247 From Eq, (12.27) Arg)» 23104 J008 ogo suai = 025i 12" pe oi avy) 20.0701 = 30-1999 9.2404 121,79" = 0.062458.21° per unit 30. Change in -phase voltage at bus @ is AV = AV). AVS? -+Av = (0.1508~0.0650—0 8624) , 121.70" = 0.092 per unit Chapter 13 Problem Solutions 13.1 For a generating unit the fuel input in millions of Btu/h is expressed as a function of output P, in megawatts by 0.032P? + 5.8P, +120. Determine (a) the equation for incremental fuel cost in dollars per megawatthour as a function of P, in megawatts based on a fuel cost of $2 per million Btu. the average cost of fuel per megawatthour when P, = 200 MW. ) (c) the approximate additional fuel cost per hour to raise the output of the unit from 200 MW to 201 MW. Also find this additional cost accurately and compare it with the approximate value. Solution: (2) The input-output curve in dollars per MWh is f = (0032P?+58P, +120) x2 = 0.064P? + 11.6P, +240 8/MWh The incremental fuel cost is af 0.128P, +116 S/MWa (8) ‘The average cost of fuel when P, = 200 MW is Hi £ 0.064(202)? + 11.6(200) + 240 Prleyatoo —- 25.6 $/MWh. (c) The approximate incremental cost for an additional 1 MW generation when P, = 200 MW is, 0.128(200) + 11.6 = 37.28/h248 ‘The additional cost per hour to raise the output from 200 MW to 201 MW can be calculated accurately as follows: 201 1 _ (0.128P, +116), = 0064? +116R,| = 37.264 s/n hn ho 13.2 The incremental fuel costs in $/MWh for four units of a plant are he = 0.01271 +9.0 do = 42 = 0.0096P,. +60 Pa _ th _ _ dhe _ . As = Fey = 00B%SH80 A = FI = 0.0068F,. + 100 Assuming that all four units operate to meet the total plant load of 800 MW, find the incremental fuel cost of the plant and the required output of each unit for economic dispatch. Solution: é 1 1 1)" _ a Lota dt _ ~ (core oom * ons oom) = 227600 baba Bs br = on (24 242+) art ant ag Ta — rr 2are x20" (555+ acons + gues *oome) > 88 Py = S00 Tl A = arPyr + br = 2.176 x 10°? x 800 + 8,368 10,1088 $/MWh Using Eq, (134), for each unit we have Act _ 1.1088 _ yor088 9 , 2088 =2 _ oa yw 1o.1088 — Pa = 102088 —6 ; 3088=8 _ oe mw 10.1088 Pys = 101088 — 8 - 1088 = 8 oe3.6 MW 10.1088 — 10 | --. 7 0.0068 OLN 13.3 Assume that maximum load on each of the four units described in Prob. 13.2 is 200 MW, 400 MW, 270 MW and 300 MW, respectively, and that minimum load on each unit is 50 MW, 100 MW, 80 MW and 110 MW, respectively. With249 these maximum and minimum output limits, find the plant A and MW output of each unit for economic dispatch Solution: The solution to Prob. 13.2 shows that each unit’s output would be 92.4 MW, 428 MW, 263.6 MW and 16 MW, respectively, if there were no maximum and minimum limits on unit outputs. It is seen that the output of Unit 2 violates its upper limit, and the output of Unit 4 violates its lower limit. This fact does not necessarily mean that the outputs of Units 2 and 4 should be set at their upper and lower limits, respectively. In fact, these limits should be checked individually. First, assume that Unit 2 is operating at its upper limit of 400 MW. Using the remaining Units 1, 2 and 3, we calculate the plant 2 as follows: (t+2+2) -( a : ) 2.813793 x 10% oe Nat as tae 0.012 0.008 * 0.0068 oo ) = asumecao (p85 vghg + gid) = sosnee Sine P,2 = 400 MW, the total output of Units 1, 3 and 4 should be 400 MW. Therefore, A = orPyr tbr = 2813793 x 10-* x 400 + 9.062069 = 10.187536 $/MWh Using this plant A, each unit’s output is Arb: _ 10.187586-9 Py = AS = SE = 989655 Mw = Acts _ 10187586-8 : Py = SB = SO 7 ade Mw Axbe _ 10187586 —10 ; Py = St = Seo 27 5802 Mw Py & 400 MW It is seen that the outputs of Units 3 and 4 violate their respective upper and lower limits. Consequently it is concluded that other units besides Unit 2 need be operating at their limits if the output of Unit 2 is specified to be 400 MW. ‘This time assume that Unit 4 is operating at its lower limit of 110 MW. Using Units 1, 2 and 3 only, the plant 2 is calculated as follows: aa) = 32x 107% i: i (sw 0.008, tr = or(Sa Bab) = ance (58s racist hg) = 16 1 or = Tor * Toa96 * T.008, Since P,« = 110 MW, the total output of Units 1, 2 and 3 should be 690 MW. Therefore, A = arPyr tbr = 3.2x 10"? x 690+7.6 = 9.808 $/MWh Using this plant A, each unit's output is Anbr _ 9808-9 Pa a OID 3333 MW250 13.4 Axbe _ 9.808~6 Py = AGB = SEES 2 sop.6667 sw Ant _ 9308 7 Py = ASD = SORES Sos mw Pe S OMW Apparently there’are no limit violations here. Therefore, economic dispatch requires that the output of Unit 4 be set to its lower limit of 110 MW and that the outputs of the remaining units be those obtained above Solve Prob. 13.3 when the minimum load on Unit 4 is 50 MW rather than 110 MW, Solution: It was shown in Prob. 13.3 that iffthe output of Unit 2 is set to its maximum limit of 400 MW, some other units will also have to be operating at their limits. We now examine whether load limit constraints will be violated if Unit 4 is operating at its new lower limit of 50 MW. Using Units 1, 2 and 3, the plant Ais caleulated as follows *l(4 42,2)" 3 ) > (sem * arom com) = 22+ bs b) _ soyw?( 2-46, 8)_2 (S++) ~ 920° (oat oom tai) = 78 Since P;« = 50 MW, the total output of Units 1.2 and 3 should be 750 MW. Therefor, 10 8/MWh y 0 d= orPyrtbp = 32x 10-9 x 750476 Each unit's output is calculated as follows: Ash _ 10-9 Pas Shs Oop = ean Mw _ Ack _ w-6 Po = ASB = agg a Ais.e807 Mw a _ Py = 10—8 = 250 MW a 0.008 It is noted that the output of Unit 2 exceeds its maximum load limit. It follows from the above analysis that both Units 2 and 4 should be operating at their upper and lower limits, respectively. Therefor, let Py 5 400 MW and Py 2 50 MW, and find the plane Aas follows: or or = The plant 2 in this case is a arPyr tbr = 48x 10-9 350+84 = 10.08 3/MWhag gor Fe see A hr pty 13.5 251 ‘The outputs of Units 1 and 3 are calculated to be 1008 90 MW “O01 1008 — 8 : Py = = tee 7 280 MW In summary, soMW —P,s = 260 MW 00MW Py, = 50 MW The incremental fuel costs for two units of a plant are aft 8 dfs = = 0.012P, x > 0.008P,2 + 9.6 N= 41 +80 0 ae 8Pya where f is in dollars per hour and P, is in megawatts. If both units operate at all times and maximum and minimum loads on each unit are 550 MW and 100 MW, plot A of the plant in $/MWh versus plant output in MW for economic dispatch as total load varies from 200 to 1100 MW. Solution: At their lower limit of 100 MW, the incremental costs of the units are calculated as af = = oompread) = a2 th : Be a = 104 AAs the plant output exceeds 200 MW, initially the incremental fuel cost A of the plant is determined by Unit 1 alone and the additional power should come from Unit 1. This will continue until the incremental fuel cost of Unit 1 becomes $10.4/MWh, (i., 0.012P, 1 +8, 10.4) from which the values of P, ; = 200 MW. Therefore, for 200 < Pyr < 300, A = 0.0127), 48.0 = 0.012(Pyr ~ 100) +80 = 0.012P,7 +68 For Pyr > 300, both units will increase their outputs simultaneously, To determine which ‘unit will reach its upper limit first, we calculate incremental costs at the upper limit as follows: af 5 Briley 7 OU+8O], = 146 dfs Af = 0.008P,2 + 9.6| = 40 aya p, .=ss0 IPpa=ss0 ‘The result shows that Unit 2 will reach its maximum load limit earlier than Unit 1. The value of P,, for which the incremental cost becomes $14.0/MWh is computed from 0.012P, +8.0 = 140 which yields P, 1 = 500 MW. For 300 < P,r < 1050, the plant 2 is calculated. Since the incremental fuel costs of Units 1 and 2 should be the same, we have 0.012P,, +80 = 0.008P,2+9.6252 from which Py = 1.5P,~200, Since P,1 + Py2 = Pyr, P, can be represented in terms of Por as Py1 + 1.5Py1~ 200 = Pyr from which P,: = 0.4P,r-+ 80. The plant 2 is then given by A = 0.012P,;+8.0 = 0.012(0.4P,r +80) +8.0 = 0.0048P,7 +8.96 For Pyr > 1050,.only Unit 1 will have an excess capacity, and the plant A is determined by Unit 1 alone as » 0.012P,:+8.0 0.012 (Pyr ~ 550) +8.0 = 0.012P_r-+ 1.4 The results are summarized as follows: For 200 L, > Ls. Given t factor of plant 1, Ls, can be calculated from 10.0L; system 9 of $12.0/MWh, the penalty 2.0 from which we have ty = 12 |A power system has two generating plants and B-coefficients corresponding to Eq. (13.37) which are given in per unit on a 100 MVA base by x 10-8 The incremental fuel costs in $/MWh of the generating units at the two plants are aft dh aPyi aP; If Plant 1 presently supplies 200 MW and Plant 2 supplies 300 MW, find the penalty factors of each plant. Is the present dispatch most economical? If not, which plant output should be increased and which one should be decreased? Explain why. y= = 0.012P,1+6.6 de = 0.0096P,2 + 6.0 Solution: The power loss Pz, is given by the equation 5x 107? -0.03 x 10-8 | 0.15 x 10" ] [Ps Pim [Pov Prat] [coosxio® —_sxio-2| 0210-9 || A: | 0.15 x 10° 0.2 x 107 [0.06 x 10-> 1 5 x 10-8PS, ~2 (0.08 x 10°%) Py :Py2 +8 x 10-PZ. + 0.15 x 10-2 Py $0.2 x 107P,2 +0.06 x 10-2 where P, and Py2 are in per unit on the 100 MVA base. Penalty factors are calculated as 1 1= {2x (6x 10-4) Py, ~2(0.08 x 10-4) Pye +0.15 x 10-9}| Vane Tear xtos ~ 10705769 1 1 = {2 x (8 x 10-3) Ppp ~2(0.08 x 10-8) Py, $0.2 x 10-5}| | Tramps wios ~ 108008toseeee = a esveoot = a ‘aze suoryenbs anoge aya anjos Yoryse 24 pur ty jo sonrea ou, 1 soorses'e y860266'1 *4gviazez0't + *70686000'0- ©£70686000°0 ~ '€9¢99296°0 Se uonumal ze Yq ¢-01 * P9686'0 x o1 ~ OF + F9— = "a («01 x 895896'9 x 01 xZ+96°0) + 4a (c-0T x (BHrORD'O-) x OF Xz) 01 X POTOSL'O x OL OL +8— 4g (<8 x (gpparoo-) x oF xz) +z (¢-01 x gsteer'e x OF xz +80) ‘axey am ‘sonqea zadozd Zurnmansqns uody, (tay +4-) (tay +4) 2g (teagye +) + 84a ("2@Y2) eg (Haye) + a (Saxe +) se uaramar are ypryae (0tq + Mattge + Mattga)y +x — (9+ Maz) (hg + Mattae t+ Mattaa)y ty — (lat Mate) Ag pi 4g 40} pastos aq prnoys suorzenbe ow Sunnortos aya aexa (EH'eT) ba wos} swonO} 41 E G=RS uoatd St Q°01 = (y)X ZEMS yun aad ¢ og TERS ‘\ payepdn ue ureago 07 uoryeso3t 3s1y 243 BuLMp suo|reTMoTED AzesszoIU ay unoyred ‘F-¢T afdurexg ut y waysts jo anpea Sunres ayy se YMI/O'OIS BUS] 6'eT ‘ypredsip oqwouose axaryoe 01 pasvarsep aq prnous z aeyd Jo vey yr Posvarout aq plnous T wejd Jo indano aya “(= "arp/2f p) °7 wow solreus st (P/F p) HT ug Ha, ‘gp piseze'6 = a grisece’s = 7p *7 —leeeer'6 ‘axe am ‘s3S09 Jang feuauro.oU ay ou! pareiodsoout aze sroWey Arpeuad Hay tgp oP Coen) «Sp IY¥p 88°8 = 09 +008 x 96000 = ‘oe = 994002 x 7100 aq 01 porejnojes are sosnq auejd oma 942 28 s1800 any Jerous.OUT at, vst13.10 255 ‘Step 4: The transmission loss is computed to be Bu iPhy + 2Bi2PoiP)2 + Ba2Pia + BioPs1 + B2oPy2 + Boo 8.383183 x 107* x (2.062489)? + 2 x (0. 049448) x 10-3 x 2.062489 x 3.333861 + 5.963568 x 10° x (3.333861)" + 0.750164 x 1073 x 2.962480 + 0.38994 x 10-* x 3.333861 + 0.090121 x 10-? = 0. 104201 per unit ‘Step 5: Since Po = 5 per unit, Po+P.— (PX? + PE) = 5+0.10:201 - (2.062480 + 3.393861) 0.292148 per unit ‘The incremental change in Ais calculated from Eq, (13.58) as follows: Po + Pi? ~ (Pi) + PY) PY P+ = 0.202149 (10-9) 5.396350 — ax = (x9) and the updated \ becomes @ AM + Ar = 10—0.5413826 = 9.4586174 Suppose that bus @ of a four-bus system is a generator bus and at the same time a load bus. By defining both a generation current and a load current at bus @ as shown in Fig. 13.5c, find the transformation matrix C for this case in the form shown in Eq. (13.31). Solution: Let the generator and load currents at bus @ be denoted by JY and If, respectively. ‘The total system load current is given by In = H+h+h Constants dz, ds and dy are then obtained to be i a de Ib “BD &- Since the net current injection at bus @, Io, is If + If, from Eq, (13.23) we have & Min = Zirh + Zio (+18) + Zish + Zale = Zirh + Zialh + (e212 + dslis + diZi) Io from which we also get, eae =2i2 he =2i1 a birt list GZig ) * hZiat list Gly @2athZjs+ al, wth ~telf -t22 Ip256 13.11 where J2 = —Vin/Z1s. Now the load currents can be represented in terms of generator currents and the no-load current as follows Hf = ~detyhy — dato! - dts? cy mdstyl; — dgtol? = dgty 12 Ig = ~4etyhy ~ dete} ~ dats The transformation C of old currents J, If + If, 13 and Ig to the generator and no-load currents I,, Jf and [2 is defined as i 1 o 0 oe h h HHH) | mds toate ht || ac] Is wht dite at || 8 Pp h waits data dts | 1 * The four-bus system depicted in Fig. 13.5 has bus and line data given in Ta- ble 13.2. Suppose that the bus data is slightly modified such that at bus © P-generation is 4.68 per unit, and P-load and Q-load are 1.5 per unit and 0.9296 per unit, respectively. Using the results of Table 13.3, find the power-flow solu- tion corresponding to this modified bus data. Using the solution to Prob. 13.10, also find the B-coefficients of this modified problem in which there is load as well as generation at bus @. Solution: ‘The power flow solution should be the same as that of Table 18.3 except that P- and Q- generation at bus (2) needs to be modified to account for the load at that bus. Using P- and G-generation of 2.8 per unit and 1.325439 per unit as shown in Table 13.3, and adding to those P- and Q-load at bus Q), we get Pyp = 3.18415 = 4.68 per unit Qy2 = 1.225490+.0.9296 = 2.255090 per unit ‘The bus voltages and P- and Q-generation at bus () should remain the same, The load currents are calculated from the power flow results as follows: Ps 1.5 + 70.9296 To ,=2.49905° ~2.2 + 5136340 0.9605) ¢ 1.07932", 28+ j1.73500 943042 2e2E5e> ~1.5382150 + 70.8648990 = -2.2633193 + j1.4623529 2.881685 + 71.9741431 = 0.216412 + j0.0132372 = 0.3390421 ~ j0.0005983 0.4393167 — 70.012638913.12 257 Constants t; and t; defined in Prob. 18.10 are calculated to be — lat alata Zu Zia t delist Glia hoe = 0,9930664 + j0.0013435 = = 1.020780 — j0.0004610 Using the constants de, dg, d4, ty and t2, the transformation C defined in Prob. 13.10 is obtained: 1 o 0 =0.2200866 — 0.013432 0.778921 ~ j0.0131625 —0,2200866 ~ j0.0134432 0.336921 + 70.0001386 —0.3397464 + 70.0007558 —0.3366921 + j0.0001386 —0.4362876 + 70.0119610 -0.4402238 + j0.0128677 -0.4362876 + 70.0119610 c Using Rous given in Example 13.3, we then find 4543134 + 30 0.892027 + 70.076 0.938793 + 0.028045 0.892927 ~ 0.076535 0.938793 — 0.023045, BOTSAL4 +50 0.194991 + 70.054548 0.194991 — 70.054548 0.246415 + 50 CTRauwC” = x10-® ‘The no-load current is calculated, as given in Example 13.3, to be i Rt 0.000436 — 0.387164 Also using the power-flow solution, we have - 5 (ES) a = Le = 10 joore LFS) _ 10 josrenis an 1-30) 1.0196070 — 70.4388369 10, 2,43095" ‘The matrix T., of Eq, (13.36) is then calculated to be 8.894036+ j0 —- —1.336685+ 0.988211 0.364217 + 70.355146 T. =1,336685 - 0.388211 3.789452+j0_0.011490-+ j0.086254 | x 10-* 0.364217 ~ 0.355146 0.011499 — 0.086254 0.036937 + 70 ‘The Becoefficients are the real parts of the matrix Ta. Finally, the power loss is calculated as follows: Leiaise P, = [1gisis2 468|1][B]|__ 468 0.098153 per unit aes Tee ‘Three generating units operating in parallel at 60 Hz have ratings of 300 MW, 500 MW, and 600 MW and have speed-droop characteristics of 5%, 4% and 3% respectively. Due to a change in load, an increase in system frequency of 0.3 Hz is experienced before any supplementary control action occurs. Determine the amount of the change in system load, and also the amount of the change in generation of each unit to absorb the load change.258 Solution: ‘Using Eq, (18.65), the change in the system load is calculated to be Sas) Af ap Sas) Af Raa) Fa 00) 03 : 9) $8 = -1905 mw Sri Af _ _ 30003 ary = ~S2BL 093 oy py eee eee are = ~Z2 AF 8093 _ eos gw aps = —S28f _ 800.03 69 x¢y Rasta ~~ 003 60 13.13 A 60-Hz system consisting of the three generating units described in Prob. 13.12 is connected to a neighboring system via a tie line. Suppose that a generator in the neighboring system is forced out of service, and that the tie-line flow is observed to increase from the scheduled value of 400 MW to 631 MW. Determine the amount of the increase in generation of each of the three units and find the ACE of this system whose frequency-bias setting is 58 MW/0.1 Hz. Solution: ‘The increase in the total generation resulted in the increase in the tie line flow by 231 MW. Therefore, it follows from Eq, (13.65) that at = -( 500 so0) af 005 * 008 * 0.03) Go from which we have Af = 21x60 = -0.36 Hz ‘The thee units should have increased their outpute according to Eq, (1.6) as follows: 202 (038) AP, =~ = MW APy2 = “g7 = MW = 800 (-0.36) _ 4Pis = -Tag ee = UMW ‘The ACE of the system is determined from Eq. (13.68) as ACE = (631 ~ 400) ~ 10(-58)(-0.36) = 22.2 MW 13.14 Suppose that it takes 5 minutes for the AGC of the power system of Prob. 13.13 to command the three units to increase their generation to restore system fre- A13.15 259 quency to 60 Hz. What is the time error in seconds incurred during this 3- minute period? Assume that the initial frequency deviation remains the same throughout this restoration period. Solution: ‘The frequency error in per unit is 0.36 =6 x 107? per unit ‘Therefore, the time error incurred during the 5-minute 6 x 107 x5 x 60 = Solve Example 13.8 when the system load level is 1300 MW. Solution: ‘Among the four combinations z1, 22, £3 and zo, combination zs is infeasible since the total generation from units 1 and 2 cannot exceed 1250 MW, (4) Combination zs: With units 1, 2 and 4 operating, we use Eqs. (13.7) and (13.8) to calculate the coefficients or = * = (0.008-' +.0.0096-" + 0.0117#)"* = 3.1243 x 107° by oh he) (8, 64, TS fae «(2+ 2+) a or (sis mats 7) eat ‘The incremental fuel cost at the load level of 1300 MW is A = arPyr tbr = 3.1243 x 10-* x 1300+ 7.3374 = 11.9990 ‘The corresponding economic dispatch outputs are ed _ 12.3000 - Py 11999052 = coiss a 25 MW 3900 - Pra HAWO= 84 _ soo79 & 521 MW 11.3080 7 Pye — =f 354.45 = 354 MW = “oor The hourly production casts of the three units are calculated to be fy = 000473, 4807,:+800] = 4628 S/h fy = 0.004872, +6.4P,2 +400] 5087 $/h loyentts 0.0055P2,+75Fy4 +400] = 3T44S/h260 (4) Combination z2 Using the values of ar and ér found in Example 12.8, \ is ealeulated to be A= opPyr-+br = 3038 x 107? x 1950+ 7.4634 = 11.4128 The corresponding economic dispatch outputs are Azbr _ 11.4128- 80 Sa 426.60 = 427 MW 11.4128 - 6.4 = DANES = sonar = 522 MW TU 522.17 & 522 MV 11.4128 - 7.9 = 351.28 = 351 MW = oor ‘The hourly production costs of the three units are calculated to be aA 0.00475, + 807; + 500 = 4645 8/h fa = 0.0048P%, +647, +400) = 5049 $/h |eanena fs = 0.005P2,+ 797,24 ooo = 3989 $/h 3st (#8) Combination 2 Using the values of ar and by found in Example 13.8, A is calculated to be A = aPyr tbr = 2.3805 x 10-* x 1300474712 = 10.56585 ‘The corresponding economic dispatch outputs are Axb: _ 1056585 — 80 Pac SS85—80 7 501 MW 1058585 — 64 : _ 105688 = 64 soaoy a ase MW 1056585" 79 _ jon en = Pps = SSE— 79 255.59 = 256 MW Pac 1OSEEE5 78 arery x oro sw ‘The hourly production costs of the four units are calculated to be fi: = 0.004P2, +80P, . + 500| = 3480 S/h eyumsat fa = 0.0048P?, + 6.4P;2 + 400| = 4082 $/h rant J = 0.005P2, +7.9P, + 600| = 3055 8/h 206 Sa = 0.0085P2, +757; 4 +400] = 2021 $/h -2913.16 If the start-up costs of the four units of Example 13.9 are changed to $2500. $3000, $3400, and $2600, and the shut-down costs are changed to $1500, $1200. $1000, and $1400, respectively, find the optimal unit commitment polic; sume that all other conditions remain unchanged. Solution: ‘At stage 6, the condition remains the same. Therefore, Fa(6) = $45,868 At stage 5, by using different transition costs T, we have Fi(5) = {P,(5)+Ti9(5) + Fo(6)} = [58,428 + 1000 + 1400 + 45,868] = $ 105, 696 Fi(5) = {Pz (6)+Tho(5) + Fo(6)} = [59,356 + 1000 + 45,868] = $106,224 Fy(5) = {Ps(6)+Ts.0(5) + Fo(6)} = [58,236-+ 1400 + 45,868] = $105,504 At stage 4, Pua) = min [Ps(4) +Th1(4) + Fi(S)- P(A) + Te 24) + Fa(5); Pu(4) + Tia(4) + Fa(8)) ‘min [76,472 +0 + 106, 696; 76,472 + 1400 + 106, 224; 76, 472 + 1000 + 105, 504) min [183, 168; 184, 096; 182, 976] = $182,976 Fa(4)_ = min [Pa(4) + To.2(4) + Fi(5); Pi(4) + Taa(4) + Fa(5); Pa(4) + To.9(4) + F(5)] min (79, 184 + 2600 + 106, 696;79, 184 + 0+ 106, 224;79, 184 + 1000 + 2600 + 105, 504) ‘min (188, 480; 185, 408; 188, 288] = $185,408 At stage 3, FQ) = min{P,(8) +7318) + (M5 Pi) + T.2(8) + (4) rin 70, 908 +0 + 182, 976; 70, 908 + 1400 + 185, 408} = min 253,884; 257, 716] = $253, 884 Fx) = min(P2(3) +7208) + Fi(4); Pa(@) + To2(8) + Fa()] min [68, 976 + 2600 + 182, 976; 68, 976 +0 + 188, 408) tin [254,552;254, 984] = $254,384 AQ) = min[Pe(8) +208) +A) P03) + 728) + F(A] ‘min (67, 856 + 3400 + 182,976; 67, 856 + 3400 + 1400 + 185, 408] = min (254,232; 258,064] = $254,232 At stage 2, Fy(2) = min{Ps(2) + Th.0(2) + Fu(3); Pi(2) + Ti a(2) + Fa); Py(2) +Ti9(2) + Fa(3)} min [58,428 + 0-+ 253, 884; 58,428 + 1400 + 254, 384;58, 428 + 1000 + 254,292] mun 812, 312;314,212;313, 660] = $312,312 sin P(2) + Ta.(2) + Fu(3); Pa(2) + Ta.a(2) + Fe(3); P2(2) + Te.3(2) + F(3)] rin [59,956 + 2600 + 253, 884; 59,356 + 0 + 254, 384; 50,356 + 1000 + 2600 + 254,222] rin 315, 840;313, 740;317, 188] = $313,740 sin [Ps(2) + To.a(2) + Fi(8); Ps(2) + Ta a(2) + Fa(3); Pa(2) + To.3(2) + Fa(2)] ‘min 98, 236 + 3400 + 253, 884,58, 236 + 3400 +1400 + 254, 384; 58,296 +0 + 254,232] = min(315,520;317,420;312, 468) = $312,468 FQ)262 At stage 1, Fy) = min [Po(1) + To.2(1) + Fi(2); Pot) + Toad) + Fa(2); Pall) + Taal) + Fa(2)] ‘= min (45, 868 + 3400+ 2600 + 312, 312; 45, 868 + 3400 + 313, 740; 45, 868 + 2600 + 312, 468] =, min (364, 180; 363,008; 360,936] = $360,936 ‘When the least cost path is retraced, the optimal unit commitment is found to be the same as that of Example 13.9. The associated total operating cost in this case is $360,936, which is $600 Jess than $361,536 obtained in Example 13.9. 13.17 Due to a 400 MW short-term purchase request from the neighboring utility, the demand during the second interval of the day is expected to increase from 1400 MW to 1800 MW for the system described in Example 13.9. Assuming that other conditions remain unchanged, find the optimal unit commitment policy and the associated total operating cost for the day. Solution: In applying dynamic programming to this problem, the process up to stage 3 should be the same as that given in Example 13.9. At stage 2 no other combinations besides z; and z2 have sufficient capacity to serve the increased load of 1800 MW; therefore, we only have to consider combinations 2; and 22. It was found at stage 4 that to serve the load of 1800 MW, the minimum production costs of combinations z; and 22 would be $76,472 and $79,184, respectively. Stage 2 can now be handled as follows: FQ) = min[F,(2)+Ti22) + Fi(3);Pi(2) +7122) + Fa(3);Pi(2)+ T.s2)+F)] min (76, 472 + 0 + 254, 484; 76,472 + 1500 + 254, 884; 76, 472 + 1500 + 254,432] min [330, 956; 332, 856; 332, 404] = $330,956 Fy(2) = min {P(2) + Ta.a(2) + Fu(3);Pa(2) + Taa(2) + Fa(3); Pa(2) + Taal2) + Fa(3)] min [79, 184 + 3000 + 254, 484; 79, 184 + 0 + 254, 884; 79, 184 + 4500 + 254, 432} = min (236, 668; 334, 068;338, 116) = $334, 068 Fa) = min [PS(1) + Tal) + Fi(2);Po(1) + Boalt) + Fe2)] ‘min [45, 868 + 6000 + 330, 956; 45, 868 + 3000 + 334, 068] min [382, 824; 382, 036] = $382,824 ‘The optimal unit commitment is found by retracing the least cost path in the forward direc- ton, and 1s given by Stage | Optimal combination | Load level T Es 1100 MW 2 n 1800 MW 3 a 1600 MW. 4 mn 1800 MW 5 3 1400 MW 6 2 1100 MW, From F)(1), the total operating cost is found to be $382,824. a263 13.18 Suppose Unit 4 of Example 13.9 will have to be taken off line for 8 hours begin- ning at the fifth interval of the day to undergo minor repair work. Determine the optimal unit commitment policy to serve the system load of Fig. 13.11 and the increase in the operating cost for the day. Solution: Combinations =; or z3 are no longer a viable option during the 5** and 6° intervals, There. fore, the only F,(k) that needs to be evaluated at stage 5 is F,(5) = 106,724 which was given in Figure 13.14. Since 2» is the only combination available at stage 5, F,(4), for i= 1 and i= 2, are to be recalculated (using the already available information in Figure 13.14) as Fy(4) = 184,696 Fa(4) = 185,908 Now, recalculation of Fk) is required from stage 3 onwards. At stage 3, (3) = min(P.(8) + Ta) + AMA) + Tal) + A) rin (70,908 +0 + 184, 696;70, 908 + 1500 + 185,908} min (255,604; 258,316] = $255,604 Fy) = min[Ps(3) + F113) + Fi(4);Pa(S) + T2208) + F(4)] ‘min (68, 976 + 3000 + 184, 696; 68, 976 + 0 + 185, 908] min [256, 672; 254, 884] = $254, 884 F4(3) ‘min [Ps(3) + Ts,1(3) + Fy (4); Pa(3) + T3.2(3) + Fa(4)) = min|67,856 + 3000 + 184, 696; 67, 856 + 4500 + 185, 908] amin [255, 552; 258,264] = $255,552 At stage 2, FQ) = min{P,(2)+7.,2) + AG)sP.2) +722) + ABA) +732) + 0) min [58, 428 + 0 + 255, 604; 58, 428 + 1500 + 254, 884; 58, 428 + 1500 + 255, 552) min [314,032;314, 812;315, 480] = $314,032 amin [P3(2) + Ta.2(2) + Fi(3);P,(2) + Ta(2) + Fa(3);Po(2) +722) + Fa(3)] min (59, 356 + 3000 + 255, 604; 59, 356 + 0 + 254, 884; 59, 356 + 4500 + 255, 552) min (317, 960;314, 240;319, 408] = $314,240 min [P3(2) + Ta,1(2) + Fi(3); Ps(2) + Ta,2(2) + Fo(3); Pa(2) + Ta,a(2) + Fa(3)) rin (58, 236 + 3000 + 255, 604;58,236 + 4500 + 254, 884; 59,296 + 0 + 255,552] = min 316,840;317, 620;313, 788] = $313,788 At stage 1, Fa(1) = min [P3(1) + To.a(1) + Fi(2);Po(2) + Taal) + Fa(2); Pol) + Taal) + F2(2)] min [45, 868 + 6000 + 314, 032; 45, 868 + 3000 + 314, 240; 45, 868 + 3000 + 313, 783] = min 365, 900; 363, 108; 362, 656] = $362,656 ‘Tracing the process in the forward direction, we find the optimal policy as (79 3 23 21 22 9) from stage 1 to stage 6. The increase in the operating cost is 362, 656 - 961,536 = $1120264 13.19 A diagram similar to Fig. 13.14 is shown in Fig. 13.15 in which directed branches represent transitions from one state, represented by a node, to another. Associ- ated with each directed branch (i, j) is the cost fi; (k), as defined in Eq, (13.7 The values of fj (k) are given in Table 13.7. Note that index k of fij(k) does not play any role here, and consequently will now be omitted. If the value of fi is interpreted to be the distance between states i and j, then the unit commitment problem becomes that of finding the shortest path from the ori- gin, represented by node (), to the destination, represented by node @. The problem of this nature is called the stagecoach problem. Write the backward re- currence equation similar to Eq. (13.75), and solve the problem by commencing calculations at the destination and then moving toward the origin. In forward recurrence the process starts with the origin and moves toward the destination. Write the forward recurrence equation, solve the problem and check the result with that of the backward dynamic programming procedure Table 13.7 Matrix of costs (or distances) fi; between states (or nodes) @ and © of Fig. 13.15 i ® G®S®OOO®OOD 20°15 17 35 31 38 39 42 33 36 40 34 26 22 29 25 41 44 15 18 ° 1S @8988600 Solution: ‘The backward recurrence equation can be written as FON) = min (fy + FN) Starting with node 10, the destination, we have at stage 5 Fro(5) = 0265 At age 4, min {feo + Fio(5)] = min{15 +0] = 15 in [fo10 + Fio(6)] = min[18 +0] = 18 Ac sage 3 F(3) = min foo + Fo(4): Soo + Fo(4)] = min [26 + 19:22 + 18) = minj4t;40) = 40 ‘F,(3) min [fes + Fa(4); feo + Fo(4)] = min [29 + 15;25 + 18] min (44;43) = 43, F(3) =. min[fre + Fo(4); fro + Fo(4)] = min [Al + 15:44 + 18) = min|[56;62] = 56 At sage 2 F(2) = min [fos + Fa(8); foe + Fo(8); far + F(3)] = min|35 + 40;31 + 49,38 + 56] min |75;74;94] = 74 = min fas + Fo(3):fae-+ Fo(S); for + Fr(@)] = min[39-+ 40;42 + 43533 + 56] min (79;85;89] = 79 FA(2) =. min (fes + Fs(8); fos + Fol) far + Fr(@)] = min [96-+ 40,40 + 43,34 + 56) min [76;83;90] = 76 At stage 1 F(t) = min [fir + Fa(2)ifos + Fa(2)s fie + Fa(2)] = min 20+ 74:15 + 79:17 +76] min [94;94;93] = 93 Retracing the path in the forward direction, the shortest path is found to be 14445—-9-10 and the shortest distance is 93. “The forward recurrence equation can be written as BM) = min {y+ BU Starting with node 1, we have at stage 1, FQ) =0 ‘At stage 2, FQ) = min(f2+ Fi(t)] = min[20+0) = 20 FQ) = min{fis + Fi(1)] = min(15 +0] = 15 Fy(2) = min [fis + Fi()] = min[17 +0] = 17 At stage 3, (3) = min [for + Fa(Qhi fas + Fa(2)ias + Fa(2)] min 35 + 20; 39 + 15;36+ 17) = min (65;54;53] = 53 Fe(3) = min (foe + Fa(2):foe-+ Fs(2hi fas + Fa(2)] ‘min 31 + 20,42 + 18;40-+ 17] = min[51;57;57] = 51 (3) = min [for + Fa(2)i far + Fs(2)i far + Fal2)) min (38 + 20;33 + 1 34+ 17] min (58; 58:51)266 14.1 At stage 4, Fala) min [foe + F5(3); fea + Fe(3)i fra + (3) sin 26-+ 53;29 + 51;41 +51] = min{79;80;92} = 79 F(4) min [fs0 + Fs(3); foo + Fo(3)s fro + Fx(3)] sin [22 + 53;25 +51;44 +51] = min(75;76;95] = 75 At stage 5, Fyo(5) = min [foro + Fa(4); foro + Fo(4)) min (15 +79;18 +75] = min|94;93] = 93 Retracing the process from node 10 in the backward direction, the shortest path is identified to be wo—9-5—4—1 land the distance of the corresponding path is 93. This result is identical to that by the backward dynamic programming procedure. Chapter 14 Problem Solutions A four-bus system with Zpys given in per unit by ® ®@ ® @ ® [50.041 70.031 70.027 50.018 @ | 70.031 70.256 70.035 70.038 @ | 70.027 30.035 0.158 70.045 @ |jo.018 70.038 70.045 70.063 has bus voltages Vi = 1.020°, Vo = 0.98708, Vs = 0.960% and Vj = 1.0420°. Using the compensation current method, determine the change in voltage at bus @ due to the outage of line )-@ with series impedance j0.3 per unit. Solution: 1 (0.041 ~ 0.027) 0.014 2-9 2 2, _ | 3(0.031 -0.035) | _ | -Jo.00 = Moe} 1 | © | 5(0.027-0.158) | ~ | —70.131 ° 3(0.018 - 0.045) —30.027 By Eq. (14.14), hg = ce (0-096) 2581 18 = Zaist zis ~ JO0Id—j(-0181)—J08 ~ =F By Eq, (14.7), ) x (-y0008) = ~00108 per uni