From b18347ffc9db9641e215995edea1c04c363b2bdf Mon Sep 17 00:00:00 2001 From: Angelo Rossi Date: Wed, 21 Jun 2023 12:04:16 +0000 Subject: Initial commit. --- benchmarks/dcn22.dat | 720 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 720 insertions(+) create mode 100644 benchmarks/dcn22.dat (limited to 'benchmarks/dcn22.dat') 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 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 -- cgit v1.2.3