BEGIN NEW DATA CASE C BENCHMARK DC-37 C Trivial test of ZnO modeling. Single phase, single exponential, no gap. C For documentation, see the EMTP Newsletter, Vol. 1, No. 2, pages 6-9. C i = 2500 * ( v / V-ref ) ** 26 where 2500 = COEF and 26 = EXPON. C The lack of a gap follows from V-flash being an arbitary negative value. C Note 2-column arrester (COL=2.0), and twice COEF = 1250 is 2500 total. C The idea of using COL came from Dan Durbak of PTI (Schenectady, New C York) at the end of an unrelated telephone call to BPA on July 3, 1986. C Finally, a 2nd ZnO arrester is applied right across the voltage source. C Of a total of 12 subcases, several illustrate power and energy output. PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) C MAXZNO EPSZNO EPWARN EPSTOP ZNOLIM1 ZNOLIM2 ZINC OXIDE 20 1.D-8 1.D-3 0.1 0.6 1.5 default values C ZO, , , , .9, , { To improve ZnO convergence, restrict the Newton ZnO correction .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 -1SEND REC .306 5.82 .012 200. 92REC { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 C VREF VFLASH VZERO COL 778000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) C The following arrester is applied right across the voltage source, so it is C disconnected, and requires no iteration. Newton iteration is 1-dimensional C (the matrix is 1 x 1 only). The second arrester involves "table lookup." C Note that V-ref is equal to the peak source voltage, so the peak current C is equal to the coefficient: COL * COEF = 2500 amps. The plot agrees. 92SEND { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 C VREF VFLASH VZERO COL 408000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14SEND 408000. 60. C --------------+------------------------------ C From bus name | Names of all adjacent busses. C --------------+------------------------------ C SEND |TERRA *REC * C REC |TERRA *SEND * C TERRA |SEND *REC * C --------------+------------------------------ BLANK card ending source data C Step Time REC SEND REC SEND C TERRA TERRA C 0 0.0 0.0 0.0 0.0 0.0 C 1 .5E-4 0.0 407927.5198 0.0 2488.47851 C 2 .1E-3 0.0 407710.1048 0.0 2454.223589 C 12 .6E-3 0.0 397606.9641 0.0 1278.142608 C 22 .0011 0.0 373418.3984 0.0 249.9564103 C 32 .0016 0.0 336001.2998 0.0 16.05612212 C 33 .00165 0.0 331579.2191 0.0 11.37746957 C 34 .0017 446716.5798 327039.3298 .0013592657 7.950150616 1 { Request for all node voltage outputs C Final step : 400 .02 53295.59263 126078.9337 .5103896E-5 .2302356E-4 C Variable maxima : 693010.7427 407991.9464 123.5121341 2498.717272 C Times of maxima : .00175 .01665 .00175 .01665 C Variable minima : -711052.757 -407991.946 -240.945005 -2498.71727 C Times of minima : .00765 .00835 .00765 .00835 PRINTER PLOT 144 3. 0.0 20. REC { Axis limits: (-7.111, 6.930) 194 2. 0.0 20. SEND { Axis limits: (-2.499, 2.499) CALCOMP PLOT C The following plot card illustrates automatic plotting from zero through C the end time TMAX of the study. Columns 5-7 give the t-axis length in C inches, and columns 12-15 being negative is the flag to plot all time. C With a 10-inch axis and a time span of 20 msec, the result is 2 msec/in. C For such all-time character plotting in units of [seconds], see DC-6. C 78901234567890123456789012345678901234567890123456789012345678901234567890 19410. -1. REC 16-Char. HeadingVertical axis Y: $STARTUP, dc37star.dat ! { Use disk file for re-initialization immediately C This will halve the number of pixels/inch (PIXPUN) for an Apollo screen C plot. Also, it will set the smoothing tolerance squared, TOLRCE, to C 1/10. The result will be a half-size and very bumpy screen plot: 194 1. 0.0 10. REC 16-byte Heading Y-axis labeling. BLANK card ending plot cards BEGIN NEW DATA CASE C 2nd of 12 subcases. Same basic network as just solved, only with modified C ZnO characteristic as derived by DC-39. One exponential with flashover. C The first (pre-flash) characteristic has near-infinite resistance, with C exponent equal to unity. Leakage current (about 1.E-20 amps) will occur C prior to flashover at v = VREF = 778 kV. The operational characteristic C is i = 29479.54 * ( v / V-ref ) ** 26.53, which is very close to the C original characteristic of 1st subcase. Change on 23 Jul 1984 ("M39.+"). C Enhancement beginning 30 July 1986. Series voltage sources for all C nonlinear elements are allowed, so let's take the simplest possible C case of a 10-KVolt battery (variable "BATTER" defined within TACS). C Before the pulse arrives, this only draws leakage current from the C linear representation of the surge arrester (R-leakage). But when C the 10 kV are added to the surge, the answer changes significantly. PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 TACS HYBRID { We use TACS only to produce series voltage "BATTER" of ZnO 99BATTER = 10000. { Small battery (dc source) is inserted in series with ZnO 33BATTER { Output the only this one TACS variable that controls ZnO source 77BATTER 10000. { Initial condition required for smooth electrical step 1 BLANK card ending all TACS data -1SEND REC .306 5.82 .012 200. 1 92REC 5555. 1 C ============================================================================= C 92REC TYP11 5555. 1 C The preceding comment card is just for verification of solution. See the C explanation on comment cards below the blank card ending switch cards. C ============================================================================= C VREF VFLASH VZERO COL 0.778000000000000E+06 1.0 C COEF EXPON VMIN 0.294795442961157E-20 1.0 .900000E+01 9999 { Bound on exponentials of 1st, pre-flash v-i curve 0.294795442961157E+05 0.265302624185338E+02 0.545050636122854E+00 9999 { Bound on exponentials of 2nd, post-flash v-i curve TACS CONTROLBATTER { Only 1st of three A6 names, for series voltage, is used BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches C ============================================================================= C The easiest way to verify correct operation of the series voltage BATTER C of the Type-92 element is to cancel it out using an electrical network C battery. The following Type-11 source will do the job. But then the C nonlinear element must have the second name changed to "TYP11", note. C 11TYP11 -10000. { This battery cancels series voltage of ZnO "BATTER" C ============================================================================= 14SEND 408000. 60. BLANK card ending source data C The following beginning shows leakage current between steps 1 and 33 when C only the TACS voltage "BATTER" is exciting the arrester. Since the line C impedance (Z-thev) is so much smaller than the ZnO, the receiving voltage C REC is very, very small. But it is nonzero, note --- and correctly so: C First 3 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 1 output variables are branch currents (flowing from the upper node to the lower node); C Next 1 output variables belong to TACS (with "TACS" an internally-added upper name of pair). Updated 2 Nov 00 C Step Time SEND REC SEND REC TACS C REC TERRA BATTER C 0 0.0 0.0 0.0 0.0 0.0 10000. C 1 .5E-4 407927.5198 .269681E-19 407927.5198 -.37891E-22 10000. C 2 .1E-3 407710.1048 .269681E-19 407710.1048 -.37891E-22 10000. C 12 .6E-3 397606.9641 .269681E-19 397606.9641 -.37891E-22 10000. C 22 .0011 373418.3984 .269681E-19 373418.3984 -.37891E-22 10000. C 32 .0016 336001.2998 .269681E-19 336001.2998 -.37891E-22 10000. C 33 .00165 331579.2191 .269681E-19 331579.2191 -.37891E-22 10000. C 34 .0017 -119678.217 446717.5472 327039.3298 .165479E-20 10000. 1 { Request for all node voltage outputs C 400 .02 8093.382341 117985.5514 126078.9337 .7639597E-3 10000. C Variable maxima : 474201.1091 793085.1587 407991.9464 225.9075813 10000. C Times of maxima : .0063 .0159 .01665 .0162 0.0 C Variable minima : -511598.378 -768087.191 -407991.946 -333.924608 10000. C Times of minima : .00295 .0063 .00835 .00765 0.0 PRINTER PLOT 194 3. 0.0 20. REC { Axis limits: (-3.339, 2.259) CALCOMP PLOT 194 2. 0.0 20. REC 16-Char HeadingVertical axis Y. BLANK card ending plot cards BEGIN NEW DATA CASE C 3rd of 12 subcases. Same basic network as just solved, only with modified C ZnO characteristic as derived by DC-39. There is a 2-exponential curve, C the pre-flashover curve of 2nd subcase. V-flash < 0 means no gap here. $STARTUP, (ATPDIR)startup ! { Re-initialize with original disk file of 1st subcase C This will restore the default parameters, after illustrating (see the C 2nd subcase) what happens in the absence of such restoration for later C stacked subcases. The second subcase continued with the mini CalComp C plots that began at the end of the 1st subcase with dc37star.dat. C Note about "(ATPDIR)" as used on preceding $STARTUP. This is C optional. If it appears, ATP will replace these 8 bytes by the C content of symbol ATPDIR (as used at BPA, "C:\ATP"). PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) ZINC OXIDE 20 .000050 .02000 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 -1SEND REC .306 5.82 .012 200. 92REC 5555. 1 C VREF VFLASH VZERO COL 0.778000000000000E+06 -1.0 1.0 C COEF EXPON VMIN 0.505584788677197E+07 0.464199973324622E+02 0.632754084797274E+00 0.122767153039007E+05 0.166775903445228E+02 0.816748018907843E+00 9999 { Bound on two exponentials of single v-i curve BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14SEND 408000. 60. BLANK card ending source data C Step Time REC SEND REC C TERRA C 0 0.0 0.0 0.0 0.0 C 1 .5E-4 0.0 407927.5198 0.0 C 2 .1E-3 0.0 407710.1048 0.0 C 12 .6E-3 0.0 397606.9641 0.0 C 22 .0011 0.0 373418.3984 0.0 C 32 .0016 0.0 336001.2998 0.0 C 33 .00165 0.0 331579.2191 0.0 C 34 .0017 446715.6097 327039.3298 .0027223116 1 { Request for all node voltage outputs C Final step printout: 400 .02 104698.8727 126078.9337 .6380412E-3 C Variable maxima : 626465.3591 407991.9464 217.0115979 C Times of maxima : .00175 .01665 .00175 PRINTER PLOT 144 3. 0.0 20. REC { Axis limits: (-6.329, 6.265) CALCOMP PLOT 194 2. 0.0 20. REC BLANK card ending plot cards BEGIN NEW DATA CASE C 4th of 12 subcases C Modify 1st subcase of DC-37 to illustrate scaling of voltage and current C by 1000 for ZnO surge arrester. C 1 October 2000, add illustrations of 3-digit exponents: numbers C larger than 1.E+100 and smaller than 1.E-100. It was necessary C to find a subcase without plotting, since otherwise the limit of C FLTINF (set in STARTUP to 1.E19) would be imposed. To avoid this C limit, we set IPLOT = -1 below (columns 9-16). Node OVER is to C demonstrate near overflow of the Intel limit of around 1.E+308, and C node UNDER is to demonstrate near underflow of the limit 1.E-308 C Each involves series R-L with R = 12.7 and L = .45 mH (or the C negative of this for overflow), which implies a time constant of C Tau = L / R = .45E-3 / 12.7 = 3.54E-5 sec. The simulation lasts for C 20 msec, so EXP ( T-max / Tau ) = EXP ( .020 / 3.54E-5 ) = C EXP ( 564.44 ) = 1.365E+245. This 245 is close to the 303 observed C (close enough for engineers to understand the physics involved). C A final detail concerns the .DBG file. For Salford EMTP, this C is a separate file as long as output is buffered (typically LU6VRT C has value 32768 for an output buffer of size 32 Kbytes). As a C separate file, the user might never look to see the message that C is produced for each optimal encoding that requires a 3-digit C exponent. For example, the first two occur to produce the output C of step 150. The associated diagnostic becomes highly visible if C output is not buffered (if LU6VRT = 0). Then, from DC37.LIS : C 100 .005 .316184E-12 0.0 0.0 .577391E-22 -.887674E75 .267212E-74 C FLTOPT. Wierd number D9 = 1.234180423134E+0113 SPYCD2(1:35) = 0.1234180423133828610000+114 C FLTOPT. Wierd number D9 = 1.921900653658E-0113 SPYCD2(1:35) = 0.1921900653658086780000-112 C 150 .0075 -288.499567 0.0 0.0 -.305176E-3 -.12342E114 .19219E-112 C There is nothing wrong. The number is wierd only in the sense C that Salford omitted the "E" during encoding, so ATP logic was C forced to restore the "E" manually. C 14 January 2011, add illustration of the new protection against C overflow of node voltage. DC-61 illustrated use of the new optional C second parameter LOGB10 on the PEAK VOLTAGE MONITOR declaration. C But the protection there was invisible since there never was any need C for it. Here, there is need. With a node voltage limit of 1.E+304, C execution will be terminated with the KILL = 264 error message: C KILL = 264. ATP halts execution because some node voltage has C exceeded the bound of 1.00000000+304 as defined by the optional C second parameter of the user`s NODE VOLTAGE MONITOR request. This C is at node "MIDO " on step number 399. WSM. C Ruler for next card: MAXVLT HALTNV { Format is: ( 32X, I8, E8.0 ) PEAK VOLTAGE MONITOR 1 1.E304 { Peak node voltage = 1.E+304 VOLTAGE SOURCES IN KV PRINTED NUMBER WIDTH, 10, 2, { Request minimum precision, since explosive .000050 .020 1 -1 1 0 1 -1 0 5 10 33 1 40 10 100 50 395 1 92VOLT 5555. 1 C VREF VFLASH VZERO COL 408000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) VOLT 1.E8 { Avoid warning to weak connection to ground C Network modification. We must split the unstable series R-L into two C separate, series branches. This is to add a node at which the voltage C will increase without bound (exponential growth). The old data escaped our C new protection because terminal node voltages were identically zero (it was C only the branch current that grew uncontrolled): C OVER 12.7 -.45 { Series R-L with negative L } 1 OVER MIDO 12.7 { Series R with positive R } MIDO -.45 { Series L with negative L } 1 UNDER 12.7 .45 { Series R-L with positive L } 1 BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14VOLT 408. 50. 14OVER 1.0 50. { Voltage source is shorted at .2 ms } 2.E-4 14UNDER 1.0 50. { Voltage source is shorted at .2 ms } 2.E-4 BLANK card ending source data C First 4 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 3 output variables are branch currents (flowing from the upper node to the lower node); C Step Time MIDO VOLT OVER UNDER VOLT MIDO UNDER C TERRA TERRA TERRA C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 C 1 .5E-4 3.395807 407.9497 .9998766 .9998766 2.491993 -.188656 .0325693 C 2 .1E-3 19.6688 407.7987 .9995066 .9995066 2.468123 -1.47002 .0707492 C 3 .15E-3 113.9285 407.5471 .9988899 .9988899 2.428834 -8.89209 .0773084 C 4 .2E-3 656.5329 407.1949 0.0 0.0 2.374852 -51.6955 .0458835 C 5 .25E-3 3802.936 406.7423 0.0 0.0 2.307161 -299.444 .0079213 1 { Request for output of all node voltages C 395 .01975 .125E302 406.7423 0.0 0.0 2.307161 -.98E300 .24E-299 C 396 .0198 .723E302 407.1949 0.0 0.0 2.374852 -.57E301 .42E-300 C 397 .01985 .419E303 407.5471 0.0 0.0 2.428834 -.33E302 .72E-301 C 398 .0199 .243E304 407.7987 0.0 0.0 2.468123 -.19E303 .12E-301 BLANK card ending plot cards BEGIN NEW DATA CASE C 5th of 12 subcases C Modify 2nd subcase of DC-37 to illustrate scaling of voltage and current C by 1000 for ZnO surge arrester. Note that VOLTAGE SOURCES IN KV is C not needed because the declaration of the preceding subcase remains in C effect for this one. PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 -1GEN VOLT .306 5.82 .012 200. 92VOLT 5555. 1 C VREF VFLASH VZERO COL 0.778000000000000E+06 1.0 C COEF EXPON VMIN 0.294795442961157E-20 1.0 .900000E+01 9999 { Bound on exponentials of 1st, pre-flash v-i curve 0.294795442961157E+05 0.265302624185338E+02 0.545050636122854E+00 9999 { Bound on exponentials of 2nd, post-flash v-i curve BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14GEN 408. 50. BLANK card ending source data 1 { Request for all node voltage outputs BLANK card ending plot cards BEGIN NEW DATA CASE C 6th of 12 subcases C Modify 4th subcase of DC-37 to illustrate scaling of voltage and current C by 1000. Note data (specifically, source voltage) is unscaled. Instead C of VOLTAGE SOURCES IN KV, we here use output scaling of BVIV and C BCIA (following two declarations). Output of the time-step loop is C identical to that of 4th subcase. See October, 1997, newsletter. C 28 December 1998, add first column-80 punch in excess of 4. A story C in the April, 1999, newsletter should introduce use of 0 plus the 16 C choices that are summarized immediately before the blank card that C ends branch cards. Details of this mapping remain an ATP secret. C 34567890123456789012345678901234567890 V-base is read from cols. 33-40: BASE VOLTAGE IN VOLTS 1000. BASE CURRENT IN AMPERES 1000. { I-base is read from cols. 33-40 VOLTAGE SOURCES IN KV -1.0 { Cancel usage of preceding subcases PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 C 28 Dec 98, original col-80 punch of 1 is changed to B to add power output: 92VOLT 5555. B C VREF VFLASH VZERO COL 408000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) C Documentation of new column-80 choices that begin 28 December 1998: C 5 ==> Append power & energy while leaving voltage & current C 6 ==> Append power & energy while omitting current (but not the voltage) C 7 ==> Append power & energy while omitting voltage (but not the current) C 8 ==> Append power & energy while omitting both voltage and current C 9 ==> Append power while leaving voltage & current C A ==> Append power while omitting current (but not the voltage) C B ==> Append power while omitting voltage (but not the current) C C ==> Append power while omitting both voltage and current C D ==> Append energy while leaving voltage & current C E ==> Append energy while omitting current (but not the voltage) C F ==> Append energy while omitting voltage (but not the current) C G ==> Append energy while omitting both voltage and current BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14VOLT 408000. 50. BLANK card ending source data C Column headings for the 3 EMTP output variables follow. These are divided among the 5 possible classes as follows .... C First 1 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 1 output variables are branch currents (flowing from the upper node to the lower node); C Next 1 output variables are either power or energy or both as a pair (column 80 punches > 4). C Step Time VOLT VOLT VOLT C TERRA TERRA C 0 0.0 0.0 0.0 0.0 C 1 .5E-4 407.9496661 2.491993465 1016.607902 C 2 .1E-3 407.7986766 2.468123475 1006.497487 1 { Request for all node voltage outputs C 200 .01 -408. -2.5 1020. C 250 .0125 -288.499567 -.305176E-3 .0880430807 C 300 .015 -.80639E-11 -.14726E-20 .118745E-31 C 350 .0175 288.4995667 .3051758E-3 .0880430807 C 400 .02 408. 2.5 1020. C Variable maxima : 408. 2.5 1020. C Times of maxima : .02 .02 .01 C Variable minima : -408. -2.5 0.0 C Times of minima : .01 .01 0.0 BLANK card ending plot cards BEGIN NEW DATA CASE C 7th of 12 subcases illustrates power & energy output of nonlinear element. C Apparently this was never documented as Orlando Hevia found it to be in C error during late May of 1998, following overhaul of branch data input. C KISS: 2 volts across R = 2 gives 1 amp. Then, nonlinear has 1 ohm and C one volt, so power is 1 on 1st step. Trapezoidal rule then gives the C energy as dT ( E(0) + E(1) ) / 2 = .001 * 1 / 2 = .0005 (about). PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) .001 .001 1 1 1 0 0 0 0 0 SEND REC 1.0 { Half of 2 ohms total is this linear branch 92REC 4444. { ZnO is piecewise-linear } 4 C VREF VFLASH VZERO 0.0 -1.0 0.0 1.0 1.0 { First point of i-v curve. 10. 10. 9999. { Terminator for piecewise-linear characteristic BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14SEND 2.0 1.0 { 1 Hz is close enough to dc for this test BLANK card ending source data C Step Time REC REC SEND REC C TERRA TERRA C 0 0.0 0.0 0.0 0.0 0.0 C 1 .1E-2 .9999605221 .9999802609 1.999960522 .4999803E-3 1 { Request for all node voltage outputs BLANK card ending plot cards BEGIN NEW DATA CASE C 8th of 12 subcases is added 13 October 2006 to demonstrate operation of C extended power and energy output (punches 5-16) for switches. Of course C the same meanings apply to switches as to branches -- with the exception C that power & energy are flow _through_ a switch rather than consumption C within a branch. Data for this illustration of extended switch outputs C is copied from the 6th subcase. One switch is added in series with the C N.L. element. This is (GEN, VOLT). The current, power and energy flow C through it obviously should be identical to those of the branch. Since C the branch has "B" in column 80, only current and power will be shown. C The switch has "7" to add energy to these two. Since the 2 columns of C current are side by side, and the 2 columns of power are side by side, C it is possible to see by casual inspection that they agree. As for the C energy, it seems believable; printed values are monotone increasing. C 34567890123456789012345678901234567890 V-base is read from cols. 33-40: BASE VOLTAGE IN VOLTS 1000. BASE CURRENT IN AMPERES 1000. { I-base is read from cols. 33-40 VOLTAGE SOURCES IN KV -1.0 { Cancel usage of preceding subcases PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 92VOLT 5555. B C VREF VFLASH VZERO COL 408000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) BLANK card terminating branch data GEN VOLT MEASURING 7 BLANK card terminating all switches 14GEN 408000. 50. BLANK card ending source data GEN { List of bus names for node voltage output C Column headings for the 6 EMTP output variables follow. These are divided among the 5 possible classes as follows .... C First 1 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 2 output variables are branch currents (flowing from the upper node to the lower node); C Next 3 output variables are either power or energy or both as a pair (column 80 punches > 4). C Step Time GEN GEN VOLT VOLT GEN GEN C VOLT TERRA TERRA VOLT VOLT C *** Switch "GEN " to "VOLT " closed before 0.00000000E+00 sec. C 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 C 1 .5E-4 407.9496661 2.491993465 2.491993465 1016.607902 1016.607902 .0254151975 C 2 .1E-3 407.7986766 2.468123475 2.468123475 1006.497487 1006.497487 .0759928323 BLANK card ending node voltage outputs C 400 .02 408. 2.5 2.5 1020. 1020. 3.078116283 C Variable maxima : 408. 2.5 2.5 1020. 1020. 3.078116283 C Times of maxima : .02 .02 .02 .01 .01 .02 C Variable minima : -408. -2.5 -2.5 0.0 0.0 0.0 C Times of minima : .01 .01 .01 0.0 0.0 0.0 BLANK card ending plot cards BEGIN NEW DATA CASE C 9th of 12 subcases will illustrate power and energy output along with a C phasor solution. Since previous subcases all involved one nonlinear C element, make this linear and easy to verify by hand. Add a switch as C was done for the preceding subcase. Illustrate 3 column-80 punches > 4. C Note that the 8-punch results in the same outputs that a 4-punch would, C but they are located differently : power and energy are appended rather C than overlay the corresponding voltage and current outputs. As for the C switch, it is permanently closed in spite of T-open = 19 msec < T-max. C Finally, turn off V and I scaling. Date of addition: 13 October 2006 BASE VOLTAGE IN VOLTS 1.0 { Cancel usage of preceding subcase BASE CURRENT IN AMPERES 1.0 { Cancel usage of preceding subcase .000500 .020 1 1 1 2 1 -1 5 5 GEN MID 0.5 { power and energy output } 8 LOAD 0.5 { power, energy, and voltage output } 6 BLANK card terminating branch data MID LOAD -1. .019 { current and energy output } F BLANK card terminating all switches 14GEN 1.0 50. -1. BLANK card ending source data C Total network loss P-loss by summing injections = 5.000000000000E-01 C Solution at nodes with known voltage. ... C Node Source node voltage Injected source current Injected source power C name Rectangular Polar Rectangular Polar P and Q MVA and P.F. C GEN 1.0 1.0 1.0 1.0 0.5 0.5 C 0.0 0.0 0.0 0.0 0.0 1.0000000 C Column headings for the 7 EMTP output variables follow. These are divided among the 5 possible classes as follows .... C First 1 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 1 output variables are branch currents (flowing from the upper node to the lower node); C Next 5 output variables are either power or energy or both as a pair (column 80 punches > 4). C Step Time LOAD MID GEN GEN LOAD LOAD MID C TERRA LOAD MID MID TERRA TERRA LOAD C *** Phasor I(0) = 1.0000000E+00 Switch "MID " to "LOAD " closed in the steady-state. C 0 0.0 0.5 1.0 0.5 0.0 0.5 0.0 0.0 C 1 .5E-3 .4938441703 .9876883406 .4877641291 .246941E-3 .4877641291 .246941E-3 .246941E-3 C 2 .1E-2 .4755282581 .9510565163 .4522542486 .4819456E-3 .4522542486 .4819456E-3 .4819456E-3 BLANK card ending node voltage outputs (none for this subcase) C 40 .02 0.5 1.0 0.5 .005 0.5 .005 .005 C Variable maxima : 0.5 1.0 0.5 .005 0.5 .005 .005 C Times of maxima : 0.0 0.0 0.0 .02 0.0 .02 .02 C Variable minima : -.5 -1. .496087E-31 0.0 .496087E-31 0.0 0.0 C Times of minima : .01 .01 .005 0.0 .005 0.0 0.0 PRINTER PLOT 184 2. 0.0 10. GEN MID { Axis limits on branch power: (0.000, 5.000) BLANK card ending plot cards BEGIN NEW DATA CASE C 10th of 12 subcases will illustrate passing of switch power and energy C to TACS. New (as of 31 October 2006) TACS sources are Type-94 for power C and Type-95 for energy. Otherwise, they are like Type-91 for current. C There are no dynamics, so dT is immaterial except that we want smooth C curves, so take 100 steps over one cycle. Pass one sinusoidal ampere C through a 1-ohm resistor. Note the column-80 punch "8" on the switch C to append both switch power flow and energy flow. Of course, the TACS C sources should be identical. Output shows this. Numbers are simple C enough to be verified with a pocket calculator. WSM, 3 November 2006 PRINTED NUMBER WIDTH, 12, 2, { Request maximum precision (for 9 output columns) .000200 .020 1 1 1 1 1 -1 5 5 TACS HYBRID 94GEN { Power flow through the switch having A6 terminal node "GEN " 95SWIT { Energy flow through the switch having A6 terminal node "SWIT " 77GEN 1.0 { Initial condition on the Type-94 power source avoids zero 33GEN SWIT { Output the values (power and energy) of these two new sources BLANK card terminates all TACS data SWIT 1.0 { 1-ohm resistor connects the switch to ground BLANK card ending all BRANCH cards GEN SWIT MEASURING 8 BLANK card ending all SWITCH cards 14GEN 1.0 50. -1. C --------------+------------------------------ C From bus name | Names of all adjacent busses. C --------------+------------------------------ C SWIT |TERRA *GEN * C GEN |SWIT * C TERRA |SWIT * C --------------+------------------------------ BLANK terminates the last SOURCE card C Total network loss P-loss by summing injections = 5.000000000000E-01 C Output for steady-state phasor switch currents. C Node-K Node-M I-real I-imag I-magn Degrees Power Reactive C GEN SWIT 1.00000000E+00 0.00000000E+00 1.00000000E+00 0.0000 5.00000000E-01 0.00000000E+00 GEN { Just one node voltage output C Column headings for the 5 EMTP output variables follow. These are divided among the 5 possible classes as follows .... C First 1 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 2 output variables are either power or energy or both as a pair (column 80 punches > 4). C Next 2 output variables belong to TACS (with "TACS" an internally-added upper name of pair). C Step Time GEN GEN GEN TACS TACS C SWIT SWIT GEN SWIT C *** Phasor I(0) = 1.0000000E+00 Switch "GEN " to "SWIT " closed in the steady-state. C 0 0.0 1.0 1.0 0.0 1.0 0.0 C 1 .2E-3 .998026728 .996057351 .199606E-3 .996057351 .199606E-3 C 2 .4E-3 .992114701 .984291581 .397641E-3 .984291581 .397641E-3 C 3 .6E-3 .982287251 .964888243 .592559E-3 .964888243 .592559E-3 C Column headings for the 5 EMTP output variables follow. These are divided among the 5 possible classes as follows .... BLANK card ends OUTPUT variable requests C 100 .02 1.0 1.0 .01 1.0 .01 C Variable maxima : 1.0 1.0 .01 1.0 .01 C Times of maxima : 0.0 0.0 .02 0.0 .02 C Variable minima : -1. .80245E-31 0.0 .80245E-31 0.0 C Times of minima : .01 .005 0.0 .005 0.0 PRINTER PLOT 194 4. 0.0 20. TACS GEN Axis limits: (0.000, 1.000) 194 4. 0.0 20. TACS SWIT Axis limits: (0.000, 10.000) BLANK card ending all batch-mode PLOT cards BEGIN NEW DATA CASE C 11th of 12 subcases will illustrate alternative [Z]-based iteration C that first is being made available to others in January of 2007. Data C is identical to the 1st subcase, and so is the solution, in spite of C the large number of output digits (typically 10). The Newton iteration C is that accurate (this proves it for 1 arrester with 1 exponential). C Note the new declaration immediately after the opening Type-92 branch C card. This is new. Column position and case of the request words are C critical, so to avoid mistakes, paste this to other data of interest. C The declaration is required only for the 1st NL element of a subnetwork C that is to be solved with impedance rather than admittance formulation. C Beware of reference branch use, however, as the flag goes with data. C I.e., a different subnetwork that requests a copy of this element for C its first will automatically be solved using [Z] rather than [Y] whether C or not this is the user's desire. The choice of solution method follows C the original data. Finally, only use [Z] with Type-92 exponentials as C service begins 8 January 2007. WSM. PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (for 8 output columns) NO Y-BASED NEWTON { Every subnetwork is to be solved using [Z] rather than [Y] .000050 .020 1 1 1 0 1 -1 0 2 10 33 1 40 10 100 50 -1SEND REC .306 5.82 .012 200. 92REC { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 [Z]-based Newton iteration { Column and case matter. Declare not use of [Y] C VREF VFLASH VZERO COL 778000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14SEND 408000. 60. BLANK card ending source data C Step Time REC SEND REC SEND C TERRA TERRA C 33 .00165 0.0 331579.2191 0.0 11.37746957 C 34 .0017 446716.5798 327039.3298 .0013592657 7.950150616 1 { Request for all node voltage outputs PRINTER PLOT 144 3. 0.0 20. REC { Axis limits: (-7.111, 6.930) 194 2. 0.0 20. REC { Axis limits: (-2.499, 2.499) BLANK card ending plot cards BEGIN NEW DATA CASE C 12th of 12 subcases is appended 5 February 2009 to illustrate two new C request words that prevent reverse current flow. Until now, the third C quadrant (having negative voltage and current) is assumed to be a copy C of the 1st quadrant (having positive voltage and current) except for the C reversal of signs. But Prof. Hans Kr. Hoidalen in Trondheim, Norway, had C a desire to prevent reverse current flow. He wanted to model a physical C diode as an ideal diode in series with an exponential characteristic to C account for the forward drop. All of this is contained within the Type- C 92 exponential ZnO element provided one of two new special request words C is used. Either 1) Diode model allows reverse leakage or C 2) Diode prohibits reverse current C To illustrate, copy that single exponential of the 1st subcase. First, C in its original form, 2nd with a leaky diode, and 3rd with an ideal C diode. These are nodes 1) ZNO, 2) LEAKY, & 3) IDEAL, respectively. C The leakage branch is a linear resistor that draws the same current as C the exponential at voltage V-min = 0.5 per unit = 778 kV / 2 = 389 kV. C Drive all 3 alternatives from the same sinusoidal voltage source and C compare the resulting currents. For positive voltage, all 3 currents C will agree. But for negative voltage, all 3 never will agree. For C negative voltage, IDEAL will carry no current, of course. The other C two will carry current, and for voltage less than 389 kV the currents C will agree. This is for ZNO and LEAKY. But for negative voltages C higher than 389 kV, the current of ZNO will be larger. The source C amplitude of 400 kV is purposely reduced to make both the linear and C the exponential portions easily visible on a resulting plot. WSM. PRINTED NUMBER WIDTH, 12, 2, { Request maximum precision (for 8 output columns) C MAXZNO EPSZNO EPWARN EPSTOP ZNOLIM1 ZNOLIM2 ZINC OXIDE 20 1.D-8 1.D-3 0.1 0.6 1.5 default values .0002 .010 { Half a 50-Hz cycle will vary voltage from max + to max - 1 1 C Let's connect resistors of 10K ohms in series with the 3 arrestors: SEND ZNO 10000. SEND LEAKY 10000. SEND IDEAL 10000. 92ZNO { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 C VREF VFLASH VZERO COL 778000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) 92LEAKY { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 Diode model allows reverse leakage { 3rd quadrant consists of leakage-R only C VREF VFLASH VZERO COL 778000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 C This V-min = 0.5 was per unit voltage for end of linear segment. Reduce it: C 1250. 26. 0.1 9999. { Bound on exponential segments (only one precedes) 92IDEAL { Type 92 is for v-i curve } 5555. { 5555 flag is for exponentials } 1 Diode prohibits reverse current C VREF VFLASH VZERO COL 778000. -1.0 0.0 2.0 C COEF EXPON VMIN 1250. 26. 0.5 9999. { Bound on exponential segments (only one precedes) BLANK card terminating branch data BLANK card terminating all (in this case, nonexistent) switches 14SEND 400000. 50. BLANK card ending source data SEND IDEAL ZNO LEAKY { Output source voltage followed by 3 arrester volt BLANK card ends node names for selective voltage output PRINTER PLOT 194 1. 0.0 10. BRANCH Compare leakage Arrester current ZNO LEAKY IDEAL BLANK card ending plot cards BEGIN NEW DATA CASE BLANK