269 5.000000000 { Row 3 of [R]
+ 13.04968473 10.29304025 21.22065908 { Row 3 of [L]
+51GENA OPEN3 MODEL [R][L]
+52OPHB FLT3
+53OPHC FLT3
+ 5.0000000 -3.0235027 8.5235027
+ 21.2206591 11.8091947 11.5335303
+ 8.5235027 5.0000000 -3.0235027
+ 11.5335303 21.2206591 11.8091947
+ -3.0235027 8.5235027 5.0000000
+ 11.8091947 11.5335303 21.2206591
+BLANK card ending branches
+ GENB SEQB MEASURING 1
+ GENB PHSB MEASURING 1
+ GENB OPHB MEASURING 1
+ GENC SEQC MEASURING 1
+ GENC PHSC MEASURING 1
+ GENC OPHC MEASURING 1
+BLANK card ending switches
+14GENA 1000. 50. -0.00 -1.
+14GENB 1000. 50. -120. -1.
+14GENC 1000. 50. -240. -1.
+BLANK card follows the last source card
+ FLT1 FLT2 FLT3 { Show that the fault voltage of OPH (FLT3; 3rd of 3) differs
+C Column headings for the 9 EMTP output variables follow. These are divided among the 5 possible classes as follows ....
+C First 3 output variables are electric-network voltage differences (upper voltage minus lower voltage);
+C Next 6 output variables are branch currents (flowing from the upper node to the lower node);
+C Step Time FLT1 FLT2 FLT3 GENB GENB GENB GENC GENC GENC
+C SEQB PHSB OPHB SEQC PHSC OPHC
+C *** Phasor I(0) = -1.8475209E+02 Switch "GENB " to "SEQB " closed in the steady-state.
+C *** Phasor I(0) = -1.8475209E+02 Switch "GENB " to "PHSB " closed in the steady-state.
+C *** Phasor I(0) = -1.8475209E+02 Switch "GENB " to "OPHB " closed in the steady-state.
+C *** Phasor I(0) = 1.8475209E+02 Switch "GENC " to "SEQC " closed in the steady-state.
+C *** Phasor I(0) = 1.8475209E+02 Switch "GENC " to "PHSC " closed in the steady-state.
+C *** Phasor I(0) = 1.8475209E+02 Switch "GENC " to "OPHC " closed in the steady-state.
+C 0 0.0 -333.3333 -333.3333 572.66667 -184.7521 -184.7521 -184.7521 184.75209 184.75209 184.75209
+C 1 .1E-3 -333.1685 -333.1685 547.50887 -180.3089 -180.3089 -180.3089 180.30888 180.30888 180.30888
+C 2 .2E-3 -332.6749 -332.6749 521.81078 -175.6877 -175.6877 -175.6877 175.68774 175.68774 175.68774
+C 3 .3E-3 -331.853 -331.853 495.59776 -170.8932 -170.8932 -170.8932 170.89322 170.89322 170.89322
+BLANK card ending node voltage outputs
+C 400 .04 -333.3372 -333.3372 572.64196 -184.7478 -184.7478 -184.7478 184.74783 184.74783 184.74783
+C Variable maxima : 333.33754 333.33754 977.26389 230.90147 230.90147 230.90147 230.90154 230.90153 230.90153
+C Times of maxima : .01 .01 .017 .028 .028 .028 .018 .018 .018
+C Variable minima : -333.3372 -333.3372 -977.2634 -230.9015 -230.9015 -230.9015 -230.9015 -230.9015 -230.9015
+C Times of minima : .04 .04 .027 .018 .018 .018 .028 .028 .028
+ CALCOMP PLOT { Display some waveforms to show that all is smooth & sinusoidal
+ 144 4. 0.0 40.-1.E31.E3FLT1 FLT2 FLT3
+ 194 4. 0.0 40.-250.250.GENB SEQB GENC PHSC
+BLANK termination to plot cards
+BEGIN NEW DATA CASE
+BLANK
+$EOF
+
+
+ In order that it not be lost, let's append Orlando Hevia's FORTRAN
+program. This confirms the preceding 5th and 6th subcases as well as
+other fault types not illustrated by preceding ATP data:
+ COMPLEX A,A2,Z0,Z1,Z2,AIR,AIS,AIT,E,J
+ PI=4.0*ATAN(1.0)
+ A= CMPLX(-0.5,SQRT(3.0)/2.0)
+ A2=CMPLX(-0.5,-SQRT(3.0)/2.0)
+ J=CSQRT(CMPLX(-1.0,0.0))
+ Z0=CMPLX(10.5,14.0)
+ Z1=CMPLX(3.0,4.0)
+ Z2=CMPLX(1.5,2.0)
+ E=CMPLX(1000.0,0.0)
+ AIS= -J*SQRT(3.0)*E*(((1.0,0.0)+A2)*Z2+Z0)/
+ 1 (Z1*Z2+Z2*Z0+Z0*Z1)
+ AIT= J*SQRT(3.0)*E*(((1.0,0.0)+A )*Z2+Z0)/
+ 1 (Z1*Z2+Z2*Z0+Z0*Z1)
+ AIR=AIS+AIT
+ AS=CABS(AIS)
+ AT=CABS(AIT)
+ AR=CABS(AIR)
+ FS=ATAN2(AIMAG(AIS),REAL(AIS))*180.0/PI
+ FT=ATAN2(AIMAG(AIT),REAL(AIT))*180.0/PI
+ FR=ATAN2(AIMAG(AIR),REAL(AIR))*180.0/PI
+C WRITE(*,*)Z0
+C WRITE(*,*)Z1
+C WRITE(*,*)Z2
+C WRITE(*,*)A,A2,J
+ WRITE(*,*)'PHASE-PHASE-GROUND FAULT'
+100 FORMAT('IB ',2F12.4)
+101 FORMAT('IC ',2F12.4)
+102 FORMAT('IRES ',2F12.4)
+ WRITE(*,*)'RECTANGULAR'
+ WRITE(*,100)AIS
+ WRITE(*,101)AIT
+ WRITE(*,102)AIR
+ WRITE(*,*)'POLAR'
+ WRITE(*,100)AS,FS
+ WRITE(*,101)AT,FT
+ WRITE(*,102)AR,FR
+C
+ AIS= -J*SQRT(3.0)*E/(Z1+Z2)
+ AIT= J*SQRT(3.0)*E/(Z1+Z2)
+ AIR=AIS+AIT
+ AS=CABS(AIS)
+ AR=CABS(AIR)
+ AT=CABS(AIT)
+ FS=ATAN2(AIMAG(AIS),REAL(AIS))*180.0/PI
+ FT=ATAN2(AIMAG(AIT),REAL(AIT))*180.0/PI
+ FR=ATAN2(AIMAG(AIR),REAL(AIR))*180.0/PI
+ WRITE(*,*)' '
+ WRITE(*,*)'PHASE-PHASE FAULT'
+ WRITE(*,*)'RECTANGULAR'
+ WRITE(*,100)AIS
+ WRITE(*,101)AIT
+ WRITE(*,102)AIR
+ WRITE(*,*)'POLAR'
+ WRITE(*,100)AS,FS
+ WRITE(*,101)AT,FT
+ WRITE(*,102)AR,FR
+C
+ AIR= E/Z1
+ AR=CABS(AIR)
+ FR=ATAN2(AIMAG(AIR),REAL(AIR))*180.0/PI
+C
+ WRITE(*,*)' '
+ WRITE(*,*)'PHASE-PHASE-PHASE-GROUND FAULT'
+ WRITE(*,*)'RECTANGULAR'
+ WRITE(*,202)AIR
+ WRITE(*,*)'POLAR'
+ WRITE(*,202)AR,FR
+202 FORMAT('IA ',2F12.4)
+C
+ AIR=3.0*E/(Z0+Z1+Z2)
+ AR=CABS(AIR)
+ FR=ATAN2(AIMAG(AIR),REAL(AIR))*180.0/PI
+C
+ WRITE(*,*)' '
+ WRITE(*,*)'PHASE-GROUND FAULT'
+ WRITE(*,*)'RECTANGULAR'
+ WRITE(*,302)AIR
+302 FORMAT('IA ',2F12.4)
+ WRITE(*,*)'POLAR'
+ WRITE(*,302)AR,FR
+ STOP
+ END
+
+Output, as produced using GNU FORTRAN compilation, linking, and execution,
+is as follows:
+
+ PHASE-PHASE-GROUND FAULT
+ RECTANGULAR
+IB -196.3879 -114.6822
+IC 165.0836 156.4214
+IRES -31.3043 41.7391
+ POLAR
+IB 227.4208 -149.7169
+IC 227.4208 43.4567
+IRES 52.1739 126.8699
+
+ PHASE-PHASE FAULT
+ RECTANGULAR
+IB -184.7521 -138.5641
+IC 184.7521 138.5641
+IRES 0.0000 0.0000
+ POLAR
+IB 230.9401 -143.1301
+IC 230.9401 36.8699
+IRES 0.0000 36.8699
+
+ PHASE-PHASE-PHASE-GROUND FAULT
+ RECTANGULAR
+IA 120.0000 -160.0000
+ POLAR
+IA 200.0000 -53.1301
+
+ PHASE-GROUND FAULT
+ RECTANGULAR
+IA 72.0000 -96.0000
+ POLAR
+IA 120.0000 -53.1301
|